Citation
I Can Explain...It's Something About the Way You Look Tonight

Material Information

Title:
I Can Explain...It's Something About the Way You Look Tonight Using Advertising Visuals to Transfer and Manipulate Multidimensional Brand Personality
Creator:
Grumbein, Adriane Kaye
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (484 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Mass Communication
Journalism and Communications
Committee Chair:
GOODMAN,JENNIFER ROBYN POTTER
Committee Co-Chair:
TREISE,DEBORAH M
Committee Members:
SUTHERLAND,JOHN C
SLAWSON,BRIAN L
Graduation Date:
8/9/2014

Subjects

Subjects / Keywords:
Advertising campaigns ( jstor )
Advertising research ( jstor )
Binocular vision ( jstor )
Brand image ( jstor )
Brands ( jstor )
Consumer advertising ( jstor )
Graphics ( jstor )
Images ( jstor )
Personality ( jstor )
Soaps ( jstor )
Journalism and Communications -- Dissertations, Academic -- UF
advertising -- bhpi -- brand -- brand-personality -- branding -- communication -- group-image-sort -- images -- mbti -- mixed-method -- personality -- scale -- survey -- visual
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Mass Communication thesis, Ph.D.

Notes

Abstract:
As the concept of branding has grown in popularity, experts disagree about whether the company or the consumer ultimately directs and determines a brand and its meaning. This study contends that brands are not completely dependent on consumer whims, but, rather, that companies still have the ability to guide and influence their brands through advertising visuals. According to meaning transfer theory, advertising is a dominant vehicle for the provision and movement of cultural meanings (McCracken, 1986; 1990), and, within advertisements, visuals play a significant role in imbuing brands with cultural meanings. Specifically, this study extends McCracken's meaning transfer theory to examine advertising visuals' ability to transfer multidimensional brand personality from culture to a brand. Using a mixed-method approach, this study includes a survey reduction, qualitative phase and quantitative phase. First, the Brand and Human Personality Indicator (BHPI) scale was created by shortening a brand personality scale, which is based on the Myers-Briggs Type Indicator (MBTI). Second, qualitative group image sorts were conducted to better understand the visual symbolic elements participants (n = 30) associated with the MBTI personality dimensions included in this study - extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). Finally, a survey (n = 526) was conducted using advertising images selected during the group image sorts. The survey sought to examine how respondents assigned multidimensional brand personalities to images. Attitude toward the ad, attitude toward the brand and purchase intent were also measured. According to this study's results, extreme similarities within the visual profiles of extraversion, intuition and feeling dimensions, and introversion, sensing and thinking dimensions made ENF and IST personalities fairly easy to transfer using only visuals. Results also suggested that the extraversion/introversion (E/I) dimension might be able to cross aesthetic borders most easily, allowing companies to create EST and INF personalities with careful attention to visuals. However, other multidimensional brand personalities might require the addition of text to anchor a company's intended meaning (Barthes, 1977). Additionally, this study provided a list of specific visual symbolic elements associated with particular individual and multidimensional MBTI brand personalities to aid researchers and practitioners in creating future brand personalities through images. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: GOODMAN,JENNIFER ROBYN POTTER.
Local:
Co-adviser: TREISE,DEBORAH M.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-08-31
Statement of Responsibility:
by Adriane Kaye Grumbein.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
8/31/2016
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

ABOUT THE WAY YOU LOOK TONIGHT: USING ADVERTISING VISUALS TO TRANSFER AND MANIPULATE MULTIDIMENSIONAL BRAND PERSONALITY By ADRIANE KAYE GRUMBEIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014

PAGE 2

© 2014 Adriane Kaye Grumbein

PAGE 3

To Matt, who stuck with me from beginning t o end, married me in the middle, and w as proud of me in every chapter

PAGE 4

4 ACKNOWLEDGMENTS This dissertation would not have been possible without the support of many people. First, I would like to thank my husband, Matt, and my family, who have loved and cheered me on th roughout my graduate career, roo ting for the day when I could stop going to school. I would like to thank my amazing dissertation advisor, Dr. Robyn Goodman, who treated me with kindness and respect, but always pushed me to do my best. I wo uld also like to thank my committee, Dr. Debbie Treise, Dr. John Sutherland and Mr. Brian Slawson, who worked side by sid e with me through this process. From plane rides to Aggie whoops and design chats , they have been full of both fun and guidance. And, m any thanks must go to Jody, Kim and Sarah, who answer ed ever y question with a smile and kept a c razy grad student just a little bit saner. Finally, I w ould like to honor the memory of my mom, who always loved and believed in me , giving me the confidence to start this journey. AMDG.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ......................... 10 LIST OF FIGURES ................................ ................................ ................................ ....................... 13 ABSTRA CT ................................ ................................ ................................ ................................ ... 14 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 16 Importance of the Study ................................ ................................ ................................ .......... 19 Purpo se of This Study ................................ ................................ ................................ ............. 22 2 LITERATURE REVIEW ................................ ................................ ................................ ....... 27 Brand Personality ................................ ................................ ................................ .................... 27 Brand Personality and Consumer Identity ................................ ................................ .............. 32 Measuring Brand Personality ................................ ................................ ................................ . 44 ................................ ................................ ............................. 46 Response to Brand Personality Scale Critiques ................................ ............................... 51 The Myers Briggs Type Indicator (MBTI) ................................ ................................ ............. 52 MBTI History ................................ ................................ ................................ .................. 52 MBTI Dimensions ................................ ................................ ................................ ........... 53 MBTI Application ................................ ................................ ................................ ........... 56 Study Brand Personality Dimensions ................................ ................................ .............. 57 Study Brand Personality Type Descriptions ................................ ................................ .... 58 ................................ ................................ ................ 66 Brand Personality & Visuals ................................ ................................ ................................ .. 70 Visuals Over Text ................................ ................................ ................................ ................... 74 History of Advertis ................................ ................................ .... 75 Advantages of Advertising Visuals Over Text ................................ ................................ 76 Visual Devices ................................ ................................ ................................ ........................ 83 Visual Techniques ................................ ................................ ................................ ........... 84 Visual Syntax ................................ ................................ ................................ ................... 88 Visual rhetoric ................................ ................................ ................................ .......... 90 Visual metaphors ................................ ................................ ................................ ...... 95 Visual exaggeration ................................ ................................ ................................ 100 Visual Symbolic Elements ................................ ................................ ............................ 101 Visual symbolic elements in advertising ................................ ................................ 103 Visual symbolic elements in gender studies ................................ .......................... 106 Meaning Transfer Theory ................................ ................................ ................................ ..... 110 Phase One of Meaning Transfer Process ................................ ................................ ....... 114

PAGE 6

6 Phase Two of Meaning Transfer Process ................................ ................................ ...... 119 Studies Using Meaning Transfer Theory ................................ ................................ ...... 121 Meaning Transfer Theory & the Brand ................................ ................................ ................ 123 Visual s in Meaning Transfer Theory ................................ ................................ .................... 126 Proposed Modifications to Meaning Transfer Theory ................................ .......................... 128 Research Questions and Hypotheses ................................ ................................ .................... 131 3 METHOD ................................ ................................ ................................ ............................. 137 Mixed Method Approach ................................ ................................ ................................ ...... 137 Dependent Variables ................................ ................................ ................................ ............. 139 Brand Personality ................................ ................................ ................................ .......... 139 Appropriateness of Single Item Measures ................................ ................................ .... 140 At titude Toward the Ad ................................ ................................ ................................ . 141 Attitude Toward the Brand ................................ ................................ ............................ 142 Purchase Intent ................................ ................................ ................................ .............. 143 Confounding Variables ................................ ................................ ................................ .. 144 Human personality ................................ ................................ ................................ . 144 Gender ................................ ................................ ................................ .................... 145 Scale Reduction ................................ ................................ ................................ .................... 145 Specific MBTI Scale ................................ ....................... 146 Specific MBT I Scale ................................ .... 146 Reliability ................................ ................................ ................................ ...................... 149 Pretest Data for Scale Reduction ................................ ................................ ................... 153 ................................ ...... 155 MBTI Dimensions in Study ................................ ................................ ................................ .. 156 Independent Variable ................................ ................................ ................................ ............ 158 Image Deck ................................ ................................ ................................ ........................... 158 Group Image Sorts ................................ ................................ ................................ ................ 168 Group Interview Definition ................................ ................................ ........................... 169 Advantages of Group Interviews ................................ ................................ ................... 17 0 Advantages of Group Interviews Over Individual Interviews ................................ ...... 171 Image Sort Definition ................................ ................................ ................................ .... 173 Advantages of Image Sorts ................................ ................................ ............................ 174 Qualitative Participants ................................ ................................ ................................ . 176 Number of Group Image Sorts ................................ ................................ ...................... 178 Screener Questionnaire ................................ ................................ ................................ .. 181 Moderator ................................ ................................ ................................ ...................... 182 ................................ ................................ ................................ ......... 183 ................................ ................................ ............................. 185 Procedure ................................ ................................ ................................ ....................... 186 Analysis ................................ ................................ ................................ ......................... 190 Survey ................................ ................................ ................................ ................................ ... 192 Advantages of Surveys ................................ ................................ ................................ .. 193 Choice of Survey Over Experiment ................................ ................................ .............. 194 Study Product ................................ ................................ ................................ ................ 197 Population ................................ ................................ ................................ ...................... 198

PAGE 7

7 Sample ................................ ................................ ................................ ........................... 199 Test Advertisements ................................ ................................ ................................ ...... 200 Questionnaire Pilot Test ................................ ................................ ................................ 201 Questionnaire ................................ ................................ ................................ ................. 202 Statistical Analyses ................................ ................................ ................................ ........ 203 4 RESULTS ................................ ................................ ................................ ............................. 205 Phase One Scale Reduction ................................ ................................ ............................... 205 Computing Overall Dimension Scores ................................ ................................ .......... 207 Point Biseria l Correlations ................................ ................................ ............................ 209 ................................ ................................ ................................ ........ 209 Reducing the Scale ................................ ................................ ................................ ........ 210 Reducing the extraversion/introversion (E/I) scale ................................ ................ 211 Reducing the sensing/intuition (S/N) scale ................................ ............................ 214 Reducing the thinking/feel ing (T/F) scale ................................ .............................. 217 Reducing the judging/perceiving (J/P) scale ................................ .......................... 221 The reduced scale ................................ ................................ ................................ ... 223 Phase Two Group Image Sorts ................................ ................................ .......................... 224 Group Image Sort Participants ................................ ................................ ...................... 224 Research Question 1 Visual Symbo lic Elements for Single Dimension MBTI Brand Personalities ................................ ................................ ................................ .... 228 Extraverted (E) ................................ ................................ ................................ ....... 231 Introverted (I) ................................ ................................ ................................ ......... 241 Extraversion/introversion (E/I) summary ................................ ............................... 251 Sensing (S) ................................ ................................ ................................ ............. 252 Intuition (N) ................................ ................................ ................................ ............ 261 Sensing/intuition (S/N) summary ................................ ................................ ........... 271 Thinking (T) ................................ ................................ ................................ ........... 272 Feeling (F) ................................ ................................ ................................ .............. 282 Thinking/feeling (T/F) summary ................................ ................................ ............ 293 Research Question 2 Visual Symbolic Elements for Multidimensional MBTI Brand Personalities ................................ ................................ ................................ .... 294 Extraverted/Sensing/Thinking (EST) ................................ ................................ ..... 297 Extraverted/Sensing/Feeling (ESF) ................................ ................................ ........ 298 Extraverted/Intuitive/Thinking (ENT) ................................ ................................ ... 299 Extraverted/Intuitive/Feeling (ENF) ................................ ................................ ...... 299 Introverted/Sensing/Thinking (IST) ................................ ................................ ....... 300 Introverted/Sensing/Feeling (ISF) ................................ ................................ .......... 301 Introverted/Intuitive/Thinking (INT) ................................ ................................ ..... 301 Introverted/Intuitive/Feeling (INF) ................................ ................................ ........ 302 Research Question 3 Understanding Images ................................ .............................. 303 Group image sort themes ................................ ................................ ........................ 303 MBTI personality dimension themes ................................ ................................ ..... 307 Image related themes ................................ ................................ ............................. 309 Research Question 3 (RQ3) summary ................................ ................................ .... 322 Research Question 4 Most Representative Images ................................ .................... 324

PAGE 8

8 Extraverted/Sensing/Thinking (EST ) ................................ ................................ ..... 325 Extraverted/Sensing/Feeling (ESF) ................................ ................................ ........ 325 Extraverted/Intuitive/Thinking (ENT) ................................ ................................ ... 326 Extraverted/Intuitive/Feeling (ENF) ................................ ................................ ...... 326 Introverted/Sensing/Thinking (IST) ................................ ................................ ....... 328 Introverted/Sensing/Feeling (ISF) ................................ ................................ .......... 328 Introverted/Intuitive/Thinking (INT) ................................ ................................ ..... 329 Introverted/Intuitive/Feeling (INF) ................................ ................................ ........ 329 Research Question 4 (RQ4) summary ................................ ................................ .... 331 Phase Three Survey ................................ ................................ ................................ ........... 332 Survey Respondents ................................ ................................ ................................ ...... 335 Scale Reliability ................................ ................................ ................................ ............. 336 Individual and Multidimensional MBTI Personality Dimensions ................................ 340 Confounding Variables for Hypotheses ................................ ................................ ........ 342 Hypothesis 1 Personality Within Advertisements ................................ ...................... 347 Supported hypotheses ................................ ................................ ............................. 347 Partially supported hypothesis ................................ ................................ ............... 350 Unsupported hypotheses ................................ ................................ ........................ 351 Hypo thesis 1 summary ................................ ................................ ........................... 355 Hypothesis 2 Personality Across Advertisements ................................ ...................... 355 Supported hypotheses ................................ ................................ ............................. 356 Partially supported hypotheses ................................ ................................ .............. 362 Hypothesis 2 summary ................................ ................................ ........................... 363 Research Question 5 Attitude To ward the Ad ................................ ............................ 363 Research Question 6 Attitude Toward the Brand ................................ ....................... 364 Research Question 7 Purchase Intent ................................ ................................ ......... 365 5 DISCUSSION ................................ ................................ ................................ ....................... 367 Phase One Discussion Scale Reduction ................................ ................................ ............. 368 Phase Two Discussion Group Image Sorts ................................ ................................ ........ 372 Phase Three Discussion Survey ................................ ................................ ......................... 397 Theoretical Implications ................................ ................................ ................................ ....... 411 Limitations ................................ ................................ ................................ ............................ 414 Scale Reduction ................................ ................................ ................................ ............. 414 Group Image Sorts ................................ ................................ ................................ ......... 415 Survey ................................ ................................ ................................ ............................ 417 6 CONCLUSION AND PRACTICAL APPLICATION ................................ ........................ 419 APPENDIX A SAMPLE ADS FOR SCALE REDUCTION DATA ................................ ........................... 428 B IMAGE DECK IMAGES ................................ ................................ ................................ ..... 429 C ................................ ................................ ................................ .... 434

PAGE 9

9 D DIMENSION DESCRIPTI ONS FOR GROUP IMAGE SORTS ................................ ........ 441 E SURVEY ADVERTISEMENTS ................................ ................................ .......................... 442 F HAND SOAP LOGO AND BOTTLE ................................ ................................ ................. 444 G QUESTIONNAIRE ................................ ................................ ................................ .............. 445 H NUMBERED IMAGE DECK IMAGES ................................ ................................ .............. 463 LIST OF REFERENCES ................................ ................................ ................................ ............. 467 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 484

PAGE 10

10 LIST OF TABLES Table page 2 1 Brand Personality Scale (BPS). ................................ ................................ ......................... 46 3 1 Dichotomous adjective pairs from the original brand specific MBTI scale. ................... 146 3 2 specific MBTI s cale. ........................... 148 3 3 Correlations between judging/perceiving (J/P) and other dimensions. ........................... 157 4 1 Original Alpha Range for Six Item Su bscales. ................................ ................................ 206 4 2 Item Scales. ................................ ......... 210 4 3 Extraversion/Introversion (E/I) Point Biser ial Correlation Results. ................................ 212 4 4 ................................ ............ 212 4 5 Sensing/Intuition (S/N) Poi nt Biserial Correlation Results. ................................ ............ 215 4 6 ................................ ........................ 215 4 7 Thinking/Feeling (T/F) Poin t Biserial Correlation Results. ................................ ............ 217 4 8 ................................ ........................ 218 4 9 Judging/Perceiving (J/P) Poin t Biserial Correlation Results. ................................ .......... 221 4 10 ................................ ...................... 222 4 11 Brand and Human Personali ty Indicator (BHPI). ................................ ............................ 224 4 12 Overall Demographics for Group Image Sort Participants. ................................ ............. 225 4 13 Individual Demographics for Gro up Image Sort Participants. ................................ ........ 227 4 14 Themes for Individual Brand Personalities in Images. ................................ .................... 228 4 15 Multidimensional Brand Pers onality Visuals. ................................ ................................ . 295 4 16 Images sorted as extraverted/sensing/thinking (EST). ................................ ..................... 325 4 17 Images sorted as extraverted/sensing/f eeling (ESF). ................................ ....................... 326 4 18 Images sorted as extraverted/intuitive/feeling (ENF). ................................ ..................... 327 4 19 Images sorted as introverted/sensing/thi nking (IST). ................................ ...................... 328

PAGE 11

11 4 20 Images sorted as introverted/sensing/feeling (ISF). ................................ ......................... 329 4 21 Images sorted as introverted/intuitive/thin king (INT). ................................ .................... 329 4 22 Images sorted as introverted/intuitive/feeling (INF). ................................ ....................... 331 4 23 Images selected for phase three survey. ................................ ................................ ........... 332 4 24 Survey advertisements. ................................ ................................ ................................ .... 334 4 25 Overall Demographics for Survey Respondents. ................................ ............................. 336 4 26 ........... 337 4 27 . ....................... 338 4 28 ........................ 338 4 29 /perceiving (J/P) subscale on survey. ..................... 339 4 30 ................................ ......... 340 4 31 Correct respo ndent choices over 50% for individual MBTI personality dimensions. ..... 340 4 32 Correct respondent choices over 50% for multidimensional MBTI personalities. .......... 342 4 35 MBTI personalities within the extraverted/intuitive/feeling participant advertisement (ENF participant ad). ................................ ................................ ................................ ....... 348 4 36 MBTI personalities within the extraver ted/intuitive/feeling researcher advertisement (ENF researcher ad). ................................ ................................ ................................ ........ 349 4 37 MBTI personalities within the introverted/sensing/thinking advertisement (IST ad). .... 349 4 38 MBTI personalities within the introverted/intuitive/feeling participant advertisement (INF participant ad). ................................ ................................ ................................ ........ 350 4 39 MBTI personalities within the introverted/intuitive/feeling researcher advertisement (INF researcher ad). ................................ ................................ ................................ ......... 351 4 40 MBTI personalities within the extraverted/sensing/thinking advertisement (EST ad). ... 352 4 41 MBTI personalities within the extraverted/sensing/feeling advertisement (ESF ad). ..... 352 4 42 MBTI personalities within the extraverted/intuiti ve/thinking advertisement (ENT ad). ................................ ................................ ................................ ................................ ... 353 4 43 MBTI personalities within the introverted/sensing/feeling advertisement (ISF ad). ....... 354

PAGE 12

12 4 44 MBTI personalities within the introverted/intuitive/thinking advertisement (INT ad). .. 354 4 45 Extraverted/sensing/thinking personality (EST) across advertisements. ......................... 356 4 46 Extraverted/intuitive/thinking personality (ENT) across advertisements. ....................... 357 4 47 Extraverted/intuitive/feeling personality (ENF) acros s advertisements. ......................... 358 4 48 Introverted/sensing/thinking personality (IST) across advertisements. ........................... 359 4 49 Introverted/sensing/ feeling personality (ISF) across advertisements. ............................. 360 4 50 Introverted/intuitive/thinking personality (INT) across advertisements. ......................... 360 4 51 Introverted/intuitive/feeling personality (INF) across advertisements. ........................... 361 4 52 Extraverted/sensing/feeling personality (ESF) across advertisements. ........................... 362 4 53 Between subjects effects for attitude toward the ad. ................................ ....................... 364 4 54 Between subjects effects for attitude toward the brand. ................................ .................. 365 4 55 Between subjects effects for purchase intent. ................................ ................................ .. 366 5 1 Brand and Human Personality Indicator (BHPI). ................................ ............................ 368

PAGE 13

13 LIST OF FIGURES Figure page 2 1 Proposed modifications to meaning transfer theory. ................................ ....................... 131 4 1 Continuums for overall dimensio n scores. ................................ ................................ ....... 209 5 1 Modifications to meaning transfer theory. ................................ ................................ ....... 412 6 1 Image of a man washing his dog, which was used during the group image sorts. .......... 425

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy I CAN EXPLAI ABOUT THE WAY YOU LOOK TONIGHT: USING ADVERTISING VISUALS TO TRANSFER AND MANIPULATE MULTIDIMENSIONAL BRAND PERSONALITY By Adriane Kaye Grumbein August 2014 Chair: J. Robyn Goodman Major: Mass Communication As the concept of branding ha s grown in popularity, exper ts disagree about whether the company or the consumer ultimately directs and determines a brand and its meaning. This study contends that brands are not completely dependent on consumer whims, but, rather, that companies still h ave the ability to gu ide and influence their brands through advertising visuals. According to meaning transfer theory, advertising is a dominant vehicle for the provision and movement of cultural m eanings (McCracken, 1986; 1990), a nd, within advertisements , visuals play a significant role in imbuing brands with cultural meanings. Specifically, this study multidimensional brand personality from culture to a brand. Using a mixed method approach, this study include s a survey reduction, qualitative phase and quantitative phase . First, the Brand and Human Personality Indicator (BHPI) scale was created by shortening a brand personality scale , which is based on the Myers Briggs Type Indicator (MBTI). Second, qualitative group image sorts were conducted to better understand the visual symbolic elements participants ( n = 30) associated with the MBTI personality dimensions included in this study extraversion/ introversion ( E/ I ), sensing/ intuition ( S/N) and

PAGE 15

15 thinking / feeling ( T/ F) . Finally, a survey ( n = 526) was conducted using advertising images selected during the group image sorts . The survey sought to examine how respondents assigned multidimensional brand personalities t o images. Attitude toward the ad, attitude toward the brand and purchase intent were also measured. with in the visual profiles of extraversion, intuition and feeling dimensions , and introversion , sens ing and thinking dimensions made ENF and IST personalities fairly easy to transfer using only visuals. R esults also suggest ed that the extraversion/introversion (E/I) dimension might be able to cross aesthetic borders most easily, allowing companies to cre ate EST and INF personalities with careful attention to visuals. However, other multidimensional brand personalities might require the addition of text to anchor intended meaning (Barthes, 1977). Additionally, this study provided a list of spec ific visual symbolic elements associated with particular individual and multidimensional MBTI brand personalities to aid researchers and practitioners in creating future brand personalities through images.

PAGE 16

16 CHAPTER 1 INTRODUCTION As the concept of brandin g has grown in popularity, many brand scholars, practitioners and gurus have preach ed about the power of the consumer to ultimately direct and determine the brand and its meaning (e.g., Davis, 2000; Elliot, 2006; Neumeier, 2006, Plummer, 1984/2000). While Neumeier (2006) argued that whole extent of a brand could be summed up as the omy, everything from the logo design (Hampp, 2010; Zmuda, 2013) and commercial con tent (Elliot, 2006) to pricing and discounts (Brewer, 2010). However, in the excitement, these voices forgot the power of the advertiser to shape a brand and mold consumer perceptions. Amidst the turmoil, talented agency creatives continued to produce pers uasive advertisements that convinced consumers Apple was for the sophisticated, artistic mind and Pepsi was th e voice of the next generation. Rather than being forced to relinquish brand control to consumer whims, this study suggests that a company may be largely in command of its own brand. As a touchpoint where the brand and consumer meet, advertising not only provides a company with the means to define its brand, but also to sway consumer perceptions. This argument is supported by Jewett and (2013) research that showed advertisements, particularly advertising visuals, could image only advertisement resulted in the vast majority of consumers (80% t o 92%) ascribing the

PAGE 17

17 brand with the marketer intended brand personality. Thus, advertisers defined the brand. explanation for why consumers are more lik message as it is rather than reject it. Hall (1980) asserts that media messages, such as advertisements, are encoded (i.e., given meaning) by the producer and then decoded (i.e., interpreted) by the receiv er. When consumers decode media messages, they do so in one of three ways: a dominant (or preferred) reading, a negotiated reading or an oppositional reading. In a exact intended meaning. A negotiated reading indicates that the receiver understands and periences and interests temper their interpretation (Chandler, 2007; Hall, 1980). Lastly, in an oppositional reading, the media message in a completely contrary wa y with an alternate frame of reference (e.g., feminist, radical) (Chandler, 2007; Hall, 1980). For example, suppose Starbucks produced an advertisement to promote its fair trade coffee and convince consumers that it is a socially responsible company. A co nsumer engaged in a dominant reading will view the ad and agree that Starbucks is a socially responsible company. Alternately, a consumer employing a negotiated reading may view the ad and agree that, for the most part, Starbucks is socially responsible co mpany, although he/she knows the company uses a large number of throw away paper cups. Finally, a consumer using an oppositional reading may view the ad, understand that Starbucks wants him/her to think that it is a socially responsible

PAGE 18

18 company, but choose s instead to think of it as an overgrown, money grubbing corporation who is exploiting its store employees by underpaying them while purchasing the majority of its coffee from poverty stricken farmers. However, true oppositional readings such as this are i ncredibly (i.e., hegemonic) perspective, which results in a dominant or negotiated reading (Hall, 1980). As members of a particular culture with specific sets o f widely held beliefs, it is almost impossible for an individual to completely step outside of those near universal views, which, having been learned at such an early age, seem natural (i.e., real or true) (Hall, 1980, p. 167). Although Hall (1980) recogni zed that it was possible for individuals to decode different meanings from a single media message, he asserted that most individuals would decode a message within the preferred meaning framework. He stated, We say dominant always possible to order, classify, have the institutional/political/ideological order imprinted in them and have the whole social order embedded in them as a set of meanings, practices and practical purposes in this culture[.] (Hall, 1980, p. 169) According to Hall (1980), the culturally based preferred meaning acts as a guiding force, (1980) three readin gs dominant and negotiated consumers both understand and agree with the engage in the oppositional reading understand the preferred meaning, though they choose not to accept it. So, although a consumer may inevitably be influenced by his/her own, unique life experiences and engage in a negotiated reading of company advertisements, the strength of the preferred meaning will strongly influence his/her decoding. Th

PAGE 19

19 encoding/decoding framework further reinforces the idea that a company can use advertising to create a desired brand personality, which will be read, understood and accepted by the vast majority of consumers with little or no alterat ion. However, in order for consumers to decode the brand personality messages through any reading, the advertiser must first encode it. According to Hall (1980), messages are encoded through the production process, which includes drawing from a variety of socio cultural and political meanings. For example, Hall (1980) asserted that television newscasts cannot transmit a ough audio and visual components (Hall, 1980, p. 164). In the same way, advertising creatives are responsible for encoding an ad with preferred brand personality messages through culturally understood audio and visual elements during the production process . McCracken (1986; 1 988) referred to the culturally based elements that communicate meaning as symbolic elements . In particular, advertising visuals are largely responsible for imbuing products and brands with specific cultural meanings (McCracken, 1986). During production, the creative team carefully chooses the visual images and symbols that make up an ad (McCracken, 1990). In turn, these visuals then provide consumers with precise, predetermined meanings a preferred reading about topics such as gende r, status, age, lifestyle, time, place and personality (Jewett & Sutherland, 2013; McCracken, 1986; 1989; 1993). The current study will refer to these tools of visual meaning creation as visual symbolic elements . Importance of the Study Understanding how a those messages, and who controls a brand becomes increasingly important considering that modern consumers are bombarded every day with advertisements to buy this, wear that, eat here

PAGE 20

20 and go t here. According to some estimates, American consumers may see as many as 5,000 15,000 ads each day, a dramatic increase from the estimate of only 500 per day in the 1970s (Berger, 2011; Danesi, 2008; Johnson, 2009; NAA, 2009). Additionally, in 2011, advert isers spent more than $158 billion in the U.S. market, and ad spending is projected to surpass $189 bil lion by 2016 (eMarketer, 2012). study, it only takes a walk down the ai sles of the nearest supermarket or megastore to see the Ludlow, 2002; So rrell, 2000). Products (such as new car models) that previously took five years to develop can now be produced in only 18 months (Sorrell, 2000). These dramatic decreases in initial production time mean competitors can replicate new ideas and produce compe ting products much quicker, making it difficult for a brand to stand out based on product features alone. Instead, marketplace differentiation requires strong brands that take advantage of emotional, psychological and lifestyle differences, such as brand p ersonality (Sorrell, 2000). Not only are consumers today seeing an unprecedented number of advertisements from an innumerable number of brands, but an increasing percentage of modern print advertising is 2008, p. 277). In fact, the last century has seen pictures squeezing text out of advertising designs as the space devoted to images grows proportionally larger (McQuarrie, 2008). According to Barthes (1977), there is no doubt that advertising images have i ntentional meanings imbedded in them a priori by the advertiser. Further, advertisers carefully choose visuals that will most clearly transmit their desired meanings about a product or brand. In fact, like Hall (1980), Barthes (1977) exp licitly stat ed that

PAGE 21

21 image is frank advertising images are so carefully chosen to transfer specifi c meanings, text may not be necessary to communicate brand messages. In fact, Barthes (1977) suggested that text, when aspects of the image are most important a nd guide them toward a dominant reading of the message. He suggested that anchoring text in an ad allows the advertiser to direct the reader to a controls [the view of an advertising message is fully contained in the visual, with text serving only to reiterate an already existing message, then the swing in modern advertising toward visual do minant (or visual only) advertising should not prohibit advertisers from clearly communicating a message about their brand or consumers from clearly understanding it. Text may also be unnecessary deeply embedded in a shared through either a dominant or negotiated reading. Related to the idea that text is not necessary, Williamson (1978/2010) went as fa r as to question whether language in advertisements might be seen as a barrier to reaching the idea or something , they do not directly , italics in original). And, if the filled picture, perhaps image only ads present the most logical advertisin g solution.

PAGE 22

22 nicating brand messages on advertising visuals. Currently, there is a seri ous lack of academic scholarship focused on understanding how advertising visuals communicate brand messages, and scholars have begun to question why research on visual communication, particularly images, is so much less frequent than research on verbal co mmunication (Janiszewski, 2008; McQuarrie, 2008; Wedel & Pieters, 2008). According to McQuarrie (2008), the near domination of images in advertisements positions marketing and advertising scholars as those most capable of exploring and understanding visual persuasive power. Therefore, this study will attempt to begin filling this personality with no anch oring text to guide the viewer. Purpose of This Study When taken t increasing budgets all provide companies with substantial influence on consumers it is this influence that scholars must endeavor to understand. Therefore, this study seeks to ex plore the power advertising visuals have to transfer and manipulate multidimensional brand personalities. Little empirical research has been conducted to specifically test the power of visuals as vehicles for communicating brand personality (e.g., Boudreau x & Palmer, 2007; Park, Choi & Kim, 2005), and even fewer scholars have considered the possibility of imbuing brands with multidimensional personalities through visuals ( e.g., Cunningham, Thach & Thompson, 2007). Therefore, the current study will address t hese much neglected areas of advertising and brand personality research.

PAGE 23

23 brand personalities, the current study will investigate the specific symbolic visual elements con sumers ascribe to a particular personality. Similarly, like the other areas of scarce visual scholarship mentioned above, little academic research exists to guide advertisers as they choose specific images to communicate their brand personality to consumer s (Akay, 2001; Cunningham et al., 2007; Jewett & Sutherland, 2013). Because visuals are arguably the most important part of modern advertisements (e.g., Barthes, 1977; Janiszewski, 2008; McQuarrie, 2008; Williamson, 1978/2010), it is vital to understand ex actly how visuals communicate brand messages. By understanding how visuals translate to personality, this study endeavors to provide both scholars through advertising images. The little past research available shows that advertising visuals can transfer personality from an image to a brand (Jewett & Sutherland, 2013). But, even more powerfully, it also shows that advertising visuals are able to imbue brands with specif ic, predetermined personalities and Jewett and Sutherland (2013) showed that a single advertising image could effectively manipulate a single dimension of personali ty (e.g., extraversion (E) or introversion (I)), this study will expand this line of inquiry and examine whether a single advertising image can effectively communicate multiple personality dimensions simultaneously (e.g., extraversion/sensing/thinking (EST ) or introversion/intuition/feeling (INF)) and how those multidimensional personalities influence consumer attitudes, opinions and behaviors toward the measured with a brand personality scale modeled on the Myers Briggs Type Indicator (MBTI)

PAGE 24

24 (Strausbaugh, 1998; Sutherland, Marshall & Parker, 2004). Freling and Forbes (2005) found consumers recognized multiple brand personalities in a product even when only one dimension was manipulated. This finding provides additional support for the application of multidimensional personality scales (such as the MBTI) to study brand personalities as multifaceted rather than composed of only a single dimension. Based largely on Jungian method of interacting with the world around them based on four bipolar dimensions extraversion/introversion (E/I), sensing/intuition (S/N), thinking/feeling (T/F) and judging/perceiving (J/P) (J ung, 1923/1971; Myers, 1995; The Myers & Briggs Foundation, n.d.). Extraverts focus on the world around them and gain energy from other people, while introverts draw ene rgy from their inner world and prefer to spend time alone. The sensing/intuition (S/N) dimension analyzes how an individual takes in information. Sensing individuals concentrate on concrete information from their five senses, while intuitives interpret and add meaning to information, looking at big picture patterns. The thinking/feeling (T/F) dimension represents how individuals make decisions. Thinking individuals put more weight on objective principles, impersonal facts and logic, while feeling individual s place more emphasis on people and approach to structure. Judging individuals prefer structured, ordered environments and like to make decisions, while perceiving individuals would rather stay open and adaptable to new ideas and information. Using a mixed methods approach including scale reduction, qualitative group image sorts nd

PAGE 25

25 manipulate specific, multidimensional brand personalities through advertising visuals. Because study will work to refine and shorten the current scale, ma king it possible for survey participants to rate the brand personality of multiple ads without survey fatigue. Image sorts will also be conducted in small, qualitative group interviews , which will be referred to as group image sorts . Images and feedback fr om the group image sorts will be used to better understand what visual symbolic elements best represent multi dimensional MBTI personalities (e.g., EST, INF) and why. Finally, a quantitative survey will be conducted using the shortened MBTI brand scale and the most representative images chosen in the qualitative group image sorts. The top image for each multi dimensional personality will be used to create an advertisement for a single brand of hand soap. A quantitative survey will then be conducted to deter mine if a larger population understands the brand personality specific visual symbolic elements identified in the qualitative group interviews in the same way when they are shown in an advertising frame . The survey will also investigate the effects of diff attitudes, opinion and behaviors toward the brand. To augment the sparse amount of past methods approach will provide a uni que and robust look at the how companies can communicate specific brand messages to consumers through advertising visuals. The following chapter provides a review of literature, including a detailed look at the concept of brand personality and how it is me asured, an examination of the role visuals play in advertising and brand personality, and an overview of meaning transfer theory (and its ive and quantitative aspects. Chapter four presents the research results in three

PAGE 26

26 parts 1) scale reduction, 2) qualitative group interviews with image sorts and 3) quantitative survey. Chapter five then discusses the combined implications and limitations of the current mixed method study and suggests potential future research. Finally, chapter six offers conclusions and practical applications.

PAGE 27

27 CHAPTER 2 LITERATURE REVIEW Scholars have produced a substantial amount of research, seeking to understand brand behaviors, and function in advertising. However, an often overlooked aspect of both brand personality and advertising research is the role visuals play in comm unication. The following specifically on the role visuals play in its creation, transfer and manipulation. This chapter will also review the inherent advantages of using visuals to communicate, particularly in advertising. Finally, the following literature review will also explore role of visuals in communication and meaning transfer, looking specifically through the lens of meaning transfer theory. Brand Personality Simp used to describe Delbaere, McQuarrie & Phillips, 2011; J. Aaker, 1997; Plummer, 1984/2000). Furthermore, past research shows consumers have no trouble understanding and attaching personalities to brands because they largely recognize that brand personalities are similar to human personalities (Ramaseshan & Tsao, 2007). Just like an individual, a brand can be described in a variety of ways, including demographically (e.g., gender, age, race), psychographically (e.g., activities, interests, opinions) and in terms of personality traits (e.g., warmth, concern, sentimentality) (D. Aaker, 1996). Moreover, like human personalities, brand

PAGE 28

28 personalities are distinctive, enduring and multifaceted (Freling & Forbes, 2005), and they are created from a wide range of factors some related to the product (e.g., category, package, price, att ributes) and others unrelated (e.g., imagery, sponsorship, symbol, ad style, spokesperson/character, personification, marketing mix) (Batra et al., 1993; D. Aaker, 1996; p richer vocabulary of metaphorical and symbolic meanings (Biel, 1993, p. 73 74) or what McCracken (1986; 1988) terms symbolic elements . This fluidity also encourages consumers to personality, scholars have also shown that brand personality can influence consumer behavior. Freling and Forbes (2005) dubbed this phenomenon the brand personality effect (p. 404). In an experiment with undergraduate students ( n = 192), Freling and Forbes (2005) found a strong, positive brand personality lead to more favorable, unique and congruent brand associations and also increased brand equity 1 of bottled water, Freling and Forbes (2005) randomly assigned students to one of six treatment groups product i nformation only, product information + a sincere brand personality, product information + a competent brand personality, product information + a n excited brand personality, product information + a sophisticated brand personality, and product information + a rugged brand personality. First, participants were asked to provide their thoughts about bottled water in general. Next, participants reviewed a vignette about the fake brand of bottled water, 1 (p. 15). More simply put, brand equity is the value contained in a brand.

PAGE 29

29 specific to their experimental condition and completed a seri es of thought listing exercises and brand related scales (i.e., purchase intent, attitude toward the brand). Freling and Forbes (2005) found that brands with personality, regardless of the specific dimension, had significantly more favorable brand attitude s and significantly higher purchase intent. Additionally, participants in the brand personality conditions also had significantly more brand associations, including a higher proportion of positive brand associations, unique brand associations and congruent brand associations, than participants who received no brand personality information. Furthermore, when tested a week later, brand name recall was also significantly higher for participants who received brand personality information (82%) than those who di d not (28%). According to Freling and Forbes (2005), these combined results suggested that the initial impressions and brand associations developed by participants who were given brand personality information were n those developed by participants who received only long as it is perceived as being strong and favorable, is likely to be associated with positive consequence Sung and Kim (2010) also found that brand personality could increase brand related undergraduate students ( n = 135) to evaluate a s et of 30 brands across three product categories on brand personality, brand trust, brand affect and brand loyalty. To prevent participant fatigue, each respondent was asked to evaluate two brands for each of the three product categories, resulting in 646 t some brand personality traits (i.e., sincerity and ruggedness) were more strongly linked to an itement) were

PAGE 30

30 affect). Furthermore, Sung and Kim (2010) found higher levels of brand personality increased brand trust and brand affect, which, in turn, increased b rand loyalty. Therefore, increasing brand personality could also have the indirect result of increasing brand loyalty. experimental study of wine bottle labels using Aa (2007) found that specific brand personalities correlated with high purchase intent. To test the influence of design on brand personality, the authors created 90 wine bottle labels for a Cabernet using all possible co mbinations of three layouts (traditional white, traditional color and modern), six colors (burgundy, navy, bright red/orange, black/brown, wasabi green and pink) and five illustrated subjects (e.g., vineyard, grapes, coat of arms, neutral animals (e.g., de er) and unusual animals (e.g., platypus, fish)). For the study, a combination of student and community member participants ( n = 262) evaluated nine of the possible 90 wine labels. The online questionnaire displayed each test wine label and asked participan ts whether they would purchase the wine, whether they liked the label, how much they thought the wine would cost and how well they thought each of ten brand personalities described the label. According to the results, the top two brand personalities ( succe ssful and charming) explained 45% of the variance in purchase intent. Adding the next two brand personalities ( spirited and up to date ) only increased the explanation of variance to 47% a percentage almost as high as when all 10 brand personality options were combined. The authors found that that brand personalities outdoorsy and tough , were the least hypothesis that, within a single product category (e.g., wine), a small subset of brand personality attributes could significantly predict purchase intent.

PAGE 31

31 In addition to scholars demonstrating that consumers were capable of recognizing and tilization of one the ability to create brand differentiation (Biel, 1993; D. Aaker, 1996; Park, Jaworski & Maclnnis, 1986; Ramaseshan & Tsao, 2007). According to Freling and Forbes (2005), brand personality can where functional differences are marginal and quickly and easily overcome by modern technology (Biel, 1993). Fo rtunately, brand personality is especially effective at creating differentiation in categories where brands have similar product attributes and may even be used as a surrogate for difficult to evaluate or similar product attributes (Biel, 1993; D. Aaker, 1 996; Freling & Forbes, 2005). With brand personality serving as an important point of differentiation, scholars and practitioners emphasize that it is important to understand who controls its creation. While some experts suggest that the consumer is in co ntrol of defining the brand (Davis, 2000; Neumeier, 2006), other scholars and practitioners argue that the company still plays a significant role in defining its brand and brand personality (Huang, Mitchell & Rosenaum Elliot, 2012; McCracken, 1986; 1989; 1 993; Plummer, 1984/2000). Plummer (1984/2000) labeled the two input (what the company wants consumers to think and feel) and out take (what consumers actually think and feel) (p. 28). According to Plummer (1984/ 2000), a brand is defined when consumers examine the company input and produce their own out take. Although, Plummer (1984/2000) asserted that these meanings are rarely transmitted with absolute clarity and without ambiguity, the work of both Barthes (1977 ) and Hall (1980) suggests a rather different picture one that places more control of brand personality

PAGE 32

32 in the hands of the company and its advertisers. In particular, when viewed through the lens of take is quite Brand Personality and Consumer Identity Once established through company input, brand personality can also influence ch brand is the one for them (J. Aaker, 1999; Plummer, 1984/2000). More and more, modern consumers are looking for products and brands that are not merely utilitarian but that also express their own personalities (Postrel, 2003). In fact, Postrel (2003) co ntends that good design and, arguably, good branding and advertising who confirms that consumers often prefer brands with personalities similar to their own, often choosing a product because they view themselves as the type of person who would use the product (J. Aaker, 1997; 1999; Govers & Schoormans, 2005; Huang et al., 2012). For example, Govers and Schoormans (2005) conducted a study to examine how Specifically, the authors examined the effects of a product having the same personality as the consumer. The authors chose 12 stimuli products three screwdrivers, three coffeemakers, three soap dispensers and three bottles of wine that varied in appearance. In the first phase of the study, the authors conducted face to face interviews with community participants ( n = 48), asking them to describe the personality of each stimuli product. Participants then completed questionnaires designed to determine product personality congruence (e.g., this product is not

PAGE 33

33 like me/is like me, this product does not match me/matches me) and evaluate the stimuli (e.g., I think this product is not beautiful/is beautiful, I would not like/I would like to have this product). In the second phase of the study conducted 10 months later, a questionnaire was mailed to the same participants ( n = 37), containing pictures of the same 12 stimuli product s. For each product, participants were asked to complete a user image congruence scale (e.g., If you consider the types of people who prefer this product, do you identify with these people?, If you consider the types of people who prefer this product, are (2005) results indicated that product personality congruence determined a significant part of a personality similar to th eir own. Although weaker, the authors also found that user image congruence had an independent and positive effect on consumer preference, suggesting that consumers preferred products with users who they felt were like them. Thus, Govers and Schoormans (20 05) recommended that companies design products with predetermined In another study, Huang et al. (2012) examined whether consumers preferred products with personalities similar to their own. Fir st, the authors conducted exploratory interviews with white, British undergraduate students ( n = 11) to examine how they perceived their favorite brands, frequently used brands and famous brands. The interviews also paid special attention to the applicatio n of neuroticism and openness to new experience two human personality traits that were not captured in most brand personality scales. The individual interviews began by personalities of products than interested each participant. Huang et al. (2012) found that participants were only able to describe relatively superficial personalities for the famous brands.

PAGE 34

34 However, when discussing their favorite brands, participants were able to discuss both personality and personal relationships. Huang et al. (2012) also noted that, when asked to describe brands, participants used adjectives that fit into all of the Big Five human personality dimensions (i.e., extraversion, agreeableness , conscientiousness, neuroticism and openness to experience). Therefore, the authors determined that both neuroticism and openness to experience, along with the other three human personality dimensions, were also applicable to brand personalities. For the second part of the study, Huang et al. (2012) conducted a pretest with British undergraduate students ( n = 152) to ensure they included both utilitarian and symbolic products in the final experiment. Using a questionnaire, the authors measured participant and feeling toward 16 products from different product categories (e.g., beer, jeans, mobile phones). The authors used the questionnaire results to create a two axis grid (i.e., low to high involvement and low to high feeling). Four products, one from the extreme part of all quadrants, were then selected for use in the final experiment (i.e., laptop computers, dishwashing detergent, jeans and soft drinks). For the final stage of the study, Huang et al. (2012) conducted an experiment with white , British undergraduate students ( n = 468). Each participant was randomly assigned one of the four product categories selected in the pretest and asked to evaluate their favorite brand within that product category. Participants indicated both their own and scale a shortened version of the Big Five human personality scale. Results showed significant positive relationships between the personality of a consumer and their favorite brand and suggested that consumers did us 343). Further, the experiment found no differences between product categories, confirming

PAGE 35

35 product category and utilitarian vs. symbolic product differences have no impact on the relations hip between a consumer and brand personality. Overall, Huang et al. (2012) confirmed that consumers prefer to use brands that have personalities similar to their own (p. 345). Based on their findings, the authors suggested that practitioners should move be yond using brand personality as a mere image generation or segmentation tool. Instead, like Govers and Schoormans (2005), Huang et al. (2012) asserted that companies should work to determine the personalities of their target market and match their personality to that of their consumers. Moreover, this matching process was made simpler by the results of Huang et al. (2012), which suggested that the same human personality measure could be used for both consumers and brands. o help a consumer express his/her personal identity, a brand and its personalities can also provide a consumer with a means of expressing multiple selves (e.g., D. Aaker, 1996; J. Aaker, 1999; Huang et al., 2012; Malar, Krohmer, Hoyer & Nyffenegger, 2011; McCraken, 1989; Sirgy, 1982). Psychologists suggest that humans identify multiple selves, including actual (who they are), ideal (who they want to be) and social (who they want others to think they are) (D. Aaker, 1996). Individuals often use brands to def ine, fulfill and communicate these selves to others. For example, someone may wear Nike because he is athletic (actual self), he wants to be more athletic (ideal self), or he wants others to think he is athletic (social self). Because brands play a vital r ole in self expression, brands that align with a do not fit can cause discomfort (D. Aaker, 1996). Nienstedt, Huber and Seelmann (2012) found further evidence that consumers may choose different products based on different selves (actual vs. ideal). In an online survey,

PAGE 36

36 participants ( n = 736) indicated their own actual and ideal personalities, as well as the brand personalities of three German news magazines us ing a German market specific brand scale. The questionnaire also included scales to measure brand relationship quality, credibility and loyalty. The overall results from Nienstedt et al. (2012) showed that the congruence (i.e., similarity in a positive effect on credibility, which lead to an increase in brand relationship. Ideal congruence also had a significant positive effect on brand relationship, which lead t o an increase in loyalty. Additionally, the authors found differences in congruency effects on loyalty when they examined ideal congruence had a 2.7 times stronger effect for the print component, whil e actual congruence had a 3.2 times stronger effect for the online component. Based on these findings, Nienstedt et al. (2012) concluded that self esteem motivated print readers while self consistency motivated online readers. In general, these findings su ggested that consumers used different aspects of their selves and their personalities to interact with different media a fact marketers should consider as they determine where to place brand pers onality infused advertisements. The findings from Nienstedt et al. (2012) also concluded that marketers should monitor the congruence between the personality of their brand and their consumers. The authors higher congruence by altering product features or the brand personality communication (Nienstedt et al., 2012, p. 20). Alternately, Nienstedt et al. (2012) proposed that a lack of personality congruence could also lead a brand to identify and direct commu nication

PAGE 37

37 In a study on the brand personalities of perfume and cologne brands, Hamilton and Sun (2005) found further evidence that marketers should consider the perceive d self image of the target audience when creating brand personality through advertising. In a survey of undergraduate students ( n = 400), participants were asked to describe the personality of 10 different perfume/cologne brands using a 15 item, perfume/co logne specific brand personality scale compiled by the authors. Using the same scale, participants also indicated their own actual their ideal brand image was very close to their own perceived self image, suggesting that the might not be general ca tegory (i.e., perfume/cologne). Alternately, rather than considering actual, ideal and social selves, Swaminathan et al. (2009) examined brand personality using attachme nt theory. According to the authors, the two primary dimensions of attachment theory anxiety and avoidance influenced the type of relationships a person engaged in and the potential for forming interpersonal attachments (Swaminathan et al., 2009, p. 98 (Swami nathan et al., 2009, p. 986). Swaminathan et al. (2009) used a series of three experiments to examine brand attachment, purchase likelihood and brand choice. In study one, participants ( n = 182) were randomly assigned to one of eight conditions that manipu lated relationship anxiety (high vs. low), relationship avoidance (high vs. low) and brand personality (sincere vs. exciting).

PAGE 38

38 Participants in each group were primed for either high or low anxiety and high or low avoidance, then asked to view a research cr eated advertisement for athletic shoes, which was designed to demonstrate either a sincere or exciting brand personality, and indicate their brand attachment. Swaminathan et al. (2009) found that consumers with high anxiety demonstrated a more significant sensitivity to brand personality. Therefore, in a second study with a similar procedure, Swaminathan et al. (2009) focused on only high anxiety attachment styles, manipulating avoidance (high vs. low), brand personality (sincere vs. exciting) and consumpt ion situation (public vs. private). Participants ( n = 159) were randomly assigned to one of eight groups, primed for public or private consumption and asked to view a research created sincere or exciting ad for a fictitious clock brand. Participants then i ndicated their purchase likelihood. Study two supported previous findings about the anxiety attachment style, but identified a boundary condition, type of consumption. Thus, brand personality influenced high anxiety individuals when purchases were made for public consumption. Finally, like the second study, study three also focused on high anxiety attachment styles, manipulating avoidance (high vs. low) and relationship expectation (low vs. high). In this study, participants ( n = 99) were assigned to one o f four groups, primed for high or low relationship expectations by reading a consumption scenario and then asked to choose one of two jean brands (one with an exciting brand personality and one with a sincere brand personality) they would most likely wear in the preceding scenario. Based on their results, the authors found that, in high relationship expectation situations, those with high avoidance were significantly more likely to choose the exciting brand than those who were low in avoidance (63% vs. 18%) and also

PAGE 39

39 significantly less likely to choose the sincere brand than those who were low in avoidance (38% vs. 82%). Based on the combined results from the three experiments, Swaminathan et al. (2009) concluded that high anxiety types may use brand persona lity as a way to manage their relationships with others rather than as a true expression of self. Unsurprisingly, the authors found this effect to be even stronger when brand consumption took place publicly rather than privately. Further, the authors note that high anxiety consumers may also chose brands with a specific personality because they are more concerned with avoiding a mismatch between their findings re inforce the importance an individual consumer plays in a consumption decision, but they also reiterate their role brand personality choice. However, once established, brand personalities are not fixed or unmovable, and as the marketer provides new brand pe rsonality information, brand personalities can shift over time (Johar, Sengupta & Aaker, 2005). Because consumers inevitably view a brand and its personality through the lens of their own life experiences, self concept and brand interactions or what Hall (1980) terms a negotiated reading and McCracken (1986; 1988) calls culture brand perceptions, including personality, are constantly evolving (Johar et al., 2005). For example, using three experiments, Johar et al. (2005) explored how chronic (individual s who often activate and use specific personality traits) and nonchronic (individuals who do not often activate and use a specific personality trait) consumers updated their brand personality impressions based on new marketing information (p. 459). Experim used a sophisticated brand personality and fictitious clothing brand, Bondi , while manipulating both chronic trait ability (high vs. low) and time of measurement (before new information and

PAGE 40

40 after new information). Participants ( n = 40) re ad a series of sophisticated brand claims about Bondi activity, participants were given additional information on Bondi , including two statements that contained negative e Bondi clothing is both the brand and themselves (Johar, et al., 2005, p. 462). The results indicated that chronics and nonchronics updated brand personality differently. When presented with negative evaluation/positive trait claims, chronics trait ratings were not affected while nonchron ics trait ratings were reduced. brand and a fictitious travel agency that specialized in adventure travels ( Best Adventure Tour Outfitters ) while manipulating chronic trait ability (high vs. low), additional information type (negative evaluation/positive trait vs. positive evaluation/ne gative trait), and time of measurement (before new information and after new information). Participants ( n = 78) were asked to form an impression about Best Adventure Tour Outfitters a kind, not both exciting and sincere scales and then rate their own involvement. As in experiment one, participants were given additional information about Bes t Adventure Tour Outfitters that had evaluative implicat utility on both

PAGE 41

41 excitement and sincerity and themselves on excitement. Like experiment one, experiment two also showed a difference in the response of chronics and nonchronics for both the explicit (excitement) and underlying (sincerity) brand personality. Person ality trait ratings increased for chronics who received negative evaluation/positive trait information and decreased for chronics who received positive evaluation/negative trait information. Alternately, personality trait ratings decreased for nonchronics who received negative evaluation/positive trait information and increased for nonchronics who received positive evaluation/negative trait information. With only minor adjustments, experiment three used the same procedures and questionnaires as experiment two (Johar et al., 2005). However, experiment three used only positive evaluation/negative trait brand information and manipulated chronic trait ability (high vs. low), time of measurement (before new information and after new information), and prime at ti me 2 (present vs. absent). Half the participants ( n = 144) were with the exciting trait before exposure to the additional brand information by asking them to unscramble five synonyms of the word exciting . As the authors expected, when neither chronics nor nonchronics were primed, the findings remained the same as the first two experiments. However, while priming made no difference to chronics for whom the personality trait being tested is already easily accessible and frequently used, priming nonchronics pr ovided the same trait accessibility as chronics, resulting in the same decrease in brand personality trait ratings when exposed to positive evaluation/negative trait brand information. As a result of the three experiments, Johar et al. (2005) concluded th at consumers renegotiate brand personality based on their own personality and the type of incoming brand personality information (positive vs. negative). Consumers who share a specific personality trait with a brand are more likely to disregard evaluation information and change their perception of

PAGE 42

42 Alternately, consumers who do not share a specific personality trait with the brand are likely to disregard brand personality based on positive or negative evaluation information. Overall, results from Johar et al. (2005) nego nfluenced by an inferences consumer s made about hidden or unaddressed personality traits based on their own personality. The authors conducted three experiments for a fictitious athletic shoe brand using f self (Puzakova et al., 2013, p. 17). In the first experiment, Puzakova et al. (2013) asked undergraduate students ( n = 65) to view an advertisement and rate it on e xcitement (the observable trait), sincerity and competence (both inferred traits). Participants were then asked to rate themselves on the same excitement, sincerity and competence scales. Study one showed that perceptions of brand personality. In study two, undergraduate students ( n = 55) viewed an exciting ad and rated the ad on excitement (observable personality trait) and sincerity (unobservable personality trait) (Puzakova et al., 2013). Next participants read an article with ambiguous information (some positive and

PAGE 43

43 elf and initial impression of the unobservable brand personality trait, sincerity, influenced their interpretation of ambiguous brand information. Finally, study three mimicked the second study, manipulating competence as the unobservable trait rather than sincerity ( n = 55) (Puzakova et al., 2013). Like study two, the and the amb iguous brand information. Overall, Puzakova et al. (2013) found that consumers use personality traits to infer unobservable traits based on how those traits are present in their own personality. For example, if an advertisement portrayed a brand as clearly exciting, a consumer who thought of him/herself as both exci ting and competent would transfer a high level of considered him/herself competent but not exciting would also assume (based on his/her own personality) that because the brand was exciting it must not be competent in other words, it must be different than him/her. Based on these findings, it is probable that consumers may also adv ertisement shows it to be clearly extraverted, consumers might make assumptions about the . Furthermore, Puzakova, et al. (2013) found that consumers often compared brands b ased on multiple brand personality traits, particularly in highly competitive product markets. Therefore, Puzakova et al. (2013) suggested secondary or tertiary brand personality traits might

PAGE 44

44 multidimensional personality profiles may be ideal for providing brands with more robust, advertisement to present a full MBTI personality , leaving nothing for the consumer to assume. simultaneously manipulate multiple brand personalities via advertising images even more important for advertisers. Measuring Brand Personality Despite the significant number of findings regarding the usefulness of brand personality, scholars continue to debate the best way to measure it. In fact, past research has provided a variety of qualitative and quantitative methods for me asuring brand personality (e.g., J. Aaker, 1997; Avis, 2012; Costa & McCrae, 1992; Geuens, Weijters & De Wulf, 2009; Huang et al., relied almost solely on qu alitative methods to determine brand personality. And, while many practitioners still rely on qualitative free association methods, academics now almost exclusively employ quantitative scales that allow them to run more complex statistical tests and make c omparisons both within and between studies (Avis, 2012; Romaniuk, 2008). Following the trend of recent academic brand personality research, this study will employ a quantitative scale, which allows statistical comparison s among personality dimensions. Help ing to usher in the era of quantitative brand personality research, J. Aaker (1997) was amon g the first to develop a widely used, quantitative Brand Personality Scale (BPS). To create her scale, J. Aaker (1997) began by generating personality traits from r esearch studies on the Big Five human personality traits (i.e., openness, conscientiousness, extraversion, agreeableness and neuroticism), other academic/practitioner scales and a free association task, resulting in a total of 309 non redundant personality traits. In a second trait generation phase,

PAGE 45

45 participants ( n = 25) rated how descriptive each of the 309 traits were of brands in general on a scale from 1 (not at all descriptive) to 7 (extremely descriptive). Retaining only those traits that scored a 6 ( very descriptive) or above left J. Aaker (1997) with 114 personality traits. Next, a sample of 37 well known brands was chosen because it represented a wide spectrum of personality types and both symbolic and utilitarian product categories. To minimize pa rticipant fatigue and boredom, the 37 brands were then split into four groups composed of similar brand groupings. Questionnaires randomly containing one of the four brand groups were then mailed to a national, nonstudent sample. Participants ( n = 631) wer e asked to rate each of point Likert scale (1 = not at all descriptive, 5 = extremely descriptive). Using the survey results, J. Aaker (1997) ran a factor analysis, which resulted in fi ve distinct factors or personality dimensions: sincerity, excitement, competence, sophistication and ruggedness. By conducting additional factor analyses for each of the five traits, J. Aaker (1997) produced 15 facets four for sincerity (down to earth, h onest, wholesome and cheerful), four for excitement (daring, spirited, imaginative and up to date), three for competence (reliable, intelligent and successful), two for sophistication (upper class and charming), and two for ruggedness (outdoorsy and tough) . J. Aaker (1997) then created the Brand Personality Scale (BPS) by assigning the three highest correlating personality traits from the original trait list to each of the 15 facets, resulting in a 45 both test alpha measurements. Based on test retest data, three traits were dropped because of low retest scores (.74 s (1978) criterion (> . 7) for reliability. Using the reduced

PAGE 46

46 42 trait scale, J. dimensions indicated high levels of internal reliability ( .90 .95). A final confirmatory pha se of research was then conducted to ensure that the brand stimuli or participants did not bias the BPS. Using the new 42 trait scale, J. Aaker (1997) conducted another survey with a new sample of participants ( n = 180) and a different set of brands. Based on the results, a confirmatory analysis supported the stability of the five factor model and the 42 traits (see Table 2 1). Table 2 1. Brand Personality Scale (BPS). Sincerity Excitement Competence Sophistication Ruggedness 1 Down to earth Daring Relia ble Upper class Outdoorsy 2 Family oriented Trend Hard working Glamorous Masculine 3 Small town Exciting Secure Good looking Western 4 Honest Spirited Intelligent Charming Tough 5 Sincere Cool Technical Feminine Rugged 6 Real Young Corporate Smooth 7 Wholesome Imaginative Successful 8 Original Unique Leader 9 Cheerful Up to date Confident 10 Sentimental Independent 11 Friendly Contemporary ave of brand personality research and is the most commonly used brand personality scale, it is not without question both its usefulness and application. Critiq questions about its quantitative methodology (Arora & Stoner, 2009; Avis, 2012; Romaniuk, 2008), what it actually measures (e.g., Azoulay & Kapferer, 2003; Lee, 2013), its proposal that brand personalities differ from h uman personalities (Geuens et al., 2009; Huang et al., 2012), its generalizability to individual brands (Austin, Siguaw & Mattila, 2003), and its inability to cross

PAGE 47

47 cultures (e.g., Avis, 2012; Geuens et al., 2009; Huang et al., 2012). While the differences in qualitative and quantitative approaches were addressed above, the other critiques will be detailed below. the scale actually measures (e.g., Avis, 2012; Azoulay & Ka pferer, 2003). Specifically, Azoulay (or brand image) rather than just personality. Azoulay and Kapferer (2003) blamed this problem largely on brand personality r personality. The authors noted that human psychology worked for many years to strictly define items rel traits that were directly associated with each of these categories, such as intelligent , feminine and upper class . Illustrating the confusion about what brand personalit y really entails, in a recent qualitative study, Lee (2013) suggested that brand personality was composed of more than psychological personality traits and should also include socio cultural components, specifically narratives of socio economic variability (e.g., hobbies, consumption patterns, personal values), narratives of life scene variability (e.g., economic status, education level, occupation) and narratives of physical variability (e.g., gender, age, appearance). These observations are in direct oppo sition with how psychologists define human personali ty (Azoulay & Kapferer, 2003). Further, based on other scholarly literature (e.g., Azoulay & Kapferer, 2003) and widely used ppropriately

PAGE 48

48 cultural categories contained variables similar to well known methods of audience segmentation taugh t in popular advertising curriculum, were valuable, like J. inclusive, confusing the concepts of brand image and brand personality. ge and brand personality. As such, academics should further note the need to distinguish between the more encompassing idea of brand image and the more specific, psychologically based component of brand personality . In response to their concerns about what (2003) referred scholars to previous human psychology research in an attempt to redefine personality. The current study took advice by utilizing a brand speci fic version of a human personality test, the MBTI. should more closely align with human personality rather than creating an entirely new set of brand specific personalities . Like Azoulay and Kapferer (2003), other scholars agree that brand personality should strongly resemble human personality (e.g., Geuens, Weijters & De Wulf, 2009; Huang et al., 201 2). Geuens et al. (2009) noted, Consumers also appear to experience no prob lems in assigning human characteristics to brands ( J. Aaker, 1997) or in building a relationship with brands (Fournier, 1998). Therefore, it is possible that the Big Five [human personality] structure also extends to brand personality. (p. 98) Indeed, like has five dimensions. However, because J. Aaker (1997) used traits outside the definition of

PAGE 49

49 human personality (e.g., intelligence, gender and social status), results from her BPS became al. (2009) also contended that because human personality had been studied longer and more in depth than brand personality, it was more develope d and consistent. Therefore, by aligning the of work perfecting human personality research and measurement. Further, Geuens et al. (2009) argued that because p eople used brand personalities to both relate to companies and show their own personalities, a brand personality scale that resembled a human personality scale would ions to scale establishes a different set of personalities for brands rather than aligning the two. Additionally, Huang et al. (2012) questioned why, if human consumers were giving 334). In a compilation of various brand personality scales, Huang et al. (2012) found that many brand scales had at least some overlap with the Big Five brand personality scale (see Huang et al., 2012, p. 336 ). Further, based on their study results, which asked participants to rate both themsel ves and their favorite brands on the same Big Five human personality scale, Huang et al. (2012) concluded that both brand personality and human personality could be measured by the same scales and generate the same dimensions. Therefore, it appears that br ands can be measured using human personality scales and may not require a unique personality scal (1997) BPS .

PAGE 50

50 work for product categories, it does not generalize to individual brands (Austin, Siguaw & Mattila, 2003; Geuens et al., 2009). Both Austin et al. (2003) and Geuens et al. (2009) suggested variance at th e individual brand level, leaving only differences between product categories. Therefore, while J. Aaker (1997) claimed the scale was generalizable, Austin et al. (2003) argued from what and to what the brand personality framework is U.S. students ( n = 247) evaluated nine individual restaurant brands (three quick service, three item BPS. Then, unlike J. Aaker (1997), who combined data for each product category, Austin et al. (2003) cond ucted confirmatory factor analyses for each of the nine restaurant brands. According to the authors, the model did not provide a satisfactory fit for any of the nine brands (p. 83). Next, Austin et al. (2003) aggregated the data at three different levels: 1) quick service, 2) quick service and casual dining, and 3) quick service, casual dining and upscale dining. However, after preforming confirmatory analyses on each level of aggregated data, the model still did not provide a satisfactory fit. The authors likely due to their investigation of only one product category, as opposed to multiple product J. may work between produ ct categories, it does not appear to be effective at t he individual brand level.

PAGE 51

51 replicated cross culturally (e.g., Avis, 2012; Geuens et al., 2009; Huang et al., 2012). A ccording creation in different cultures, they found significantly different brand personalities. Moreover, in their compilation of different brand personality studies , Huang et al. (2012) identified multiple brand personality scales from a variety of different countries, each with a different set of brand personality dimensions (see Figure 2 1). In fact, unique Japanese and Spanish brand personality scales were develop ed by J. Aaker and her research partners (e.g., J. Aaker, Benet Martinez & several researchers to construct a country ). Because (McCrae, 2002), are consistent across culture, it follows that brand personality scales should also ) scale was not valid across cultures. Response to Brand Personality Scale Critiques examine s qualitative roots (e.g., Avis, 2010; Romaniuk, 2008), others worked to develop better quantitative scales. For example, Huang et al. (2012) further linked human and brand personality ). Huang et al. (2012) also produced an abridged 19 item scale that uses the Big Five human personalities dimensions to measure both human and brand personalities. Alternately, Geuens et al. (2009) proposed a new five factor scale (activity, responsibility , aggressiveness, simplicity and emotionality) composed of 12 items, which more closely aligned with the Big Five human personality dimensions of openness, conscientiousness, extraversion, agreeableness and neuroticism (Costa

PAGE 52

52 & McCrae, 1992). However, whil e these scales addressed many of the issues that plagued J. asking participants to rate brand personality on a 5 or 7 point Likert scale, five factor scales suggest t hat more of some trait is inherently better or worse. When, in actuality, no research has been done to determine whether higher or lower levels of these factors benefit a bran d. Alternately, Jewett and Sutherland (2013) and Sutherland et al. (2004) opted t o measure brand personality based on a variation of the Myers Briggs Type Indicator (MBTI) scale, a Foundation, n.d.). Based on four polar dimensions rather than a five factor model, this scale also answers many of the criticisms levied against previous brand personality scales, such as measuring only personality, aligning brand personality with human personality and working across cultures. Furthermore, unlike five factor models such those based on the Big Five, the MBTI, which is discussed in detail below, does not suggest that one personality is better than using J. Aaker specific version of the MBTI that overcomes many of the criticisms levied at her work. The Myers Briggs Type Indicator (MBTI) MBTI History The Myers Briggs Type Indicator (MBTI) is based on a theory of pers onality developed by Swiss psychiatrist Carl Jung in the 1920s (Hirsh & Kummerow, 1989; Jung, 1923/1971; Myers, 1995; The Myers & Briggs Foundation, n.d.). Jungian personality theory posits that seemingly random variations in human behavior can be explaine d by basic, inborn differences in how people approach life (Jung, 1923/1971; The Myers & Briggs Foundation, n.d.). Jung (1923/1971) specified three bi polar personality dimensions extraversion/introversion (E/I),

PAGE 53

53 sensing/intuition (S/N) and thinking/feel ing (T/F). According to Jung, these personality dimensions are distributed at random among the general population without regard to education, social status, gender or upbringing. Furthermore, his theory contends that differences in human personality are u nconscious and instinctive, not a matter of conscious judgment or intention (p. 180). During the 1940s and 50s, Katherine Cook Briggs and Isabel Briggs Myers, an American mother o create a practically applicable tool for determining individual personalities the Myers Briggs Type Indicator (MBTI) (Hirsh & Kummerow, 1989; Myers, 1995; Tieger & Barron Tieger, 1995; 2000). To determine personality type, the MBTI operationalized Jung added a fourth judging/perceiving (J/P) based on Briggs unpublished human personality research to create 16 possible personality types. After rising in popularity during the 1970s, the MBTI has remained popular worldwide because of its ease of use, practical application and explanatory power and has become one of the most widely used, reliable and valid personality measurements available (Hirsh & Kummerow, 1989; The Mye rs & Briggs Foundation, n.d.). Scholarly l iterature also show s the MBTI to be an accurate, reliable and valid personality measurement (Carlyn, 1977; McCarley & Carskadon, 1986). Additionally, since its development more than 50 years ago, the MBTI has been used in a variety of marketing and advertising studies (e.g., Azzadina, Huda & Sianipar, 2012; Gould, 1991; Huang & Yang, 2011; LaBarbera, Weingard &Yorkston, 1998; Shank & Langmeyer, 1994; Till, 2010). MBTI Dimensions referred to a s attitude types (Jung, 1923/1971; Tieger & Barron Tieger, 1995; 2000). Jung

PAGE 54

54 describes the difference between e xtraverts and i ntroverts in terms of how they relate to the object (e.g., people, things, external accomplishments) (Edinger, n.d.; Jung, 1923/19 71). While e xtraverts place a high value on the object and constantly use the object to orient themselves, i ntroverts tend to withdraw from the object, working to prevent the object from gaining power over them (Jung, 1923/1971, p. 178 179). According to t the extraversion/introversion ( E/I ) where he/she gets and directs his/her energy (The Myers & Briggs Foundation, n.d.; Tieger & Barron Tieger, 1995; 2000). While e xt raverts focus on the world around them and gain energy from other people, i ntroverts draw energy from their inner world and prefer to spend time alone. The difference between e xtraverts and i ntroverts can also be described in terms of breadth vs. depth. Ex traverts often know a little about a lot, preferring to explore a variety of subjects and ideas. Conversely, i ntroverts tend to have fewer but deeper interests, immersing themselves in the topics that appeal to them. Extraverts and i ntroverts also differ i n thought and communication styles. Extraverts often need to think out their ideas aloud, while i ntroverts prefer to quietly think through an idea in their head before speaking it. f unction types (Jung, 1923/1971). According to Jung, s ensors appreciate simple, matter of fact reality, while i ntuitives value big ideas and see obscure connections others may miss (Edinger, n.d.; The Myers & Briggs Foundation, n.d.; Tieger & Barron Tieger, dimension, then, specifically analyzes how an individual takes in information. Sensors concentrate on concrete information from their five senses, while i ntuitives interpret and add meaning to information, looking at big picture patterns. These preferences often allow s ensors to accurately remember details and excel at here and now practical matters. Alternately, i ntuitives

PAGE 55

55 tend to be creative and good at imagining long term possibilities for the future (Tieger & B arron Tieger, 2000). Typically, s ensors also prefer established ways of doing things, while i ntuitives enjoy creating new ways to accomplish a task or goal. function types (Jung, 1923/197 t hinking is closely tied to intellect, while f eeling places a high premium on the development and maintenance of personal relationships (Edinger, n.d.). According to Jung and other scholars, thinking/feeling (T/F) is also the only dimension correlated to gender, with a higher percentage of male t hinkers and female f eelers (Edinger, n.d.; Jung, 1923/1971; Tieger & Barron Tieger, 2000). In a simplified definition for the MBTI, the thinking/feeling (T/F) dimension represents h ow individuals make decisions and come to conclusions (The Myers & Briggs Foundation, n.d.; Tieger & Barron Tieger, 2000). Thinkers put more weight on objective principles, impersonal facts and logic, while f eelers place more emphasis on people and persona l concerns. While t hinkers are more driven to be fair and consistent, f eelers are guided by their personal values and extenuating circumstances. Additionally, t hinkers tend to take things less personally, are harder to offend and are uncomfortable with emo tions. Alternately, f eelers often wear their heart on their sleeves, are comfortable with emotions and may get their feelings hurt more easily. approach to structure and the ty pe of life he/she adopts. According to Myers (1995), the basic differences in personality concern the way people perceive (become aware of things, people, occurrences and ideas) and judge (come to conclusions about what has been perceived) (p. 1). Unlike t original personality theory. Rather, the judging/perceiving (J/P) dimension was added to the

PAGE 56

56 MBTI as a result of unpublished personality research conducted by Briggs pri or to the Psychological Types (Myers, 1995). In the MBTI, j udging individuals prefer structured, ordered environments and like to make decisions, while p erceiving individuals would rather stay open and adaptable to new ideas and infor mation. Additionally, j udgers like to make a plan and stick to it, while p erceivers are spontaneous and prefer to keep plans open, fearful of missing out if something better comes along (Tieger & Barron Tieger, 2000). Judgers are organized, follow rules to as suggestions. Further, j udgers often have a work before play attitude, while p erceiver s prefer a play before work approach. MBTI Application To fully appreciate the MBTI, it is important to understand several general principles that Barron Tieger, 1995; 2000). First, each person is born with one true personality type, which remains the same throughout their life despite changes in development, maturity or circumstances. Second, unlike some other personality and brand personality measures, the MBTI r ecognizes that all personality types are identical in value equal but different. Although each of the 16 personality types has its own set of strengths and weaknesses, no personality type or dimension is good or bad, better or worse. Third, while persona lity type can provide valuable insight into personality, every individual is unique. Though people with the same personality type share common traits, a lifetime of unique circumstances ensures that each person is a distinct original. Rather than evaluatin polar dimensions measure preferences (Hirsh & Kummerow, 1989; Jung, 1923/1971; Myers, 1995; The Myers &

PAGE 57

57 Briggs Foundation, n.d.; Tieger & Barron original dime nsions extraversion/introversion, sensing/intuition (S/N) and thinking/feeling (T/F), plus the fourth dimension added by Myers Briggs judging/perceiving (J/P) (Myers, preference to exist in varying strengths (see Tieger & Barron Tieger, 2000, p. 11 ). Therefore, the polar endpoints; the weaker the preference, the closer he/she wi ll be to the midpoint. For example, although two individuals are both t hinkers, one may have a much stronger t hinking preference and be placed at the far thinking end of the thinking/feeling (T/F) continuum, while the other individual may have only a sligh t t midpoint. It is also important to note that while individuals have a natural preference within each dimension (i.e., they prefer introversion to extraversion), every person can and does use both si des of each dimension at certain times (Tieger & Barron Tieger, 2000). So, although an individual may be a strong i ntrovert and prefer small groups of close friends that does not mean he/she is not also able to assert his/her extraverted side and effective ly network in a large crowd when necessary. Study Brand Personality Dimensions This study will examine three of the four MBTI personality dimensions extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). The fourth dimension judging/perceiving (J/P) will be excluded because it was not part of the original Jungian personality theory, and it is often shown to overlap with other MBTI dimensions (Carlyn, 1977; Hirsh & Kummerow, 1989). Furthermore, the judging/perceiving (J/P) dimension does not seem to be particularly well suited for describing a brand or being shown through visual symbolic elements because of the need for all successful brands to display continuity in their

PAGE 58

58 product or service, product category and advertising . This necessary rigidity aligns with the be considered judging brands. Therefore, the judging/perceiving (J/P) dimension will be omitted from this study. F or more details on the exclusion of the judging/perceiving (J/P) dimension, se e Study Brand Personality Type Descriptions Excluding the judging/perceiving (J/P) dimension leaves this study with a total of eight, unique multidi mensional personality types, which are created by combining the three remaining dimensions extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F) (Martin, n.d.; Myers, 1976a). Each unique human personality type is detailed be low, followed by a brand personality type. The eight human personality type descriptions were created from a combination of Myers (1976a) and Martin (n.d.). To create the eight brand personality type descriptions, references to internal concepts that could only be known or expressed by a person were removed from the human type descriptions. This exclusion left only observable or directly communicable aspects that could be expressed by a brand through some marketing channel (e.g., advertising, social media, visual identity, product packaging). Extraverted/sensing/thinking (EST). EST individuals are matter of fact. They tend to like mechanical things, have a head for business and enjoy sports with their friends. They also pay enthusiastic attention to the oute r world of people, places, things through hands on and real life experiences and feel the need to analyze it. EST individuals tend to be logical and analytical in their approach to life and have an acute sense of how objects, events and people in the world work. They are best with real things that can be worked, handled, taken apart or put together. They are not interested in subjects they see no use for but can apply themselves when necessary.

PAGE 59

59 EST individuals are oriented to current facts and realities, wh ich gives them a pragmatic quality. They may be a bit blunt or insensitive and dislike long explanations (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an extraverted/sensing/thinking (EST) brand becomes: An EST brand is ma tter of fact. It may be related to mechanical, business or sports industries. An EST brand portrays enthusiasm and excitement about the world around it and is involved or usable in the outer world of people, places and things. It can be helpful in hands on situations and/or connected with physical experiences. An EST brand appears logical and analytical and contributes to how objects, events and people in the world work. It is connected to real things that can be worked, handled, taken apart or put together . It is shown as hard working but not redundant, excessive or superfluous. An EST brand is oriented to current blunt or insensitive and be communicated in a sho rt, to the point style rather than with long explanations (Myers, 1976a; Martin, n.d.). For example, Under Armour embodies many aspects of an EST brand. Extraverted/sensing/feeling (ESF). ESF individuals are outgoing, accepting, friendly, energetic, warm h earted, talkative, popular and born cooperators. They know what is going on and join in eagerly, making things more fun for others by their enjoyment. They find remembering facts easier than mastering theories and have little interest in abstract thinking or technical subjects. ESF individuals are mainly interest in things that directly and visibly affect with people as well as things. They have a deep concern for people and show their caring in warm and pragmatic gestures of helping. ESF individuals actively and intensely care about people, need harmony and may be good at creating it. They are always doing something nice for

PAGE 60

60 someone and bring an aura of warmth to all that they do. They work best with encouragement and praise. They pay enthusiastic attention to the outer world of hands on and real life experiences and are oriented to current facts and realities, giving their feelings a hands on, pragmatic quality (M yers, 1976a; Martin, n.d.). When this human personality is applied to brands, an extraverted/sensing/feeling (ESF) brand becomes: An ESF brand appears to be outgoing, accepting, friendly, energetic, warm hearted, talkative, popular and a good cooperator. I eagerly, making things more fun for others by its participation. It is focused on concrete facts, not big picture theories. An ESF brand is not abstract or technical. It is mostly applicable to things that direct require common sense and practical ability with people as well as things. It portrays a deep concern for people and contributes actively and practically. It is focused on people and helpful for creating harmony. An ESF brand is often shown doing nice things for people and exuding warmth. It is shown as attuned to the outer world of hands on and real life experiences. An ESF brand is oriented to current facts and realities, giving its caring demeanor a hands on, pragmatic quality (Myers, 1976a; Martin, n.d.). For example, Southwest Airlines embodies many aspects of an ESF brand. Extraverted/intuitive/thinking (ENT). ENT individuals are quick, ingenious, hearty, frank, energetic, ent husiastic and good at many things. They are stimulating company, alert and outspoken. They are driven by their attention to the outer world of possibilities. ENT individuals look for patterns and meaning in the world and often have a deep need to analyze, understand and know the nature of things. They are resourceful in solving new and challenging problems but may neglect routine assignments. They are skillful in finding logical reasons for what they want.

PAGE 61

61 ENT individuals are usually good at anything that r equires reasoning and intelligent talk, such as public speaking. They are usually well informed and enjoy adding to their fund of knowledge. They may sometimes be more positive and confident than their experience in an area warrants. ENT individuals are or iented to the future, giving their thinking an abstract quality (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an extraverted/intuitive/thinking (ENT) brand becomes: An ENT brand appears to be quick, ingenious, hearty, fran k, energetic, to keep a pulse on the world around it and is ready and willing to change to avoid obsolescence . An ENT brand is attuned to patterns and meaning i n the world. It may aid in analyzing, understanding or knowing the nature of things. It is also resourceful in solving new and challenging problems, but may neglect routine tasks. An ENT brand communicates logical reasons for its stance. It is usually good at anything that requires reasoning and intelligent talk and has a good public face. It appears well informed and often adds to its fund of knowledge. An ENT brand may sometimes sound more positive and confident than its experience in an area warrants. It is oriented to the future rather than the present (Myers, 1976a; Martin, n.d.). For example, General Electric (GE) embodies many aspects of an ENT brand. Extraverted/intuitive/feeling (ENF). ENF individuals are warmly enthusiastic, energetic, high spirite d, ingenious, imaginative, sociable, popular and sympathetic. They are able to do almost anything that interests them. They are quick with a solution for any difficulty and ready to help anyone with a problem. ENF individuals often rely on their ability to improvise instead of preparing in advance, and they can usually find compelling reasons for whatever they want. They attend to the outer world of possibilities and thrive on what is possible and what is

PAGE 62

62 new. ENF individuals generally feel real concern for what others think or want and try to handle and have a strong desire to bring harmony into their relationships. ENF individuals are openly expressive and emp athic people who bring an aura of warmth to all that they do. The are oriented to the new and to the possible and, thus, often enjoy working to manifest a humanitarian vision or helping others develop their potential. ENF individuals can present a proposal or lead a group discussion with ease and tact. They are responsive to both praise and criticism (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an extraverted/intuitive/feeling (ENF) brand becomes: An ENF brand appears to b e warmly enthusiastic, energetic, high spirited, ingenious, imaginative, sociable, popular and sympathetic. It is capable of doing almost anything it chooses. It is quick with a solution for any difficulty and ready to help with any problem. An ENF brand m ay improvise or interact on the fly rather than preparing in advance. It is able to communicate compelling reasons for whatever it wants to do. It seems to attend to the outer world of possibilities and thrive on what is possible and what is new. An ENF br and generally shown as actively and intensely caring about people and desiring harmonious relationships. It is warmly expressive and empathic in its communicat ion. An ENF brand is oriented to the new and to the possible and, thus, often enjoys working to manifest a humanitarian vision or helping others develop their potential. It appears to be capable of leading a group discussion with ease and tact. An ENT bran d is responsive to both praise and criticism (Myers, 1976a; Martin, n.d.). For example, Nestle Toll House embodies many aspects of an ENF brand.

PAGE 63

63 Introverted/sensing/thinking (IST). IST individuals are cool onlookers, serious, quiet, reserved, practical, or derly, matter of fact, logical, realistic, dependable and natural troubleshooters. They observe and analyze life with detached curiosity and unexpected flashes of original humor. They are usually interested in impersonal principles, facts, cause/effect and how/why mechanical things work. They want to understand how things and phenomena in the real world work so they can make the best and most effective use of them. IST individuals naturally look for the underlying sense to any facts they have gathered. They are quiet and analytical observers of the environment (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an introverted/sensing/thinking (IST) brand becomes: An IST brand appears to be a cool onlooker, serious, quiet, reserved , practical, orderly, matter of fact, logical, realistic, dependable and a natural troubleshooter. It seems to observe and analyze with detached curiosity and unexpected flashes of original humor. It is usually attuned to impersonal principles, facts, caus e/effect and how/why mechanical things work. An IST brand seems to seek to an understanding of how things and phenomena in the real world work in order to most effectively put them to use. It gives the impression of looking for the underlying sense to fact s it has gathered and is a quiet, analytical observer of the environment (Myers, 1976a; Martin, n.d.). For example, John Deere embodies many aspects of an IST brand. Introverted/sensing/feeling (ISF). ISF individuals are retiring, quietly friendly, sensiti ve, kind and modest about their abilities. They shun disagreements and do not force their opinions or values on others. They are loyal, considerate and concerned with how other people feel. ISF individuals have a deep felt caring for living things and typi cally show their caring in very practical ways, since they often prefer action to words. They view actions that are of practical help to others as particularly important. Their warmth and concern are generally not

PAGE 64

64 expressed openly. Their realism brings an aura of quiet warmth and caring. ISF individuals may need time to master technical subjects, as their interests are usually not technical (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an introverted/sensing/feeling (ISF) b rand becomes: An ISF brand appears to be retiring, quietly friendly, sensitive, kind and modest. It seems to shun disagreements and not force its opinions or values on others. It is loyal, considerate and concerned with how others feel. An ISF brand appear s to display deep care for living things, and it is typically shown demonstrating its care in very practical ways, often preferring action to words. It seems to views actions that practically help others as particularly important. Its realism portrays an a ura of quiet warmth and caring. It is not overly technical (Myers, 1976a; Martin, n.d.). For example, Ancestry.com embodies many aspects of an ISF brand. Introverted/intuitive/thinking (INT). INT individuals are quiet, reserved, impersonal, skeptical, crit ical and independent. They enjoy especially theoretical or scientific subjects and are logical to the point of hair splitting. INT individual are analytical and detached in their approach to the world, and they naturally question and critique ideas and eve nts as they strive for understanding. They are usually interested mainly in ideas, with little liking for parties or small talk. They have original minds for their own ideas and fields that appeal to them and tend to have sharply defined interests. INT ind ividuals are driven to understand whatever phenomenon is the focus of their attention. They want to make sense of the world as a concept and they often enjoy opportunities to be creative. They attend to the inner world of possibilities, symbols, abstra ctions, images and thoughts. To them, ideas are the substance of life, and insights in

PAGE 65

65 conjunction with logical analysis are the essence of their approach to the world (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an intro verted/intuitive/thinking (INT) brand becomes: An INT brand appears to be quiet, reserved, impersonal, skeptical, critical and independent. It is useful for theoretical or scientific subjects and can be perceived as logical to the point of hair splitting. It is analytical and detached in its approach to the world. It seems to question and critique ideas and events to increase understanding. An INT brand is usually interested in big ideas, with little liking for frivolous parties or small talk. It is origina l in its own ideas and field and tends to have sharply defined interests. It is driven to understand whatever phenomenon is the focus of its attention. An INT brand appears to display creativity and strive to make sense of the world as a concept. It attend s to the world of possibilities, symbols, abstractions, images and thoughts. An INT brand is portrayed as focusing on ideas and insights in conjunction with logical analysis is its essence (Myers, 1976a; Martin, n.d.). For example, Intel embodies many aspe cts of an INT brand. Introverted/intuitive/feeling (INF). INF individuals are quietly conscientious and concerned for others. They live in the inner world of values and ideals. They are full of enthusiasms and loyalties, but seldom talk of these until they know you well. They are friendly but often too absorbed in what they are doing to be sociable. INF individuals care about learning and have a deep felt caring and idealism about people. They experience this intense caring most often in their relationships with others, but they may also experience it around ideas or projects. Their energy and attention naturally draws them to the inner world of ideas and insights, particularly those that embody a concern for human potential. INF individuals often have deep

PAGE 66

66 interests in creative expression as well as issues of spirituality and human development. They are little concerned with possessions or physical surroundings (Myers, 1976a; Martin, n.d.). When this human personality is applied to brands, an introverted/int uitive/feeling (INF) brand becomes: An INF brand appears to be quietly conscientious and concerned for others. It is portrayed as placing high importance on values and ideals. It is full of enthusiasm and loyalty but seldom expressive until you know it wel l. An INF brand is friendly but often too absorbed in what it is doing to be sociable. It seems to care about learning and portray a deep felt caring and idealism about others. It is shown expressing care most often about people it is in a relationship wit attention are inwardly focused and revolve around ideas and insights, particularly those that embody a concern for human potential. It often appears to demonstrate a deep interest in creative expression as well as issues of spirituality and human development. It depicts little concern for possessions or physical surroundings (Myers, 1976a; Martin, n.d.). For example, the American Red Cross embodies many aspects of an INF brand. Previous literature supported the existence, importance and influence of brand personality; therefore, the discussion now turns to how advertisers can harness and create personality in their brands. Plumm more commonly known as brand personality was the only facet of a brand determined largely by communication strategies and advertising, thus providing advertisers with the opportunity to strategic ally imbue their brand with carefully chosen characteristics th at attract their target market. In fact, scholars agreed that advertising factors heavily into the development of brand personality (e.g., Delbaere et al., 2011; Doyle, 1989; Meenaghan, 1995, S waminathan, Stilley &

PAGE 67

67 Ahluwalia, 2009; Ouwersloot & Tudorica, 2001; Puzakova, Kwak & Taylor, 2013; Yoo, Ban & Kim, 2009 ). According to brand with a particular personality at an emotional/symbo lic level (p. 27). Meenaghan (1995) brand personality (p. 31). Doyle (1989) agreed, asserting that advertising augments and personality, Reid and Buchanan (1979) asked 120 female cat owners to describe a fictitious female shopper based on looking at her shopping list. Participants were randomly given a list that included one of the following: 9 Lives cat food (a brand that had advertised with a persona named Morris the Cat), Brand X cat food (a national brand with no personality as determined in p retesting), or no cat food. After examining the list, participants were asked to write a brief Buchanan (1979) found that significantly more positive stateme nts were made about the fictitious shopper when 9 Lives cat food was included on the list (78%) than when the list included national Brand X (20%) or no cat food at all (26%). Furthermore, no differences were found between heavy, medium, light and non user s in regards to the number of positive statements, indicating differences were due to advertising exposure, rather than brand usage. The advertising oriented t 28). More recently, Yoo et al. (2009) conducted two experiments to examine the influence of advertisements on brand personality. In the first experiment, Yoo et al. (2009) divided Korea n

PAGE 68

68 undergraduate students ( n = 126) into two groups to view three magazine ads for a fashion brand. One group viewed a series of three consistent fashion ads from a single campaign, while the second group viewed a series of three inconsistent fashion ads fr om three different campaigns. The authors found that brand personality was significantly higher for two of the four brand personality traits (sophistication and excitement) when participants were exposed to the consistent advertising campaign. Although no statistical difference was found for the other two personality traits (boldness and credibility), Yoo et al. (2009) suggested this might be the result of the type of fashion advertising and the specific personality traits participants at tributed to the bra nds. In a second experiment, Yoo et al. (2009) examined the differences in brand personality based on a consistent advertising campaign, inconsistent advertising campaign and a single ad exposure. The authors used a bank to mitigate any product related inf luence on results. The authors repeated the same experimental process with a different set of Korean undergraduate students ( n = 160), dividing the participants into three groups. Yoo et al. (2009) found that two of the four of the brand personality traits (competence and sophistication) were statistically higher when participants were exposed to the consistent campaign, while the other two brand personality traits (expertise and sincerity) were statistically higher for participants who viewed the same ad t hree times. However, participants exposed to the inconsistent advertising campaign consistently ranked the brands as having the least amount of personality. Overall, research by Yoo et al. (2009) underscored the importance of a consistent advertising messa ge, either as a campaign or repeated exposure to a single ad for developing strong brand personality. Besides the findings from Yoo et al. (2009) showing that consistent or repeated ad messaging produced brand personalities, Delbaere et al. (2011) demonstr ated that advertisements

PAGE 69

69 could use visual personifications 2 to trigger more positive brand personality attributions. In a pilot study, Delbaere et al. (2011) found that advertisements did not need words for consumers to begin anthropomorphizing inanimate p roducts. Using these findings, Delbaere et al. (2011) conducted an experiment with undergraduate students ( n = 187) to examine whether ads with personifications would elicit more brand personality attributions than those with no personification. Four fict itious products were chosen for the experiment a moisturizer, a snack bar, a snack mix and bleach. Three ads were created for each product using the same headline but different images a personification (e.g., a nut and a piece of fruit each wearing a w edding personification visual metaphor (e.g., ge. During the experiment, each participant saw and rated one ad for each category two ads with images of product packaging (controls), one personification and one non personification visual metaphor. Participants were asked to answer a series of questio ns about each of the four ads they saw, including measures of brand personality. Based on the results, the average brand personality score was significantly higher for the personification ad than for the non personification visual metaphor ad. And, both th e personification and non personification visual metaphor ad generated significantly higher brand personality scores than the product packaging control ad. The study conducted by Delbaere et al. (2011) accentuated the role advertising and visuals in advert ising play in creating brand personality. Moreover, according to the authors, visual personification in advertising (Delbaere et al., 2011, p. 127). 2 These draw on anthropomorphism (i.e., the tendency to attribute human qualities to things).

PAGE 70

70 Brand Personality & Visuals While brand personality is influenced by every element of an ad as shown above, the nonverbal, visual aspects may be the most important contributors (Ang & Lim, 2006; Batra, Lehmann & Singh, 1993). Scholars agree that most successful brands have compelling visual components that consumers learn to automatically associate with a brand and its personality are associated with an independent cowboy and the Nik e is almost synonymous with its swoosh . In a seminal study on the effects of visuals versus verbal advertising claims on product attribute beliefs, Mitchell and Olson (1981) found consumers could convert visuals into meaningful information about a brand ev en if the visuals were not directly related to the product (p. 329). For example, participants ranked a facial tissue brand as softer when shown an advertisement pairing the brand with a fluffy kitten than when shown an advertisement pairing the brand with attributes from seemingly unrelated advertising visuals. Although Mitchell and Olson (1981) were concerned only with product attributes, such results suggest consumers might also be able to draw specific brand personality conclusions from strategically chosen advertising images. Indeed, Jewett and Sutherland (2013) found an advertisi ng image that portrayed a specific personality characteristic could transfer its particular Boudreaux and Palmer (2007) found that images had a stronger effect than e ither color or layout on brand personality and market success factors such as purchase intent, cost and likeability. Despite a number of past studies linking strong brands to strong aesthetics (e.g., Batra, Lehmann & Singh, 1993; Biel, 1993; D. Aaker, 199 6; Henderson & Cote, 1998), only a few have

PAGE 71

71 focused on the link between specific visual symbolic elements and brand personality (e.g., Akay, 2001; Boudreaux & Palmer, 2007; Cunningham et al., 2007; Jewett & Sutherland, 2013; Park et al., 2005). In fact, Ba tra, Lehmann and Singh (1993) lamented the lack of academic scholarship exploring this particular link, while Boudreaux and Palmer (2007) reported that the few studies that have focused on visuals readily acknowledge the tangible benefits of creating brand he current study will examine the link between specific visual symbolic elements and particular brand personalities. In one of only a few studies to studies found to tie specific advertising visuals to brand personalities, multi phase research design by Pa rk et al. (2005) created and tested a scale to brand personality. Results showed designers could successfully employ specific visual attributes (e.g., balance, symmetry, contrast) to shape a web brand personality. However, research conducted by Park et al . (2005) did not include specific, micro level descriptions of how to create these visual attributes. Like Park et al. (2005), Cunningham et al. (2007) also explored the overlap between b rand personality and website design, specifically in regards to e commerce websites. Based on previous literature, Cunningham et al. (2007, p. 10) suggested specific types of imagery for each of the eight MBTI dimensions, including: Extraversion : pictures of people Introversion : pictures of nature/solitude Sensing : real photos, timelines Intuitive : artistic representations, illustrations Feeling : colorful/harmonious images and backgrounds Thinking : monochromatic, black and white, tabulated data, graphs Judg ing : fewer images Perceiving : several images

PAGE 72

72 By combining their observations on individual dimensions, Cunningham et al. (2007) also developed a model of potential text, navigation and images that may appeal to each MBTI personality type (see Cunningham et al., 2007, p. 20 ). For example, to create a website with an extraverted/intuitive/thinking/perceiving (ENTP) personality, Cunningham et al. (2007, p. 20) suggested using pictures of people, artistic representations or illustrations, a monochromatic or bla ck and white color scheme, tabulated data and several images. Alternately, the authors hypothesized that a website with an introverted/sensing/feeling/judging (ISFJ) personality would have pictures of solitude or nature, real photos or timelines, colorful and harmonious tones and fewer images (Cunningham et al., 2007, p. 20). However, Cunningham et al. (2007) relied solely on previous research to compile their model and did not test their assumptions. In another study looking at specific types of imagery a nd personality, Akay (2001) found that participants who viewed television commercials were able to identify particular executional elements within an advertisement that they associated with specific MBTI personality cipants viewed television advertisements with both audio and visuals, many of the elements tied to brand personality were visual or could be represented visually. Participants noted fun, humorous and sexual elements in extraverted ads, while a romantic atm osphere and low lighting played an important role in introverted ads. Sensing ads were simple and factual with direct, rational messages, while intuitive ads used product personification and combined multiple images. Alternately, thinking ads were logical, direct and unemotional, while feeling ads incorporated images of family, emotion, accomplishment and flirtation. Akay (2001) found less conclusive results for judging and more well

PAGE 73

73 organized, while perceiving ads may include quick editing, sexuality and surprise endings, although these finding are weaker and less focused than the other personality facets. Finally, in a survey of undergraduate students ( n = 304), Jewett and Sutherland (2013) chosen images that represented single dimensions of the MBTI had the ability to transfer brand personality from an advertisement to a brand. Fur thermore, the authors found a high level of agreement more than 80% for each of the four dimensions included in the study (introversion, extraversion, thinking and feeling) between the intended brand personality chosen by the researchers and the acknow ledged brand personality recognized by the participant. The authors also proposed some specific visual traits that corresponded to each of the four personality dimensions included in the study: Extraversion : tight cropping, multiple people smiling and maki ng eye contact with the camera, public setting, bright/saturated colors Introversion : wide view, single person with back to the camera, low/romantic lighting Think : product shown, no people, white background, lines/text identifying product features Feel : p erson and animal, touching/cuddling, relaxed atmosphere (2013) study tested on ly one dimension of brand personality per advertising image. The current study plans to build on these results to simultaneously manipulate multiple personality dimensions with a single image, creating more specific and complex brand personalities (e.g., E ST, INF).

PAGE 74

74 Visuals Over Text Because both advertising and visuals were shown to play a significant role in the creation, expression and manipulation of brand personality, it is now important to specifically examine how visuals function within advertisements and how advertising images communicate with consumers. While much past scholarship examined advertising and its effectiveness, visuals erosion of the primacy of and consumer research continued to concentrate on the value of words, largely ignoring the increasing role of visuals (Pracejus, 2003, para. 1). Instead, advertising visuals often functione d as necessary, but overlooked, building blocks of advertisement stimuli that were investigated as mere afterthoughts or post hoc analyses. Further, Pracejus (2003, para. 1) noted that when advertising visuals were addressed, they were frequently portrayed as having only secondary or peripheral importance. visual advertising. In an examination of experimental, advertising related articles in the Journal of Consumer Research , Journal of Marketing and Journal of Marketing Research from 1990 1999, Pracejus et al (2003) found that 80% of the studies varied only the verbal, not the visual, information in test advertisements. The authors concluded that these Together, the historically chronicled rise in the importance of advertising visuals and the rand personality. As such,

PAGE 75

75 the following section will concentrate specifically on those studies that have purposefully examined advertising visuals, beginning with the history and advantages of using advertising visuals instead of text. History of Advertis Starting in the early twentieth century, advances in image and color reproduction technology accelerated, encouraging advertisers to favor images over text as early as the 1920s (Marchand, 1985). Around the same time, psychologists abilities to bypass logical argument, stimulate emotion and communicate messages. Today, almost a century later, modern advertisers still rely on the same advantages inherent in visuals. In fact, McQuarrie (2008) argued that contemporary mass media print advertising was made distinctive by its reliance on pictures to persuade. He further contended that, unlike in any other form of communication, images now shouldered the burden of persuasion in this genre. McQuarrie (2008) of fered two possible reasons for this trend a dismissive and a substantive explanation. The dismissive explanation viewed the increase in advertising visuals as a natural outworking of the increase in general visual stimulation consumers have experienced s ince the advent of television and web advertising, both of which are inherently visual media. However, he noted that such a dismissive explanation attempted only to explain away the change and required no substantial change in theory or attitude. Instead, McQuarrie (2008) suggested the truer answer might lie in a more substantive explanation one that saw the difference as a necessary and purposeful response to a fundamental change in consumer behavior. In his substantive explanation, McQuarrie (2008) asse rted that the modern American consumer had developed an However, he suggested that pictures did not currently suffer the same fate and were able to bypass consume r barriers, believably and effectively communicating direct advertising

PAGE 76

76 messages. Further, McQuarrie (2008) maintained that pictures in ads would be processed when words might not be (p. 109). Advantages of Advertising Visuals Over Text Like McQuarrie (200 8), other scholars also recognized the inherent advantages visuals over text. In fact, some scholars argued that an make the connection between an image and a product appear so natural that viewers accepted it without much deep questioning or counterargument (Biel, 1993; M archand, 1985; Messaris, 1997; Williamson, 1978/2010). According to Williamson (1978/2010) , This is why advertising is so uncontrollable, because whatever restrictions are at their use of images and symbols. And it is precisely these which do the work of the ad anyway[.] (p. 175) Thus, while verbal advertising claims were held to a relatively strict accountability textual arguments invited rebuttals, pictures deflected criticism and inspired belief (Marchand, 1985). This ability to byp ass questioning made images an ideal tool for making statements visually that would sound exaggerated, presumptuous or inappropriate in writing. Because consumers processed visual representations less logically than written statements, brand promises made through images were less likely to be challenged (Biel, 1993). For instance, the Green Giant character as a visual symbol of fresh from the farm by the viewer than a verbal claim stating that a brand of canned vegetables i 1993, p. 73).

PAGE 77

77 In addition to bypassing consumer logic, advertising images are also commonly used to circumvent personal (and potentially embarrassing) assumptions or expectations that a viewer may not want to directly confront through words particularly issues relating to sex and social status (Messaris, 1997; 1998). For example, by showing an attractive female fawning over an average looking male who was drinking a certain brand of beer, advertisers can create an association between female attraction and that specific beer brand. While few advertisers would premise is implicitly shown and understood every day in a multitude of ads. Or, as Earnes t no advertiser could say in words and retain his self This implicit visual argument works equally well for social status, allowing viewers to affiliate themselves with advertised products that align with a particular rung on the social ladder. In this too (Messaris, 1997, p. xxi). Building on this premise, the current study suggests that advertising allowing viewers to display their own personality and identity by aligning themselves wi th a In addition to the power contained in visuals to imply and persuade, past research also supports the superiority of visuals over text. For example, McQuarrie and Mick (2003) conducted a study on visual vers us verbal rhetorical figures. Although the authors tested both visual and verbal rhetorical figures in this study, their results divided the two groups and provided some specific insights on visual versus verbal stimuli. For the experiment, the authors cre ated a 32 page mock magazine with 16 ads (eight test and eight filler) and 10 articles. Each ad

PAGE 78

78 had a similar layout, with a pictorial image at the top center of the page and a headline and brand name at the bottom center of the page. In the experiment, pa rticipants ( n = 242) were divided into two main groups directed processing (i.e., participants were told to pay attention to the ads) vs. incidental exposure (i.e., participants were not told to pay attention to the ads). In the direct processing group, participants were told the marketing department was researching magazine advertisements. Directed processing participants were also instructed to carefully product s. The incidental exposure group was told the communication department was researching a new magazine directed at college students. Incidental processing participants were also instructed to focus on the articles and be prepared to answer questions evaluat ing the figure vs. non rhetorical figure, scheme vs. trope, and verbal rhetorical figure vs. visual rhetorical figure within the two main groups. Of the 16 ads they created for eight brands, there were eight rhetorical figure ads and eight non rhetorical control ads. Of the eight rhetorical figure ads, two contained visual tropes, two contained verbal tropes, two contained visual schemes and two contained verbal schemes (McQuarrie & Mick, 2003). In the magazine, scheme and two trope) and four control ads. The ads in each magazine were also evenly divided between visual and verbal manipulations. On arrival for the experiment, participants were given a copy of the magazine with instructions clipped to the front. After looking through the magazine, participants returned it and received a questionnaire, which included measu res of ad (2003) study showed higher ad recall for visually manipulated ads (31%) than verbally

PAGE 79

79 manipulated ads (5%). According to McQuarrie and Mick (2003), pictur e superiority was most evident in the incidental exposure group where participants were not specifically directed to pay attention to ads, creating a more real life advertising viewing experience. In the incidental exposure group, the authors reported 27% recall for visual schemes and 29% recall for visual tropes. Alternately, verbal tropes produced only 2% recall, and verbal schemes produced no appeared to have a greater impact than verbal figures, especially under conditions of incidental In another study, McQuarrie and Phillips (2005) conducted an experiment ( n = 177) to investigate the use of indirect persuasion in advertising. The study cre ated print advertising stimuli, which combined metaphorical and non metaphorical images and text to create four types of ads: 1) verbal literal (photo of product packaging and direct verbal claim), 2) verbal metaphor unanchored (photo of product packaging and metaphorical verbal claim), 3) visual metaphor (metaphorical picture and no verbal claim), and 4) anchored visual metaphor (metaphorical metaphorical claims in (McQuarrie & Phillips, 2005, p. 17). In other words, both verbal and visual metaphors were ab le message. However, according to McQuarrie and Phillips (2005), using pictorial metaphors rather than verbal metaphors strengthened this effect, suggesting that rhetorical figures (e.g., metaphors, puns, hyperboles) communicated visually in advertisements may be more effective than those communicated verbally.

PAGE 80

80 ) study on metaphoric advertisemensts versus non metaphoric advertisements further s upports the superiority of visuals over text. In their experimental design, the authors manipulated product type (symbolic vs. utilitarian), headline type (metaphoric vs. non metaphoric) and picture type (metaphoric vs. non metaphoric). Additionally, Ang a nd Lim (2006) chose a total of four different products two symbolic (i.e., cologne, designer watch) and two utilitarian (i.e., mineral water, toothpaste) resulting in a total of 16 author created advertisements. Each advertisement included a headline a t the top, an illustration in the middle and a small picture of the product with a fictitious brand name on the bottom right. Metaphoric while non metapho 44). Similarly, metaphoric images showed both the product and a picture of the object that the product was being compared to (e.g., broccoli with a bottle of mineral water), while non metaphoric images showed only the product (e.g., a bottle of mineral water). During the experiment, undergraduate students ( n = 200) were randomly assigned to one of eight treatment gr oups and asked to examine two print ads for either utilitarian or symbolic products (i.e., cologne and designer watch vs. mineral water and indicated their attit ude toward the ad, attitude toward the brand and purchase intent. attitudes toward the brand and purchase intent than metaphoric headlines. Furthermore, Ang and Lim (2006) found that when the image in an advertisement was metaphoric, the type of headline (metaphoric vs. non metaphoric) made no difference in attitude toward the ad, attitude toward es to the advertisements were

PAGE 81

81 contained wholly in the visuals, regardless of the type of textual anchoring. These findings orical picture was used (Ang & Lim, 2006, p. 50). In another experiment, Childers and Houston (1984) examined the effects of picture superiority on consumer memory. The authors showed undergraduate students ( n = 271) a series of 20 Yellow Page advertisemen ts, some of which contained both pictures and words and others composed of only words. As an additional variable, half the students were directed to process the ads at a sensory (shallow) level, while the other half was directed to process at a semantic (d eep) level. Both groups of students were asked to recall as many brand names as possible immediately after the seeing the advertisements and then again two days later. Childers and Houston (1984) found that advertisements with pictures were recalled better than word only ads in all cases, although that difference was not statistically significant for the immediate, sensory processing group. Furthermore, the picture superiority effect became even more evident over time, with the ident in the delayed recall results. Although the strength of the picture superiority effect varied based on the experimental conditions, Childers and Houston d their interaction with advertising images. Using eye tracking software, Rayner et al. (2008) assigned participants ( n = 2 4) to one of two groups and asked them to view 48 previously collected advertisements (Rayner et al., 2001) for a variety of products (e.g., cars, skin care, watches, food, laundry detergent) taken from popular British magazines. Participants assigned to t he like group were asked to rate how well they liked each ad, while the effectiveness group was

PAGE 82

82 asked how effective they thought each ad was (Rayner et al., 2008; Rayner & Castelhano, 2008). Despite their assigned group, results from Rayner et al. (2008) r evealed no significant differences in the amount of time participants spent looking at the advertisements. Similarly, in both the like group and the effectiveness group, participants spent considerably more time looking at pictures (61%) rather than text ( 39%) (Rayner et al., 20 08; Rayner & Castelhano, 2008). Although research showed that visuals play a crucial role in advertising, there is still much to be learned about how and when they are most effective. For example, Rayner et al. (2001) used eye tracki ng software to conduct another experiment about how participants viewed print advertisements when they were given a specific goal. American participants ( n = 24) were divided into two groups and asked to imagine that they had just moved to the United Kingd om and needed to buy a car (the car condition) or purchase skin care products (the skin care condition). Participants were then allowed to scroll through a series of 24 ads taken from popular British magazines (8 car, 8 skin care and 8 filler) at their own pace. Not only did participants spend more time looking at the ads relevant to their condition, but they also spent more time reading the text (71 73%) rather than looking at the pictures (27 29%). Interestingly, these results were in direct contrast to t heir later findings (Rayner et al. 2008) (detailed above), where into the role of text vs. images. py based ads were more effective than visual based ads. In a field test, the authors placed liquor advertisements in both leaving the bathroom and asked to verbally res pond to a questionnaire ( n = 146). Copy based

PAGE 83

83 bathroom ads had higher product recall and depth of information recall than visual based bathroom ads. When examined as a whole, it appears the effectiveness of advertising visuals might be closely related to o findings, it is possible that visuals are more important when participants have ge neral interactions with the advertisements (Rayner et al., 2001; Rayner et al., 2008). In contrast, text may become more central when consumers are given a specific task to accomplish (Lehmann & Shemwell, 2011; Rayner et al., 2001; 2008). Therefore, it cou ld be inferred that visuals might be superior for attracting casual consumers and influencing their initial interactions with a brand. Alternately, text might play a more significant role for serious consumers nearing purchase. Regardless, although images played a reduced role in some instances (e.g., goal oriented viewing), their overall importance, particularly in casual viewing, is evident (e.g., Ang & Lim, 2006; Childers & Houston, 1984; McQuarrie & Phillips, 2005; Rayner & Castelhano, 2008). As Postrel And, given the evidence above, the same attitude appeared to hold true for advertising brands were judged, at least in part, by the way their advertising looked and the visuals they used. Visual Devices In addition to demonstrating the superiority of visuals over text in many cases, past research also shows that consumers processed pictures more involuntarily, unconsciously and effortl essly than text, making them ideal carriers for brand messages (Arnheim, 1969; Danesi, 2008; McQuarrie, 2008). In fact, images are so powerful in advertising that it is not only the content that communicates information but also the way the content is pres ented. According to

PAGE 84

84 Messaris (1997), it is likely that viewers were less aware of these visual devices, such as visual techniques (e.g., point of view, angle, distance), visual syntax (e.g., metaphors, tropes, hyperbole) and visual symbolic elements (e.g., gender, setting, color), than they were of the Messaris (1997) further suggested that visual devices played a particularly substantial role in print advertising because creative directors spent more time meticulously designing the single, still image than the multiple, moving images of a television commercial. Therefore, payi ng visual influence. Visual Techniques Visual techniques, such as point of many visual devices for creat ing and controlling meaning within an ad. For example, advertisers have used various visual techniques, including implied distance (e.g., close up, long shot), angle (e.g., high, low) and subjective point of view (e.g., looking directly at the camera, look ing matter (Messaris, 1997, p. xiv). These visual techniques elicited different feelings and responses from the viewer depending on where they were in relatio n to the subject or scene. Advertising images also employed these visual techniques to draw attention to and elicit emotion from viewers, often by mimicking real life, face to face interpersonal human interactions (Messaris, 1997). For example, an image of someone looking directly at the viewer could innately cause him/her to return the gaze. Alternately, a low angle shot could produce feelings of inferiority or dependence (like a child looking up to an adult), while a high angle shot can product feelings o f superiority or authority (like an adult looking down at a child).

PAGE 85

85 In a two part study of automobiles and their owners, Schuldt, Konrath and Schwarz (2012) found the angle at which a product was shown in an advertisement could affect not only how viewers perceived the object but also how they perceived the owner. For the first experiment, Schuldt et al. (2012) selected eight different automobiles four sedans and four SUVs proc edure allowed the authors to manipulate both automobile type (SUV vs. sedan) and portrayal angle (head on vs. side profile). Community participants ( n = 492) were then recruited from cities across the U.S. by placing an advertisement on Craigslist.org. Dur ing the experiment, participants viewed an automobile from one angle and answered a series of seven point Likert scales related to status/power and favorability. The results of experiment one showed that automobiles depicted head on were rated significantl y higher in power/status than those viewed from the side. Additionally, consistent with their reputation as dominant vehicles, SUVs were ranked as significantly higher in power/status than sedans. However, when the results were examined more closely, Schul dt et al. (2012) found that head on portrayals increased the status/power rating of SUVs but did not influence the status/power rating of sedans. The results also revealed that the portrayal angle had no effect on favorability. According to Schuldt et al. (2012), the link between head on portrayals of SUVs and status/power underscored that visual (Schuldt et al., 2012, p. 708). In a second experiment, Schuldt et al. (2012) examined the link between associations about the product and associations about the owner. For this experiment, undergraduate students ( n = 106) read a brief profile of a target person, Adam , viewed several pi ctures of his daily life including a picture of his SUV, and reported their impressions of him using a power/status scale.

PAGE 86

86 The only element manipulated in the experiment was the SUVs portrayal angle (head on vs. side ed that Adam received a significantly higher status/power rating when his SUV was depicted head on rather than in side profile, suggesting that a Overall, accor ding to Schuldt et al. (2012), participants who viewed an SUV head on rather than from a side profile rated both the vehicle and its owner higher in status/power. These findings further supported the importance of advertisers paying close attention to not only the content of advertising images, but also to specific visual cues within the images, such as horizontal and vertical angles. The findings also suggested that subtle visual manipulations of a product might shift its meaning, which could then be trans ferred on to the owner. In a discussion of semiotics and visual meaning, Jewitt and Oyama (2004) also argued that the same image could have different meaning based on how it interacted with the viewer, namely through contact, distance and point of view . examined whet her the individual in a visual wa s looking directly at the viewer. The authors identified images with direct eye contact as demand images because they were symbolically demanding something from the vi ewer. Exactly what they were demanding could be determined by facial expressions and gestures. For example, an image could demand deference by Oyama, 2004, p. 145). Adding gestures also modified or further enhanced the meaning, as individuals in an image were not engaging the viewer through contact, Jewitt and Oyama (2004) re ferred to them as offers because they simply offered information. In these images, viewers tended to observe subjects in a more detached, impersonal way.

PAGE 87

87 Jewitt and Oyama (2004) also examined the symbolic meaning of distance between the viewer and the indi viduals in an image, asserting that this distance mirrored the norms of real life social interactions. The authors suggested that close up shots mimicked the physical closeness of Oyama, 2004, p. 146). Conversely, far away shots reflected the way an individual normally saw a stranger impersonal and in outline. Distance was also shown to visually communicate group belonging, suggesting to a viewer that they belong ed in the same group as people in close up shots, while people in far away shots differed from them. Jewitt and Oyama (2004, p. 146) noted a variety of distances and meanings that existed in between close up and far away by using film terminology, includin g: Close up shot (head and shoulders or less): intimate/personal relationship Medium shot (cuts off the human figure between waist and knees): social relationship Long shot (full human figure whether barely fitting in frame or much farther off): impersonal relationship Finally, point of view indicated the position from which the viewer interacted with an image. According to Jewitt and Oyama (2004), in point of view, the vertical angle determined the symbolic power relationship. If a viewer was looking down on an object, it suggested they were in a symbolic position of power. In contrast, positioning the viewer as looking up at an object implied they were symbolically subordinate to the object. Eye level viewing, then, created a symbolic equality. Alternately , the authors proposed that the horizontal angle determined the amount of viewer involvement. A front on view created maximum involvement, directly literally and

PAGE 88

88 addition to these discussed points of view, the authors noted that there were many degrees and meanings in between (e.g., three quarters view, birds eye view). Visual Syntax I n addition to using the aforementioned visual techniques (e.g., point of view, distance, also provide products with meaning through visual syntax (e.g., casual connections, metaphors, relationships in visual syntax could occur either se quentially over time (e.g., TV, movie) or simultaneously in a single visual display (e.g., print advertising) (p. 75). To correspond with the focus of the current study, the following section will examine visual syntax as it occurs in single visual display s, specifically print advertising. According to Messaris (1997), advertisers employ visual syntax by juxtaposing a product meaning for the product. Messaris (19 97) suggested that ads often used visual syntax to connect a product to people (e.g., typical users, models, celebrities), lifestyles (e.g., setting, social context) or images, allowing the product to take on the meaning of the image sharing its frame (i.e ., advertise ment) (Messaris, 1997, p. 184). More specifically, Messaris (1997) proposed that visual syntax included: casual connections, contrasts, analogies and generalizations. Casual connections contained implicit promises that if the consumer used this product, then they would be associated with some of the same qualities, interactions, attributes or satisfactions (Messaris, 1997, p. 184). Contrasts, typified by product comparisons and before and after advertisements, juxtaposed the product with an imag e in order to illustrate differences and show the viewer what the product was not.

PAGE 89

89 Analogies attempted to reinforce some quality of the product by showing similarities between the product and the image. For example, an ad might use visual analogy to reinfo speed by showing both an image of the car and an image of a cheetah. With generalizations, ads showed specific instances of a visual concept to illustrate a more general point. Or, alternately, they might use generalities to further a particula r single instance. Often, ads using generalizations featured multiple images or vignettes with a single commonality. Zuckerman (1990) conducted a study to examine the ability of visual syntax to transfer an work, Zuckerman (1990) tested real world ads rather than researcher designed stimuli. In an experiment with high school students, Zuckerman (1990) collected advertisements that used visual syntax from magazines the students were known to read. In particular, the author chose 1997; Zuckerman, 1990). For example, one of the ads chosen was for Marlboro cigarettes. Because Marlbor o cigarettes were targeted at women before the appearance of Marlboro man in the 1950s, the authors concluded that there was nothing inherently masculine about the product. Therefore, any masculine associations with Marlboro could be reasonably assumed to come from uline imagery with the product. To further ensure that any meaning found in the study came from the advertising and not ges. For example, Zuckerman (1990) paired Marlboro cigarettes with Benson & Hedges cigarettes. At the time, Benson & Hedges cigarettes was running a campaign filled with images of upscale people and wealth in urban settings. Zuckerman (1990) then used the sample ads to create two

PAGE 90

90 slide shows one with pictures of the products extracted from the ad (e.g., a pack of Marlboro cigarettes), and the other the advertising image (e.g., a cowboy on horseback). The high school students were then divided into two gr oups one was shown the product only slides, and the other was shown the image only slides. Both groups rated each slide using a 23 item, bi polar adjective scale (e.g., urban/rural, masculine/feminine). Zuckerman (1990) found that students tended to rate the products and their associated images similarly. For example, students rated both the package of Marlboro cigarettes and the cowboy image as masculine. Additionally, the study showed that students rated product pairs differently. For example, on the ur ban/rural scale, students rated Marlboro cigarettes as rural, but rated Benson & Hedges cigarettes as urban. Because both parts of the ad (i.e., the product or the image) were assigned similar descriptions despite being rated by different sets of students, Zuckerman (1990) including the forms discussed above, was an effective meaning to a product (Janiszewski, 2008; Jeong, 2008; Messaris, 1997; Mitchell & Olson, 1981; Morgan, 2005). Visua l rhetoric two visuals, other scholars began working toward an understanding of visual rhetoric as a tool to communicate meaning in advertising images. In their much cited taxonomy for classifying rhetorical figures (also known as figures of speech) in advertising (s ee McQuarrie & Mick, 1996, p. 426 identified two figurative modes: 1) schemes , which contain excessive order or regularity, and 2)

PAGE 91

91 tropes , which contain a deficiency of order or regularity. Within the two figurative modes (i.e., schemes and tropes), McQuarrie and Mick (1996) defined multiple rhetorical operations. Schemes contained repetitions (e.g., rhyme, alliteration) and reversals (e.g., antimetabole, antithesis). Alternately, tropes contained substitutions (e.g., hyperbole, ellipsis) and destabilizations (e.g., metaphor, pun). 20 03) research was on verbal rhetorical figures, their conclusions specifically noted the existence come from setting aside verbal materials altogether and examining th e visual component of ad In fact, McQuarrie and Mick (1999) soon turned their own scholarship to the exploration of experiments and interviews, McQuarrie and Mick (1999) examined the effect of visual rhetorical figures on interpretations/inferences drawn from it) and pleasure (the enjoyment a consumer gets from an actual magazine ads with a variety of visual rhetorical figures were collected two schemes (rhyme and antithesis) and two tropes (meta phor and pun). In addition to the original ads, a parallel set of non rhetorical ads was created by digitally removing or breaking the visual rhetorical figures. The headlines for all eight ads were also removed, and a matter of fact headline was placed be low the image so that it did not overlap or interfere with the visual. Additionally, in all eight ads, the brand name was replaced with a fictitious name. For example, in the original metaphor ad, a box of motion sickness medication replaced the buckle to seatbelt. In the parallel, non rhetorical ad, the product box was moved away from the seatbelt

PAGE 92

92 sickness? Take Pramnol McQuarrie & Mick, 1999, p. 43). In the first experiment, undergraduate students ( n = 72) reviewed a subset of four ads (two rhetorical and two non rhetorical) and completed a six item elaboration scale for each ad (McQuarrie & Mick, 1999). Results from experiment one established the ability of visual rhetorical figures to alter elaboration, showing significantly more elaboration for visual rhetorical figures compared to the non rhetorical contr ol in each of the four ad sets. In the second experiment, undergraduate stud ents ( n = 181) repeated the same process as experiment one, but completed scales for attitude toward the ad, difficulty of comprehension, preference for visual processing and answered a variety of additional personal questions (e.g. sex, original culture, whether they ever suffer from motion sickness) (McQuarrie & Mick, 1999). Based on findings from experiment two, McQuarrie and Mick (1999) demonstrated visual tropes (e.g, metaphors, puns) demonstrated an even more pronounced increase in positive attitude toward the ad compared to more simplistic schemes (e.g., rhyme, antithesis). ults, suggesting that consumers needed no special visual propensity to understand and appreciate visual rhetorical advertisements for potentially gender spec ific products, indicating that consumers did not necessarily need to be a product user to appreciate visual rhetorical devices. However, to understand visual rhetorical figures, leading the authors to investigate further.

PAGE 93

93 In a third part of the study, individual interviews were conducted with undergraduate students ( n = 12), including four American females, four American males and four international, female stu dents from Asian countries (McQuarrie & Mick, 1999). Each interviewee was shown the four ads that used visual rhetorical figures and asked about their meanings. After providing meanings for each ad, the interviewees were then asked about which particular a spects of the ad contributed to their ascribed meaning. While all three groups were able to understand and draw similar meanings from the visual schemes (i.e., rhyme and antithesis), the Asian female students had marked trouble interpreting and understandi ng the visual tropes (i.e., metaphor and pun). Alternately, both male and female American students easily interpreted similar meanings from the visual tropes, despite their differences in gender and potential product use (e.g., mascara, weight loss product s). Overall, McQuarrie and Mick (1999) found that, when compared to their non rhetorical counterparts, ads that used visual rhetorical figures resulted in more elaboration dominant and power of visual rhetorical figures to provide meaning in ad preferred meaning. Although McQuarrie and rhetorical figure from the matter of fact headline under the guise of an ad in progress, the verbal text might still have performed an anchoring function (Barthes, 1977), directing participants to a particular meaning. To avoid the possibility of verbal interference with the visual meaning, the current study will avoid using any headlines.

PAGE 94

94 In a second study (discussed in detail above), McQuarrie and Mick (2003) revisited the impact rhetorical figures had on con sumer responses to advertising, this time specifically examining the differences between directed processing (i.e., participants were told to pay attention to the ads) and incidental exposure (i.e., participants were not told to pay attention to the ads). processing was not directed (i.e. the incidental exposure group), recall within the incidental exposure group was almost twice as likely for ads that contained rhetorical figures (25%) than ads without rhetorical figures (14%). Further, recall for rhetorical figure ads increased in both the directed processing and incidental exposure groups, suggesting that ads with rhetorical figures were more memorable despite the amount of processing effort consumers expend. Additionally, within rhetorical figures, tropes (e.g., metaphors, puns) had higher recall than sch emes (e.g., rhyme, antithesis). According to McQuarrie and Mick (2003), visual rhetorical figures achieved better reca ll than their controls, with visual tropes outperforming visual schemes. Rhetorical figures also had by both directed processing and incidental exposure gro ups. Similar to ad recall, visual tropes also significantly outperformed visual schemes in both the direct processing and incidental exposure groups. Overall, McQuarrie and Mick (2003) findings supported past evidence of the positive impact of visual rheto rical figures, particularly tropes (e.g., metaphors, puns), on consumers. 2003) original work, McQuarrie later teamed up with another researcher to develop a typology specific to visual rhetoric in advertising (Phillips & McQuarrie, 2004). Phillips and McQuarrie

PAGE 95

95 rhetorical figures (e.g., McQuarrie and Mick, 1996) do not adequ ately capture important visual rhetorical figure are physically pi resulting in a nine part matrix (Phillips & McQuarrie, 2004, p. 116). The visual structure was divided into by operation was (Phillips & McQuarrie, 2004, p. 116). Further more, Phillips and McQuarrie (2004) predicted that in cognitive elaboration, belief formation/change, ad liking and recall although, the article does not t est these predictions. Visual m etaphors insights on the effects and effectiveness of visual rhetorical figures and inspired other researchers to produce their own typologies. Ho wever, more recently, scholars have built on McQuarrie and Delbaere et al., 2011; Gkiouzepas & Hogg, 2011; Jeong, 2008; Ma, 2008; Van Mulken, Le Pair & Forcevi Quarrie & Mick, 1996, p. 431).

PAGE 96

96 Mulken et al. (2010 ) specifically examined three categories of visual metaphors based on the spacial distribution of the source image (i.e., the image that contains the original meaning) and the target image (i.e., the image that takes on new meaning from the source image), including: similes (the target and the source are presented as visually separate), hybrids (the target and the source are visually fused together), and contextual (either the target or the source is visually absent). According to Van Mulken et al. (2010), their visual metaphor categories were similar to called similes , hybrids and contextual metaphors , Phillips and McQuarrie (2004) called juxtapositions , fusions and replacements . Using an experiment, Van Mulken et al. (2010) sought to examine the validity of their life automobile advertisements, which were selected by a panel of experts ( n = 17) in a pretest, including five simile ads, five hybrid ads, five contextual ads and five non metaph orical ads. Logos, slogans and brand names were removed from the advertisements. Participants ( n = 212, 75 Dutch, 68 French and 69 Spanish) were gathered though snowball sampling via the Internet. Each participant viewed all 20 advertisements and completed an online questionnaire that had been translated into their own language. Using a series of scales, the questionnaire measured deviation from expectation (i.e., not straightforward), perceived complexity and appreciation. After the scales had been complet ed for each of the 20 test advertisements, participants were asked to reexamine each advertisement and indicate whether they had recognized a comparison the first time they had viewed the

PAGE 97

97 advertisement. If they answered yes, participants were prompted to c this advertisement, the automobile is comp understand how other people interpreted it (Van Mulken et al., 2010, p. 3423). R esults from Van Mulken et al. (2010) showed that participants viewed advertisements with all types of visual metaphors (i.e., simile, hybrid and contextual) as more deviant (i.e., not straightforward) and more complex than advertisements with no visual metaphors. In terms of devia nce, hybrids were considered the most deviant, followed by similes and then contextual metaphors. According to Van Mulken et al. (2010), participants might have found hybrids most deviant because they fused two pictorial elements into a single element in a way that was unlikely to occur in reality. In terms of complexity, contextual metaphors were considered more complex than both similes and hybrid metaphors. Additionally, findings indicated that participants had a higher appreciation for all types of visu al metaphors if they recognized and understood the comparison being made. Therefore, Van Mulken et al. (2010) suggested that increased deviation from expectation, increased comprehension and decreased perceived complexity combined as good predictors of a c However, the authors did not rule out additional factors that future research might find also potentially, brand person ality. Although Van Mulken et al. (2010) found some small cultural differences in how participants viewed visual metaphors (e.g., French and Dutch participants perceived no

PAGE 98

98 difference in the complexity of similes and hybrids, while Spanish participants con sidered hybrids to be more complex than similes), the authors noted that most variances were only relative and, therefore, inconsequential. Instead, findings from Van Mulken et al. (2010) showed that despite differences in language and culture, the three W estern European nationalities (i.e., Dutch, French and Spanish) viewed visual metaphors in largely the same way, further supporting advertisement through visuals (Ha ll, 1980). Finally, in regards to their visual metaphor typology, Van Mulken et al. (2010) confirmed that visual metaphors could be distinguished based on the space between the target and the source, which also influenced deviation from expectation and per ceived complexity. More importantly, the authors also found that consumers responded differently to visual metaphors based on the spacial distribution of the source and the target images (i.e., similes, hybrids and contextual metaphors). In another study, n = 261), which found that advertisements using metaphorical images were more persuasive than those using non metaphorical images. During the experiment, the author asked participants to vi ew and respond to one of three types of advertising stimuli non metaphorical (literal product with verbal argument), metaphorical with verbal anchoring (metaphorical image with verbal argument), and metaphorical without verbal anchoring (metaphorical ima ge with no verbal persuasive power stemmed from requiring more cognitive processing and elaboration from consumers i n short, asking consumers to invest more in simply being told with words. Furthermore, Jeong (2008) argued that the higher level of consumer participation required in extracting meaning from a metaphorical image might make

PAGE 99

99 consumers more likely to adopt an ad (2008) results, not only could consumers extract meaningful messages from visual only metaphorical adverti intended message. Although the current study will not use metaphorical images, like Jeong intended meanings from visual o nly, advertising images. In another study of visual metaphors, Ma (2008) surveyed Chinese students ( n = 50), Nike advertisements that contained a pictorial (i.e., visual) metaphor. Because Ma (2008) w as interested in both the meaning participants ascribed to the visual metaphor and the culture from which it originated, the survey metaphor and their percep tions of and exposure to American culture. Participants were first asked to describe each advertisement, and then answer a series of open their interpretation of the visual metaphors, when the metaphor was clear, their negotiated readings aligned largely with a single preferred reading. For the first ad, which showed a Nike shoebox that opened to reveal a soccer stadium, 94% of the particip ants agreed that the Nike an arrow leading straight from the Nike l ogo to the center with no regard for the twists and turns, Nike rejected the

PAGE 100

100 (p. 14). Alternately, Ma (2008) admitted that the metaphor in the third ad, which showed a tiny man flying high above rugged mountain terrain, was more complicated, leading to etaphorical implicature [as in the first two ads], the more likely the consumers are able to reach an agreement about its meaning; otherwise, the weaker the implicature [as in the third ad], the greater 6). In other words, by carefully selecting strong visual metaphors, advertisers are capable of clearing communicating their preferred meaning. It is only when advertising visuals are poorly chosen and weakly executed that the preferred meaning in lost for consumers. found to examine brand personality in terms of metaphoric advertising images. Findings showed that metaphoric images resulted in higher attitudes toward the ad , attitudes toward the brand and purchase intent and enhanced responses to the advertisement despite the type of headline used. In terms of brand personality, Ang and Lim (2006) observed that ads with metaphoric images had higher levels of sophistication a nd excitement and lower levels of sincerity and competence compare to nonmetaphoric images. Overall, Ang and Lim (2006) noted that pictorial metaphors could influence brand personality perceptions and suggested that practitioners could use metaphorical ima ges to create or modify brand personality. Visual e xaggeration Past research has recognized and categorized visual metaphors as persuasive advertising tools, and scholars have established their ability to influence consumer response (e.g., elaboration, att itude toward the ad, recall, liking). However, although loosely related to visual metaphors, visual exaggeration (e.g., visual puffery , visual hyperbole ) has received little academic attention. For example, researchers have only recently begun to investiga te the concept

PAGE 101

101 of visual puffery visual cues and imagery in advertisements that result in higher product expectations than actual product evaluations (Toncar & Fetscherin, 2012). While academics have investigated the effects of verbal puffery and U.S. co urts have handed down definitions and rulings about its use, visual puffery has been largely overlooked. Despite court rulings in the last visual p. 147), only one study specifically addresses the concept of visual puffery. In a multi generate sensory expectations, Toncar and Fetscherin (2012) found evidence of visual puffery. The authors began by conducting a semiotic analysis of three perfume advertisements to identify adjective pairs (e.g., light/understated, bold/powerful, arousing/sensual, romantic/femini ne) that described the range of meanings present in the ads. The authors then conducted structured interviews where potential consumers completed a survey, which asked them to evaluate each of adjecti ve pairs. After evaluating the advertisements, the interviewees were asked to blindly smell each perfume and use the same set of attribute adjective pairs to evaluate the smell. Toncar and Fetscherin (2012) found that puffery did exist and might help reinforce intangible product benefits like brand personality. Visual Sym bolic Elements In addition to examining the meanings associated with a position in the frame and the influence of visual syntax on consumer responses, other authors have explored the meaning of specific visual symbolic element within an image. As discussed earlier and in line

PAGE 102

102 visual elements that communicate culturally based meanings as visual symbolic elements . According to Pracejus (2003), visual elements accentuate or interfere with words, visual elements themselves func tioned as information (para 2). McCracken (1990) asserted that during the process of creating an advertisement, t he creative team chooses specific symbolic elements to communicate precise (and predetermined) cultural meanings, often through visuals. According to McCracken (1986), these visuals are imbued with cu ltural meanings, such as gender, status, age, lifestyle, time and place. Research also showed that ad creators clearly understood the meaning of specific visual elements (Pracejus, 2003). By placing a product or brand in the same frame as an image, adverti sers invite consumers to form nonsensical connections between the meaning filled, visual symbolic element and the product or brand. In other words, because things are shown in the same space, consumers assume they are connected. For example, by simply juxt aposing a perfume beautiful women and their brand of perfume. Visual symbolic elements also work to tell consumers what a product or brand is not. In this examp le, the perfume is not for athletes or men or ugly women. In addition to ad creators understanding the meanings embedded in visual symbolic elements, Pracejus (2003, para. 2) argued that consumers were also able to decode visual meanings. Based on past res

PAGE 103

103 In the brand personality research presented above, several studies that examined visual symbolic elements in terms of brand personality were discus sed ( e.g., Akay, 2001; Boudreaux & Palmer, 2007; Cunningham et al., 2007; Jewett & Sutherland, 2013; Park et al., 2005 ). The following section will present literature that investigates symbolic visual elements in terms of advertising, both outside of and w ithin gender studies. Visual symbolic elements in advertising In one of the few advertising studies found to examine symbolic visual elements not related to gender or brand personality, Pracejus et al. (2003) examined the meaning of white space in adverti sements according to both the ad creators and consumers. First Pracejus et al. (2003) surveyed creative directors at North American advertising agencies ( n = 31) to discover their opinions on the meaning of white space in an advertisement. According to the results, creative directors most frequently proposed that white space communicated prestige, market power, trustworthiness, industry leadership and brand quality. Next, to examine whether viewers understood these same meanings, the authors asked consumers ( n = 63) to examine two advertisements for the same product one manipulated to include a high amount of white space and the other to include a low amount of white space. Consumers were asked which ad best conveyed prestige, market power, trustworthiness , industry leadership and brand quality. Based on results, high amounts of white space conveyed prestige, trustworthiness and brand quality significantly more than low amounts of white space, indicating that consumers were able to understand several of the meanings ad creators i ntended white space to express. explicitly pointed out (Pracejus et al., 2003). Participants ( n = 130) were exposed to only one ad (either high o r low in white space) with no explicit mention of the white space in either case. Similarly to the previous study, prestige, trustworthiness and brand quality were all significantly

PAGE 104

104 increased for the high white space ad. Additionally, the third study also showed an increase in attitude toward the brand and purchase intention but still not market pow er or brand leadership. A fourth study was conduc ted to ensure that the lack of size context in the previous studies had not influenced results. For this study, the authors manipulated both white space (high vs. low) and ad size (three sizes) and placed the test ads in a mock newspaper. Like the previous studies, results showed that white space had a significant positive effect on prestige, trustworthiness, brand quality, attitude toward the brand and purchase intention. Additionally, the fourth study also showed white space to significantly affect market power and brand leadership. However, ad size only significantly affected perceived company size. Based on findi ngs from Pracejus et al. (2003), white space clearly functioned as a meaningful symbolic visual element, which contained information. Furthermore, Pracejus et al. (2003) study demonstrated that, despite size, consumers were able to correctly interpret the specific white space meanings intended by ad creators, particularly when advertisements were placed in context (e.g., a newspaper). Beyond white space, the results from Pracejus et al. (2003) suggested that consumers were visually literate and capable of u nderstanding the meanings contained in symbolic visual elements. Because the current study will use full bleed images, there will be no traditional white space in the advertisement designs. However, white space has been considered within the context of the images. It is hypothesized that introverted (I) images will have wider crops with more background and empty visual space (i.e., white space) than extraverted (E) images, which are hypothesized to be closely cropped with subjects filling most of the frame. In another study of symbolic visual elements, Cline and Young (2004) examined the meaning contained in direct to consumer advertising (DTCA) visuals for prescription drugs. Cline and Young (2004) conducted a content analysis of 994 total DTCA across 18 po pular

PAGE 105

10 5 magazines from January 1998 to December 1999. The authors looked for several rew ards categories: identity rewards, activity re wards an d relational rewards. Identity re wards included whether the models were healthy (e.g., healthy, ill), friendly (e.g. , smiling, not smiling) and active. The activity category was further divided into physical activity (e.g., climbing stairs, dancing, work related clothing), social activity (e.g., picnics, parties, embracing), taking medicine and inactive/passive (e.g., r eading, head shot, body part). Relational rewards included social context and relational context. Social context included family (e.g., people from two generations, baby), romance (e.g., only two people embracing or gazing at each other), work (e.g., work related clothing or equipment), recreational (e.g., relaxing), and other. Relational context was based on the number of people shown in the ad, and included alone (one person), dyad (two people) or group (more than two people). Although Cline and Young (20 04) did not detail the cultural meanings associated with each visual symbolic element they coded, the visual symbolic elements revealed that DTCA portrayed pos itive personal characteristics (e.g., (Cline & Young, 2004, p. 150). Cline and Young (2004) reported that, although DTCA were typically associated with serio us, debilitating medical conditions, more than 90% of the ads depicted exclusively healthy people with more than half engaged in physical or social activities. advertise d conditions could be attractive, healthy and lively with prescription drug treatment (Cline & Young, 2004, p. 151) even though medical results showed this is not always true. Additionally, Jewitt and Oyama (2004) looked at how meaning was conveyed through the people, places and things in images. The authors suggested viewers might recognize a young

PAGE 106

106 male as a wimp ive pose, averted gaze, glasses, [and] 2004, p. 141). Furthermore, the was also represented by such symbolic attributes as size, position, color, lighting and gesture. Props could also provide visual meaning, for example cars/motorcycles were often symbo ls of male virility/sexual prowess and sports equipment/settings might signal heterosexu ality and male competitiveness. Visual s ymb olic elements in gender studies Perhaps more than any other line of research, gender studies understand s the meaning contain ed in specific advertising visual symbolic elements and regularly identif ies the particular visual symbolic elements that lead to specific conclusions (e.g., Goffman, 1979; Lindner, 2004; Schroeder & Zwick, 2004). If fact, in one study of gender meanings i n advertising, Schroeder and Zwick (2004) astutely acknowledged, Images, saturated by a long cultural history, constitute an engaging and deceptive culturally and historically bound visual langu age system. Advertising imagery as a subset within this syst em interacts with it, borrowing from and influencing the larger world of visual culture. (p. 45) However, gender researchers are often narrowly focused on how visual symbolic elements influence interpretations of masculinity and femininity, ignoring othe r possible cultural meanings. For example, in a seminal book examining gender portrayals in advertisements, Goffman (1979) focused on hands, eyes, knees, facial expressions, head posture, relative size, positioning/placing, head eye aversion and finger bit ing/sucking or what this study would term visual symbolic elements. By analyzing these individual elements in more than 500 advertisements, Goffman (1979) argued that larger assumptions could be made about gender roles in society. Goffman (1979) broke hi s gender findings into the following categories: Relative size : Differences in size correlated to differences in social weight, with more importance given to the relatively larger/taller individual. Men were almost always

PAGE 107

107 shown as taller than women, unless the man was clearly subordinate in social class and often dressed in a servitor costume (e.g., chef, butler, server). Feminine touch : Women were often shown just barely touching, cradling, caressing or tracing the outline of an object in an ad but rarely grasping, manipulating or holding. underscoring her body as a delicate, precious thing. Function ranking : When men and women were shown working together, the male was likely to perform the executive, guiding or instructional role. When men were shown as under the instruction of women, it was most often in a traditionally female domain (e.g., the kitch was limited to domestic duties. Alternately, ads may choose to picture men as incompetent or childlike as they attempt traditionally female duties (cooking, cleaning, ironing, chi ld rearing), or mildly successful under the close watch of the female. Family : Most often, the nuclear family was portrayed as a father, mother, one female and further depicted the dominant male slightly outside the family circle, playing the role of protector. Ritualization of subordination : Visually, lowering or prostrating indicated deference, while superiority was depicted by holding the head high and body erect. Typ ically, women were more likely to be positioned lying on a bed, sofa or floor, while men were placed physically above the women and higher on the page. Goffman (1979) also noted ne knee is bent slightly forward or a canting posture where the body or head is lowered, signifying submission. Furthermore, women were more likely to smile and exhibit playful, childlike or clowning poses, while men were more seriously portrayed. This uns erious behavior suggested that women were more naïve, flippant and changeable than men, who were serious, committed and formal. Goffman (1979) also identified the male extended arm (creating a boundary around a female), arm lock (male lending support to a female), shoulder hold (potentially a sign of sexual proprietorship), hand holding (an indication of sexual potential and an exclusive relationship) as additional signs of male domination and female submission. Licensed withdrawal : Women, more often than m en, were shown to have psychologically withdrawn from the social situation, leaving them dependent on others. This withdrawal can be visually shown through covering the face often with hands out of fear, shyness or laughter. Sucking or biting the finger ca n also be seen as partially covering the face and may indicate anxiety, thoughtfulness or disengagement. Similarly, finger to finger positioning also implied disengaged rumination. As other forms of withdrawal, Goffman (1979) also addressed head/eye aversi on (to conceal current feelings) and withdrawn gaze (indicating a wandering mind).

PAGE 108

108 However, despite identifying specific visual symbolic elements and the gender meaning they contained, Goffman (1979) did not investigate any non gender related, wider cultur al meanings they may potentially contain. Similarly, in a longitudinal content analysis of how women were portrayed in advertising, Lindner (2004) examined advertisements ( n = 1,374) in Time and Vogue from 1955 to 2002. To analyze the ads, Lindner (2004) u touch, ritualization of subordination and licensed withdrawal categories supplemented with additional categories from other researchers, including body display (Kang, 1997), movement and locati on (Umiker Sebeok, 1996) and objectification (a new category developed for the study). The body display category examined the amount of clothing a women was wearing (e.g., revealing clothing, hardly any clothing, no clothing at all), which is often associa ted with the sexualization of women. The movement category considered whether a woman was inhibited in her ability to move (e.g., wrapped in a blanket), symbolically limiting her ability to exert control over her environment. The location category noted wh ere a women was pictured (e.g., kitchen, bedroom, bathroom, nowhere), looking for women in domestic environments or in decontextualized environments that suggested lack of purposeful activity (e.g., self absorbed grooming, no action). The objectification c ategory looked at whether a woman was portrayed as merely an object to be looked at in the ad but no specific visual symbolic element examples were listed for this category. Overall, Lindner (2004) found that women were depicted more stereotypically in Vog ue than in Time , and that the stereotypical depictions of women had only slightly decreased over time. Similarly, examining male gender stereotypes in advertising, Schroeder and Zwick

PAGE 109

109 doing the laundry, (p. 22). In addition to examining the male representation in historical, ontological and photographic realms, Schroeder and Zwick (2004) analyzed several purposefully chosen advertisements. In their assessment, Sc hroeder and Zwick (2004) provided some specific visual symbolic elements and interpretations. For example, in one alcohol advertisement, a somewhat unattractive man look ed into a mirror after ordering a particular liqueur. Reflected back at him, he saw a s looking twenties, tall, dark, a rakish curl of hair falling seductively down h handkerchief and clean, white dress shirt. The authors also noted that when looking in the mirror, the unattractive man had lost his visual implies there are certain male characteristics (e.g., young, tall, dark, well tailored) th at are universally attractive and other less desirable attributes that require alcohol to overcome (e.g., eyeglasses, pointed nose, oversized chin). for the idea that visual symbolic elements convey meaning that is rooted in culture. The current study will take advantage of these culturally infused meanings to embed a particular, preferred brand personality into the test advertisements through carefully selected visual symbolic elements. Additionally , as

PAGE 110

110 a whole, the above studies highlight the potential meaning present in every visual aspect of an image. As such, it also is important to be aware of the potential to convey unintentional meanings through visual symbolic elements, which may influence th e ability to transfer the preferred brand personality. Therefore, not only is it important to choose images that include visual symbolic elements that convey the preferred brand personality, but also to choose images that do not include visual symbolic ele ments that communicate other, unintended brand personalities. Meaning Transfer Theory Past research demonstrated that visuals contained meaning, which advertisers could pair with products through visual techniques (e.g., point of view, angle, distance), vi sual syntax (e.g., metaphors, tropes, hyperbole) and visual symbolic elements (e.g., gender, setting, color). In fact, scholars have long accepted that goods contain meaning, and that individuals conspicuously consume goods, borrowing that meaning to tell the world something about themselves (e.g., Danesi, 2008; McCracken, 1986; 1989; 1993; Veblen, 1899/1994). As an early pioneer of this idea, Veblen theorized that individuals judge and are judged based on the goods and services they conspicuously consume ( Veblen, 1899/1994). Veblen (1899/1994) noted that conspicuous monetary st atus based on his/her visible consumption. Although the venues for conspicuous day schools, workplaces and malls, individuals are still routinely evaluated based on the brand s they purchases meant to exhibit weal th and status (Messaris, 1997 ).

PAGE 111

111 Indeed, according to Veblen (1899/1994), individuals conspicuously consume to show 2005) noted an additional, and perhaps more powerful, motivation for modern longer conspicuously consume only to demonstrate their status and wealth. Rather, McCracken to define themselves, fashioning an identity by capturing the meaning that resides in goods to construct their own lives. Thus, 54). Today, individuals conspic uously consume to create and define meaning for themselves as well as the watching eyes of others. Combined, the high premium placed on freedom of choice and the disintegration of institutions that previously supplied individuals with meaning and definitio n (e.g., family, church, community) has left modern individuals to construct their identity from the ground up (McCracken, 2005). Particularly in Western cultures, the combination of individualism and alienation gives individuals almost boundless freedom t o define matters of gender, class, age, personality and lifestyle. However, when left to create 2005, p. 112), placing significant pressure on individuals to cons ume goods that provide the desired cultural meanings. Williamson (1978/2010) agreed that individuals are no longer identified by the work that they do or the products they produce , but, rather, b y what they consume . With consumption providing the me ans and meaning consumers need t o construct an identity, it beco me s essential to understand where that meaning originated, how it came to reside in goods, and how consumers extracted the meaning for self definition. In other words, for goods

PAGE 112

112 identity needs, there must be a process for both infusing and extracting meaning. McCracken (1986; 1988; 1990; 2005) offered one explanation for this process in his meaning transfer theory, which modeled the movement of meaning from a culturally constitut ed transferred to consumers via a two step process (see McCracken, 1986, p. 72 ). Cultural meaning is first transferred from a culturally constituted world to a consum er good via adver tising or fashion. It then moves from the good to the individual consumer through one of four ritual types possession, exchange, grooming or divestment. Because this study is particularly interested in the role advertisements play in tra nsferring and communicating culture, it will focus primarily on the first phase of the meaning transfer process and its relevance for advertisers. McCracken (1986) posited that the meaning transfer process began in the culturally constituted world, which h , where an beliefs and assumptions. In other words, the culturally constituted world was where culture and the individual met to shape his/her unique view of the world. If meaning begins in a world shaped by culture, then understanding culture is key to understanding the meaning transfer process. Defining culture, in itself, is no small feat. In their seminal book on culture, Kroeber and Kluckhohn (1952) compiled 164 definitions for culture (although the number climbs closer to 300 if footnotes are included). These definitions ranged r t, law, morals, c ustoms, and any men and which is capable of transmission from generation to generation or from one country to

PAGE 113

113 what a society d Kluckhohn, 1952, p. 141, italics in original). Yet, despite their interest in and research on culture, Kroeber and Kluckhohn (1952) shied away from proposing their own formal definition for the te rm, citing the difficulty of encapsulating such a vast concept in any succinct statement. For his purposes and the purposes of this study, McCracken (1988) defined culture as . According to McCracken (1986; 1988), culture was both a lens through which individuals saw the world and a blueprint for how they shaped it. As a lens, culture determined how individuals understood and internalized daily life experienced through the five senses (things they saw, heard, smelled, tasted and touched). As a blueprint, culture provided a plan for how individuals would build and interact with the world around them, projecting the influence of human effort. In short, culture molded both an indiv In meaning transfer theory, culture is the source of meaning (McCracken, 1990). In fact, McCracken (1990) insisted the first step to understanding the relationship between culture and consumer behavior was t o treat culture as meaning (p. 5). McCracken went on to cite culture as the source of marketing strategies that infuse goods with meaning and the source of consumer strategies that extract and use meaning for self construction (McCracken, 1990). Danesi (20 08) marketplaces of the past However, individuals do not create meaning in a vac uum. They looked to and depend on those around them fo r help. Friends and organizations alike can develop what zoologist and

PAGE 114

114 author Desmond Morris called costume echo , adopting uncanny similarities in a variety of surfaces from personal dress to company logos (Postrel, 2003). While costume echo explained th e purposeful alignment and expression of similar tastes, values and meanings, aesthetics can also inadvertently align an individual or organization with others. Making aesthetic choices align s an individual or organization with others who also had the same taste. In some cases, particular aesthetic choices may lead to a false identity or unwanted alignment, forcing the individual or organization to adjust its aesthetic signals. Perhaps more important was the recognition that every aestheti c choice has Phase One of Meaning Transfer Process After meaning is established by culture and created in the c ulturally constituted world, it is then transferred to a consumer good phase one of the meaning transfer process (McCracken, 1986; 1988; 1990; 2005). According to McCracken (1986; 1988; 1990) , advertising was one of the primary vehicles for transferring meaning from the culturally constituted world to a consumer good. In fact, with their proliferation and appeal, Danesi (2008) argued that advertisements might be instrumental in shaping consumers thoughts, personalities and lifestyles one need that distinguishes the human species from all others the need for Williamson (1978/2010) agreed with McCracken that advertising had an important function beyond merely selling things that of creating structures of meaning an d attaching that meaning to the products it promotes. Advertising work s to translate information about a good or product into information about a person. More simply, advertisements attempt to translate what individuals buy into who they are . Furthermore, once set in motion, advertising work s in a circular, self

PAGE 115

115 made to represent other, non material things we need; the point of exchange between the two is pre ng making, Hirschman (2007) examined marketplace metaphors related to hair and hair care products, arguing that metaphor, not advertising, was the principal vehicle for meaning creation. Rather, according to Hirschman, successful advertising was the foremo st user of meaning, rather than creator. Despite Hirschman is clear that advertising plays an important role in conveying meaning, whether it is as a meaning creator or a meaning user. is to equate tangible objects with intangible concepts (e.g., feelings, moods, social status, sexuality), thus making the unattainable attainable, convincing consumers their dr eams, desires and goals are reachable through consumption (McCracken, 1986; 1988; 2005; Messarias, 1997; Williamson, 1978/2010). Once adve rtising successfully transfers meaning from the culturally constituted world to the consumer good, it is then up to th e consumer to recognize and embrace the meaning, completing the transfer process (McCracken, 1986; 1990). For, as long as it remains in an advertisement, a product or brand can only represent the intangible. However, upon consumption, a product or brand ca n come to be or create the intangible it once merely stood for (Williamson, 1978/2010). In this way, consumers have act of consumption.

PAGE 116

116 s ultimate meaning. Similarly to the brand research discussed above, much of current communication scholarship emphasized the role of the consumer as sole meaning maker, overlooking the power of the creator (e.g., marketers, advertisers, creative teams) (M essaris, 1997). Although it is possible for viewers to draw a variety of conclusions from the same advertisement or visual, it does not follow that the creator has no role in shaping how the majority of consumers will respond. In fact, limited only by budg et and brand image, the advertising creative team is free to choose from almost an infinite number of symbolic elements, each with their own culturally constituted meanings and symbolic properties (McCracken, 1986, p. 75). This choice not only places the c reative team in direct control of the tools for conveying cultural meanings but also the degree to which the ad effectively transfers these meanings, allowing them to imbue it with a culturally preferred meaning. As Hall (1980) argued, most consumers will engage in a dominant reading of the who do not follow the dominant reading, will most likely participate in a negotiated meaning, generally agreeing with the ad pers onal experiences and interests. measured by his/her ability to affect a particular viewer response. And, a lthough his comments referred specifically to film directors and the cinema, the same argument holds true for creative directors and advertisements. Therefore, despite the sometimes seemingly diverse viewer interpretations of both film and advertising, Mes us to the possibility that, in the hand of a skillful director, the conclusions at which the viewers arrive may be very substantially those that the director has

PAGE 117

117 Thus, the cre ator is largely responsible for generating and imbuing goods with meaning through advertising. More specifically, in practice, advertisers transfer meaning by placing a consumer good and a representation of the culturally constituted meaning (also called a symbolic element ) in the same frame an advertisement (McCracken, 1986; 1988; Williamson, elements (meaning) in the same frame, thus encouraging the consumer to glimp se the similarity between them. When done correctly, the good soon comes to stand for the desired meaning in the unbounded expanses of open countryside in their advertis ements, a Harley Davidson motorcycle Before an advertisement is created, the marketing team must first identify the cultural meanings it wishes to transfer to the product (McCracken, 1986; 1989; 1993). These meanings may be based on market research or other advice and can include areas such as gender, status, age, lifestyle, time, place and personality (Jewett & Sutherland, 2013; McCracken, 1986; 1988; 1990). In other words, it is up to the marketing team to determine who they want the product to be and what they want it to say (McCracken, 1990, p. 6). For example, based on focus groups and competitor research, a marketing team may decide that a particular laundry detergent should represent h igh class, modern females interested in environmental issues. The responsibility now shifts to the creative team who surveys the culturally constituted world and carefully chooses symbolic elements (particular objects, people, contexts and motifs) to commu nicate the intended cultural meanings (McCracken, 1990). According to McCracken (1986), the creative team is free to deliver the meaning through an almost infinite number of symbolic element combinations, constrained only by budgetary limitations and the n eed for a

PAGE 118

118 about place (e.g.. fantasy/naturalistic, indoor/outdoor, urban/rural), time (e.g., of day, of year), people (e.g., number, sex, age, class, status, occ upation, clothing, posture) and more. For example, to represent the laundry detergent described above, the creative team may choose to show a slender, nicely dressed woman wearing pearls while using the detergent in a red, front loading, high efficiency wa shing machine as t he sun sets outside the window. However, because specific symbolic elements often house multiple meanings, the creative team must exercise extreme care, working to simultaneously highlight the desired meanings and suppress the unwanted on es (McCracken, 1986; 1988; 1990). Based on the it is here that a good creative team distinguishes itself from a bad or mediocre one. If the symbolic eleme nts in an advertisement are poorly chosen or combined, consumers can transfer undesired, alternate meanings to the product or fail to make any transfer at all. This renders the advertisement ineffective, or worse, detrimental. And yet, McCracken (1986) not ed that creative always fully cognizant of how and why a selection is made, even when this selection presents Although s ome goods may have conventional meanings or lend themselves more easily to particular meanings, there is no predetermined meaning inherent in any particular good. In can be to clearly communicate it. While one team may choose to transfe r the qualities mentioned above

PAGE 119

119 to its laundry detergent, another team may wish to communicate that its laundry detergent is budget conscious, easy to use and unpretentious. This advertisement would employ much different symbolic elements than the first la undry detergent, perhaps showing a relaxed mom wearing a t shirt and jeans helping her child scoop detergent from an economy sized package into a plain Jane, white washing machine. However, advertising is equally capable of transferring either meaning to w hat is essen tially the same product soap. divesting old meaning, adding to current meaning or claiming new meaning altogether (McCracken, 1986; 1988). A particular mea ning may also be claimed by more than one good (or brand) either simultaneously or subsequently. According to McCracken (1986), this meaning merry go te of cultural meanings. Phase Two of Meaning Transfer Process Once meaning is transferred to a good through a carefully prepared advertisement, it is free to move to its final resting place in the consumer (McCracken, 1986; 1989; 1993). According to McCra cken (1986), the final phase of the meaning transfer process (from good to consumer) is accomplished through one of several types of ritual possession, exchange, grooming and divestment. Possession rituals consu mer good and its meaning as his/her own (McCracken, 1986, p. 79). For example, consumers might spend time cleaning, discussing, comparing, reflecting on, showing off or photographing their belongings in order to signify that a good and its meaning repr esented them (McCracken, 1986, p. 79). Alternately, in exchange rituals , one individual chose, purchases and presents a consumer good to another individual. For example, one friend gives another a birthday gift, which results in the purchaser presenting th e receiver with both a good and a meaning.

PAGE 120

120 Individual consumers participate in grooming rituals time, patience or anx iety ( McCracken, 1986, p. 79). For instance, before going out for the night, a consumer often participates in an elaborate rituals to get ready, coaxing out the meaning contained in a certain style of makeup, hair, clothes, perfume, etc. However, because of the brief nature of both the activity (e.g., going out for an evening) and the goods (e.g., makeup that washes off), the claimed meanings are temporary and the grooming rituals must be repeated divestment ritual , i ndividual consumers, as either the previou s or future owner, attempt to rid a consumer good of its former associations in preparation for the meaning to be claim ed by a new owner . For example, after purchasing a previously owned sports car, a man may clean it meticulously to rid the car of all remnants the old owner and prepare to take ownership of its meaning for himself (e.g., masculinity, virility, excitement). Alternately, the original owner of a good may also perform a divestment ritual, attempting to For example, a woman may clean her old, fur coat before donating it to a charity to distance herself from the good and prepare to let go of its meaning (e.g., sophi stication, a ffluence, luxury). After searching consumer goods for a particular, desired meaning (e.g., gender, class, age, lifestyle, personality), individual consumers who are looking to construct their self image can then take possession of this meaning through one of the rituals discussed above (McCracken, 1986; 1988). It is when c onsumers move p. 8). Therefore, while it i s the job of an advertisement to transfer meaning from the culturally

PAGE 121

121 constituted world to the consumer good by showcasing the similarities between the meaning found in the symbolic elements presented and the good itself, it is the consumer who takes the f meaning from the consumer good to himself/herself and claim it as his/her own. Studies Using Meaning Transfer Theory Thus far in advertising research, McCracke primarily applied to studies exploring celebrity endorsements (e.g., Byrne, Whitehead & Breen, 2003; Campbell & Warren, 2012; Halonen Knight & Hurmerinta, 2010). In fact, McCracken (1989) proposed a meaning transfer mod el specifically to illustrate how the process explains the meaning given to products by celebrity endorsements (see McCracken, 1989, p. 315 ). general process of meaning elements, which resided in a celebrity, could be moved from the celebrity to a consumer good (1986; 1988) original model of meaning transfer, the celebrity could be passed on to the consumer via consumption. For example, Campbell and Warren (2012) applied meaning transfer theory to the study of celebrity endorsements in a three part experimental study with undergraduate students, finding that a brand was more likely to acquire a cele than the positive traits (p. 184). More specifically, Campbell and Warren (2012) found that: 1) endorser

PAGE 122

122 both a congruent brand (i.e., a brand that had a similar perceived meaning/personality as the celebrity) and incongruent brand (i.e., a brand that had a dissimilar perc eived the endorser not sincerely liking the brand while the end with sincere endorsements. The authors proposed that their findings added nuance to meaning (Campbell & Warren, 2012, p. 184 ). Practically, Campbell and Warren (2012) suggested marketers needed to seriously consider both the positive and the negative meanings associated with a potential celebrity endorser. endorsement The case study conducted by Byrne et al. (2003) examined what happened when a large British grocery retailer (Sainsbury Supermarkets) used The Naked Chef (a British celebrity cook named Jamie Oliver) to endorse their brand. After conducting field research with consumers and interviewing key advertising agency personnel, the authors concluded that the process of meaning transfer via celebrity endorsement became more complex for retailers. Because a retail store had physical properties and experiences (e.g., shopping, checking out, looking for products), many aspects of e ndorser, including in store personnel, stock and employee knowledge. For example, in their case study, Byrne et al. (2003) noted that Jamie Oliver endorsed buffalo mozzarella. Therefore, it became vital that all in store employees were aware of the product , where to find it and how it could be used in order to maintain congruency between the brand and the endorser. Additionally,

PAGE 123

123 the product must be constantly stocked and the checkout lines must be short and efficient to align the store with the endorser ompetent, easy going style. Several years later, Halonen Knight and Hurmerinta (2010) also studied the alliance between British grocery retailer Sainsbury Supermarkets and celebrity endorser The Naked Chef (Jamie Oliver) using meaning transfer theory. Unli ke the previous study, Halonen Knight and Hurmerinta (2010) explored the concept of mutual meaning transfer, both from celebrity endorser to brand and from brand to celebrity. Using a case study approach, Halonen Knight and Hurmerinta (2010) analyzed newsp aper articles from a 2004 campaign where Jamie Oliver partnered with Sainsbury Supermarkets to endorse farmed salmon. Halonen Knight and Hurmerinta (2010) found evidence of reverse meaning transfer, noting that negative meanings originating with the brand (Sainsbury Supermarkets promotion of farmed salmon) were assigned to the celebrity endorser (Jamie Oliver). Furthermore, according to the authors, in many cases the newspaper articles directly associated negative meanings with the endorser (Jamie Oliver) a nd only implicitly with the brand (Sainsbury Supermarkets). Based on their findings, Halonen Knight and Hurmerinta (2010) proposed a reciprocal model of meaning transfer for celebrity endorsements, underscoring their nature as mutual brand alliances with b oth parties providing and receiving meaning. Meaning Transfer Theory & the Brand More recently, meaning transfer theory has evolved to include brands as containers for culturally constituted meaning. McCracken (1993) admitted that meaning transfer theory h ad not carefully considered the brand, instead focusing exclusively on goods as carriers of meaning. However, seeking to remedy the oversight, McCracken acknowledged meaning was often housed in the brand, rather than the good (see McCracken, 1993, p. 128 ). Like consumer goods,

PAGE 124

124 brands are capable of soaking up cultural meanings (McCracken, 1993). In fact, McCracken Just as it does for co nsumer goods, advertising imbues en tire brands with culturally constituted meaning, which can then be transferred to consumers (McCracken, 1993). In fact, according to McCracken (1993), brands had meaning only because advertising put it there. Additionally, McCracken (1993) asserted that st rong brands succeed because they contained appealing, useful and powerful meanings, while weak brands failed because their meanings were have value because they add value and they add value by adding aning (McCracken, 1993). From a marketing perspecti ve, the idea of mea ning provides a starting point for assessing particularl y its soft value, which includes non functional aspects of a brand such as brand personality, brand image, brand relationship and emotional appeal (Biel, 1993; McCracken, 1993). The first step is to examine and understand the meanings contained in the brand and the power those meanings ho ld for consumers. However, to understand the full compare meanings contained in a brand, marketers are (McCracken, 1993, p. 139). McCracken (1993) reminded industry professionals, Our obje ctive is to determine the meanings of the brand and then to manage these meanings through better marketing. We want to augment the meanings that give us competitive advantage and jettison those that do not. A determination of these meanings enables us to d esign better advertising and other marketing instruments to capture and maintain the cultural meanings that give us brand loyalty. (p. 139)

PAGE 125

125 Updating meaning transfer theory to include brands as containers for meaning also allowed scholars to begin explorin g the role brands play in housing and transferring meaning. For example, in a qualitative study, McCracken (1993) used a series of open ended questions to examine the cultural meanings contained in beer and beer brands. He found that beer generally contain ed masculine cultural meanings. However, based on the brand of beer, the type of masculinity changed and included such categories as wild college partier, stylish young professional and conservative ladder climber. Therefore, young men likely sought out sp ecific beer brands in order to possess and convey a particu lar masculine cultural meaning. Alternately, Gwinner and Eaton (1999) used meaning transfer theory to examine the transfer of brand personality via sporting events. In an experiment with undergradu ate students ( n = 360), participants examined photos of sporting events that were manipulated to look like sponsored magazine ads for one group or unsponsored images for the other. Those exposed to sponsored events were then divided into sub groups where t he events and brands had no similarity, functional similarity (the brand is actually used during the event), or image based based conditions, the authors found a higher level of congruence between the sponsored events and brands as opposed to the unsponsored events and concluded that a transfer of meaning had taken place. Furthermore, Gwinner and Eaton (1999) found the transfer process was enhanced when the event and brand had matching functional or image based (b rand personality) similarities. In another two part experimental study, Escalas and Bettman (2005) used meaning ) and outgroups (i.e., a grou For the first study, the authors asked

PAGE 126

126 undergraduate students ( n = 243) to identify a group that they felt a part of (ingroup) and a group they did not feel a part of (outgroup). The participants were then asked to list one brand consistent with the image of each group and one brand inconsistent with image of each group. brand connections were measured using a seven item scale. The second study ( n = 75) used a method si milar to the first study, differing only in the scale used (a 24 item independent/interdependent scale), the order of the questionnaire, the inclusion of brand ratings and an additional thought protocol at the end that asked participants to retrospectively reflect on their thoughts during the process. In both studies, Escalas and Bettman (2005) found that consumers reported stronger self brand connections with ingroup associated brands and weaker self brand connections with outgroup associated brands, sugge sting that brands should be aware of the groups consumers connect with their brand. Visuals in Meaning Transfer Theory (1986) suggested that within advertisements, visu als often functioned as the most powerful and meaning the images we see in ads which give them significance, which transfer their significance to the product description of products an (p. 8), Postrel (2003) suggested visual design (e.g., the product itself, packaging, visual identity, advertising). However, thus far, the role of images in meaning transfer theory had been largely ignored and remains woefully untested. The current study will address this research dearth by focu sing specifically on the role images and their visual symbolic elements play in the transfer of brand personality.

PAGE 127

127 Desp ite the lack of research examining visuals through the lens of meaning transfer theory, scholars do agree that images have the potential to create and transfer meaning on a variety of levels, including content, tone, affect and personality ( e.g., Janiszewski, 2008; Jewett & Sutherland, 2013). Therefore, all aspects of a visual must be considered for their potential cultural meanings, from s clothing and body posture (McCracken, 1986). Yet, although studies have made broad assumptions and judgments about the meaning in advertising images, only a few have attempted to analyze the particular visual symbolic elements that lead to their conclusion (e.g., Akay, 2001; Cline & Young, 2004; Goffman, 1979; Jewett & Sutherland, 2013; Jewitt & Oyama, 2004; Pracejus et al., 2003). Although research has clearly shown that sy mbolic visual elem ents contain specific meaning s (e.g., Akay, 2001; Jewett & Sutherland, 2013; Jewitt & Oyama, 2004; Pracejus, 2003; Pracejus et al., 2003), McCracken (1990) and others (e.g., Postrel, 2003; Zuckerman, 1990) asserted that visual symbolic elements possess no inherent meaning, cautioning that visual meaning, like all meaning, is dependent on time, place, experience and associations in short, culture. Zuckerman (1990) stressed that the most basic assumption of meaning transfer was that there was some meaning t o be transferred. Therefore, to succeed, an image must first have meaning to the viewer through some system of real world familiarity. Color, for example, has different connotations and signifies different meanings based on cultural context (Labrecque, Pat context, blue can mean all sorts of things, from IBM to the United Nations. Blue has historical Therefore, advertisers must

PAGE 128

128 remember that the meaning embedded in visuals is dependent on the culture of those interpreting them. Proposed Modifications to Meaning Transfer Theory In its current state, meaning transfer theory provides one explanation for t he movement of meaning from culture to goods/brands through advertising and on to the consumer through research, some scholars have proposed modifications. For exam ple, in an ethnographic study of British high school students, Ritson and Elliot (1999) employed meaning transfer theory to exam ine how advertising wa s used in daily social interactions. Ultimately, the authors suggested ) model based on their interview findings: 1) the potential role of advertising ritual in transferring meaning from advertising to individual consumers, and 2) the cyclical influence of individual consumers on the meaning that resides in the culturally con s tituted world (see Ritson & Elliot, 1999, p. 271 ). First, Ritson and Elliot (1999) suggested that advertising could be used to create meaning outside of its relationship to a particular brand or product. In his original meaning transfer theory, McCracken (1986; 1988) listed possession, exchange, grooming and divestment rituals as the vehicles for transferring meaning from a consumer good to the individual consumer. Ritson and Elliot (1999) proposed adding an additional type of ritual advertising ritual to that list. Unlike the other rituals, however, the authors suggested that advertising ritual has the potential to bypass the consumer good completely, positioning advertising as a link directly from the culturally constituted world to the individual con sumer. The authors observed that, unlike other rituals, when two individuals formed rituals around advertisements, verbal exchanges in the form of advertising interpretations took the place of physical exchanges According to the authors, while meaning tran

PAGE 129

129 Ritson & Elliot, 1999, p. 271 272). Therefore, consumers did not need to interact with or co nsume the advertised good in order for advertising t o transfer and provide meaning. and Elliot (1999) proposed that individuals use advertising metaphor to add mea ning back into the culturally constituted world. Traditionally, a metaphor is the juxtaposition of two normally unalike ideas, with one lending meaning to the other. With the term advertising metaphor , Ritson and Elliot (1999) suggested that advertising co uld lend its meaning to other objects, ideas or situations, providing context, familiarity or understanding. In this way, the creation of meaning was not only top down but also bottom up, allowing individuals to add meaning back into cultur e through advert ising metaphor. In a separate study of American Indian images used to advertise national brands, Merskin from the consumer back to the culture where stereotypes are experienced and recirculate d Merskin, 2001, p. 162 ). However, while Merskin (2001) proposed a recirculation of m eaning back to into culture, she did not use her results to specifically support this theoretical modification. Furthermore, although both Ritson theory based on the ir own scholarly observations, neither new model has been tested. The above sections offer examples of academic investigation into the process of meaning making. However, scholars have only begun to explore human meaning making, leaving much left to learn about how individuals create and derive meaning from the world around them. In

PAGE 130

130 fact, Janiszewski (2008) suggested that the act of meaning creation was arguably more important and deserving of scholarly attention than the more frequently studied act of info rmation analysis (i.e., how people receive and process information, usually studied by examining text based information). By focusing on meaning creation, advertising research could emphasize the contributions of social identity, environmental context and culture in meaning creation (Janiszewski, 2008, p. 283). While Ritson and Elliot (1999) and Merskin (2001) highlighted the role of rituals in culture (a proposed thi rd step in the process), the current study intends to focus on the role advertising plays in transferring meaning from culture to brand (the first step in the process). In particular, the current study will investigate the function of advertising images an d the visual symbolic elements contained within in the transfer process. Although McCracken (1986; 1988) noted the preeminent importance of visuals in meaning transfer, little scholarly work has addressed the role images play in the process. As such, the c urrent study plans to expand specific role advertising images and their visual symbolic elements play in the meaning transfer ring process (see Figure 2 1).

PAGE 131

131 Figure 2 1. Proposed modifications to meaning transfer theory. The new model is based on Research Questions and Hypotheses Therefore, based on the meaning transfer theory, past visual communication research, previous brand personality literature and a brand specific version of the Myers Briggs Type Indicator (MBTI), this study will qualitatively explore the connection between the visual symbolic elements contained in images and multi dimensional MBTI personalities, asking:

PAGE 132

132 RQ1: Which visual symbolic elements in advertising images transfer meaning consistent with each MBTI dimension, including: a) extraverted/introverted (E/I), b) sensing/intuitio n (S/N) and c) thinking/feeling (T/F)? RQ2: How do visual symbolic elements that represent each MBTI dimension combine in advertising images to portray multidimensional MBTI personalities, including: a) extraverted/sensing/thinking (EST), b) extraverted/s ensing/feeling (ESF), c) extraverted/intuitive/thinking (ENT), d) extraverted/intuitive/feeling (ENF), e) introverted/sensing/thinking (IST), f) introverted/sensing/feeling (ISF), g) introverted/intuitive/thinking (INT) and h) introverted/intuitive/feeling (INF)? RQ3: How do consumers read and negotiate per sonality information in images? RQ4: What image did group image sort participants select as most representative for each multidimensional MBTI personality, including: a) extraverted/sensing/thinking (ES T), b) extraverted/sensing/feeling (ESF), c) extraverted/intuitive/thinking (ENT), d) extraverted/intuitive/feeling (ENF), e) introverted/sensing/thinking (IST), f) introverted/sensing/feeling (ISF), g) introverted/intuitive/thinking (INT) and h) introvert ed/intuitive/feeling (INF)? Additionally, meaning transfer theory names advertising as one of the primary modes of transferring meaning from culture to a product or brand (McCracken, 1986; 1988; 1993). Furthermore, past research showed that images were pa rticularly effective containers for meaning within an advertisement (McCracken, 1986; Williamson, 1978/2010), and that the number of visuals in advertising was on the rise (Janiszewski, 2008; Marchand, 1985; McQuarrie, 2008). Past scholars also recognized (Cooper, 2009; Danesi, 2008; Janiszewski, 2008, McQuarrie, 2008; Messaris, 1997; 1998; Morgan, 2005) oftentimes even more effectively than text (Childers & Houston, 1984;

PAGE 133

133 Janiszewski, 2008; McQuarrie & Phillips , 2005; Rayner, et al., 2008; Williamson, 1978/2010). Finally, it was shown that specific visual symbolic elements within images could contain particular meanings (e.g., Cline & Young, 2004; Goffman, 1979; Jewitt & Oyama, 2004; Lindner, 2004; Pracejus, 200 3; Schroeder & Zwick, 2004). Other studies have focused on brand personality as one specific meanings that images could transfer from culture to a brand, and scholar s agreed that advertising plays a major role in developing brand personality (e.g., Delbaer e et al., 2011; Doyle, 1989; Meenaghan, 1995 ; Yoo et al, 2009) . Research has also show n that brand personality could increase levels of brand trust, affect and loyalty (Sung & Kim, 2010) and help to create brand differentiation (Biel, 1993; D. Aaker, 1996; Park et al., 1986; Ramaseshan & Tsao, 2007). But, perhaps more importantly, modern consumers have demonstrated their reliance on brands and their meaning, such as personality, to define and show the world who they are (e.g., D. Aaker, 1996; J. Aaker, 1999 ; Huang et al., 2012; Malaret al., 2011; McCraken, 1989; 2005; Pl ummer, 1984/2000; Sirgy, 1982). One way to operationalize and measure personality is with the MBTI (Hirsh & Kummerow, 1989; Myers, 1995; The Myers & Briggs Foundation, n.d.). With Strausbaugh (1998) brand specific MBTI scale, the longstanding history, tradition, reliability and validity of the MBTI is available for brand researcher as they study brand personality (e.g., Carlyn, 1977; Jung, 1923/1971; McCarley & Carskadon, 1986; Myers, 1995; The Myers & Briggs Foundation, n.d.; Tieger & Barron Tieger, 1995; 2000). Additionally, in prior studies, Akay (2001), Cunningham et al. (2007) and Jewett and Sutherland (2013) suggested specific symbolic visual elements that transferred specific MBTI bran d pe rsonality dimensions and types.

PAGE 134

134 contribution to the creation of brand personality through advertising, and the growing number of visuals in advertising, it is vital for scholars to understand the role advertising visuals play in the ability to transfer cultural meanings to larger populations through symbolic visual elements that represent specific, multi dimensional brand personalities, hypothesizing: H1: Within each tested advertisement, an advertisement using an image with visual symbolic elements specific to a particular MBTI personality will be assigned that particular brand personality more often than any other brand personality, including: a) An advertisement using an image with extraverted/sensing/thinking (EST) visual symbolic elements will result in a higher percentage of extraverted/sensing/thinking (EST) brand personal ity assignments when compared to all other personalities. b) An advertisement using an image with extraverted/sensing/feeling (ESF) visual symbolic elements will result in a higher percentage of extraverted/sensing/feeling (ESF) brand personality assignmen ts when compared to all other personalities. c) An advertisement using an image with extraverted/intuitive/thinking (ENT) visual symbolic elements will result in a higher percentage of extraverted/intuitive/thinking (ENT) brand personality assignments when compared to all other personalities. d) An advertisement using an image with extraverted/intuitive/feeling (ENF) visual symbolic elements will result in a higher percentage of extraverted/intuitive/feeling (ENF) brand personality assignments when compared to all other personalities. e) An advertisement using an image with introverted/sensing/thinking (IST) visual symbolic elements will result in a higher percentage of introverted/sensing/thinking (IST) brand personality assignments when compared to all oth er personalities. f) An advertisement using an image with introverted/sensing/feeling (ISF) visual symbolic elements will result in a higher percentage of introverted/sensing/feeling (ISF) brand personality assignments when compared to all other personalit ies. g) An advertisement using an image with introverted/intuitive/thinking (INT) visual symbolic elements will result in a higher percentage of introverted/intuitive/thinking (INT) brand personality assignments when compared to all other personalities. h) An advertisement using an image with introverted/intuitive/feeling (INF) visual symbolic elements will result in a higher percentage of introverted/intuitive/feeling (INF) brand personality assignments when compared to all other personalities.

PAGE 135

135 H2: Across all tested advertisements, an advertisement using an image with visual symbolic elements specific to a particular MBTI personality will be rated highest for that particular brand personality, including: a) An advertisement using an image with extraverted /sensing/thinking (EST) visual symbolic elements will result in a higher percentage of extraverted/sensing/thinking (EST) brand personality assignments compared to advertisements that do not have images with extraverted/sensing/thinking (EST) visual symbol ic elements. b) An advertisement using an image with extraverted/sensing/feeling (ESF) visual symbolic elements will result in a higher percentage of extraverted/sensing/feeling (ESF) brand personality assignments compared to advertisements that do not hav e images with extraverted/sensing/feeling (ESF) visual symbolic elements. c) An advertisement using an image with extraverted/intuitive/thinking (ENT) visual symbolic elements will result in a higher percentage of extraverted/intuitive/thinking (ENT) brand personality assignments compared to advertisements that do not have images with extraverted/intuitive/thinking (ENT) visual symbolic elements. d) An advertisement using an image with extraverted/intuitive/feeling (ENF) visual symbolic elements will result in a higher percentage of extraverted/intuitive/feeling (ENF) brand personality assignments compared to advertisements that do not have images with extraverted/intuitive/feeling (ENF) visual symbolic elements. e) An advertisement using an image with intro verted/sensing/thinking (IST) visual symbolic elements will result in a higher percentage of introverted/sensing/thinking (IST) brand personality assignments compared to advertisements that do not have images with introverted/sensing/thinking (IST) visual symbolic elements. f) An advertisement using an image with introverted/sensing/feeling (ISF) visual symbolic elements will result in a higher percentage of introverted/sensing/feeling (ISF) brand personality assignments compared to advertisements that do n ot have images with introverted/sensing/feeling (ISF) visual symbolic elements. g) An advertisement using an image with introverted/intuitive/thinking (INT) visual symbolic elements will result in a higher percentage of introverted/intuitive/thinking (INT) brand personality assignments compared to advertisements that do not have images with introverted/intuitive/thinking (INT) visual symbolic elements. h) An advertisement using an image with introverted/intuitive/feeling (INF) visual symbolic elements will result in a higher percentage of introverted/intuitive/feeling (INF) brand personality assignments compared to advertisements that do not have images with introverted/intuitive/feeling (INF) visual symbolic elements.

PAGE 136

136 Finally, in addition to examining how the visual symbolic elements in advertising images transfer brand personality, the current study will also investigate the influence of brand personality on attitude toward the ad, attitude toward the brand and purchase intent. Therefore the current study asks: attitude toward the ad? attitude toward the brand? oice of MBTI brand personality influence his/her purchase intent?

PAGE 137

137 CHAPTER 3 METHOD The current study combined past work exploring the connection between specific symbolic visual elements and particular MBTI brand personality dimensions (Akay, 2001; Jewet t & Sutherland, 2013) with previous brand personality and visual communication research to test Specifically, the current study used a mixed method composed of two initial phases a scale reduction and qualitative group image sorts that culminated with a final quantitative survey. The preliminary scale reduction and group image sorts provided the content to be tested in the final survey. Mixed Method Approach Scholars have noted the importance of using mixed methods in both advertising ( Morrison, Haley, Sheehan & Taylor, 2002) and brand personality (Arora & Stoner, 2009) research. Therefore, the current study employed a mixed method design composed of both qual itative and quantitative research. During the first phase, (1998) brand specific MBTI scale was reduced using a series of statistical methods outlined later in this chapter. The second phase included qualitative group image sorts of potential advertising images that represent specific brand personalities. Finally, a quantitative survey, which uses the reduced brand specific MBTI scale and findings from the group image sorts, was co nducted during the third phase. Scholars suggest several reason s for using a mixed methods research design, which include: a better understanding of the topic, increased validity, decreased methodological limitations, and more complete findings. One of the most compelling reasons is to gain a better understanding of t he subject of study (Greene, 2007). Mixed method scholars propose that

PAGE 138

138 approaching a subject from multiple perspectives inherently leads to a more thorough understanding of the topic (Babbie, 2010; Johnson & Christensen, 2004). Because mixed methods resear ch is not confined to a single method or approach, it can often answer a broader, more inclusive range of research questions (Johnson & Christensen, 2004). The second reason for triangulating research through a mixed method approach is that it allows the r esearcher to check the results of each method, ensuring that all techniques lead to similar conclusions (Maxwell, 2013). Mixed methods studies can also present stronger evidence for a conclusion when findings converge, providing corroboration (Johnson & Ch ristensen, 2004). As such, many scholars have also touted triangulating research through a mixed method approach as a way to increase validity (e.g., Balnaves & Caputi, 2001; Denzin, 1978; Greene, 2007; Maxwell, 2013). According to Denzin (1978), triangula tio no single method ever adequately solves the problem of rival causal factors [i.e., internal validity issues]...[b]ecause each method reveals different aspects of empirical reality, multiple methods of observat ions must be Additionally, mixed methods can help to overcome the limitations and biases of a single method (Babbie, 2010; Greene, 2007; Maxwell, 2013). By layering multiple study methods, researchers can reap the benefits of both qualitative and qu antitative inquiry (Johnson & Christensen, 2004). Further, Johnson and Onwuebuzie (2004) suggested that mixed methods 15) of both qualitative and quantitative approaches in a singl e study. Johnson and Christensen (2004) referred to this concept as the principle of complementarity, suggesting that mixed methods researchers strategically use the strengths of one method to overcome the weaknesses of another.

PAGE 139

139 However, perhaps most impor tantly, mixing methods can lead to broader, deeper, more inclusive and more complex findings that better represent the intricacies of humans and the world they inhabit (Greene, 2007). Unlike solely qualitative or quantitative research, mixed methods resear ch can add insight and understanding that might otherwise be missed (Johnson & Christensen, 2004). Specifically, mixed methods research allows words, picture and narratives to add meaning to numbers while also allowing numbers to add precision to words, pi ctures and narratives (Johnson & Christensen, 2004, p. 414). Thus, when used together, a mixture of qualitative and quantitative research can produce a more comprehensive knowledge and understanding of a particular topic, which is necessary to inform both theory and practice (Johnson & Christensen, 2004). Dependent Variables The current study evaluated four dependent variables brand personality, attitude toward the ad, attitude toward the brand and purchase intent. As discussed in the previous chapter, pa st literature showed that images had the ability to manipulate brand personality (e.g., Ang & Lim, 2006; Batra, Lehmann & Singh, 1993; Boudreaux & Palmer; 2007; Jewett & Sutherland, 2013; Mitchell & Olson, 1981), which, in turn, could affect attitude towar d the ad, attitude toward the brand and purchase intent (e.g., Ang & Lim, 2006; Gkiouzepas & Hogg, 2011; McQuarrie & Mick, 1999; 2003). Therefore, the current study examined the ability of advertising images to transfer specific MBTI brand personalities to consumers and investigate the influence of those personalities on consumer attitudes and behaviors. Brand Personality set of human characteristics associated with a b rand (J. Aaker, 1997, p. 347). Specifically, the

PAGE 140

140 brand specific MBTI scale. Because the scale was reduced as part of the current study, both the original scale and t he reduction process are discussed in detail below. The scale reduction results, including the shortened scale in its entirety, are provided in detail in the next chapter. The up image sorts and survey. Appropriateness of Single Item Measures Although the first dependent variable, brand personality, was measured with a multiple item scale, the remaining three dependent variables used single item measures. According to Bergkvist item measures are necessary for eliciting attributes 1 , such as personality. However, Bergkvist and Rossiter (2007, 2009a; 2009b) found that single item measures were more appropriate for measuring doubly concrete constructs 2 , such as attitude toward the ad, attitude toward the brand and purchase intent. For doubly concrete constructs, Bergkvist and Rossiter (2007; 2009b) demonstrated that single item measures that used one good item: 1) had equal predictive validity as multi ple item measures, 2) captured the same amount of information as multiple item measures, 3) did not suffer from common method bias 3 , and 4) increased validity by eliminating off attributes 4 . Therefore, as suggested by Bergkvist and 1 attribute that is an internal t rait or state that has outward manifestations, which are mental or physical activities, . 2 A doubly concrete construct consists of a concrete singular object (e.g., the advertisement, the branded product) and a concrete attribute (e .g., attitude, purchase intent) (Bergkvist & Rossiter, 2007; 2009a; 2009b). 3 ommon methods b ias occurs when the correlation between two or more constructs is inflated because they were measured in the same way 4 Off attributes are items in a multiple item measure that do not truly measure the intended construct (e.g., using useful to measure attitude toward the brand) (Bergkvist & Rossiter, 2009b). According to Bergkvist and Rossiter with ratings of the focal attribute, as they are with multiple item measures, the resulting scor

PAGE 141

141 Rossiter (2007; 2009a; 2 009b), the current study used single item measures (discussed in detail below) to measure attitude toward the ad, attitude toward the brand and purchase intent. Attitude Toward the Ad , A Ad , which disposition to respond in a favorable or unfavorable manner to a p. 49). According to scholars, affect, or likability, i attribute (Bergkvist & Rossiter, 2009b, p. 9; MacKenzie & Lutz, 1989). It is also important to meant to gauge consu mer attitudes toward advertising in general (MacKenzie & Lutz, 1989, p. 49). In the past, academics usually relied on multiple item scales (e.g., bad/good, unpleasant/pleasant and unfavorable/favorable) to measure attitude toward the ad (e.g., MacKenzie & Lutz, 1989). However, recent studies have shown that a single item measure was more appropriate for measuring attitude toward the ad (Bergkvist & Rossiter, 2007; 2009a; 2009b). Specifically, Bergkvist and Rossiter (2009b) found that the single item like to dislike was the best measure of attitude toward the ad. Although the use of a single item measure is relatively new (Bergkvist & Rossiter, 2007; 2009a; 2009b), many academics have embraced a single item measure of attitude toward the ad (e.g., Bakalash & Riemer; 2013; Berkvist & Rossiter, 2008; Bergkvist, Eiderbäck & Palombo, 2012; Pomering, Johnson & Noble, 2013). Therefore, participants indicated their attitude toward the ad using a single item measure like to dislike (Bergkvist & Rossiter, 2007; 2009a ; 2009b). Participants were instructed to

PAGE 142

142 seven statements: I dislik e it extremely, I dislike it quite, I dislike it slightly, I neither liked or disliked it, I liked it slightly, I liked it quite, and I liked it extremely (Berkvist & Rossiter, 2009a). Items were coded from 1 to 7, where 7 was a positive response (Berkvist & Rossiter, 2009a). Attitude Toward the Brand The third dependent variable measured was attitude toward the brand, A Brand . Fishbein attitude toward a brand is caused by beliefs about the brand (Fishbein & Ajzen, 1975; Mitchell & Olson, 1981). Unlike attitude toward the ad, which measures affect , both the definit ion and measure of attitude toward the brand focus on brand evaluation (Bergkvist & Rossiter, 2009b). However, similar to attitude toward the ad, attitude toward the brand has traditionally been measured using multiple item measures (e.g., bad/good, unplea sant/pleasant and dislike/like) (e.g., Gardner, 1985). Yet, like attitude toward the ad, Bergkvist and Rossiter (2007; 2009a; 2009) found it was more appropriate to measure attitude toward the brand with a single item measure. In particular, Bergkvist and Rossiter (2009b) demonstrated that the single item good to bad was the best measure of attitude toward the brand. Additionally, since the employed a single item measur e of attitude toward the brand (e.g., Berkvist & Rossiter, 2008; Bergkvist et al., 2012; Fennis, Das & Fransen, 2012; Hartmann & Apaolaz Ibáñez, 2012; Kapoor & Heslop, 2009; Pomering et al., 2013; Sabri & Obermiller, 2012; Suh, 2009).

PAGE 143

143 Therefore, participan ts indicated their attitude toward the brand using a single item measure good to bad (Bergkvist & Rossiter, 2007; 2009a; 2009b). Participants were instructed statemen seven statements: I think it is extremely bad, I think it is quite bad, I think it is slightly bad, I think it is neither good nor bad, I think it is slightly good, I thin k it is quite good, and I think it is extremely good (Berkvist & Rossiter, 2009a). Items were be coded from 1 to 7, where 7 was a positive response (Berkvist & Rossiter, 2009a). Purchase Intent The final dependent variable for the current study was purchas e intent, PI. According to Lutz, MacKenzie and Belch (1983), purchase intent refers to the likelihood that a consumer will pu rchase the brand in the future. item measures also extended to pur chase intent. Again, Bergkvist and Rossiter (2007; 2009a; 2009b) found that a single item scale was a more effective measure of purchase intent than multiple item measures (e.g., unlikely/likely and uncertain/certain). Specifically, Bergkvist and Rossiter (2009b) found that the single item unlikely to likely was the best measure of purchase intent. Like both attitude toward the ad and attitude toward the brand, many other scholars have also used single item measures to evaluate purchase intent (e.g., Bakala sh & Riemer; 2013; Berkvist & Rossiter, 2008; Chrysochou & Grunert, 2013; Hartmann & Apaolaz Ibáñez, 2012; Sabri & Obermiller, 2012; Swaminathan et al., 2009). Therefore, participants indicated their purchase intent on a seven point range using a single it em measure unlikely to likely (Bergkvist & Rossiter, 2007; 2009a; 2009b). Participants

PAGE 144

144 ments: No chance or almost no chance, Slight possibility, Some possibility, Fairly good possibility, Probable, Very probable and Certain or practically certain (Berkvist & Rossiter, 2009a). Items were coded from 1 to 7, where 7 was a positive response (Ber kvist & Rossiter, 2009a). Confounding Variables In addition to the four dependent variables being measured in the current study brand personality, attitude toward the ad, attitude toward the brand and purchase intent it was possible that other factors might have influenced the role of advertising images in the manipulation of brand personality. In particular, human personality and gender had the potential to confound the brand personality findings. Therefore, both of these additional confounding variabl es were measured in the current study. H uman p ersonality Past research (e.g., Govers & Schoormans, 2005; Hamilton & Sun, 2005; Huang et al., 2012; Huber & Seelmann, 2012; Nienstedt et al., 2012; Plummer, 2000) found that an cou ld personality. To control for this variable, both group image sort participants and survey respondents indicated their own human personality using the same personality measure that was used to measure brand personality specific MBTI scale. Using the same scale to measure both human and brand personality allowed for ability to results.

PAGE 145

145 Gender Gender also had the potential to be a confounding variable. Men and women are different in a variety of ways that may influence how th ey view and interpret the symbolic visual elements in advertising images. Additionally, differences in the gender distribution of some MBTI personality types (e.g., there are more female feelers and male thinkers) also have the potential to confound study Barron Tieger, 2000). Therefore, gender was measured by asking participants to indicate their gender and including gender as a variable. Scale Reduction (1998) brand specific MBTI scale. Once shortened, the new scale was used to measure one of potential confoundin g variabl es, human personality. specific MBTI scale contained 24 items six item pairs for each of the four personality dimensions. made it cumbersome for measuring the personality of multiple brands or advertisements in one quantitative questionnaire. Multiple applications of such a lengthy instrument could contribute to participant fatigue, which could negatively affect internal validity (Campbell & Stanley, 1963). Therefore, this st udy attempted to decrease the number of dichotomous adjective pairs while maintaining the integrity of the scale. Using pre test data gathered in a previous survey (discussed in detail below), the researcher used statistical data reduction techniques Cro correlations to reduce the number of scale components in the brand specific MBTI scale

PAGE 146

146 developed by Strausbaugh (1998). This shortened version of the brand specific MBTI scale is referred to as the BHPI (Brand an d Human P ersonality Indicator). Specific MBTI Scale specific MBTI scale is based on the Myers Briggs Type Indicator, which determines human personalities based on four dimensions: extraversion/introversion (E/I ), sensing/intuition (S/N), thinking/feeling (T/F) and specific MBTI scale contains 24 dichotomous adjective pairs six for each of the four dimensions (see Table 3 1). Each dichotomous adjective pair i s composed of two opposite adjectives, and, when applied, participants are asked to choose the adjective that best describes the brand in question. For example, when presented with the adjective pair assertive/mild , the participant might choose mild as the best description of the brand. Table 3 1. Dichotomous adjective pairs from the original brand specific MBTI scale. Extraversion/ Introversion (E/I) Sensing/ Intuition (S/N) Thinking/ Feeling (T/F) Judging/ Perceiving (J/P) 1 Assertive/ Mild Systematic/ Imaginative Logical/ Emotional Clear Cut, Definite/ Undecided, Variable 2 Active/ Reserved Realistic, Down to Earth/ Idealistic, Visionary Firm/ Soft Hearted Deliberate/ Adaptable 3 Dynamic/ Moderate Practical, Functional/ Creative, Theoretical Cold/ Wa rm Dependable/ Changeable 4 Sociable/ Shy Sensible, Factual/ Instinctual Indifferent/ Sympathetic Steadfast/ Innovative 5 Outspoken/ Quiet Narrow Focus/ Wide Interests Rational, Reasonable/ Passionate, Perceptive Decided/ Flexible 6 Energetic/ Calm Cons ervative/ Unconventional Stoic/ Excitable Well defined/ Indistinct Specific MBTI Scale Strausbaugh (1998) created her initial version of the brand specific MBTI by choosing six adjectives from the Adjective Checkl ist (ACL) to represent each pole of the four MBTI

PAGE 147

147 dimensions (a total of 48 adjectives, or 24 pairs). The ACL was originally developed in 1949 at the Institute of Personality Assessment and Research to allow researchers to gather data on personalities and has grown to include 300 personality specific adjectives (Strausbaugh, 1998). The ACL also integrated the theoretical personality perspectives of leading personality psychologists, including Jung, making it an ideal starting point for personality scale ite ms. In an earlier study, Thorne and Gough (1991) had previously compared the ACL to the MBTI and starting point for creating her brand specific MBTI scale by choosing their highest correlating adjectives for each MBTI dimensi on to create her initial scale. To pretest the scale, Strausbaugh (1998) asked students ( n = 80) to evaluate three cola brands Coca Cola, Diet Coke and Cherry Coke using the initial scale. The pretest results were then examined using a combination of R (Strausbaugh, 1998). To refine the scale, Strausbaugh (1998) noted the pairs with the high est reliability and selected additional synonymous adjectives from the ACL. These additional adjective pairs were used to replace 10 inconsistent or unreliable pairs in the pretest scale (e.g., confident/cautious became dynamic/moderate). Seven other pairs were also modified by replacing one half of a pair with a more appropriate opposite (e.g., energetic /gentle became energetic/calm). Completing this process provided Strausbaugh (1998) with the current 24 item brand specific MBTI scale. To test the brand s pecific MBTI scale, Strausbaugh (1998) conducted in class surveys with undergraduate students ( n = 244) at a large southeastern university. The scale was evaluated using a total of 10 brands across five product categories (two brands per category).

PAGE 148

148 In orde r to reduce survey fatigue, respondents were split into two groups, each evaluating a set of five brands. After gathering data, Strausbaugh (1998) tested the internal consistency of each of half me thod. While creating the scale, Strausbaugh (1998) reported relatively high internal consistency across all the brands tested in her study, with all dimension averages exceeding .66 (see Table 3 2). Although many scholars re commend alpha values of at least .7 (DeVellis, 2003; Nunnally, 1978; Pallant, 2010), other scholars feel only a minimum of .5 or .6 is necessary to meet standards of internal reliability (Caplan, Naidu & Tripath, 1984; Nunnally, 1967; Strausbaugh, 1998) . A dditionally, the small number of items per dimension (six) may contribute to slightly lower alpha values (Pallant, 2010). Therefore, all four dimensions were deemed reliable. n (E/I) dimension was the most tightly constructed, followed by sensing/intuition (S/N), thinking/feeling (T/F) and, finally, judging/perceiving (J/P). Strausbaugh (1998) also noted that the variance could be attributed to nuanced differences between the c ollection of similar, but not synonymous, adjectives. Table 3 specific MBTI scale. Dimension Alpha Range Average Scale Alpha Extraversion/Introversion (E/I) . 7371 .8618 . 8124 Sensing/Intuition (S/N) . 6 128 .7773 . 7085 Thinking/Feeling (T/F) .5309 .8019 . 6980 Judging/Perceiving (J/P) . 5399 .7724 . 6674 The validity of the original brand specific MBTI scale was confirmed using a series of correlations, t tests and cross tabulations (Strausbaugh, 1998) research suggested the brand

PAGE 149

149 192) of brand personality. Further, the study argued that the scale was soundly constructed and met all the necessary reliability an d validity conditions (Strausbaugh, 1998). Additionally, review (e.g., Sutherland et al., 2004; Jewett & Sutherland, 2013). Finally, according to Strausbaugh (1998), the brand specific MBTI scale was a nd (p. 193). Reliability 0, p. 6), and has three divisions: stability, equivalency and internal consistency (Wimmer & Dominick, 2011). simplest and most intuitive approach to stability testing is the test retest method, which requires measuring a group of people twice with the same measure and correlating the results. However, because the same group of people must complete the same measure twice, this method of testing reliability has significant shortcomings, including participant fatigue (caused by taking the measure multiple times), participant access (gaining access to the same participants at two d ifferent times), carry over effects (participants remembering their answers from the first test) and genuine change in behavior or attitude if time passes between the two tests (Pallant, 2010; Pedhazur & Schmelkin, 1991). Because of the inherent problems w ith test retest reliability and because this study was more interested in the reliability within a scale, stability testing was unsuitable. between two parallel forms of a

PAGE 150

150 implies, equivalency testing requires developing and administering two different measurement techniques to measure the same concept, then comparing the results. However, Wimmer and Dominick (2011) reduction phase was to fine tune a single scale, equivalency testing was not appropriate. Finally, the third component of reliability is internal consistency, also known as inter item consistency (Wimmer & Dominick, 2011). According to Wimmer and Dominick (2011), items that retest method, internal consistency allows researchers to me asure reliability with a single administration of the measure (Pedhazur & Schmelkin, 1991; Wimmer & Dominick, 2011). Additionally, internal consistency testing only requires one scale, unlike equivalency testing. Moreover, the scale reduction phase of the current study aimed to shorten an existing scale while ensuring that the items continue to measure the same concept. Therefore, the current study used 998) brand sp ecific MBTI scale. There are two common methods for measuring internal consistency. One method for measuring internal consistency is the split half method. Although the split half method is correlated like other internal consistency measures, it requires s plitting the scale in half, correlating one half with the other. For example, a 10 item scale would be split into two groups of five and compared. Therefore, the test is measuring the reliability of two five item scales rather than one 10 item scale. Furth er, scale items may be split in many ways. With most

PAGE 151

151 researchers using ad hoc split methods (e.g., odd even, first second), there is no guarantee the two split groups are parallel, causing additional issues. As such, some scholars decry the split half meth od as a severely limited method for measuring internal consistency and suggest avoiding its use (Pedhazur & Schmelkin, 1991). Moreover, the split half method is best for rten the brand specific MBTI scale. Therefore, although Strausbaugh (1998) used the split half method during the development of the original scale, the current study did not apply the split half method. Another method for measuring internal consistency is alpha is the most common method for measuring the internal reliability of a scale (Pallant, 2010; Pedhazur & Schmelkin, 1991; Wimmer & Dominick, 2011) and 2) this method was used by values ranging from 0 to 1, a higher value indicates a higher reliability (Pallant, 2010). Typically, Pallant, 2010). However, other scholars feel only a minimum of .5 or .6 is necessary to m eet standards o f internal reliability (Caplan et al. , 1984; Nunnally, 1967; Strausbaugh, 1998). Alternately, Pedhazur and Schmelkin (1991) suggested that there is no authoritative, universal t amount of error he or she is willing number of scale items, and scales with a small number of it ems (less than 10) can produce

PAGE 152

152 smaller values (e.g., .5). Therefore, Pallant (2010) suggests that for small scales it may be better to base internal reliability on mean inter item correlations. When using correlations for this purpose, scholars recommend o ptimal inter item correlation values between .2 and .4 (Briggs & specific MBTI scale has 24 items, the scale is really a combination of four smaller sub scales one for each of the four MBTI dimensions. Each of the dimension sub scales is composed of six items, making them scale, potenti ally causing specific MBTI. As the scale was reduced, the cu alpha discussed below to guide the in ternal consistency of the BHPI. consistency. Althoug h the average alpha value for each dimension was marginally acceptable, the low end of the alpha ranges for three of the four MBTI dimensions dipped significantly below the commonly suggested level of .7 (see Table 3 2). The low end of the thinking/feeling dimension and judging/perceiving J/P dimension were both below .6 (.5309 and .5399, respectively), and the sensing/intuition (S/N) dimension was below .7 (.6128). And, although the average alpha values were higher, neither the thinking/feeling dimension n or the judging/perceiving J/P dimension reached .7 (.6980 and .6674, respectively). Thus, addressing these low internal consistency values provided further value for the current reduction phase.

PAGE 153

153 Also, since each of the four sub scales has sma ll number of items, point biserial biserial correlation (r pb ) is a special case of the Pearson product moment correlation (r) (Calkins, 2005). The calculation, results and interpretation for a point biserial correlation are similar to a Pearson correlation (Brown, 2001). However, unlike a Pearson correlation, which requires two continuous variables, a point biserial correlation allows the researcher to use one dichotomous and one continuous vari able (Brown, 2001; Calkins, 2005; Lowry, 2013). Therefore, the combination of dichotomous (the poles of an MBTI personality dimension) and continuous (number of participants) data being compared in the current study made point biserial correlations appropr iate (Brown, 2001; Bruning & Kintz, 1968; Calkins, 2005; Lowry, 2013). Pretest Data for Scale Reduction The scale reduction was conducted using previously gathered pretest data collected 24 item brand specific MBTI scale to examine the effects of advertising visuals on the brand personality of a fictitious brand of cola, Fresh Cola . Unlike the current study, the pretest manipulated only two of the extra version/introversion (E/I) and thinking/feeling (T/F) in the visuals. However, to keep from skewing results, the full, 24 item scale was used and participant responses for all four personality dimensions were recorded. To examine the effects of advertisi ng visuals on brand personality, the researcher created four brand specific advertisements for Fresh Cola one extraverted ad, one introverted ad, one thinking ad and one feeling ad. Each advertisement was portrait, full color and visually dominant, with a large, full and advertising background allowed for the creation of profession looking advertisements n to

PAGE 154

154 represent each of the four tested personalit ies and purchased from Depositp hotos ® (de positphotos.com) or iStock ® (istockphoto.com). Purchasing stock photos allowed the researcher to use professional quality images for minimal cost while providing the appropriate copyright licensing to alter and republish the advertisements. The Extraverted (E) image showed a tightly cropped, intimate group of people smiling and dancing in a nightclub environment. Two of the three people were looking at the viewer, and the image used bright, saturated colors. Alternately, the Introverted (I) image showed a woman walking alone and barefoot on the beach at sunset. The lighting was low and romantic with the ocean horizon ahead of her and only a single line of footprints beh ind her. The Thinking (T) image showed a bottle of Fresh Cola on a white background. Black lines led from specific features of the product to brief explanatory text to the right of the bottle (e.g., lid Conversely, the Feeling (F) image showed a young woman cuddling her dog. Both the woman and the dog were relaxed with their bodies and faces touching. For consistency, the Fresh Cola logo and top third of the product bottle were shown in the bottom right corner of all four advertisements. Survey participants ( n = 304) were undergraduate college students recruited from four intro level mass communication courses at a large southeastern university. Undergraduates were offered extra credit for their participa tion. To conduct the survey, an electronic questionnaire was created using Qualtrics ® (www.qualtrics.com), online survey software, and distributed by the course professors. After receiving consent, the questionnaire first asked students to indicate their o wn personality. Then, to ensure students did not confuse the fictitious brand with a real brand, participants were asked their familiarity with the brand Fresh Cola on a five point Likert scale item (not at all familiar to extremely familiar) and whether t hey had ever tasted the product (yes or no). Next, participants

PAGE 155

155 were shown advertisements for Fresh Cola on the images they saw. Additionally, participants were asked to indicate their brand attitude base d on each advertisement using a three item semantic differential scale bad/good, pleasant/unpleasant and unfavorable/favorable with a seven point range (MacKenzie & Lutz, 1989). Purchase intent was also evaluated on a five point Likert scale item (not at all likely to extremely likely). Finally, participants were asked to provide some basic demographic information about themselves. During the survey, participants saw two of the four possible ads either the Extraverted (E) ad or the Introverted (I) ad and either the Thinking (T) ad or the Feeling (F) ad. Limiting participants to only two advertisements one from each personality dimension being tested decreased the potential for participant fatigue. This strategy also ensured that participants provid ed personality feedback for only one adjective from each personality dimension. Because the MBTI personality dimensions are bi polar pairs, survey logic was employed to prevent a single participant from rating both opposing poles. udging/Perceiving (J/P) Dimension While the judging/perceiving (J/P) dimension was not manipulated visually in the current study (see below for rationale), it was included in the scale reduction process and the BHPI. Thus, while the current study measured all four personality dimensions, it was only concerned with the ability to transfer three dimensional personalities through visuals. However, retaining the judging/perceiving (J/P) dimension in the new, shortened scale allows future research to be conducte d on all four dimensions using the BHPI, including research on the role of the judging/perceiving (J/P) dimension in brand personality.

PAGE 156

156 MBTI Dimensions in Study Although all four MBTI dimensions were included in the scale reduction and were measured each time the scale was used, the current study visually manipulated only three of the four MBTI dimensions extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). The fourth dimension judging/perceiving (J/P) was excluded bas ed on five reasons. First, this decision was made, in part, because the judging/perceiving (J/P) dimension is not part of the original Jungian personality theory. Rather, Myers and Briggs added the judging/perceiving (J/P) dimension during the creation of the MBTI model during the 1940s 50s (Myers, 1995), and it was based o Second, the judging/perceiving (J/P) dimension often functions in tandem with other dimensions. According to Hirsh and Kummerow (1989), in the MBTI framewo rk, judging is linked to the thinking/feeling (T/F) dimension because both thinking and feeling are both ways to make a decision or reach a conclusion (p. 52). Alternately, the authors suggested that the MBTI framework linked perceiving to the sensing/intu ition (S/N) dimension because both sensing and intuition provide individuals with ways to gather and process information (Hirsh & Kummerow, 1989, p. 52). Carlyn (1977) also compiled results from multiple studies that found high correlations between judging /perceiving (J/P) and sensing/intuition (S/N) and between judging/perceiving (J/P) and thinking/feeling (T/F) (see Table 3 3). Based on these findings, it is questionable whether the judging/perceiving (J/P) dimension is truly ind ependent.

PAGE 157

157 Table 3 3. Cor relations between judging/perceiving (J/P) and other dimensions . Samples Study S/N & J/P T/F & J/P 36 University Students Porter & Roll (1992) .52** .36* 300 Female Prospective Secondary Teachers Richek (1969) .28* .14 65 Female Prospective Secondary Teachers Richek (1969) .29* .29* 403 Students at Ohio State Schmidt & Fretz (1968) .39** .17** 300 Male Freshmen at Long Island University Stricker & Ross (1963) .33** .18** 184 Female Freshmen at Long Island University Stricker & Ross (1963) .47** .02 397 Male High School Students Stricker & Ross (1963) .26** .12* 614 Female High School Students Stricker & Ross (1963) .33** .20** 225 Male Freshmen at Wesleyan Stricker & Ross (1965) .34** .24** 201 Male Fr eshmen at Caltech Stricker & Ross (1965) .24** .23** 422 Freshmen at Emory Webb (1964) .25** .06 Note: * p < .05 , ** p < .01 Third, while judging/perceiving (J/P) is able to provide insight into human personality, the dimension is not particu larly well suited for explaining brand personality. The MBTI judging/perceiving (J/P) dimension describes how a person lives their outer life (The Myers & Briggs Foundation, n.d.). A judging person prefers a structured and decided lifestyle, while a percei ving person favors a more flexible and adaptable lifestyle. Because all brands fit into a structured and pre determined product category, all brands could be perceived as judging brands. Additionally, because successful brands have continuity in their prod ucts, service and advertising, they are not free to constantly change or adapt their offerings like the perceiving dimension would suggest. This inflexibility further suggests that, to some degree, all brands could be considered judging brands by the natu r e of good marketing practices. Fourth, the judging/perceiving (J/P) dimension also indicates how people become aware of things and how they form conclusions about those things. Because brands are not sentient beings, they can neither become aware of things nor form conclusions about them. Rather, a marketer works to convince people to become aware of things about a brand and make

PAGE 158

158 conclusions about a brand namely, to purchase. Therefore, the judging/perceiving (J/P) dimension did not seem applicable to bra nds. Fifth, the judging/perceiving (J/P) dimension is difficult to portray using only advertising images. Although the judging/perceiving (J/P) dimension has potential to be communicated via headlines or body copy, this study was interested in the role of visuals in communicating brand personality and thus excluded judging/perceiving (J/P). Independent Variable and manipulating brand personality. Therefore, advertising images, as visualizations of brand personality, functioned as the independent variable. Two steps, which are discussed in detail below, were required to choose the specific advertising images for use in the final survey. First, the researcher used her desi gn background and knowledge to choose images hypothesized to fit each of the eight personality dimensions. These images formed an image deck . Second, the researcher used the image deck to conduct group image sorts to determine which image best represented each of the eight personality dimensions. Those eight images then functioned as the Image Deck Creating an image deck was the first step necessary to select the particular advertising images that functioned as the curr image deck is a researcher compiled set of images that participants sort during group image sorts (Martin & Gourley of 24 images three images hypothesized to fit each of the eight personality types included in the study. All images were cropped to 5x7 inches and printed on cardstock. Additionally, all chosen images were portrait orientation to mimic the feel of a full page magazine adver tisement.

PAGE 159

159 Images were gathered from several stock ima ge websites, including: Depositp hotos ® ( www.depositphotos.com ) , Dreamstime ® ( www. dreamstime .com ), iStock ® ( www.istockphoto.com ) and Shutterstock ® ( www.shutterst ock.com ). With a background in advertising, graphic design and brand personality research, the researcher chose three specifi dimensions EST, ESF, ENT, ENF, IST, ISF, INT and INF. Based on the brand personality descriptions (detailed in the literature review) and past studies (Akay, 2001; Cunningham et al., 2007 ; Jewett & Sutherland, 2013), the researcher compiled general descriptions (below) using visual symbolic element hypothesized to fit each pole of the three personality dimensions included in the current study extraverted/introverted (E/I), sensing/intuit ion (S/N) and thinking/feeling (T/F). Extraverted (E). Extraverted images should show a group of people (preferably 3 or more) in a social setting (e.g., nightclub, party, event, workplace). The people should be happy, smiling and possibly looking at the camera. The images should have bright, saturated colors and be fairly tightly cropped around the people. The images should convey activity, energy and fun. Introverted (I). Introverted images should show only one (possibly two people) in a private or semi private location (e.g., home, empty beach, uncrowded restaurant) with low or romantic lighting. Alternately, images may show a single person with a pet. The shots should be soft or muted colors and convey a quiet, calm and mellow atmosphere and/or attitude. Sensing (S). (e.g., participating in activities, fixing things, helping people). The shots shoul d show people using their five senses to engage with the world around them. The images should be realistic and

PAGE 160

160 down to earth, and reflect a practical, hands on attitude. The images may employ more muted or primary colors. Intuitive (N). Intuitive images sh ould show people planning or daydreaming for a big project or life altering circumstance (e.g., architect with plans for a skyscraper, pregnant mother daydreaming about her baby). The images may be brightly colored, have a more white/modern/futuristic look , or mix the two. Images may also mix realistic photography and illustration. The image may include abstract art, modern sculpture, etc. Alternately, the image may be fanciful with mythical creatures and daydream like qualities (e.g., child daydreaming, un icorns, rainbows , fields of flowers, sunbeams). Thinking (T). Thinking images should show a cerebral environment (e.g., lab, classroom). The images may show the product and emphasize or point to particular features (e.g., V8 engine, easy twist cap, micro s crub particles). People in the images should be thoughtful and may be studying, reading, measuring or examining. Feeling (F). Feeling (F) images should show subjects touching (e.g., bodies, faces, hands). At least one person should be shown, and should be touching or interacting with another person or a cuddly pet. The subjects should be relaxed and convey warmth, emotion and love. The images should show people who are comfortable with each other and have some element of sentimentality (family, close friend s, couples). Images should have a warm, fuzzy feeling. The setting should be bright, warm and non sterile (e.g., outside, comfy living room, front porch). Next, the researcher used the general descriptions to produce three scenarios for each of the eight p ersonality types included in the current study (below). The scenarios were written by the researcher as a starting point to find images that would exhibit multidimensional personalities. Because little previous research examined visuals and multidimensiona l

PAGE 161

161 personalities, the researcher had to begin with relatively little initial data. By imagining how each personality would look with different settings, surroundings and people, the researcher was able to create scenes that fit each multidimensional persona lity. While as much of each individual required modification to fit the combined personality profile. Extr averted/Sensing/Thinking (EST): Extraverted/Sensing/Thinking (EST) Scenario 1: Multiple scientists in a lab (possibly looking at the camera, may be wearing round glasses), lots of instruments, crowded around with a test in progress or measuring something (but not touching each other), close crop Extraverted/Sensing/ Thinking (EST) Scenario 2: Group of coworkers (possibly looking at the camera, may be wearing round glasses), in action, bright/saturated colors, working on project (building a structure, conducting surgery, etc.), hands busy/dirty, people not touching, cl ose crop Extraverted/Sensing/Thinking (EST) Scenario 3: Group of professors/students (possibly looking at the camera, may be wearing round glasses), wearing bright colors, working on project, hands busy/dirty, in action, design classroom, room relatively c lean, people not touching, close crop Ex traverted/Sensing/Feeling (ESF): Extraverted/Sensing/Feeling (ESF) Scenario 1: Family (parents and kids) in a kitchen (possibly looking at camera), lots of cooking tools, crowded around a bowl touching each other, pa close crop Extraverted/Sensing/Feeling (ESF) Scenario 2: Group (possibly looking at camera), bright/saturated colors, in action, hands busy/dirty, working on project (gardening/et c.), people touching/showing emotion, smiling/enjoying the outdoor environment, close crop Extraverted/Sensing/Feeling (ESF) Scenario 3: Group of people (possibly looking at camera), bright/saturated colors, working on project, hands busy/dirty, in action, painting studio, walls colorful, messy painting tools, people touching and showing emotion, close crop

PAGE 162

162 Extra verted/Intuition/Thinking (ENT): Extraverted/Intuition/Thinking (ENT) Scenario 1: Group of scientists in a lab (may be wearing square or thick ri mmed glasses, possibly looking at the camera), colorful swirly illustrated fumes coming out of vials, not touching, close crop Extraverted/Intuition/Thinking (ENT) Scenario 2: Group of coworkers (may be wearing square or thick rimmed glasses, possibly look ing at camera), bright/saturated colors, people not touching, daydreaming about project possibilities (looking out over large head (possibly), looking at building plans, close crop Extraverted/Intuition/Thinking (ENT) Scenario 3: Group of professors/students (may be wearing square or thick rimmed glasses, possibly looking at the camera), wearing bright colors, daydreaming about project possibilities, hands idle, illustrat classroom, walls white, computer areas clean, art on walls framed if any, people not touching, close crop Extr averted/Intuition/Feeling (ENF): Extraverted/Intuition/Feeling (ENF) Scenario 1: Family (parents and kids) (possibly looking at out, crowded together, touching each other, close crop Extraverted/Intuition/Feeling (ENF) Scenario 2: Couple/family (possibly looking at the camera), daydreaming about project p ossibilities (looking at large garden/imagining a playground), bright/saturated colors, hands idle (not actively building/constructing), close crop Extraverted/Intuition/ Feeling (ENF) Scenario 3: Teacher and student(s) (possibly looking at camera), wearing bright colors, daydreaming about project possibilities, hands idle, touching and showing emotion, close crop Int roverted/Sensing/Thinking (IST): Introverted/Sensing/Thinking (IST) Scenario 1: One scientist alone in a darkish lab, muted colors, lots of instruments, concentrating on a test in progress, possibly measuring something or reading instruments, wide crop Introverted/Sensing/Thinking (IST) Scenario 2: One person (may be wearing round glasses), muted colors, sunrise/sunset/low light, working on project (gardening/building/etc.), hands busy/dirty, in action, wide crop

PAGE 163

163 Introvert ed/Sensing/Thinking (IST) Scenario 3: One student, wearing muted colors, low light, working on project, hands busy/dirty, in action/studying, walls white, computer areas clean, art on walls framed if any, wide crop In troverted/Sensing/Feeling (ISF): Introv erted/Sensing/Feeling (ISF) Scenario 1: Couple alone in low lit kitchen (e.g., candles on small table set for two), possibly wearing apron, muted/neutral colors, lots of cooking tools, man or woman concentrating on measuring something, people touching and showing emotion, wide crop Introverted/Sensing/Feeling (ISF) Scenario 2: One person, muted/neutral colors (if possible), low light (if possible), interacting with pet (washing/petting/etc.), hands busy/dirty, in action, enjoying/soaking in the outdoor envi ronment/beauty, pet touching and showing emotion, wide crop Introverted/Sensing/Feeling (ISF) Scenario 3: One person, muted/neutral colors (if possible), low light (if possible), working on project, hands busy/dirty, in action, pet touching and showing emo tion, wide crop Introv erted /Intuition/Thinking (INT): Introverted/Intuition/Thinking (INT) Scenario 1: One person (may be wearing square or thick framed glasses), not looking at the camera, neutral/muted colors, creative illustrated wide crop Introverted/Intuition/Thinking (INT) Scenario 2: One person (may be wearing square or thick framed glasses), not looking at the camera, muted colors, sunrise/sunset/low light, hands idle, reading a book, wide crop Introverted/Intuition/Thinking (INT) Scenario 3: One student (may be wearing square/thick framed glasses), not looking at the camera, muted colors, low light (if computer, hands idle, computer areas cl ean, art on walls framed if any, wide crop Intr overted/Intuition/Feeling (INF): Introverted/Intuition/Feeling (INF) Scenario 1: Couple, not looking at the camera, no other people/animals, low light, muted/neutral colors, touching and showing emotion, wide crop Introverted/Intuition/Feeling (INF) Scenario 2: One person, muted colors, sunrise/sunset/low light, daydreaming about possibilities, hands idle, enjoying/soaking in the outdoor environment/beauty, pet touching and showing emotion, wide crop

PAGE 164

164 Introverte d/Intuition/Feeling (INF) Scenario 3: One person, wearing muted/neutral colors, sunrise/sunset/low light, daydreaming about project possibilities, hands idle, pet touching and showing emotion, wide crop Next, stock image websites were searched for the imag es that most closely aligned with aspects, Adobe Photoshop CS5 was used to combine multiple images, creating a single image that more closely fit the descript ion. Below is a description of each selected image (see Appendix B to view the selected images). Extr averted/Sensing/Thinking (EST): Extraverted/Sensing/Thinking (EST) Image 1: Two scientists (one female and one male) dressed in white lab coats and gloves smile at the camera. They are holding vials filled with multi colored liquids and seem to be caught in the act of working. The female scientist is wearing protective glasses, and the male scientist is wearing a hair cover. The photo is closely cropped aro und the two individuals. Extraverted/Sensing/Thinking (EST) Image 2: Four surgeons crowd close together as they work together to operate on a patient. All the individuals are concentrating and looking down at their work. Several of the individuals are hold ing surgical instruments and are in the process of operating. No blood is shown, and the patient is cut almost completely out of the frame with the focus completely on the surgeons. All of the surgeons are wearing light blue protective clothing, hairnets, masks and gloves. The faces. Extraverted/Sensing/Thinking (EST) Image 3: Two individuals (one male and one female) smile at the camera as they crowd around a pottery whe el. Both are touching the clay with their hands dirty and are in the process of throwing a pot. The male appears to be helping/teaching the female pottery student. Both individuals are wearing overalls and the girl has on a bright green shirt. The area aro und the two individuals is lightly coated in clay, but relatively tidy. The image is closely cropped around the two potters. Ex traverted/Sensing/Feeling (ESF): Extraverted/Sensing/Feeling (ESF) Image 1: A young mother and father are in the kitchen with th eir elementary aged son. The smiling father has his arms around the boy, helping him chop brightly colored vegetables on a cutting board. The mother leans over cropped arou nd the family.

PAGE 165

165 Extraverted/Sensing/Feeling (ESF) Image 2: A young mother and father kneel outside in the garden with their toddler aged daughter. The mother smiles at the camera as she helps the little girl water a plant. The father crouches closely behind the pair, smiling as he watches. All three individuals are wearing brightly colored, summer clothing. The surrounding garden is lush and green, and the image is cropped tightly around the family. Extraverted/Sensing/Feeling (ESF) Image 3: A family of four mother, father, young boy and young girl work together to paint a wall in their house. All four individuals are looking at the camera and smiling. Each family member has a paint brush/roller and is in the middle of a paint stroke. The two children are standing on a small ladder to help their parents, who each have one arm safety supporting a child. The family is painting a formerly white wall a bright, chartreuse shade of green. They are standing on a gleaming, wooden floor and the whole family is dres sed in brightly colored casual clothes. The image is cropped around the family, although it does show th e full body of each individual. Extra verted/Intuition/Thinking (ENT): Extraverted/Intuition/Thinking (ENT) Image 1: A group of young scientists (two mal es and one female) wearing white lab coats, gloves and protective eye glasses smile and look eagerly as the girl holds up a vial of green liquid. Brightly colored, illustrated fumes (added by the researcher) swirl out of the vial. The group is seated at a wooden table in a clean, bright room. On the table, a collection of large beakers and vials hold a variety of brightly colored (e.g., yellow, green, blue, red) liquids. The image is cropped closely around the individuals. Extraverted/Intuition/Thinking (EN T) Image 2: A group of volunteers (four females and two males) stand in a group and smile at the camera as they take a break from clean up trash. Two members of the group hold trash bags while another holds an instrument for picking up litter. The group is standing outside under a large, green tree on a sunny day. Above their heads is a thought bubble (added by the researcher) that contains a hand cropped around the vol unteers. Extraverted/Intuition/Thinking (ENT) Image 3: A group of college aged students (three females and two males) study together. Two females and one male are seated at a table in the foreground surrounded by a laptop, books and several cups. The group is smiling and looking at the book the male is pointing to. Behind them, the other female and male stand at a large white board wall covered in marker brainstorming and sticky notes. They appear to be discussing one of the ideas on the wall. The students are wearing brightly colored clothing, and the image is closely cropped around the group. Extr averted/Intuition/Feeling (ENF): Extraverted/Intuition/Feeling (ENF) Image 1: A family of four (mother, father, young boy and young girl) crowd around the stove m aking dinner. The kitchen is modern, with a gas range, stainless steel hood and bright orange walls. The parents smile as the father

PAGE 166

166 lifts the lid off of a pot and colorful smell fumes (added by the researcher) escape. In the foreground, the two children s mile and laugh as the young boy stirs something in a small bowl on the counter. The image is cropped closely around the family. Extraverted/Intuition/Feeling (ENF) Image 2: A couple (one male and one female) sits daydreaming about their future. The young m an smiles at the camera with one arm drawn illustrations of their future (e.g., a city, a home, a dollar sign, a question mark). The illustrations are drawn in turquoise and couple is dressed in grays and bright yellows. The image is closely cropped with the couple situated in the lower left corner. Extraverted/Intuition/Feeling (ENF) Image 3: A young woman (perhaps a teacher or a mother) sits beside a small boy at a table. The table is covered in white paper and a smiles down at him. The small boy holds a red pe ncil in the air and appears to be thinking about what to draw. An illustrated cloud of colorful shapes (added by the researcher) escapes from the tip of the pencil. The image is tightly cropped around the pair. Int roverted/Sensing/Thinking (IST): Introvert ed/Sensing/Thinking (IST) Image 1: A lone, female scientist stands in a lab measuring blue liquid in a beaker. The woman has her hair pulled back and is wearing a white lab coat, gloves and protective eyeglasses. She looks seriously down at her work as she leans against a counter with shelves full of bottles behind her. The picture is taken through a second set of shelves, which are blurred in the foreground. The image is In troverted/Sensing/Thinking (IST) Image 2: An older, male farmer squats alone in his field and looks down to examine his wheat. He is wearing a light blue, collared, work shirt with worn jeans. He also has on an olive green ball cap pulled down over his ey es. The image is somewhat loosely cropped showing the field wheat field and blue sky behind the farmer. Introverted/Sensing/Thinking (IST) Image 3: A single student sits on the floor against the wall in a library. The girl has legs stretched out in front o f her with a laptop on her knees and a book open beside her. She is looking at the screen and typing. The library walls, floor and shelves are mostly white, and the girl wears a white cardigan over a dark dress. The image is relatively loosely cropped with the girl in he bottom, left corner and a long row of library shelves stretching down the corridor behind her. In troverted/Sensing/Feeling (ISF): Introverted/Sensing/Feeling (ISF) Scenario 1: A couple (one male and one female) stands alone in a kitchen. Th e man tosses a salad on the counter as the woman leans over to kiss his cheek with a wine glass in her hand. An empty bottle of wine sits on the counter in the background. A chandelier in the foreground casts yellow, romantic light. The kitchen is

PAGE 167

167 modern a nd neutral with white cabinets and beige counters. The couple is also dressed in neutral grays. The image is somewhat loosely cropped with the couple centered in the scene. Introverted/Sensing/Feeling (ISF) Image 2: A young man washes his dog outside on a sunny day. The man bends over the dog, scrubbing its fur with soap. All that can be seen of the man is the top of his head and his red shirt. The image is somewhat tightly cropped, although the dog is the focus of the image with the young man half cropped out of the image on the left and a large tree in the background. Introverted/Sensing/Feeling (ISF) Image 3: A middle aged woman washes the dishes in the sink as a cat prowls on the counter. The woman has her hands in soapy water scrubbing a plate as she lo oks lovingly at the cat. The cat stands to the left of the woman wearing a white, patterned shirt. The cat mimics the coloring of the woman and is white with large gray patches. The kitchen is modern and almost completely white. Two white vases sit on the counter filled with light pink flowers. The image is somewhat loosely cropped, showing the woman and cat entirely, along with the kitchen background. Introv erted /Intui tion/Thinking (INT): Introverted/Intuition/Thinking (INT) Image 1: A middle aged man stands alone in front of a chalkboard that is covered in white chalk illustrations and doodles. The man looks like a professor with glasses, a long sleeved, collared shirt rolled up to his elbows, and a loosened tie. The man is turned sideways in the frame and has one hand on his hip and the other on his chin in a thoughtful pose as he examines the board. The image is somewhat loosely cropped with the man in the lower half of the frame and the chalkboard filling the entire background. Introverted/Intuition/Thinking (INT) Image 2: A young woman sits with her back against a tree reading a thick book. The image shows the woman from the back. She is wearing light colored jeans, black sneakers, a black and white, horizontal striped shirt and a black sun hat. The image is somewhat loosely cropped with the girl in the bottom, left corner and an long view of a lush, green field and blurred trees in the background. Introverted/Intuiti on/Thinking (INT) Image 3: A young man stands alone on a white background holding a tablet. The man is looking up at gray, illustrated circles that are escaping the tablet, which are filled with various illustrations (e.g., briefcase, calculator, piggy ban k, light bulb). The man has dark, gelled hair and is wearing a light blue, collared dress shirt buttoned all the way to the top. The image is somewhat loosely cropped with the man in the bottom, right corner and the illustrations centered in the frame. Int r overted/Intuition/Feeling (INF): Introverted/Intuition/Feeling (INF) Image 1: A couple (one male and one female) sit, lounging under a large tree. The man is leaning with his back against the tree trunk, while

PAGE 168

168 the woman leans back on his chest. The woman has one arm lifted above her head and green, romantic light over the image. The couple is wearing neutral colored clothing that b lends in with the light. The image is loosely cropped with the couple taking up a relatively small portion of the frame. Introverted/Intuition/Feeling (INF) Image 2: A woman sits alone on the beach at sunset with her dog. The image is backlit by the sun, c reating a low, romantic light. The woman is wearing sunglasses as she looks out over the ocean. Her yellow lab sits leaning up against her body also looking out over the water. The image is somewhat loosely cropped with the woman and the dog centered in th e frame, and the beach and ocean visible around them. Introverted/Intuition/Feeling (INF) Image 3: A young woman stands next to her saddled horse as it grazes in a pasture in the evening light. The woman is wearing jean shorts and a neutral colored shirt. The house is a dark, rich reddish brown. The image is loosely cropped with the woman and the horse in the bottom, left corner of the frame. The green pasture and bright blue sky extend far behind the pair. Group Image Sorts A series of qualitative group im age sorts was the second step necessary to select the image sorts are the fusion of two qualitative research methods group interviews and image sorts (both dis cussed in detail below). While past research has regularly used both group interviews and image sorts, group interviews do not typically include image stimuli, and image sorts are most often conducted with individuals rather than groups. However, for reaso ns discussed below, combining these techniques provided the rich est data for the current study. In their book about qualitative research in advertising, Morrison et al. (2002) noted that account planners rely heavily on both traditional qualitative approac hes (such as interviews and focus groups) and innovative qualitative approaches (such as visual prompts and word associations) to gain consumer insights. Therefore, it was appropriate for the current study to use group image sorts, which combined group int erviewing (traditional) and image sorting (innovative) to investigate how consumers visually determine brand personality.

PAGE 169

169 During the group image sorts, participates sorted the previously compiled image deck to determine which image best represented each MB TI dimension included in the current study extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). Further, participants were asked to explain their choices, providing the researcher with micro level descriptions of the visua l symbolic elements associated with each dimension. Group Interview Definition Because the current study combined group interviews and image sorts to create a new method, group image sorts, it is important to understand the details of how each individual m ethod works. Therefore, the following sections define and discuss the advantages of group interviews, followed by a definition and discussion of the advantages of image sorts. The group image sort method and process used in the current study are then discu ssed in detail. For this study, group interviews were composed of three participants each. According to some scholars (Babbie, 2010; Morrison et al., 2002), the terms focus group and group interview are interchangeable. Berger (1998) even called focus grou 89), and Gaskell (2000) noted that there were many similarities between the two methods. However, other scholars suggested that focus groups typically have five to 15 participants (Babbie, 2010) or six to eight particip ants (Gaskell, 2000), while group interviews might have only two to four (Bernard & Ryan, 2010; Thornberg, 2008). Although the number of participants varies, the process of conducting in depth interviews, group interviews and focus groups is similar (Gaske ll, 2000). Therefore, because this study used groups of three participants, the term group interview was most accurate. For the current study, a group interview was defined as a moderated discussion with three participants. The current study used three per son group interviews as the setting for the group image sorts for several reasons. First, because the true purpose of this format was to conduct group

PAGE 170

170 image sorts, it was vital that all participants could see, hear, touch and manipulate the physical materi als presented. A group of three allowed the participants to sit normally at a table without being overcrowded, but still allowed them to be close enough to the materials so that they could contribute to the sort. Second, the group image sort exercise requi red that the participants work together and agree on which pile each image should be placed. Three participants per group was a large enough number to allow ample discussion between members. But, three participants was also a small enough group to allow ea ch participant to see/touch all the visual stimuli, and allow the group to ultimately come to a conclusion on where to place each image. Finally, an odd number of participants was desirable to prevent the possibility of a tie. With three participants, a si ngle dissenting member could be overruled. Advantages of Group Interviews Group interviews have a number of advantages that are similar to those of other qualitative interviewing techniques, such as individual interviews and focus groups. These advantages include: understanding attitudes, producing concentrated topical data, an d allowing follow up questions. Like other forms of qualitative interviewing, one advantage provided by group interviews ard a brand. According to Gaskell (2000), an interview (individual or group) was most valuable because it provided data to help of individual or group inter views also include s attitudes, feelings, values, reactions and motivations (Bernard & Ryan, 2010; Gaskell, 2000; Gibbs, 1997). Further, Berger (1998) suggested focus groups were usually conducted to discov er

PAGE 171

171 images for potential advertisements, taking advantage of t relationship between individuals and potential marketing materials. Qualitative interview techniques, such as group interviewing, also allow researchers to gather a large amount of data specific to their research questions (Morgan, 1997). According to Morgan (1997), interviewing was observations) that gathered a wide variet y of information, Morgan (1997) noted that interviewing techniques (e.g., focus group, group interviews) allowed researchers to directly target the information they were most interested in. The current study also took advantage of the inherent flexibility in qualitative forms of interviewing, such as group interviews (Babbie, 2010; Berger, 1998). Group interviews provide a high level of flexibility, allowing the moderator to easily follow up on comments, ask additional questions, clarify responses and reque st opinions on specific issues (Berger, 1998; Bernard & Ryan, 2010; Krueger, 1988). Additionally, in a group interview, a moderator can follow participants through unexpected twists and turns that may provide valuable, but unforeseen, information. The flex ibility of group interviews was an especially important advantage for the current study as it sought to understand what specific visual symbolic elements in each image were responsible for creating brand personality. With the ability to ask participants qu estions about why they placed images in each category, the researcher could follow their train of thought and uncover the specific symbolic visual elements in each image that c onvey a particular personality. Advantages of Group Interviews Over Individual I nterviews In addition to their general advantages, (e.g., understanding attitudes, producing concentrated topical data, allowing follow up questions), group interviews, as a form a focus

PAGE 172

172 group, also have advantages over individual interviews. These advanta ges include: encouraging group dynamics, gathering data more quickly, and complemen ting other forms of research. Because group interviews allow the researcher to interview multiple people simultaneously, they have the added advantage of group dynamic. Acco rding to Morgan (1997), it is the group interaction that produces data. More specifically, Morgan (1997) suggested, valuable source of insights into complex behavio (Morgan, 1997, p. 13) and became more than the sum of their parts (Gaskell, 2000). Therefore, the current study used group interviews, rather than one on one interviews, to take advantage of group dynamics. According to Bernard and Ryan (2010), another aspect of the group dynamic is a group interviewee and a moderator. Babbie (2010) suggested that these unforeseen discussions could also uncover aspects of the topic that were unanticipated by the researcher. For example, the dynamics of a multi person group could encourage participants to dis cuss similarities and differences of opinion, providing additional research data (Bernard & Ryan, 2010). Grumbein (2013) also noted this aspect of group dynamic; she found that when students worked as a team to sort brand personality adjectives, the negoti ation and discussion involved was valuable to the outcome. Because team members had to agree on the definition and final placement during the exercise in loud, w orking to convince the others in their group. Therefore, by asking multiple individuals to collaborate in the current study, the moderator was afforded a glimpse into an otherwise silent

PAGE 173

173 thought process. As such, this study encouraged and took special note of discussion among participants during group image sorts as they negotiated which category an image belonged to and why. Like focus groups, group interviews also allow researchers to gather a large amount of data more quickly and easily than conducing in dividual interviews (Morgan, 1997). Morgan Additionally, among scholars, the use of focus groups (or here , group interviews) is common in tandem with other methods, particularly the development of surveys (Babbie, 2010; Bernard & Ryan, 2010; Gaskell, 2000; Gibbs, 1997; Morgan, 1996). Because focus groups were questions, Berger (1998) noted that many researchers used the technique in pilot studies and as complements to other kinds of studies (p. 91). Morgan (1996) also reported that many researchers specifically used group interviews to develop content for ques tionnaires. Further, according to Morgan (1996), combining surveys and curre nt study to gather and develop data for a later quantitative survey. Image Sort Definition Because the current study fused group interviews and image sorts to create group image sorts, the sections above discussed the definition and advantages of group int erviews. The following section now discusses the definition and advantage of image sorts the second method used to create group image sorts. Although past researchers have performed image sorts in a one on one context, the current study asked participant s to partake in group image sorts. For the current study, a group

PAGE 174

174 image sort was defined as moderated group composed of three participants who work together to sort an image deck into pre determined categories. participants were asked to collectively conduct several image sorts. This study chose to use the term image sort because it best represented the process used. However, similar techniques have been known by many names, including photo sort (Gaskell, 2000; Van Riel, Stroeker & Maathuis, 1998), picture/issue sort (Gaskell, 2000), image bank (Moriarty & Rohe, 2005), card sort (Nielsen, 2004) and pile sort (Bernard & Ryan, 2010; Harman, 2001; Potter & Trotter, 1993). The term image sort was also selected to dis of people to help the agency determine brand personality (Morrison et al., 2002). Additionally, the current study avoided t he name pile sort because the term most often referred to sorts using words rather than visuals (Harman, 2001; Potter & Trotter, 1993). three included MBTI personality dimensions extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). Weller and Romney (1988) referred to the process of asking participants to conduct several sorts of the same items as a variation on the sort technique calle d a multiple sort (as opposed to a single sort ). According to Weller and Romney (1988), in a multiple sort condition, participants were given the opportunity to sort and re sort the same items using a different criterion each time (p. 24). In the current s tudy, the criterion for each re sort was specified by the moderator to fit the individual MBTI dimensions. Advantages of Image Sort s The group image sorts used in the current study drew on a number of the advantages of image sorts to help provide a better advertising visuals and gather insights on how they viewed specific visual symbolic elements in

PAGE 175

175 relationship to particular brand personalities. These advantages include: allowing non verbal feedback, pr omoting discussion, increasing participant enjoyment, providing brand personality information and identifying images for particular purposes. One of the biggest advantages of an image sort may be the additional level of non verbal feedback it provides (Van Riel et al., 1998). Traditional qualitative interviews allow participants to provide only verbal feedback. However, in addition to verbal responses, image sorts also allow participants to provide non verbal feedback by examining and sorting the images. Ac cording to Van Riel et al. (1998), the technique was developed as a result of dissatisfaction with existing methods for measuring images that required participants to only verbally respond s are non verbal, they should be better measured by non to asking participants to discuss their thoughts out loud as they sort, the current study also gathered non verbal data from par ticipants th The ability to stimulate discussion is another advantage of image sorts. According to Gaskell (2000), stimulus materials (e.g., pictures, drawings, photographs) were useful to eans of getting people to use their imagination and to help participants overcome any natural inhibitions, such as embarrassment, shyness or a desire to please. less conscious, and more deeply As an additional advantage, image sorts are enjoyable for participant s and create a sense of fun during the group interviews (Bernard and Ryan, 2010; Van Riel et al., 1998). Van Riel et al. (1998) also found that, when compared to attitude scales and a Q sort, image sort results

PAGE 176

176 evoked the most positive reactions from respo ndents. In turn, because respondents enjoyed participating in the image sort, they not only completed the activity, but also provided many spontaneous comments about the airlines and the photos, which were valuable for the researchers results (Van Riel et al., 1998). Image sorts can also provide researchers with brand personality information. Compared to other methods (e.g., Q sorts, attribute scales), Van Riel et al. (1998) noted that image sorts Because the current study investigated brand personality, this image sort advant age was particularly important. Finally, image sort s also provide researchers with the added advantage of identifying specific images for particular purposes. According to Van Riel et al. (1998), image sorts also r, advertising agencies often use image sort techniques to create distinctive campaigns (Van Riel et al., 1998) and match message strategies to the target audience (Moriarty & Rohe, 2005). provide the researcher with eight unique images, each representing one multidimensional MBTI brand personality EST, ESF, ENT, ENF, IST, ISF, INT and INF this technique effectively aligned with the goals of the current study. Qualitative Participants F or the qualitative group image sorts, a purposive sample of students was selected from a large, public southeastern university. A purposive sample allowed the researcher to purposefully choose participants that were most able to meaningfully contribute to the study (Creswell, 2007).

PAGE 177

177 extra credit for their participation. They also had access to researcher provided snacks and beverages during the group image sort sessions. More specifically, in order to participate in the study, students needed to function as experts who were able to provide an informed opinion (Moriarty & Rohe, 2005). Therefore, ADV 3203 Advertising Design and Graphics . After completing this course, advertising students had in depth knowledge and understanding of brand personal ity and the symbolic nature of advertising images. Alternately, because students must apply for entrance to the graphic design major after completing four semesters of department prerequisites and undergoing a rigorous review and selection process, all gra phic design majors were able to contribute meaningfully to the discussion about the meanings of spe cific symbolic visual elements. By limiting group image sort participants to advanced level advertising and graphic design students, the researcher was able to collect more detailed and nuanced information about how particular visual symbolic elements relate d to a specific MBTI personality dimensions. B ecause advertising or graph ic design students . In addition to requiring participants to be advanced advertising and graphic design students, qualitative participants were also limited to American students in order to achieve a more homogenous population. Limiting participants to Ame rican students also allowed the researcher to examine a single culture. Because culture plays a significant role in the interpretation of symbolic elements, conducting group image sorts within one culture provided

PAGE 178

178 the current study with a more in depth und erstanding of a single group. For this study, American students were defined as those who have lived in the U.S. for 18 years or more. Number of Group Image Sorts Because group image sorts are a unique combination of group interviews and image sorts, no pr evious data existed to guide the current study in terms of the number participants necessary. participant numbers for qualitative interviews and individual image sorts wer e used as guidelines for the current study. In accordance with qualitative tradition, group image sorts were conducted until saturation was reached. Guest, Bunce and Johnson (2006) noted that the most commonly referred to type of saturation wa s theoretical saturation . Theoretical saturation is typically defined as the point at which the researcher is uncovering the same information and no longer identifying new themes (Guest et al., 2006; Ryan & Bernard, 2004). The point of saturation cannot be predicted wi th a simple, fixed number of participants. Rather, when and how saturation is reached depends on the number and complexity of the texts, investigator experience and fatigue and the number of investigators examining the text (R yan & Bernard, 2004, para. 61) . When examining past literature regarding qualitative interviewing, some scholars suggested that increasing the number of interviews might not always result in better findings (Gaskell, 2000; Guest, et. al, 2006). According to Gaskell, (2000), this was tr ue because: 1) there were limited versions of reality, and 2) the data must be analyzed. Although each individual had a unique set of experiences, Gaskell (2000) argued that those experiences grew out of a set of shared social processes, creating a shared cultural lens through which individuals view the world around them. Because many people share a limited number of cultural lens, a limited number of realities exist, particularly when participants are homogenous and from a single culture as in the

PAGE 179

179 current study. Therefore, Gaskell (2000) suggested that as more and more interviews were conducted, the researcher would continue to see the same shared experience and overlap of insights. In addition, once the interviews were conducted, the data must still be ana lyzed, requiring both time and energy from the researcher (Gaskell, 2000). Thus, by virtue of memory and time constraints, there is a limit to the amount of data a single researcher can synthesize (Gaskell, 2000). Scholars also indicated that the number of interviews needed to reach saturation varied from study to study. However, according to Morgan (1997), theoretical saturation could be achieved with three to five focus groups in most cases. Similarly, Kruegar (1988) recommended beginning with three focus groups and adding an additional group if the third group provides new insights. Wimmer and Dominick (2011) also advocated beginning with three groups and adding additional groups as necessary. Alternately, when conducting individual interviews, Guest et a l. (2006) suggested that six interviews might be enough to develop high level, overarching themes with both meaning and useful interpretations. Neilsen (2000) agreed, proposing that there were significant diminishing returns in new data as the number of us ability testers (participants) increased and advised using five usability testers, which could uncover 85% of usability issues and maximize the cost to benefit ratio. Finally, Gaskell (2000) suggested the limit for a single researcher was between 15 and 25 individual interviews or six to eight focus groups. Unlike research on qualitative interviewing, past literature on image sorts, which are done on individual bases rather than in groups, shows an enormous range. According to Van Riel et al. (1998) the typ ical number of respondents for an image sort was between 30 and 100 (their image sort study included 92 participants). Alternately, in his study assessing the knowledge Mayan

PAGE 180

180 person sample to sort illustrated cards. Martin and Gourley grades students sorting photographs of potentially math research on the cultural nature of mathematics asked a total of 24 children to sort photos based whether they thought the subjects were good at math. Further, for their study on creating cultural palettes, Moriarty and Rohe (2 007) conducted image sorts with eight experts. Nielsen (2004) also investigated the ideal number of participants for card sorting and suggested conducting card sorts with 15 participants. Because there is such a wide range of past image sort samples rang ing from eight (Moriarty & Rohe, 2007) to 92 (Van Riel et al., 1998), this study followed the literature on qualitative interviewing to guide the number of participants. In line with the qualitative interviewing recommendations of Gaskell (2000), Morgan (1 997), Guest et al. (2006), Krueger (1988) and Nielsen (2000), the current study began with three group image sorts for the qualitative portion of this study while monitoring the stability of uncovered themes and adding additional group image sorts as neces sary. A total of 10 group image sorts were conducted to reach saturation . Further, unlike individual interviews, this study used group image sorts with three participants in each. Therefore, 10 group image sorts resulted in data collection from 30 particip ants more than the minimum of 6 recommended by Guest et al. (2006) and beyond the fi ve suggested by Nielsen (2000). Further, with 10 group image sorts of three participants each, the images were be sorted collectively by at least 30 participants. This nu mber was more the number of image sort had a process similar to the current study in many respects.

PAGE 181

181 Screener Questionnaire Prior to attending the group image sorts, participants were asked fill out a short online screener questionnaire to provide basic descriptive information (e.g., name, age, year in school, major, gender, race). The questionnaire also included weed out questions for graphic design and adverti sing majors and for non American students. Students indicating that they were advertising majors were asked if they had completed ADV 3203 Design and Graphics . As discussed above, only students who had completed ADV 3203 Design and Graphics were included i n the group interviews. Also as discussed above, students who indicated that they had not lived in the United States for at least 18 years were also excluded from the group image sorts. In addition to basic descriptive information, the screener questionnai re also asked participants to determine their own personality using the BHPI 5 brand specific MBTI scale, which the BHPI was based on, is applicable to both human and brand personalities (Strausbaugh, 1998; Sutherland et al., 2 004), it was appropriate to have participants measure their own personality using the BHPI. In fact, Sutherland et al. (2004) MBTI for measuring individual and bra (2004) also used the brand although this scale was created with brands in mind, the scale items were designed to correlate with th e human MBTI, measuring the same personality dimensions in a s imilar way (Strausbaugh, 1998). Measuring human personality with the BHPI has several additional advantages, including: increasing participant familiarity, shortening survey length and decreasin g research 5 The BHPI scale will be explained in detail in the following chapter. Because the current study plans to reduce specific MBTI scale to create the BHPI, the instrument does not yet exist.

PAGE 182

182 pairs. By asking participants to indicate their own individual personality type using the BHPI, the researcher was able to introduce them to the adje ctive pairs prior to the group image sorts. Second, the full length human MBTI, which consists of 126 questions, would take a significantly longer commitment from the participants prior to the focus group, which could result in poor data or participant att rition. Finally, the full length human MBTI is costly and must be sent off for scoring. However, the BHPI could be scored quickly and cost free by the researcher. el ectronically and scored later by the researcher. This data was used to provide descriptive information about the group interview important that participants be homoge nous (i.e., American students) and sufficiently knowledgeable about advertising visuals and brand personality (i.e., advanced advertising majors, graphic design majors) rather than a particular personality type (e.g., ESTJ, INFP). Therefore, information on forming group image sorts, but individual personality data was not. Moderator moderator. Scholars agr ee that moderators play a critical role in the success of a group interview promoting debate, keeping the conversation moving, focusing the discussion and allow ing all participants the chance to speak (Gibbs, 1997). With past experience moderating focus groups and an intimate knowledge of the topic, the researcher was best poised to guide the conversation

PAGE 183

183 during the group image sorts. An additional aid was also p resent during group image sorts to help with logistics and take notes. was a high level of interview structure and moderator involvement (Morgan, 1997). A high level of moderator involvement keeps the group focused on topics relevant to the research (Morgan, 1997). Morgan (1997) also suggested that a more structured format was desirable when one of as surveys. Therefore, because the current study had a clear research goal (i.e., to identify specific visual symbolic elements that are associated with particular personalities) and planned to use the group image sort results in a survey (i.e., participa nt chosen images will be used to create test advertisements), a high level of interview structure and moderator involvement were desirable. Appendix C for t for the group and sp ecifies their preferred order. an with an introductory statement (Morgan, 1997). The introduction should present the topic of study honestly but generally (Morgan, 1997). A general introduction ensures that the topic is presented at a level that participants can understand discussion (Morgan, 1997, p. 48 49). Therefore, the current study was introduced as part of a project to understand mor e about images and personality. The introductory statement should also include a brief summary of the ground rules (Morgan, 1997). A brief explanati on of the rules and expectations provides participants with proper group etiquette, but does not imply that the moderator will be telling the group what to do

PAGE 184

184 working as a team, the ability to disagree with other participants, respectful behavior, wanting to statement also informed the participants that the moderator had no opinion of her own and was only interested in hearing what they thought. In addition to generally presenting the topic and explaining group rules, the current ed for additional analysis. Participants were also asked to speak one at a time to provide the highest quality audio recording. After the introductory statement, participants were asked to review and sign a consent form. The consent form confirmed that the nature of the group had been explained and that the participants consented to take part in the group and to be audio recorded. The consent form also ensured that participants understood that participation was voluntary and that there would be no negative repercussions if they chose not to participate. Finally, the consent form articulated that identities would be protected. Once all the participants had read and signed the consent form, the moderator began the group by asking an icebreaker qu estion (Morgan, 1997). According to Morgan (1997), the icebreaker question allows each participant to give a brief self introduction and helps set the mood. For the current study, participants were asked to give their name, major and year in school. This q uestion established that all the participants were similar majors (advertising or graphic design) and at a similar point in their college careers. starter question (Morgan, 1997). The discussion starter qu

PAGE 185

185 to provide some meaningful statement. Because the current study was interested in particular symbolic visual elements, the discussion starter question was used to help acquaint pa rticipants discussion each participant answered, the moderator briefly expla ined the concept that specific visual elements, like color, often have meanings that are associated with personality. The moderator then explained to participants that she was interested in knowing if a set of images had any personality. After the icebreak er and discussion guide was devoted to key questions (Krueger, 1998). Key questions are those that drive the session and which the researcher is most interested in (Krueger, 1998). In the current study, the key questions included conducting the group image sorts and asking a series of follow up questions after each sort. Both the image sort process and follow up questions are discussed in detail in the procedure section below. To conclude the group discus sion, an ending question was asked (Kruegar, 1998; Morgan, 1997). An ending question signifies the close of the group discussion and allows the (1998) suggestio if anyone had anything else they would like to add. Once participants had the opportunity to answer the ending question, the moderator thanked the participants, and the group image sort was concluded.

PAGE 186

186 (Kruegar, 1988). There fore, an expert committee, who are experienced in conducting focus As an additional c heck, Kruegar (1988) also proposed that the first group serves as a final 88). In the used in during data analysis. Procedure After completing the screener questionnaire, potential participants who did not fit the previously discusse d criteria were eliminated. Once appropriate participants were identified, most group s contained either advanced advertising or graphic design students due to differenc es in The group image sor ts were conducted on campus in small, private room s . The room s were reserved ahead of time, allowing the moderator to set up prior to each image sort. By using a private room, the moderator could be su re that the necessary equipment (e.g., table, chairs) was available. A private room also ensured that the environment was quiet and conducive to group discussions. Holding the group image sorts on campus also ensured that participants (i.e., students) coul d easily access the location. On the day of the group image sort, the moderator arrived early and set up the room (e.g., set out water and food, arrange table and chairs, prepare materials, test tape recorders). As participants began to arrive, the moderat or greeted them, offered refreshments and began making small talk to build rapport. Each participant was also provided with a name tent and asked to

PAGE 187

187 write their name. Using name tents allowed the moderator and each member of the group to easily address eve ryone by name. Once all three participants arrived, but before the research activities begin, the moderator asked participants for permission to audio record the session and had them sign consent forms cluding the consent form). Once permission was obtained, the audio recorders were turned on. By audio recording the sessions, the researcher had access to the group image sort discussions in order to review and analyze the group image sort data at a later time. Only the researcher had access to the audio recordings. After the group image sorts, the audio recordings were transcribed and then the recordings were Once th e group image sort began, the moderator showed participants the image deck and explained the image sort process (Moriarty & Rohe, 2005). Specifically, the moderator explained that the participants would work as a team to sort a set of images into piles bas ed on personality descriptions, which the moderator would provide. Each card in the image deck had a full color photograph on one side and a unique number on the back so the moderator could easily record which pile an image was sorted into (Bernard & Ryan, 2010). While explaining the image sort process, the moderator emphasized that the group would complete the process as a team. The moderator also encouraged participants to discuss what they were thinking out loud to benefit both the other team members and the moderator. Immediately p rior to each sort the researcher provided the participants with a short verbal and written explanation of the dimension to be sorted for that round (see Appendix D). For example, the researcher read the following extraverted (E ) and introverted (I) descriptions: Extraverted (E): Extraverts relate easily to the outer world of people and things. They focus on the world around them and gain energy from other people.

PAGE 188

188 Extraverts often know a little about a lot, preferring to explore a variety of subjects and ideas. Extraverts often need to think out their ideas aloud. Adjectives that describe extroverts include: outspoken, active and sociable. Introverted (I): Introverts relate easily to the inner world of ideas. They draw energy from their inner world and prefer to spend time alone. Introverts tend to have fewer but deeper interests, immersing themselves in the topics that appeal to them. Introverts prefer to quietly think through an idea in their head before speaking it. Adjectives t hat describe introverts include: quiet, reserved and shy. The three dimensions extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F) were presented in random order at each group interview to ensure that presentation order did n ot affect the research results. Participants were then directed to sort the images into three piles: one for each pole of the current dimension (e.g., extraversion or introversion) or a neither pile (Gaskell, 2000; Martin & Gourley Delaney, 2013; Mori arty & Rohe, 2005; Romney, 1988; Van Riel et al., 1998). The neither pile ensured that participants were not forced to place a non applicable image into a neither pile also helped the researcher eliminate images with visual s ymbolic elements that did not align with one of the MBTI dimensions. While sorting, participants were not allowed to place an image in more than one pile (Bernard & Ryan, 2010). If participants asked whether a card could be placed in multiple piles, the mo derator first asked them to explain why they would like to place the card in multiple piles and identify the particular visual symbolic elements that make them feel the image fits in multiple piles. Then, the card that caused confusion was set aside and no t included in any pile. However, by first asking the participants to discuss and identify the particular elements that make the image fit multiple MBTI dimensions, the researcher gleaned valuable information . To start the sort, the moderator shuffled the i mage deck and placed it image side up on the table within view and reach of all the participants (Bernard & Ryan, 2010; Gaskell, 2000). Small tabletop signs with the two dimensions currently being sorted (e.g., extraversion,

PAGE 189

189 introversion) were placed on th e sorting surface to indicate where each pile should be stacked. In adjectives from the BHPI to remind participants of the dimension definition. An additional neit her sign was also placed on the table. Participants then began working together to sort each of the 24 images from the image deck into one of the two dimension piles (e.g., extraversion, introversion) or the neither pile. During the sort, the moderator ask ed participants to discuss why they chose to assign the images elements (e.g., color, number of people, poses, eye contact) translate into the specific persona lity dimensions resulted in richer data and additional understanding of the cognitive process. After all the images had been sorted, the moderator reviewed each pile with the participants and asked a series of follow up questions to ascertain which aspects of the images were most important during sorting, which specific visual symbolic elements made them place an image in a particular pile, which image most represented each dimension, what might be missing from the images and how they would describe the ide al image for each dimension (Van Riel et al., 1998). Once the first sort was complete, each pile and table top was fastened togethe r with a neither) could be noted by the moderator after the session by recording the unique number on the back of the card (Bernard & Ryan, 2010). A second, id entical image deck was then shuffled and given to the participants, allowing them to re sort the same images for the second dimension

PAGE 190

190 (Martin & Gourley Delaney, 2013). The process was then repeated a third time for final dimension. By using three separate but identical image decks, the moderator minimized down time during the group image sort and recorded the results after the session concludes. Preserving each dimension group also allowed for a more thorough review of the data and decreased the potential f or recording mistakes due to t ime constraints or distraction. Once the three sorts had been completed, the moderator asked the participants if they had any additional information they would like to add or questions they would like to ask. Participants were then thanked for their time and dismissed. After the participants had left , the moderator examined and r ecorded the image sort results. Analysis Once saturation was reached, the group image sorts provided the researcher with a better idea of 1) which imag es 2) which specific visual symbolic elements best fit each of the individual MBTI personality dimensions (i.e., extraversion (E), introversion (I), sensing (S), intuition (N), thinking (T) and feeling (F)), and 2) which specific visual symbolic elements b est fit each of the eight MBTI brand personality types included in the current study (i.e., EST, ESF, ENT, ENF, IST, ISF, INT and INF ) . However, the priority in the group interviews/image sorts was to complete all three sorts in each group, ensuring that t he best eight images one for each of the eight personality types were identified for use in the final quantitative survey. If a group completed the activities relatively quickly, there was more time for the moderator to ask probing questions and collec t more detailed information about specific visual symbolic elements. However, if the task took a group longer, less probing was possible.

PAGE 191

191 The researcher used the group im age results to answer RQ1 (a c), RQ2 (a h), RQ3 and RQ4 (a h). Analysis was conducted by reviewing a combination of audio transcriptions and notes taken during the group image sorts. More specifically, to answer RQ1, which asks about the visual symbolic elements in advertising images that transfer meaning consistent with each MBTI dimension , the researcher personality dimensions as they sort the images into distinct groups. Images from each individual Because RQ 2, which asks how the visual symbolic elements that represent each MBTI dimension combine in advertising images to portray multidimensional MBTI personalities, was interested in multidimensional personalities rather than a single dimension, the researcher combined information from the single dimension group image sorts and group discussions to identify the visual symbolic elements that represent specific MBTI personalities. For example, to identify visual symbolic elements that represent an extroverted/sens ing/thinking (EST) personality, the researcher combined the visual symbolic elements group image sort participants discussed as being extroverted (E), sensing (S) and thinking (T). The researcher also analyzed images that participants from multiple groups sorted into the extroverted (E) pile, sensing (S) pile and thinking (T) pile, using those images as model EST images. In both cases, visual symbolic elements were organized into categories and/or themes to provide other researchers and practitioners with t he unique characteristic that represent each of the MBTI dimensions and multidimensional personalities. To answer RQ3, which asked how participants read and negotiated personality information in images, the researcher holistically analyzed the group image sort data.

PAGE 192

192 Specifically, the researcher looked for large scale patterns concerning how participants interacted with and drew meaning from the images in the image deck. Finally, to answer RQ4 (a h), which asked what image from the image deck was best repres ented each multidimensional MBTI personality, the researcher selected those images that were most consistently sorted into the corresponding personality dimensions. For example, to identify the image that best represents an EST image, the researcher determ ined which image was most often sorted into the extroverted (E) pile, sensing (S) pile and thinking (T) pile . referenced with the personality assigned by the researcher. If participants sorted multiple images into the same piles an identical number of times, the resea rcher referenced which image each group chose as the most representative for each dimension 6 . Finally, if a single image still did not emerge as the most representative for a multidimensional pers onality, the researcher used her research and graphic design background to determine the most representative image. The most representati ve image for each MBTI personality included in the current study was then used to create t ts. Survey ability to transfer brand personality. Surveying is a method of gathering data, usually through interviews or questionnaires (Balnaves & Caputi, 2001; Demers , 2005; Fink, 2013). According Although the two terms are often used i nterchangeably, researchers emphasize the distinction between a survey and a questionnaire . While a survey is method of study, a questionnaire is the 6 This question was asked as a follow up after each dimension was sorted during group image sorts. See the up questions.

PAGE 193

193 actual measurement instrument (Babbie, 2010; Demers, 2005). This study gathered data using a self administ ered, online questionnaire created in Qualtrics ® ( www.qualtrics.com ). Advantages of Surveys Specifically, surveying allows researchers to collect information on who people are, how people think and what people do their opinions, actions, beliefs, values, preferences, behaviors, orientations and attitudes (Babbie, 2010; Balnaves & Caputi, 2001; Berger, 1998; Demer, 2005; Fink, 2013; Rubin, Rubin & Piele, 2010). Balnaves and Caputi (2001) also note d that surveys are about and attitudes toward advertising images and brand personality. In particular, surveys are useful for describing a large population (Babbie, 2010). With appropriate sampling and a standardized questionnaire, survey results allow researchers to make s sample, which is discussed in more detail below, was composed of college students, the survey results provide some understanding of how college students a group with considerable discretionary spending power (Marketing Charts, 2013) view personality in images . Additionally, Fink (2013) suggested that surveys were a good methodological choice for researchers to gather information that will guide future studies and programs. As such, the current study planned to use the information gathered through the survey to inform future brand personality studies. The survey results may also help guide practitioners as they build future advertising programs, helping them understand how to imbue their brand with a particular personality.

PAGE 194

194 Choice of Survey Over Experi ment ability to find correlations, or relationships, between variables. While surveys may not show causality like experiments, they are particularly adept at describing the relationship between two or more factors and providing interesting, useful material for advertisers (Berger, 1998; Rubin, Rubin & Piele, 1990). As discussed above, the current study i s particularly interested in the relationship between specific symbol ic visual elements and brand personality, making a survey an appropriate method of inquiry. Additionally, the design of the current study did not fall within the definition of an experiment (Campbell & Stanley, 1963). At most, the current study might be co nsidered a quasi experimental design with a posttest and no control group. Moreover, the current method failed to ns to advertisements, with no concern for pretesting. Furthermore, a survey was also chosen instead of an experiment based on the three purposes of the research explora tion, description and explanation (Babbie, 2010). Exploratory research can be undertaken to familiarize a researcher with a topic or explore a new subject to understanding and satisfy curiosity, 2) to test the feasibility of a more extensive study, and 3) to develop methods for later studies (Babbie, 2010, 92). Babbie also noted that exploratory social science research is particularly valuable when breaking new ground and usually leads to new insights into the research topic. The second purpose of research is description (Babbie, 2010). In the social sciences, descriptive research works to describe situations and events (Babbie, 2010). In descriptive

PAGE 195

195 studies, th e researcher uses methods of scientific observation to carefully observe and describe their observations (Babbie, 2010). Descriptive studies often answer what, where when and how questions. According to Babbie (2010), the third purpose of research is expla nation. In explanatory research, the researcher seeks to explain and answer the question of why (Babbie, 2010). In answering why something happens, researchers often look for causality (Babbie, 2010). In other words, they are looking for variables that cau se change in other variables. Although surveys have some explanatory power (Fink, 2013), researchers conducting explanatory research most often perform experiments (Ba bbie, 2010). However, unlike experimental research designs, the current study sought to e xplore and describe rather than explain for several reasons, including breaking new ground, little prior research to build on, unclear manipulation variables, multidimensional personalities tests and little general knowledge of the transfer process. First, relatively little research has been conducted to explore the relationship between images and brand personality. Even less research has examined images and brands using the MBTI as a personality measure. Thus, because the current study was attempting to br eak new ground in areas of visual research and brand personality research, exploratory survey research offered the mos t useful method of exploration. A second reason the current study chose not to use experiments was that experiments most often build on sp ecific existing knowledge when formulating testable hypotheses. For and Campbell findings. With the dearth of research on images and brand personality, the current study had little

PAGE 196

196 ody of knowledge made more sense than an experiment as a starting place. Third, a survey rather than an experiment was chosen for the current study because experiments require pre existing knowledge of which specific elements to manipulate. While group ima ge sorts provided some idea of the symbolic visual elements that transfer particular multidimensional personalities, they did not provide enough information to know all the variables in an image that may influence brand personality. If the current study we re to conduct an experiment that manipulated some, but not all, of the pertinent symbolic visual elements, the results could be confounded by an unidentified symbolic visual element. Again, more exploratory and descriptive research was required before atte mpting to manipulate images and provide support for causation. Fourth, because the current study sought to explore multidimensional personalities in images, it was unclear whether an experiment could accurately test multiple dimensions simultaneously. Alth ough different symbolic elements might influence different personality dimension, when testing multiple dimensions at once, it would be difficult to know which manipulations affected each personality leading to issues of internal validity. Therefore, if ex periments were to be conducted in the future, they would need to start by manipulating only one dimension at a time (e.g., extraversion (E)/introversion (I)) before attempting to manipulate multidimensional personalities simultaneously. Thus, in order to a nswer the proposed hypothesis concerning multidimensional brand personalities, a survey was the most appropriate method. A fifth reason for conducting a survey rather than an experiment was that the process of transferring multidimensional brand personalit ies through images had not yet been shown to

PAGE 197

197 by exploring the potential for visuals alone to transfer brand personality. Because the theory was being extended, the c urrent study must first test that the process works before beginning to manipulate particular symbolic visual elements in an experimental setting. Therefore, for the reasons discussed above, a survey not an experiment was the most appropria te method fo r the current study Study Product For the final survey, the current study used hand soap as the advertised product for several reasons. First, a hand soap brand was chosen because it was a ubiquitous product, regularly used by most consumers in both public and private settings. According to a national survey of almost 25,000 adults 18 and older, about 86% of U.S. households use liquid hand soap (Experian Information Solution, 2013). Second, hand soap was considered a neutral product category because it fell near the center of the FBC grid. Because no past research had investigated any particular product in association with MBTI personality dimensions, the FCB grid was chosen as a means of evaluating potential products for the current study. The FCB grid was developed by Richard Vaughn to integrate traditional advertising theories and learn feel do hierarchies (Vaughn, 1980). effectiveness, explaining how and why advertising wo rks. The model consisted of a four quadrant grid with two continuums 7 suggesting that different purchase decisions are made based on different criteria. In a study that included 1,800 U.S. consumers a nd 250 product categories, both complexion soap and liquid 7 ns that [require more involvement] and Vaughn, 1980, p. 30).

PAGE 198

198 household cleaner were plotted near the axis of the FBC Grid (see Vaughn, 1986, p. 58 ) . This central placement indicates that some form of soap like hand soap would meet the standards of a neut ral product category for this study, allowing its personality to be influenced by advertising. personality, a product category situated near the center of the FCB Grid ind icated people might involvement continuums to make a purchase decision. The neutrality of soaps (e.g., complexion soap, liquid household cleaners), according to the FCB Gr id (Vaughn, 1986), provided ample room for advertising to successfully transfer and manipulate a variety of viable brand personalities. Without this neutrality, it was possible that the type of purchase decision would be so strongly linked to the product c example, the high think purchase decision associated with life insurance could have conflict with new ground concerning the ability of adverting visuals to create multidimensional brand personalities, it sought to minimize confounding variables like type of purchase decision. Finally, liquid soap dispensers, similar to the hand soap bottles that were used in the current study, had been used in previous brand personality research (Govers & Schoormans, 2005). Therefore, a precedent had been set for using hand soap as a study product. Population According to the National Student Clearinghouse (2013), 19.5 million students are enrolled in U.S. colleges. Not only are college students the next generation of up and coming consumers, they are currently an important and profitable target market brands. In 2013, college students had $117 billion in

PAGE 199

199 discretionary spending power, up significantly from $90 billion in 2009 2011 (Marketing Charts, 2013). Of that money, a significant portion, $7.6 billion, went to pers onal care items like hand soap. Sample The current study used a convenience sample drawn from university students in mass communication courses. The online questionnaire was distributed to course professors, who then passed it on to their students. Professors were encouraged to provide extra credit for students who participated in the survey. According to Bartlett, Kotrlik and Higgins (2001), an appropriate survey sample size can categorical data (e.g., brand per sonality, human personality, gender), the formula for categorical data was used to determine the necessary survey sample size (Bartlett et al., 2001; Cochran, 1977): ( t ) 2 * ( p )( q ) n = --------------------(3 1) ( d ) 2 In the categorical formula shown above, t equals the value for the selected alpha level, (p)(q) equals the estimate of variance, and d equals the acceptable margin of error. As commonly used in research and recommended by Bartlett et al. (2001), the current s tudy used an alpha level of .05. Therefore, t two answer choices, Bartlett et al. (2001) recommended estimating the variance at the highest possible proportion. Therefore, both p and q = .5. Finally, according to Bartlett et al. (2001), the acceptable margin of error for research with categorical data is 5%. Therefore, d = .05. By using the above alpha levels, estimation of variance and margin of error to calculate ) categorical data formula, the current study needed a sample size of 384.

PAGE 200

200 (1.96) 2 * (.5)(.5) n = ------------------------= 384 (3 2) (.05) 2 Further, because the sample size ( n = 384) did not exceed 5% of the population of U.S. college students ( N = 19.5 million) (National Student Clearinghouse, 2013), no correction formula was necessary (Bartlett et al., 2001; Cochran, 1977). Test Advertisements Using the images chosen via the group image sorts, the researc her created a dvertisements ( see Appendix E ) of profession looking advertisement s tailored to each personality. The most representative images chosen by group image sort participants were purch ased from stock photo websites, including Depositp hotos ® ( www. depositphotos.com ), iStock ® ( www. istockphoto.com ) and Shutterstock ® ( www. shutterstock .com ) , to represent each multidimensional personality. Purchasing stock photos allowed the author to use professional quality images f or minimal cost while providing the appropriate copyright licensing to alter an d republish the advertisements. The test advertisements were designed as 5.5 x 7 inch, full color magazine ads in Adobe Photoshop CS5. This size allowed the advertisements to di splay properly in the online questionnaire while maintaining the look, feel and proportion of a full bleed magazine ad. Each image was placed in the background and sized to create a full bleed ad. For consistency, the product logo and top third of the prod uct bottle was placed in the foreground on the bottom right corner of all eight advertisements. Because the current study was interested only in the ability of advertising visuals to transfer brand personality, no additional text or copy was added to the a dvertisements. To ensure that the test advertisements had high external validity, two experts

PAGE 201

201 with professional advertising and graphic design backgrounds (not including the researcher) verified that the advertisements look like real ads. Rather than givin g the hand soap brand a specific name, a generic looking logo that read of a generic hand soap pump bottle (see Appendix F). The hand soap logo was created as a vector image in Adobe Illustrator CS5 and placed on the advertisements in Adobe Photoshop CS5. Questionnaire Pilot Test Scholars underscore the importance of pretesting a questionnaire prior to conducting the actual survey (Fink, 2013). According to Fink researcher to produce a high quality, usable questionnaire that provides the necessary research data (p. 7). Because self administered questionnaires must be clearly worded to be accurate, it is important to test whether participants understand the directions and can answer the questions (Fink, 2013). Further, Fink (2013) asserted that pilot testing will help a survey run smoothly. Therefore, the current study pilot tested the questionnaire prior to conduc ting the final survey. Fink (2013) gave several guidelines for conducting a pilot test. First, Fink (2013) suggested informally testing portions of the questionnaire (e.g., the directions, the scale). In line nformally tested portions of the questionnaire by sharing them with a committee who is experienced at conducting surveys and incorporated their suggestions. Fink (2013) also recommended anticipating the actual circumstances for the survey and mimicking the m for the pilot test. For the current study, this meant testing the survey online using the sam e software as the final survey.

PAGE 202

202 Further, without wasting resources (p. 74). Determini ng a reasonable number of pilot test respondents requires considering the length of the questionnaire (longer questionnaires need more test respondents ) and the number of improvements being made (more suggested improvements requires more test respondents ). composed of repeating a single scale, its length may not accurately reflect the number of pilot test respondents necessary. Therefore, for the current study, the number of pilot test respondents was guided by the amount of improvements required. The questionnaire was pilot tested until few improvements were necessary. results to see if respondents failed to answer questions, gave the same answer to several questions, or wrote notes in the margins or, in the case of an online questionnaire, in the comment box provided. To fulfill this guideline for the current study, the researcher reviewed the pilot test data for unanswered questions, repeated answers or notes. Questionnaire A questionnaire was created using the BHPI and researcher created advertisements (see Appendix G). The electronic questionnaire was created using online software called Qualtrics ® ( www.qualtrics.com ). Course professors at a large southeastern university then distributed the questionnaire link to their students through course management software. At the discretion of each professor, students received extra credit for thei r participation in the survey. After receiving consent, the questionnaire first asked students to indicate their own personality using the BHPI. This questionnaire used the BHPI rather than the human MBTI scale for the same reasons as the group image sorts, which included its short length, quick response

PAGE 203

203 Next, respondents The name and product packaging have been replaced by a generic logo and bottle so the encouraged respondents to focus solely on each advertising image and ignore the personality char acteristics of the product name, logo and packaging, while still maintaining the look and fe el of an actual advertisement. Then respondents saw a random hand soap advertisements that visually symbolized one of the eight brand personality types and were ask ed to use the BHPI to randomized using the Qualtrics ® survey logic to prevent the order in which the advertisements were presented from influ ses. Finally, respondents were asked to provide some basic demographic information about themselves, including gender, age, race/ethnicity, year in school and major. Respondents were me if they were receiving extra credit. However, in order to maintain confidentiality, the information respondents provide d to obtain extra credit was not linked to their survey information during analysis. Statistical Analyses The current study sought to investigate the effect of one categorical independent variable (advertising images) on a several dependent variables one categorical (brand personality) and three continuous (attitude toward the ad, attitude toward the brand and purchase intent). A serie s of differences between percentage tests were run to test H1 (a h), which hypothesized that within each advertisement, the highest percentage of respondents would assign the advertisement the intended personality , and H2 (a h), which hypothesized that the highest percentage of respondents would assign each advertisement the intended personality when compared to the other advertisements tested . A dditionally, a series of chi square tests were run for each

PAGE 204

204 confounding variabl e , human personality and gender, t o determine whether they influenced the study results. If the chi squares return non significant results (p > .05), then the confounding variables were determined not to have influence the results. Alternately, RQ3 (a b), RQ4 (a b) and RQ5 (a b), which exa mine how MBTI brand personality influences attitude toward the ad, attitude toward the brand and purchase intent, were answered using a series of three way analysis of variances (ANOVAs). Each three way ANOVA ran the dependent variable (i.e., attitude towa rd th e ad, attitude toward the brand or purchase intent) against the independent variable (i.e., brand personality) and both confounding variables (i.e., gender, human personality). These statistical tests allowed the researcher to examine whether the conf ounding variables influenced the dependent variable. Chi squares and differences between percentages tests were conducted using an online calculator created by the Emory University ( www.openepi.com ). All other statis tical analyses were performed using SPSS ® version 21.

PAGE 205

205 CHAPTER 4 RESULTS The current study used a mixed ability to transfer and manipulate multidimensional brand personalities. This chapter is broken into three sections. The first section presents the BHPI scale reduction process and results. Next, the second section details the qualitative group image sort results. Finally, the third section provides result s from the quantitative survey. Phase One S cale Reduction specific MBTI scale. To create the reduced scale, pretest data was analyzed with the intent to extrave rsion/introversion (E/I), sensing/intuition (S/N), thinking/feeling (T/F) and judging/perceiving (J/P). Before reducing the scale, the researcher outlined multiple criteria to ensure that the new scale fit the research objectives, including reducing the nu mber of scale items, avoiding equal personality results, correlating above a .6, staying within the original alpha range, limiting evaluative adjectives and considering judgmental item qualities. First, to make the scale more suitable for multiple uses in a single survey, each subscale should be reduced to fewer than six items. Beyond use in survey instruments, reducing the length of the overall scale by including fewer items per subscale would also made the scale easier and shorter to administer in a varie ty of other research settings (e.g., focus groups, screening instruments). However, because the four subscales measure unique dimensions, each needed to maintain a minimum of three items (Cook, 1981 ; Hinkin, 1995; Raubenheimer, 2004 ). Second, each reduced subscale should contain an odd number of items to avoid an equal personality result. Because the original subscales had six items each, it was possible for

PAGE 206

206 respondents to select three items that corresponded to one dimension pole and three items that corre sponded to the opposite dimension pole 1 . For example, within the extraversion/introversion (E/I) subscale, a respondent might select three items that corresponded to extraversion and three items that corresponded to introversion. When calculated, each pole canceled the other out, leaving an equal personality result. By reducing each subscale to an odd number of items, there was no potential for a tie, or an equal personality result. Third, each subscale item should have a minimum of a .6 correlation with th e overall dimension score. Although past literature suggested that correlations of .5 to 1 indicated a strong relationship (e.g., Pallant, 2010), the researcher sought to achieve items with at least a .6 correlation to ensure a strong item to scale relatio nship that exceeded the minimum scholarly requirement. range (see T able 4 tests of the origin al 24 item scale, which was composed of four, six item subscales. Her results produced a range of alpha values, which were reported in the original study (Strausbaugh, 1998). In the reduced scale, each subscale should fall within the ranges reported by Str ausbaugh (1998). Table 4 1. Original Alpha Range for Six Item Subscales . Subscale Strausbaugh (1998) 6 Extraversion/Introversion (E/I) .737 .862 Sensing/Intuition (S/N) .613 .777 Thinking/Feeling (T/F) .531 .802 Judging/Perceiving (J/P) . 540 .772 1 For example, in the pretest data used to reduce the scale, a no personality result (a score of zero) was recorded for 11.3% ( n = 103) of E/I responses, 16.1% ( n = 147) of S/N responses, 11.7% ( n = 107) of T/F responses and 13.7% ( n = 125) of J/P responses. These numbers represent a sizable portion of the total responses and further support the need for the new, reduced scale to avoid no personality results.

PAGE 207

207 Fifth, the reduced scale should limit the number of evaluative adjectives included. One of that no one personality is better or worse than another (Tieger & Barron Tieger, 1995; 2000). Therefore, the reduced scale should pay close attention to the connotations associated with each item, seeking to eliminate pairs that might lead respondents to feel that there is a right/wrong or better/worse answer. While no words are wholly without connotation, the reduced scale should keep tone, associations and inference in mind when selecting subscale items. Finally, the reduced scale should include judgmental item qualities , rather than relying entirely on intern al consistency considerations (Stanton et al., 2002). According to Stanton et al. (2002), relying solely on internal consistency could lead to an overly narrow, redundant interpretation of the scale. Therefore, Stanton et al. (2002) suggested that when red ucing a scale, researchers should consider criteria beyond statistical relationships, or what they termed judgmental item qualities . According to the authors, examples of judgmental item qualities could respondent populations, semantic redundancy with other items, perceived invasiveness, and face validity (Stanton et al., 2002, p. 173). Although scholars do not discuss judgmental item qualities as often as internal consistency measures, the authors noted that they were no less important. Stanton et al. (2002) proposed that, when reducing a scale, researchers should rely on their own expertise to help determine which items should be retained. Further, Stanton et al. (2002) concluded that, at times, interna l thoughtful compromises when choosing items to retain in a reduced Computing Overall Dimension Scores Before any statistical tests were run , the pretest data was cleaned and recoded so that all extraverted, sensing, thinking and judging data was coded as a 1 and all introverted, intuitive,

PAGE 208

208 feeling and perceiving data was coded as a 1. During the pretest, each respondent ( n = 304) completed S n = 912) once to indicate their own Next, four overall dimension scores were computed for each respondent by adding the s ix scale items: Overall E/I Score = assertive/mild + active/reserved + dynamic/moderate + sociable/shy + outspoken/quiet + energetic/calm Overall S/N Score = systematic/imaginative + realistic, down to earth/idealistic, visionary + practical, functional/cr eative, theoretical + sensible, factual/instinctual + narrow focused/wide interests + conservative/unconventional Overall T/F Score = logical/emotional + firm/soft hearted + cold/warm + indifferent/sympathetic + rational, reasonable/passionate, perceptive + stoic/excitable Overall J/P Score = clear cut, definite/undecided, variable + deliberate/adaptable + dependable/changeable + steadfast/innovative + decided/flexible + well defined/indistinct Each overall dimension scores could range from 6 to 6, indica ting the strength of a particular personality dimension on a two pole continuum (see Figure 4 1). For example, an Overall E/I Score of 6 indicated complete extraversion, while an Overall E/I Score of 6 indicated complete introversion. Alternately, an Over all E/I Score of 0 indicated neutrality neither extraversion nor introversion. Therefore, the closer an Overall E/I Score was to 6, the more strongly extraverted the person or brand, and the closer an Overall E/I Score was to 6, the more strongly intro v erted the person or the brand.

PAGE 209

209 Figure 4 1. Continuums for overall dimension scores. Point Biserial Correlations To examine the scale data before reducing it, point biserial correlations were run for each six item subscale, which corresponded to the four MBTI dimensions extraversion/introversion (E/I), sensing/thinking (S/N), thinking/feeling (T/F) and judging/perceiving (J/P). Computing overall dimension scores for each of the four MBTI dimension created a continuous variable to use in point biserial co rrelations. In each point biserial correlation, the appropriate overall original scale to see which items had the highest correlation. For example, for the extraver sion/introversion dimension, the Overall E/I Score was correlated with assertive/mild, active/reserved, dynamic/moderate, sociable/shy, outspoken/quiet and energetic/calm . The results for each dimension are presented in the sections below. s To further examine the internal reliability of the original six Alpha was also run for each MBTI dimension (see Table 4 2). Alphas for the four subscales

PAGE 210

210 l alpha averages , item subscales extraversion/introversion (E/I), sensing/intuition (S/N) and judging/perceiving (J/P) while the remaining 6 item subscale thinking/feeling (T/F) w as higher (see Table 4 2). However, all four 6 item subscales from the current study had alpha values within the range reported by Strausbaugh (1998) (see Table 4 2). Table 4 Item Scales . Subscale Current St udy 6 1 Strausbaugh (1998) 6 average Strausbaugh (1998) 6 Extraversion/Introversion (E/I) .808 .812 .737 .862 Sensing/Intuition (S/N) .646 .709 .613 .777 Thinking/Feeling (T/F) .747 .698 .531 .802 Judging/Perceiving (J/P) .639 .667 .540 .772 Note: 1 n = 912 identical 6 studies, showed a wide range of al pha values. These differences might demonstrate the need for Reducing the Scale To reduce the scale, the criteria discussed in detail above were applied to each of the four su bscales. The reduction criteria included: reducing the number of scale items, avoiding no personality results, correlating above a .6, staying within the original alpha range, limiting evaluative adjectives and conside ring judgmental item qualities. The fi rst criterion called for a reduction in the number of scale items. Therefore, each new subscale needed to have fewer than six items but at least three items (Cook, 1981 ; Hinkin, 1995; Raubenheimer, 2004 ). Additionally, based on the second criterion, the ne w subscale also

PAGE 211

211 needed to avoid a no personality result. Thus, the new subscale needed to have an odd number of items, preventing a personality tie. Combining the first and second criteria meant each new subscale needed to have either three or five items. significantly shorten the overall scale to make it more manageable for administration in a variety of research settings (e.g., surveys, focus groups, screening instruments), reducin g each subscale by only one item (from six to five) was not satisfactory. Therefore, each subscale needed to be reduced to three items rather than five items. By reducing each of the four dimensions to a three item subscale, the researcher could produce a significantly shorter, 12 item scale. Once the new subscale length was determined by the first two criteria, the remaining criteria three through six were applied individually to each subscale dimension extraversion/introversion (E/I), sensing/intuit ion (S/N), thinking/feeling (T/F) and judging/perceiving (J/P). Results for each subscale are presented below. Reducing the e xtraversion/ i ntroversion (E/I) s cale To reduce the first subscale, extraversion/introversion (E/I), the researcher applied the rema ining predetermined criteria, including correlating above a .6, staying within the original alpha range, limiting evaluative adjectives and considering judgmental item qualities. The application of each criterion to the extraversion/introversion (E/I) subs cale is discussed below. above a .6. Therefore, using a point biserial correlation, the Overall E/I Score was correlated with the individual scale items (see Table 4 3). According to the correlation results, each of the six original scale items had a correlation score of more than .6, ranging from .621 .800. Based on the correlation results, the top three extraversion/introversion (E/I) subscale items were outspoken/q uiet, active/reserved and sociable/shy.

PAGE 212

212 Table 4 3. Extraversion/Introversion (E/I) Point Biserial Correlation Results . Item Correlation with Overall E/I Score 1 Outspoken/Quiet .800** 2 Active/Reserved .746** 3 Sociable/Shy .725** 4 Energetic/Calm .7 20** 5 Dynamic/Moderate .675** 6 Assertive/Mild .621** Note: n = 912 ** p < .001 Strausb a the alpha if item deleted was calculated for each of the original six extraversion/introversion (E/I) subscale items (see Table 4 4 ). Using the C best three extraversion/introversion (E/I) subs cale items to retain were outspoken/quiet, active/reserved and energetic/calm. Table 4 . Item 1 Assertive/Mild .806 2 Dynamic/Moderate .791 3 Energetic/Calm .774 4 Sociable/Shy .773 5 Active/Reserved .768 6 Outspoken/Quiet .750 Note: n = 912 Based on the bi o subscale items outspoken/quiet and active/reserved were identical. However, the third item differed, with the point biserial correlation suggestin g the use of sociable/shy and C values suggesting the use of energetic/calm. However, in both cases, the results for each item were very close. The bi serial showed a difference of only .005, and the C showed a difference of only .001. Therefore, the researcher used the remaining two predetermined criteria to determine which ite m was most appropriate for the reduced scale.

PAGE 213

213 The fifth criterion suggested that the number of evaluative adjectives included in the subscale should be limited. In the case of the extraversion/introversion (E/I) subscale dimension, none of the top four ite ms outspoken/quiet, active/reserved, sociable/shy and energetic/calm had strong positive or negative connotative value. Therefore, all four items could serve as appropriate subscale items. The sixth and final criterion recommended the use of judgmental item qualities that relied be retained. While the first and second scale items outspoken/quiet and active/reserved were easily determined based on point biser ial correlation and C the third item required the researcher to em ploy judgmental item qualities. Because the extraversion/introversion (E/I) dimension measures where energy is gathered and directed, the researcher determ ined that the scale item sociable/shy was a better indicator than energetic/calm. Extraverts (E) relate easily to the outer world and gain energy from other people. Often, extraverts also need to think their ideas aloud. Combined, these traits indicated th at extraverts could often be perceived as sociable. However, extraverts could enjoy being around others and talking without being overly energetic. They may prefer to sit calmly among a large group of friends discussing or debating a variety of subjects. C onversely, introverts (I) relate easily to the inner world of ideas and draw energy from within themselves. They enjoy spending time alone, and prefer to quietly think fully through an idea before sharing it aloud. Together, these traits suggest that intro verts could often be perceived as shy, content to sit quietly and think. However, introverted brands and people do not have to be calm; they may be energetic, directing their energy at pursuing their own ideas and passions. Therefore, the

PAGE 214

214 adjectives sociab le and shy were determined to be better descriptors of the extraversion/introversion (E/I) dimension than energetic and calm . Using the three items determined above outspoken/quiet, active/reserved and sociable/shy the researcher then returned to crite rion four to C .758) is lower than the 6 item scale ( .862) . Thus, the new three item scale fulfills criterion four. Based on the application of the predetermined criteria, the curr ent study was able to specific MBTI scale from six items to three items. The new three item scale is composed of the items: outspoken/quiet, active/reserved and sociable/shy . Reducing the s ensing/ intuition (S/N) s cale To complete the reduction of the second subscale, sensing/intuition (S/N), the researcher applied the remaining third through sixth predetermined criteria. The application of each criterion to the sensing/intuiti on (S/N) subscale is discussed below. above a .6. Therefore, using a point biserial correlation, the Overall S/N Score was correlated with the individual scale items (see Table 4 5). According to the correlation results, two of the six original scale items systematic/imaginative and practical, functional/creative, theoretical had a correlation of more than .6. A third scale item conservative/unconventional als o neared .6 with a correlation value of .593. Therefore, based on the correlation results, the top three sensing/intuition (S/N) subscale items were systematic/imaginative; practical, functional/creative, theoretical; and conservative/unconventional.

PAGE 215

215 Table 4 5. Sensing/Intuition (S/N) Point Biserial Correlation Results . Item Correlation with Overall S/N Score 1 Systematic/Imaginative .709** 2 Practical, Functional/Creative, Theoretical .707** 3 Conservative/Unconventional .593** 4 Narrow Focused/Wide Interests .567** 5 Realistic, Down to Earth/Idealistic, Visionary .523** 6 Sensible, Factual/Instinctual .507** Note: n = 912 ** p < .001 Strausb a if item deleted were calculated for each of the original six sensing/intuition (S/N) subscale items (see Table 4 6 ). Using the C he best three sensing/intuition (S/N) subscale items to retain were systematic/imaginative; practical, functional/creative, theoretical; and conservative/unconventional. Table 4 . Item 1 Sensible, Factual/Instinctual .647 2 Realistic, Down to Earth/Idealistic, Visionary .641 3 Narrow Focused/Wide Interests .620 4 Conservative/Unconventional .600 5 Practical, Functional/Creative, Theoretical .548 6 Systematic/Imaginative .546 Note : n = 912 Based on the bi sensing/intuition (S/N) subscale items systematic/imaginative; practical, functional/creative, theoretical; and conservative/unconventional were identical. T he fifth criterion suggested that the number of evaluative adjectives included in the subscale should be limited. In the case of the sensing/intuition (S/N) subscale dimension, none of the top three items systematic/imaginative; practical, functional/cre ative, theoretical; and

PAGE 216

216 conservative/unconventional had strong positive or negative connotative value. Therefore, each of the three items could serve as appropriate subscale items. The sixth and final criterion recommended the use of judgmental item qual ities that relied be retained. Because all three scale items were easily determined based on po int biserial correlations and C r needed only to confirm that the three scale items had face validity. The sensing/intuition (S/N) dimension measures how information is taken in. Sensors (S) prefer to work with known facts and concentrate on concrete information from their five senses. T hey appreciate simple, matter of fact reality. Sensors also accurately remember details and excel at here and now practical matters. Therefore, the adjectives contained in the top three scale items systematic, practical/functional and conservative accu rately describe a sensing personality. Alternately, intuitives (N) prefer to look for possibilities and relationships. They interpret and add meaning to information, looking at big picture patterns. Intuitives also value big ideas and see obscure connectio ns others may miss, and they tend to be creative and good at imagining long term possibilities for the future. Therefore, the adjectives contained in the top three scale items imaginative, creative/theoretical and unconventional accurately describe an intuitive personality. Using the three items determined above systematic/imaginative; practical, functional/creative, theoretical; and conservative/unconventional the researcher then returned to criterion four to calculate the subs ha. Although the 3 alpha value ( .624) is less than the 6 item scale ( , it still falls well within

PAGE 217

217 .777) . Thus, the new three item scale fulfills criterion four. Based on the appl specific MBTI scale from six items to three items. The new three item scale is composed of the it ems: systematic/imaginative; practical, functional/creative, theoretical; and conservative/unconventional . Reducing the t hinking/ f eeling (T/F) s cale As with the first two subscales, to complete the reduction of the third subscale, thinking/feeling (T/F), t he researcher applied the remaining third through sixth predetermined criteria, which were discussed in detail above. The application of each criterion to the thinking/feeling (T/F) subscale is discussed below. correlation with the overall scale should be above a .6. Therefore, using a point biserial correlation, the Overall T/F Score was correlated with the individual scale items (see Table 4 7). According to the correlation results, five of the six original sc ale items had a correlation of more than .6. Based on the correlation results, the top three thinking/feeling (T/F) subscale items were firm/soft hearted; rational, reasonable/passionate, perceptive; and cold/warm. Table 4 7. Thinking/Feeling (T/F) Point B iserial Correlation Results . Item Correlation with Overall T/F Score 1 Firm/Soft Hearted .713** 2 Rational, Reasonable/Passionate, Perceptive .706** 3 Cold/Warm .689** 4 Logical/Emotional .686** 5 Indifferent/Sympathetic .650** 6 Stoic/Excitable .5 33** Note: n = 912 ** p < .001

PAGE 218

218 Strausb a if item deleted were calculated for each o f the original six thinking/feeling (T/F) subscale items (see Table 4 8 ). Using the C three thinking/feeling (T/F) subscale items to retain were firm/soft hearted; rational, reasonable/pass ionate, perceptive; and cold/warm. Table 4 . Item 1 Stoic/Excitable .748 2 Indifferent/Sympathetic .717 3 Logical/Emotional .705 4 Cold/Warm .700 5 Rational, Reasonable/Passionate, Perceptive .697 6 Firm/Soft Hearted .692 Note: n = 912 Based on the bi thinking/feeling (T/F) subscale items firm/soft hearted; rational, reasonable/passionate, perceptive; and cold/ warm were identical. The fifth criterion suggested that the number of evaluative adjectives included in the subscale should be limited. In the case of the thinking/feeling (T/F) subscale dimension, neither of the top two items firm/soft hearted and ra tional, reasonable/passionate, perceptive had strong positive or negative connotative value. However, the third item cold/warm did imply a connotative value. When applied to either a brand or human personality, the terms cold can indicate a detachedn ess or lack of emotion. According to Merriam ( www.merriam webster.com normal human emotion, friendliness, or compass guided by logic than emotions, it does not mean that it does not possess normal human emotions.

PAGE 219

219 Further, applying the term cold to a personality has a decidedly negative connotation, putting the term at odds w that no personality is better or worse than another, only different (Tiege r & Barron Tieger, 1995; 2000). Therefore, using point selected. Logic al/emotional was the fourth item according to both point alpha results. In both cases, logical/emotional scored only slightly lower than the third item, cold/warm. The point biserial correlation result for logical/emotional was only .003 less than that than that of cold/warm. Further, when applied to personality, the terms logical/emotional do not carry the negative/positive connotations f ound in cold/warm. The sixth and final criterion recommended the use of judgmental item qualities that relied be retained. While the first and second scale items firm/soft hearted and rational, reasonable/passionate, perceptive were easily determined based on p oint biserial correlation and C ploy judgmental item qualities. Because the thinking/feeling (T/F) dimension measures how decisions are made, the researcher determined that the scale item logical/emotional was a better indicator than cold/warm. Thinkers (T) base judgments on impersonal analysis and logic, putting more weight on o bjective principles and impersonal facts. Thinking is closely tied to intellect, and thinkers are more driven to be fair and consistent. Thinkers also tend to take things less personally, are harder to offend and are uncomfortable with emotions. Therefore, the adjective logical is a better descriptor of the thinking dimension than cold . While cold implies a negative, calculating

PAGE 220

220 personality, logical hints at a more analytical, objective personality, aligning more closely with the intended MBTI personality d imension. Alternately, feelers (F) base their judgments on personal values, placing more emphasis on people and personal concerns. They place a high premium on the development and maintenance of personal relationships. Feelers are also guided by their pers onal values and extenuating circumstances. Often wearing their heart on their sleeves, they are comfortable with emotions and may get their feelings hurt more easily. Therefore, the adjective emotional better describes the feeling dimension by evoking a pe rsonality who makes decisions based on emotions and better aligns with the intended MBTI personality dimension than warm . Additionally, when taken more literally, the terms cold and warm refer to temperatures. This alternate meaning might cause confusion f or participants using the scale and provide d further support for replacing the scale item cold/warm with logical/emotional. Therefore, using the three items determined above firm/soft hearted and rational, reasonable/passionate, perceptive; and logical/e motional the researcher then returned to criterion fou value ( .679) is less than the 6 item scale ( (1998) alpha range ( .802) . Thus, the new three item scale fulfills criterion four. teria, the current study specific MBTI scale from six items to three items. The new three item scale is composed of the items: firm/soft hearted and rational, reasonable/p assionate, per ceptive; and logical/emotional.

PAGE 221

221 Reducing the judging/p erceiving (J/P) s cale Finally, to complete the reduction of the fourth subscale, judging/perceiving (J/P), the researcher applied the remaining third through sixth predetermined criteria. The application of each criterion to the judging/perceiving (J/P) subscale is discussed below. above a .6. Therefore, using a point biserial correlation, the Overall J/P Score was correlated with the individual scale items (see Table 4 9). According to the correlation results, three of the six original scale items decided/flexible; deliberate/adaptable; clear cut, definite/undecided, variable had a correlation of more than .6. Therefore, based on the correlation results, the top three judging/perceiving (J/P) subscale items were decided/flexible; deliberate/adaptable; clear cut, definite/undecided, variable. Table 4 9. Judging/Perceiving (J/P) Point Biserial Correl ation Results . Item Correlation with Overall J/P Score 1 Decided/Flexible .681** 2 Deliberate/Adaptable .664** 3 Clear Cut, Definite/Undecided, Variable .657** 4 Steadfast/Innovative .559** 5 Dependable/Changeable .513** 6 Well defined/Indistinct .503** Note: n = 912 ** p < .001 Strausb a if item deleted were calculated for eac h of the original six judging/perceiving (J/P) subscale items (see Table 4 10 ). Using the C best three judging/perceiving (J/P) subscale items to retain were decided/flexible; deliberate/adaptab le; clear cut , definite/undecided, variable.

PAGE 222

222 Table 4 . Item 1 Dependable/Changeable .635 2 Well defined/Indistinct .630 3 Steadfast/Innovative .618 4 Clear Cut, Definite/Undecid ed, Variable .568 5 Deliberate/Adaptable .561 6 Decided/Flexible .552 Note: n = 912 Based on the bi judging/perceiving (J/P) subscale items decided/flexible; deliberate/adaptable; c lear cut, definite/undecided, variable were identical. The fifth criterion suggested that the number of evaluative adjectives included in the subscale should be limited. In the case of the judging/perceiving (J/P) subscale dimension, none of the top thr ee items decided/flexible; deliberate/adaptable; clear cut, definite/undecided, variable had strong positive or negative connotative value. Therefore, each of the three items could serve as appropriate subscale items. The sixth and final criterion reco mmended the use of judgmental item qualities that relied be retained. Because all three scale items were easily determined based on point biserial correlation s and C scale items had face validity. The judging/perceiving (J/P) dimension measures the approach to structure and type of life adopted. Judgers (J) prefer structured, ordered environ ments and like to make decisions. relax until deadlines are met, often demonstrating a work before play attitude. Therefore, the adjectives contained in t he top three scale items decided, deliberate and clear cut/definite

PAGE 223

223 accurately describe a judging personality. Alternately, perceivers (P) would rather stay open and adaptable to new ideas and information. They are spontaneous and like to keep plans op en may bend rules they deem unnecessary and see deadlines as suggestions, preferring a play before work approach. Therefore, the adjectives contained in the top three scale items flexible, adaptable and undecided/variable accurately describe a perceiving personality. Using the three items determined above decided/flexible; deliberate/adaptable; clear cut, definite/undecided, variable the researcher then r eturned to criterion four to calculate the subs .641) is greater than the 6 item scale ( .772) . Thus, the new three item scale fulfills criterion four. criteria, the current study specific MBTI scale from six items to three items. The new three item scale is composed of the items: decided/flexible; deliberate/adaptable; clear cut , definite/undecided, variable. The r educed s cale By combining the three scale items determined above for each of the four subscales, the researcher created a condensed 12 item scale (see Table 4 11). The new, reduced scale will be referred to as the Brand and Human Personality Indicator (BHPI).

PAGE 224

224 Table 4 11. Brand and Human Personality Indicator (BHPI) . Extravers ion/ Introversion (E/I) Sensing/ Intuition (S/N) Thinking/ Feeling (T/F) Judging/ Perceiving (J/P) 1 Outspoken/ Quiet Systematic/ Ima ginative Firm/ Soft Hearted Decided/ Flexible 2 Active/ Reserved Practical, Functional/ Creative, Theoretical Rational, Reasonable/ Passionate, Perceptive Deliberate/ Adaptable 3 Sociable/ Shy Conservative/ Unconventional Logical/ Emotional Clear Cut, D efinite/ Undecided, Variable Phase Two Group Image Sorts After completing the scale reduction, the second phase of this study was to conduct qualitative group image sorts. In the group image sorts, the researcher asked three participants to sort a dec k of images based on MBTI personality dimensions. During each group image sort, participants sorted the same image deck three times in random order, once for each of the three MBTI dimensions included in the current study. For example, participants were ha nded the image deck and asked to sort the images into categories based on a single MBTI dimension (e.g ., extraverted/introverted ). After the first sort was completed, participants were handed a second, identical image deck and asked to sort on a second MBT I dimension (e.g., sensing/ intuition ). Finally, after the second sort was completed, participants were handed a third identical image deck and asked to sort on a third MBTI dimension (e.g., thinking/feeling ). After conducting the group image sorts, the res earcher then analyzed the transcripts to presented below, followed by an analysis organized by research question. Group Image Sort Participants Prior to participat ing in the group image sorts, potential participants were asked to complete a short, online screener questionnaire. This questionnaire ensured that potential participants fit the study criteria (i.e., advanced adve rtising or graphic design major, lived in the

PAGE 225

225 U.S. for more than 18 years). Additionally, the screener questionnaire gathered potential participants PI) and basic demographic information. A series of descriptive statistics were run to examine group image sort partici overall demographic breakdown (see Table 4 12 ). The majority of participants were non Hispanic white females who were seniors in college. Participants were an average of 22 years old ( M = 21.97, SD = 2.77). The participants consisted of slightly mor e graphic design majors, and the most common three dimensional MBTI personality type was ENF. Table 4 12 . Overall Demographics for Group Image Sort Participants . Variable n % Gender 1 Female 22 73.3 Male 8 26.7 Major 1 Graphic Design 17 56.7 Advertising 13 43.3 Race 1 White (Non Hispanic) 21 70.0 Hispanic/Latino/Latina 4 13.3 Asian 3 10.0 African American 1 3.3 Other 1 3.3 Year in School 1 Senior 18 60.0 Junior 11 36.7 Master 1 3.3 Personality Type 1 ENF 1 1 3 6 . 7 ENT 5 1 6 . 7 INF 4 13.3 IST 3 10.0 ESF 2 6.7 EST 2 6 . 7 INT 2 6 . 7 ISF 1 3.3 Note: 1 n = 30

PAGE 226

226 Additionally, Table 4 13 provides individual personality and demographic info rmation for the group image sort participants. Reporting individual data provides details about both the individual members and the compos ition of each group image sort. However, a lthough the current study reports this individual demographic information, t he researcher found no discernable patterns in the data based on major, gender, age, race, year in school, or personality. Instead, themes emerged based on overall group discussion rather than specific demographic information. Therefore, because the data w as influenced by group and individual comments, participants were labeled based on their group number rather than demographic information. Participants are labeled first by the number of their group image sort, followed by their unique number within that g roup. For example, Participant 4 1 would be the first participant in the fourth group and P articipant 7 3 would be the third participant in the seventh group.

PAGE 227

227 Table 4 13 . Individual Demographics for Group Image Sort Participants . Participant # Group # P articipant ID Major Gender Age Race Grade Personality 1 1 1 1 Advertising Male 21 /Latina Senior ESF 2 1 1 2 Advertising Male 24 White (Non Hispanic) Senior ENF 3 1 1 3 Advertising Female 20 White (Non Hispanic) Junior ENF 4 2 2 1 Advertising Male 21 White (Non Hispanic) Senior EST 5 2 2 2 Advertising Male 21 White (Non Hispanic) Senior ENF 6 2 2 3 Advertising Male 26 White (Non Hispanic) Senior ENT 7 3 3 1 Advertising Female 24 White (Non Hispanic) Senior ENF 8 3 3 2 Advertising Female 22 White (Non Hispanic) Senior ENT 9 3 3 3 Advertising Female 21 White (Non H ispanic) Senior ENF 10 4 4 1 Graphic Design Female 35 White (Non Hispanic) Master ENF 11 4 4 2 Advertising Female 21 White (Non Hispanic) Junior ENT 12 4 4 3 Graphic Design Female 21 Asian Junior ENF 13 5 5 1 Graphic Design Female 21 Multi racial Junio r ENF 14 5 5 2 Graphic Design Male 20 White (Non Hispanic) Junior ENT 15 5 5 3 Graphic Design Male 22 /Latina Junior IST 16 6 6 1 Graphic Design Female 22 White (Non Hispanic) Senior ISF 17 6 6 2 Graphic Design Female 21 White (Non His panic) Junior ENF 18 6 6 3 Graphic Design Female 20 White (Non Hispanic) Junior INF 19 7 7 1 Graphic Design Female 22 /Latina Senior ENF 20 7 7 2 Graphic Design Female 22 White (Non Hispanic) Senior IST 21 7 7 3 Graphic Design Female 2 1 White (Non Hispanic) Senior ENF 22 8 8 1 Graphic Design Male 21 Asian Senior IST 23 8 8 2 Graphic Design Female 22 White (Non Hispanic) Senior INF 24 8 8 3 Graphic Design Female 22 /Latina Senior INF 25 9 9 1 Graphic Design Female 21 White (Non Hispanic) Junior INT 26 9 9 2 Graphic Design Female 21 White (Non Hispanic) Junior INF 27 9 9 3 Graphic Design Female 21 White (Non Hispanic) Junior INT 28 10 10 1 Advertising Female 20 Asian Senior EST 29 10 10 2 Advertising Female 22 Afri can American Senior ESF 30 10 10 3 Advertising Female 21 White (Non Hispanic) Senior ENT Note: n = 30

PAGE 228

228 Research Question 1 Visual Symbolic Elements for Single Dimension MBTI Brand Personalities Research Question 1 (RQ1) asked about the specific visual symbolic elements associated with each of the three MBTI dimensions included in this study extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). Because group image sort participants sorted the image deck by individual MBT I dimensions, the researcher answered RQ1 by analyzing the data from each separate dimensional sort. For example, to understand the visual symbolic elements associated with extraversion (E), the researcher examined both the images participants sorted in th e extraverted (E) pile and their verbal descriptions of why they deemed those images extraverted (E). Both common themes and less frequently mentioned themes were identified for each of the eight individual dimensions included in this study (see Table 4 14 ). A detailed description of the themes is presented for each individual dimension below. Table 4 14 . Themes for Individual Brand Personalities in Images . Dimension Common Themes Less Frequently Mentioned Themes Extraversion (E) Happiness /fun Acti veness (motion, dynamism) Choice To participant in activities To be social To be in public Collaboration Multiple people Body language Smiling Talking/communicating Laughing Eye contact Color ( bright, loud, lively, vibrant) Appe aring to feel comfortable Attire (trendy, dressy, cute, revealing) Subjects in c lose proximity to each other Dog s hown Tight cropping Introversion (I) Choosing isolation Quietness (peaceful, lack of noise/talking) Immersion In thought In activit y Relax ation Wide cropping

PAGE 229

229 Table 4 14 . Continued . Dimension Common Themes Less Frequently Mentioned Themes Introversion (I) (continued) Professions Doctors Scientists Farmers Setting Home Nature Isolation Number o f people Single person Two people Intimate couples Single person with animal No eye contact With the viewer With other subjects Color /light ing (muted, clean , low ) Sensing (S) Goal orient ation Single goal Task End result Solve a problem Short term focus Practicality/functionality Systematic process Working Science related occupation Medical related occupation Using one or more of the five senses Organization Conservative dress Color (white, clean , neutral, sterile) Technology shown Intuition (N) Big picture thinking Future thinking Experienc ing a moment Full of emotion/feelings Shown in relationships Unconventional living Imagination Creativity Children Color /lighting (bright, warm, dreamy) Illus tration included Thought bubble shown Set in nature Reflecting/daydreaming

PAGE 230

230 Table 4 14 . Continued. Dimension Common Themes Less Frequently Mentioned Themes Thinking (T) Logical/rational Process oriented Goal driven/end result Focused su bjects Working Science related occupation Medical related occupation Setting (sterile, clean, library, laboratory) Impersonal Body language Thinking pose Looking up Staring into space Not smiling Color (sterile, cool, neutral, white) Being o bjectiv e (follow rules, no emotion) Being alone Attire (business, glasses) Thought bubble shown Feeling (F) Shown in relationships Family Romantic Plutonic Animal Emotions Romantic Happy Experiencing a moment Enjoyable pa stimes Volunteering Reading Relaxation Creative endeavors Body language Touching Loving looks Smiling Laughing Set in nature Color (bright, saturated, warm, yellow) Lighting (low, soft, sunrise, sunset) Candid photos Attire (relaxed, hat, Converse)

PAGE 231

231 Extraverted (E) The first part of RQ1a asked which visual symbolic elements were consistent with extraversion (E). Therefore, the researcher examined the images that participants sorted into the extraverted pile and analyzed the verbal reasons given for the placement. Based on the findings, several common themes emerged, including the importance of happiness /fun , activeness , choice, collaboration , multiple people body language and color. Happiness/fun. During the group image s orts, participants linked the happiness, fun and energy reflected in an image to increased extraversion. In this case, participants seemed to detect the tone or spirit of the images as happy, fun or energetic rather than pointing out specifics qualities. F or example, the family in one image exuded extraversion because 3). happy and 2). O ther participants picked up on a sense of upbeat energy in extraverted images. Again, type of energy they felt coming from th 1) 1). P articipants also noted that images were more extraverted when there was a sense of fun among the subjects. Often the fun seemed to stem from a shared task o r activity people were participating in together, whether that was cooking, volunteering or brainstorming. At times, the

PAGE 232

232 (Participant 9 3). In some cases, the sense of fun even transferred from the image to the viewer , making 1) . Activeness. P articipants also noted the importance of active ness in extraverted images. For example, Participant 3 2 tied activity to movement, saying, motion Participant 4 Participant 7 3 summed up the importance of activity to extraversion, sayi Other participants also acknowledged the importance of extraverted activity including other people. Not only did the images show motion and activity, but they also portrayed groups being active together. Participants honed in on the idea of doing rather than sitting presented in extraverted images. For instance, when describing the most important attributes of extraverted images, Participant 9 Additionally, a few groups saw this activity with others as reflective of ing TV or something (Participant 6 3) . idea (Participant 6 2) . Choice. Choice also played a significant role in identifying extraverted images . T his idea of choice was reflected in three ways: the choice to be social , the choice to participa te in activities together, and the ch oice to be in a public setting. According to participants, extraverted images also often included subjects who seemed to (Participant 5 1) rather than being around other people out of necessity. if the people are what they're

PAGE 233

233 in the situation for then they're probably taking plea (Participant 1 3) . In fact, participants often pointed out that people seemed to be the focus of an extraverted image, rather than just accessories that happened to be in the frame. For example, participants in Was it something that was about being wi th the people? Or is it just that they happened to be with people? (Participant 1 2) . Subjects who chose to participate in activities were also identified as extraverted because they made a choice to simultaneously interact with the people and world aroun d them. This was particularly true when a subject was engaged in an activity that could have been done alone or 2). For example, in one image of a group of volunteers picking up trash out 2). And, in another image of a interacting 2). Other participants expressed the idea of choosing a group activity in terms of a shared experience , even if those experiences consisted of normal, everyday behavior such as painting, studying or cooking. For i nstance, H this mundane activity together. (Participant 9 that [pepper] and togeth er. (Participant 9 3) setting as important to extraversion although indo or and outdoor public settings were considered equally extraverted. For example, Participant 6 thing. the fact s and they went out of their way to go do something

PAGE 234

234 public setting s represented subjects who were adventurous, outgoing and actively looking for a way to engage other people. Although it was mentioned less frequently, a few participants also felt that an extraverted choice was something that required additional effort or guts on the part of the subject. For example, an image showing a group of volunteers picking up trash outdoors was extraverted 1). Collaboration. During their discussions, participants also highlighted the sense of collaboration often found in extraverted images. For example, participa nts identified images as 1) more of a social [activity], 3) . This spirit of collaboration was often reflected as teamwork : t the whole idea of 2). Sharing ideas, in particular, seemed to be identified as especially collaborative and was noted multip 2), 1). Multiple people. Participants also emphasized the importance of people in extraverted images and suggested that images almost always needed multiple people to be considered extraverted. (Participant 8 (Participant 4 1). Although images with t wo people could be considered extraverted, images with larger groups of people were seen as exclusively and the most extraverted. As Participant 7 1 said , I more of a sociable outing, like , Participants indicated that th is

PAGE 235

235 social factor was crucial to extraversion because they felt that people either really enjoyed group 2). Body Language. in extraversion. In particular, participants mention ed smiling, talking/communicating, laughing and eye contact. In many cases, participants aligned smiling subjects with extraversion. In extraverted images, participants noted that smiling often indicated happiness, which they associated with extraversion. So, for participants, two thirds ( n = 20) of whom identified themselves as extraverted in the screener questionnaire, happiness seemed to be derived from lots of people and socializing, while they were less likely to see being alone as a source of happines s. Smiling could 1). Also, in response to an image of a two adults helping a child water plants, Par ticipant 7 extraverted. The parents, I just assume, interacting with their daughter. Teaching her. Smiling. Images with subjects who appeared to be talking and/or communicating were also considered to be more extraverted. 3) was enough for an image to be classified as extraverted. While others linked extraversion to a more collaborati 1) or 2).

PAGE 236

236 Some participants noted that it was impossible to actually tell if subjects were talking in still images, such as those use d in the group image sorts. However, participants still acknowledged that images with 2) still communicated extraversion. Participants also classified images wi th subjects who looked like they were in the middle of talking or had just finished talk I mean king together, looking (Participant 2 1). P articipants also acknowledged that extraverted images might include subjects who were laughing. These participants often tied laughter to communication , noting that subjects who were laughing se 2). Participant 6 E ye contact was also closely linked to e xtraversion. In general, participants agreed that the subjects in extraverted images were making eye contact with another person, whether that person was another subject in the image or the viewer. In fact, Participant 3 1 suggested that the most important (Participant 3 2) whether it was another subject or the viewer. However, participants agr eed that the most extraverted type of eye contact was when one or more of the subjects in an image made eye contact with the viewer by looking at the camera. When subjects looked at the camera, participants felt they were looking directly at them as viewer

PAGE 237

237 [like (Participant 4 3). Looking directly at the camera also made 2) or 2). The direct eye contact was so strong that some g to tell you something (Participant 3 3) . For participants, t 2 2). Color. Color, furthermore, played a large role in classifying images as extraverted. Almost all of th e participants identified the association between bright, saturated colors and extraversion. Specifically, one participant saw bright colors as making the image seem more 2). Other participants also identified the extraverted imag lively, vibrant color palette s as loud , indicating that the colors almost screamed for their attention, much like an extraverted person would . For example, while looking at one photo, Participant 3 An d, in several cases, participant s saturated, lou d colors (Participant 3 1) or ( Participant 4 3 ). In particular, multiple participants pointed specifically to bright greens and oranges in the 2) 1). In dressed in bright , vibrant clothing as more extraverted of the 1).

PAGE 238

238 Participant 9 2 even expressed her perception of multiple bright colors evoking feelings lors , which make you think of more activity. And multiple colors 1 viewed bright colors as out, a few participants also ersion. For instance, one participant 3). Less frequently mentioned themes. In addition to the common t hemes discussed above, several other less frequently mentioned themes also emerged for extraversion. These less frequently mentioned themes include d : appearing to feel comfortable, attire, close proximity, tight cropping, and dogs. In a few cases, particip ants subject looking comfortable with themselves and being comfortable in a social setting. In particular, several participants identified the body posture of a girl who was sitting alone on the floor of the library with her feet stretched out in the aisle as extraverted because an extraverted 2). In reference to the same image, participants in the third group gave a more detaile d explanation of why the posture seemed extraverted: To me, I would put her as extraverted just because I feel like it takes a certain kind of personality to sit an aisle where you would be in the way of someone even if i spre 3) I would place this as extraverted because she looks comfortable in who she is. Not that

PAGE 239

239 just lik (Participant 3 2) Another less frequently discussed theme involved the attire subjects were wearing in the images. According to some participant s , trendy, dressy, cute or revealing clothing had an extraverted connotation. In one image , a gir indicated to some participants that she was extraverted even though she was standing alone in the middle of a field with her horse. While discussing another image where a young girl was alone in a library, the subject was wearing extraverted clothing, including a bold patterned dress and gold, open toe clothing. Another participant felt the girl must have dressed up intentionally because she now so I have time to go socialize (Par ti cipant 6 3). Overall, it seemed that when girls were pretty and wearing trendy, leg revealing clothing, participants assumed that both they and the image must be extraverted. Another less common theme participants mentioned was the close proximity of the s ubjects within the image. This close proximity was shown visually by the physical closeness of the subjects to one another. For example, Participant 7 2 faces are really 2 s leaning on him, and closeness translated to extraversion, perhaps because human contact suggested a human connection similar to when people made eye contact with another human. Participants could also have been reacting to the violation of personal space in images where subjects were in such close proximity.

PAGE 240

240 T here was also discussion among participants in some group s about the link between dogs and ext raversion. Participants seemed to transfer their feelings about the dogs in the images 2). Another group identified large, active dogs , like Labrador Retrievers, as particularly extraverted and assumed that people who were shown in an image with a Lab were also extraverted. Finally, s everal participants also mentioned that more tightly cropped images appeare d more extraverted. Some participants associated tight cropping with a more dynamic image and then linked that dynamism to extraversion. Alternately, other participants felt a tightly cropped image that showed only a few people indicated that there could h ave been many more people in the setting, but the photographer zoomed in to capture only a few. Because they felt that the tight crop indicated that more people were in the room, but just out of the frame, they labeled the image extraverted. One participan Yeah, for a second I like had to double take because it felt like there were more people around, but you really only see these two people (Participant 9 1) . To summarize, this analysis showed that extraverted images most often had a happy, a ctive feel and often portrayed some type of collaboration among subjects. In extraverted images, the subjects also seemed to have chosen to participate in an activity, be social and/or be in public. The images also portrayed multiple people who were smilin g, talking, laughing and making eye contact with another subject or the viewer. The colors in extraverted images were saturated, bright, vibrant, lively and loud. Participants noted greens and oranges as particularly extraverted colors that indicated activ ity and excitement. In addition, extraverted images might

PAGE 241

241 show subjects in close proximity, subjects who appear to feel comfortable and subjects wearing trendy/revealing attire. They could also could include a dog and/or be tightly cropped. Introverted (I) The second part of RQ1a ask ed which visual symbolic elements were consistent with introversion (I). Therefore, the researcher examined the images that participants sorted into the introverted pile and analyzed the verbal reasons given for the placement. B ased on the findings, several common themes emerged, including choosing isolation, quiet ness, immersion, professions, setting, the number of people, no eye contact , and color. Choosing isolation. P articipants noted that an image w as more strongly introvert ed when it seemed as though the subject (s) had chosen to be alone or to isolate themsel ves in a place that could have included other people . As Participant 8 3 explained, In response to an image sho wing an intimate couple alone under a tree, p articipants in the second group also noted th is deliberate decision to be isolated, saying, I mean, they could have done that on a park bench in a more crowded area, or they chose to go under a tree where . (Participant 2 3) Get away from everybody. (Participant 2 2) It seems like conscious choice to be away from people. (Participant 2 3) Quietness. Along with specific settings, participants also i dentified introverted images as 3) , or portraying 2). Participant 3 2 noted that image details i n introverted images, body language in introverted images: Yeah, I say introvert. He looks like he 1) I say anyone. (Participant 3 2) He looks like a really quiet person. (Participant 3 1)

PAGE 242

242 Participa nt 6 3 also tied quietness to thoughtfulness in an image of a farmer squatting in a field of wheat such as being in a library, a laboratory, alone at home, or out in nature (e.g., a field, a beach, under a tree) also suggested introversion . Participant 9 2 observed, iet, you wants to or is supposed to be talking to other people . Immersion. Participants also identified immersion as a component of introverted images. The subjects in introverted ima ges could be immersed in either thought or activity. Regarding 1). Accor cipant 3 1). Other participants For instance, Participant 1 but them being so immersed in their thought that they could care less, and The second type of immersion immersion in activity was represented by a subject 3). Thus, the idea of immersion in activity often involved a singular subject with a singular focus. For instance, participants stated,

PAGE 243

243 (Participant 9 2). While subjects did not have to be alone for participants to consider them immersed in an activity, in many cases, participants seemed to think a single subject was more immersed, perhaps because there were no other humans to distract them. Professions. In addition to immersion in thoughts and activities, participants felt that some professions were more closely tied to introversion. In some cases, participants even acknowledged that their choice to labe l an image as introverted came almost solely from the The first profession participants frequently labeled as introverted was doctor/surgeon. P articipants noted that the singular interest and passion required to become a do ctor made doctors introverted. (Participant 3 1) . Participants also noted that the profession a lo t of studying. And logical processes, like, mental processes (Participant 10 1) . Similarly , overlapping with the theme of being immersed in an activity ( discussed above ) , participants also 7 3) in the subjec t, (Participant 9 2) and were very focused: ame time, you need to be 2). So, in the case of doctors, participants seemed to fe el that focusing on the task trumped any human interaction. Scientists were another profession that was often identified as introverted in images. As with doctors, participants often perceived science as the singular focus of scientists: Their energy is co ming from science. (Participant 1 thing. (Participant 1 3)

PAGE 244

244 that they are having f un together, but they are still trying to achieve the same thing. (Participant 1 1) Participant 10 would say introverted because she works in a lab, which is someplace wh ere you really have a In their comments, participants almost seemed to be operating on a stereotype they held of scientists being science loving loners who were more focused on their work than people. In part, this stereotype might stem from the fact that the participants were all creative majors, graphic design and advertising, with, presumably, less personal experience in the sciences. Finally, participants also consistently labeled an image showing a farmer as in troverted, often due largely to his profession. In some cases, they tied farming to the introverted concept of outside by himself in the middle of nowhere 2), which is discussed further below. Participants also felt that farmers did 10 1), which they felt were more introverted characteristics. Participants in the fifth group also note d similarities between farmers and doctors being experts in their respe ctive fields: Introverted. (Participant 5 5 3) And knowing a lot about one subject. (Participant 5 dedicated his entire life to [farming]. (Participant 5 2) I would say medicine goes back to that too. (Participant 5 1) Yeah, exactly. They have to know a lot about one thing. (Participant 5 2) Participant 6 rests, immersing themselves in one topic. And, usually farmers are extremel In fact, participants seemed to identify each of the three professions discussed above doctors, scientists and farmers with extreme depth of know ledge and singular focus . This

PAGE 245

245 knowledge on a single subject, while extraversion was associated with a breadth of knowledge on many topics. This singular focus also s uggested that introverted occupations require d intense concentration on the job at hand, with little room for multitasking , suggesting that introverted images might portray subjects completing a singular task, while extraverted images might show one subjec t completing multiple projects simultaneously. Setting. In addition to specific professions , participants also noted several setting related factors tied to introversion. These settings include home, nature and isolation. First, images that appeared to be of introversion. For example, Participant 8 3 the whole thing of P articipants linked images of home even more strongly to introversion when they felt the subjects were happy at home and had made a conscious choice to be there rather than in public. In several cases, participants assumed that the Participant 9 3). For instance , Participant 1 3 suggested that images were more introverted if subjects seemed like T obviously making it a point to be in their a dinner party. Besides staying at home, images set in n ature also evoked introversion: , all of them. So, I wonder if, just being in the grandness of nature has something to do with it. (Participant 7 from all humans. (Participant 7 things. (Participant 7 1) So, your getting in touch with your natural self, I guess. (Participant 7 3)

PAGE 246

246 More specifically, o ther participants identified qualities of nature that made it seem introverted, (Participant 2 2). According to Participant 2 3, I t's just that it is isolated. It's the nature of nature . For participants, the isolation and grandeur of nature seemed to indicate that subjects were smaller in their environments, almost tryi ng to blend in with their surroundings. On the other hand, extraverts were often perceived as bigger personalities, taking up more space in the Similarly, other participants underscored the difference be tween being in nature , which was more introverted, and being in a manmade environment (Participant 9 1). While participants viewed nature as qui et, isolated, expansive, and calming, they saw manmade environments as loud, enclosed, and full of energy and people. Related to nature, an isolated natural setting the middle of nowhere seemed to have even stronger ties to introversion. Part of this r elationship might stem from participants strongly 7 3). Alternately , participant 6 2 noted the connection between the middle of nowhere and In addition to settings in home and nature, participants also identified isolated settings as (Participant 8 2) for introversion. According to other participants, isolation was 1) 2) or a lack of other

PAGE 247

247 (Participant 9 1). Number of people. Further, p articipants identified the number of peo ple in the image as one of the main determinants of an introverted image. According to participants, most 10 2). Not surprisingly, images with a single person we re most readily classified as introverted. Participants also seemed to feel that the subjects in introverted images were engaged in something of their own and happy, rather than seeming lonely or left out. Participant 4 2 ju st hanging out alone , d , while Participant 3 1 Therefore, participants seemed to identify introverted images as those where the subject was deriving their h appiness from being alone and able to engage in their own thoughts or tasks, unlike extraverted Although images with a single person were most easily identified as introverted, participa nts also noted that an image with two people could be considered introverted under certain circumstances, particularly romantic couples pictured in intimate moments . P articipant 7 1 Additionally, Participant 5 2 felt just the two of them , Part of the reason romantic couples were considered introvert ed was because participants view ed the couple as a s ingle unit. For instance, participants felt that people group couples (Participant 2 2) because ( Participant 5 2 ).

PAGE 248

248 In addition to a single person or isolated couple, participants often identif ied images with a single person and an animal as strongly introverted. Similar to romantic couples, some participants noted that a human and their pet could function as a single unit : same thing with the people and their pets. They function ( Participant 7 1 ). Some 2), indicating that the subject would rather be around a pet than a person. Other particip ants expanded on the idea of subjects choosing pets over people, calling a 1) and noting that the human in the image was 2). P articipant s identified this pattern when multiple kinds of animals , including a horse, a dog and a cat , were shown with a single human. For example, one participant identified the scene 4 1). In many cases, this pattern also held true for humans and dogs. Although dogs were sometimes explicitly linked with extraversion, a human alone with a dog evoked a more introverted feel in other situations. Participants pointed specifically to images where a subject was alone with a dog in 2). When asked about why some dogs seemed extraverted while others were introverted, participants seemed to rely on ot her visual symbolic elements in the images such as the level of activeness, interaction, body language and lighting. Brightly lit dog images where humans were playing with or interacting with their dogs while smiling or laughing were extraverted. Alternate ly, dimly lit dog images where humans were sitting calmly next to their dogs while appearing to be in thought were more often introverted.

PAGE 249

249 Although humans with horses or dogs sometimes evoked introverted feelings, participants most strongly linked introver sion with cats. Furthermore, when discussing the image of a woman and her cat, multiple participants referred to (Participant 1 2), revealing their stereotypical associations of cat owners living alone with large numbers of cat s. Participants in the sixth group even assumed the woman preferred the company of cats to other people I think an introvert. She has a cat (Participant 6 1) . (Participant 6 3) . (Participant 6 1) . However, Participant 10 2 might have summed up the link between the extraversion/introversion (E/I) I , like, not all dog owners are extraverts but all cat owners are introverts No eye contact. Acco rding to participants, the subjects in introverted images were often eye contact with any other human, including both the viewer and other subjects within the image. For example, Participant 3 2 ith anything in the image frame. we d looking at (Participant 3 just kind of sitting there, looking off into the distance 1). I f the subject in an i ntroverted image was making eye contact with something, participants noted that it was often an inanimate object. For example, Participant 3 think, again, sort of looking off into the distance or looking into like a book or at a blackboard or

PAGE 250

250 so Thus, eye contact with inanimate objects seemed to be a hallmark of introverts. This might be due to the fact that inanimate objects were not human and, therefore, had no emotion. So, introverted subjects could make eye contact with an inanimate object without forging a human connection like the subjects in extraverted images did when they looked at another subject or at the viewer. Color/light. Color and lighting also played a significant role as participants determined which images were introver ted. Participants distinguished the colors in introverted images as 1). 1), 2). Specifically, t hey also identified grays, soft blues and whites as more introverted colors. Closely related to the introverted color palette, participants also noted the effect of soft or lo w lighting in determin ing introverted images. For instance, Participant 7 rker clothes. So, it seems more sunset, also set an introverted tone in images. For example, Participant 3 Less frequently mentioned themes. In addition to the common themes discussed ab ove, two other less frequently mentioned themes also emerged for introversion. These less frequently mentioned themes include d : relax ation and wide cropping. In several cases, participants linked introversion to relaxation, which often took the form of pur suing a hobby or activity of choice in a secluded area . In particular, reading a book was cited several times It looks like she had a lot of options, and she chose that

PAGE 251

251 specifically (Participant 2 3) . Doing a hobby, she enjoys reading, but middle of nowhere to recharge (Participant 2 1) . Other times, relaxation took the form of inactivity doing nothing at all and just enjoying doing nothing 2) . Overall, it seemed that participants saw the subje cts in introverted images as being less dynamic and exerting less energy, engaging in more quiet or sedentary activities. Another less common theme participants mentioned was the wide cropping of introverted images. Some participants felt that a wide shot (Participant 4 3). Others associated the expansive backgrounds or scenery in widely cropped images with introver sion. In part, this expansiveness depicted introversion because of the smallness of a subject within the frame, which further emphasized their isolation. In summary, this analysis showed that introverted images exuded quietness, with a lack of noise or talking. In introverted images, the subjects also seemed to be immersed in their thoughts or an activity. Participants felt that professions such as doctors, scientists and farmers were particularly introverted. Most often introverted images had one person, or possibly two, particularly if they were an intimate couple. Single people shown alone with animals were also considered int roverted. The subjects in an introverted image also made conscious choices to isolate themselves from others, often in a home, in nature or in an otherwise isolated setting. Most subjects in introverted images were not making eye contact with another human either in the image or the viewer. Instead, they were looking off the image frame or at an inanimate object. The colors and lighting in introverted images were soft, muted and low, giving off a clean feel. Participants found grays, whites and blues to be particularly introverted colors. Extraversion/introversion (E/I) summary Overall, participants found extraversion (E) and introversion (I) photos to be almost polar opposites visually. While extraverted images showed multiple people who were happy and

PAGE 252

252 inte racting with each other, introverted images showed only one or two who were quiet and immersed in thought or activity. Extraverted images showed people in public, collaborating on projects, while introverted images showed a subject isolated at home or in n ature immersing themselves in their own mind or task. Additionally, while the subjects in extraverted images were smiling, laughing and looking at another person or the viewer, subjects in introverted images were making no eye contact with other humans. Th e colors for each dimension were also extremely different with bright, loud, lively and vibrant colors (i.e., green and orange) in extraverted images and muted, clean, soft, monochromatic colors (i.e., gray, white, blue) in introverted images. There may al so be some connection between extraverted images being tightly cropped and introverted images being more widely cropped. The two dimensions did have two characteristics that seemed to overlap choice and dogs. However, the application of these two ideas w as quite different. While choice seemed to be an important aspect in both dimensions, extraverts made choices that increased their interaction with other people, choosing to participate in activities, to be social and to be in public. Whereas the subjects in introverted images chose isolation. Dogs were also mentioned in both dimensions. However, extraverted images had dogs that were big and active with their owners playing or interacting with them. The dogs in introverted images were calm and comforting, s een as companions or noted for the bond they displayed with their owners. S ensing (S) The first part of RQ1b ask ed which visual symbolic elements were consistent with sensing (S). Therefore, the researcher examined the images that participants sorted into the sensing pile and analyzed the verbal reasons given for the placement. Based on the findings, several common themes emerged, including goal orientation, short term focus,

PAGE 253

253 practicality/functionality, systematic process, working , using one or more of the five senses , organization, conservative dress and color. Goal orientation. In many cases, participants identified sensing images as goal oriented. According to participants, sensing images could focus on a single goal, a task, an end goal or work toward so lving a problem. Frequently, participants mentioned that the subjects in a sensing image were working toward a common or singular goal and had an obvious purpose. doing things t hat have a purpose . (Participant 5 1) Yeah. Washing dishes. On her computer. Pulling things out of the ground. Cutting someone up [dur ing surgery]. (Participant 5 2) something that has to be do (Participant 9 2) . In some cases, participants also discussed this goal orientation in terms of being focused on an activity or task. (Participant 10 3), and i 2) such as washing a dog or painting a wall. (Participant 1 2). In these cases, it was important to participants that they could determine what the end result was going to be. For instance, while discussing an image that showed a family painting an interior wall, one participant noted, Th all . (Participant 1 2)

PAGE 254

254 Finally, participants also suggested that sensing images were goal oriented in their pursuit of a single or correct answer. For example, in response to an image showing two smiling scientists using a dropper to extract liquid from a tes t tube, Participant 3 Thus, sensing images seemed to imply a precise answer. While intuitive images nt 5 3), sensing images did. In many cases, participants seemed to equate this right or wrong, yes or no status with scientific or medic al images where there might be poor or disastrous results if something was done incorrectly. Short term focus. According to participants, sensing images also concentrated on the short forwarding into what you have to do two hours later in the 2). Participants also identified sensing images as focused on the here and now. Participants felt that tasks in sensing images were quickly accomplished: (Participant 1 3). Even if the overall task seemed to be more long term or big 2). Practicality/functionality . In addition to the short term, get it done attitude, many participants identified sensing images as those that portrayed some sort of practicality or functionality. Often, participants seemed to interpret the idea of practicality and functionality as dow n to earth, conventional or normal perhaps almost boring. For example, when sorting an image into the sensing pile, Participant 2 Other p articipants used the term functional to describe an image where subjec

PAGE 255

255 (Participant 4 1) to be 2). Another participant described one image, which showed a single male subject holding a tablet and looking up at a large number of thought bubbles floating above his head, as (Participant 2 1). Practicality and functionality were also associated with numerical and financial tasks . Fo r (Participant 4 2) and the fact that money you kinda have to sense patterns and you have to plan based on the patterns (Participant 5 3) as practical and, therefore, sensing. At other t imes, participants considered an image sensing if the subjects were engaged in a practical or necessary task. For example, participants identified the act of cleaning (e.g., a pet, a playground, dishes) as a practical task. In particular, p articipants from multiple groups equated washing a dog with sensing a v (Participant 4 2) (Participant 5 3). Linking another form of cleaning to practicality, Participant 7 1), thereby m aking it sensing. Additionally, other participants discussed the practical/functional nature of one image that showed volunteers picking up trash outdoors . For example, participant in the ninth group said, lving a problem in the community. (Participant 9 (Participant 9 3) Yeah, done. (Participant 9 2)

PAGE 256

256 Systematic process. Beyond being practical and functional, sensing images also conveyed a systematic or methodical process . Sometimes, participants pointed out that a subject like he's doing a proc ess a nd a method (Participant 2 2), like in an image of a farmer examining his crops. Other times, this systematic process was implied . For example, while discussing the most important attributes of a sensing image, Participant 1 3 noted that it felt like see the process going through their head . Another most important attribute for sensing images was that they often depicted a . For example, one image showed a man making a salad while a woman ki ssed him on the cheek. In the image, each salad ingredient (i.e., lettuce, approached making a simple salad in a very systematic way (Participant 2 3) . All thos e different bowls (Participant 2 1) . (Participant 2 3) . Although sensing subjects were systematic, that did not make them uninteresting. As one oring or a computer. They can still be fun, but they prefer 2). Therefore, the subjects in sensing images might come across as more organize and rigid, while subjects in intuitive images seemed like more go with the flow, free thin kers. Additionally, participants noted that, in many sensing images, the systematic process implied a number of steps to complete the process. For example, while discussing an image showing two smiling scientists using a dropper to extract liquid from a te st tube, Participant 6 2 an experiment is very you do this first, you do that, you do this. And you move on . A step to 2 agreed that sensing images seem to have more of a planned a

PAGE 257

257 complicated and a step by Further, while discussing the most important attributes in sensing images, Participant 8 2 pointed out that complex and structure . Working. The concept of working was also strongly associated with sensing images. Underscoring the different associations between jobs and hobbies, one group stated, Thi like, a hobby. (Participant 9 2) True, more like a job. (Participant 9 1) I think, recreation [is intuitive] versus work [is sensing]. (Participant 9 2) In particular, parti cipants also linked some specific professions with sensing images. Throughout the group image sorts, participants mentioned general office or business related jobs as well as numerically based careers in areas such as mathematics, finance and accounting as sensing . Some p articipants also identified jobs that required critical or analytical thinking as particularly sensing o, a lot of the pictures I put in sensing were doctors, test tubes, math (Participant 3 2) . Participants also identified jobs that super practical, like medicine or construction or things that take a lot of extreme, detailed know how 2) as sensing. However, the two most commonly mentioned sensing careers involved science and medicine. In part, this may be because several images in the image deck showed scientists and doctors. Yet, overwhelmingly, when images portraying science or medicine were shown, (Participant 5 1). And, i n many cases, the link between science and sensing was so strong in 3) . Specif ically, participants also noted that one image depicting two smiling scientists using

PAGE 258

258 . It looks pretty (Participant 9 2). Therefore, it seems that participants saw science related images as safe, standard and procedural, with no real elements of creativity, imagination or risk, which made them classify the images as sensing. Participants a lso identified images depicting medical careers such as doctors and 3) and 2). Furthermore, one group so strongly associated medical professions with sensing, that they seemed to find the idea of a non sensing doctor both funny and unsettling: and stuff, organized fo r me (Participant 3 3) 2) [Group 1) Using one or more of the five senses. Participants also associated sensing images with subjects using their five senses sight, hearing, taste, touch and smell and often multiple senses at the same time. For example, It f kiss the cat or something, in the sense of, like, touch. sniffing her, sort of. The flowers imply scent. Running water is sound. Yeah, I wise. (Participant 9 2) More than any other sense, however, participants linked sensing images to the sense of touch. For instance, while discussing the most sensing images in the image deck, Participant 2 2 3 also

PAGE 259

259 Orga nization. Participants also recognized a high degree of organization in sensing images . According to participants, sensing images were neat and everything seemed in its place. For example, Participant 8 1 point ed out an image with ordered tes t tubes, sayin g, really out of place. Everything just seems to follow some sort of pattern In another image, 2), 1). While discussing the most important attributes of sensing images, participants also noted 3). While discussing one image , which showed a group of scientists wearing lab coats and sitting at a table with a number of beakers filled with different colored liquids , Participant 3 s even Conservative Dress. Sensing images, moreover, often contained subjects who dressed conservatively. In the images where participants made comments about conservative dress, all of the subjects were wearing long sleeved shirts that showed little skin. For example, Participant 2 1 noted that one subject looked These conservative clothes seemed to tra interpreted as sensing. Color. In addition to , participants related particular colors to e as white, soft, clean, plain, neutral and, especially, sterile. For example, while discussing the most important

PAGE 260

260 attributes of a sensing image, participants said I often associate neutrals with practicality and function (Participant 5 1) , almost (Participant 5 2) E (Participant 7 2) . After surveying the pile of sensing images, one participant was surprised by the similarity in color, saying Wow. I just noticed how white these [sensing] photos are (Participant 3 3) . Less frequently mentioned themes. In addition to the common themes discussed above, one other less frequently mentioned theme showing technology also emerged for sensing. In particular, participants pointed out laptop computers and tablets as indicators of sensing. One (Participant 7 moved the image from a potentially intuitive image to a sensing image. During the first group, participants even noted a distinction in the brand of technology that was sensing, discussing that PC computers were sensing, while Mac c omputers aligned with intuition. To summarize, this analysis showed that sensing images were often goal oriented, focusing on a single goal, a specific task, an end result, or solving a problem. Additionally, sensing images usually had a short term focus i n the form of a task or activity that would soon be completed. Practicality and functionality were also hallmarks of sensing images, with participants often interpreting these words to mean down to earth, conventional, normal , or even boring. Sensing image s might also show a subject following a set of steps to complete a systematic process. These sensing traits, particularly practicality and a systematic process, regularly showed up in image s where subjects were working, most often in science or medical job s. Subjects who were using one or more of their five senses, especially touch, also elicited a sensing personality. More concretely, sensing images often displayed traits of organization (e.g.,

PAGE 261

261 items labeled, separate containers) and conservative dress (e. g., long sleeves, little skin showing). Participants also described the colors in a sensing image as clean, neutral and sterile, with copious amounts of white. Technology (e.g., laptops, tablets) might also have some connection to a sensing personality. In tuition (N) The second part of RQ1b ask ed which visual symbolic elements were consistent with intuition ( N ). Therefore, the researcher examined the images that participants sorted into the intuitive pile and analyzed the verbal reasons given for the placem ent. Based on the findings, several common themes emerged, including big picture thinking, future thinking, experiencing a moment, full of emotions /feelings , shown in relationships, unconventional living, imagination, creativity, children, color /lighting , illustration included and thought bubbles shown . Big Picture Thinking. According to participants, intuitive images included subjects who were thinking about the big picture. For example, while discussing the most important attributes of intuitive images, o (Participant 10 3) . This indicates that, per haps, intuitive picture that gave participants a glimpse inside their mind. Participants also noted that big picture thinking was often the result of many ideas comi ng together to form a larger, more cohesive, single idea, which they identified as a big idea or big picture idea. For some participants, intuitive images could be recognized by the fact that all the pieces in an image were coming together to form a single , bigger overall idea. For instance, in response to one image, which showed a single male subject holding a tablet and looking up at a large number of thought bubbles floating above his head, Participant 1 2 said,

PAGE 262

262 here's so many things to this that it's not like he's accomplishing one. He looks to be in thought about all of them. But it's kinda like he thought for a second and POOF, all these things came out. According to Participant 6 i , Big 1). For example, ying to find something (Participant 3 2). Intu one, giant brainstorm that just (Participant 3 2) 3). Future thinking. Closely related to the idea of big picture thinking was the theme of future thinking. Participants felt the subjects in intuitive term possibilities for the future, looking for relationships, making connecti 1). Approaching from a slightly different angle, Participant 8 2 suggested that, in an image of a romantic couple cuddling under a tree, the subjects looked like they were In some cases, participants n oted that subjects were currently engaged in an activity that they seemed to think would prosper in the long term. For example, in an image that showed a volunteer group picking up trash outside, Participant 7 1 interpreted a thought bubble, which containe magining the future and what their work is going to pay off to be 2 saw intuition in scientific

PAGE 263

263 exper imentation, which in essence is hoping for a future or different kind of Several participants noted this expectation for the future in an image of a group of volunteers who were (Participant 10 3). A few participants even saw this forward loo king thought as a sign of hope: 2) I kind of hope, I guess, makes it intuition. (Participant 1 2) Some future thinking also overlapped with the previous theme of big picture thinking, as if subjects were looking into the future to see the bigger picture of what could be. For example Participant 10 going to yield, which is Experiencing a moment. B ig pictu re or future thinking was one way in which participants identified intuitive images. However, seemingly contradictorily, they also identified intuitive images as those where the subjects were experiencing a particular moment in time . S ome of these moments were personal moments where participants felt the subject was more 2). Other times participants identified moments a s open ended or frozen in time: W ith this picture [of a couple cuddling under the tree] It indefinitely. Or, [in the image of the farmer in his field Just a moment in time kind of thing. (Par ticipant 3 2) Other participants associated intuitive images with the feeling that, although the subjects were engaged in an activity, their primary motive was an experience or connection with others,

PAGE 264

264 often family or kids. For instance, Participant 2 3 sai Full of emotion/feelings. identified intuitive images as fil 2) and (Participant 3 1) , and 1) In fact, emotion played such a strong role for some participants that it defined the intuitive image category. According to Participant 3 2, the difference between sensing and intuition could be with a more end purpose. And [intuition] is just kind of in the moment, more emot ionally Other participants also specifically identified romantic emotions as intuitive due to their rticipant 3 3). Participant 8 1 tied the theme of experiencing a moment to intimacy as well, identifying an image as intuitive because , and Shown in relationships. Similar to em otions, participants also classified images that displayed relationships as intuitive. Participant 8 1 indicated that an image could be intuitive relationships between the people inside the photos or the people and relationships depicted often overshadowed other aspects of the image . For example, Participant 5 . Other participants agreed that the activity might lean more sensing, but if multiple people were taking

PAGE 265

265 part in order to strengthen their bond of friendship, love or family, then the image was ultimately intuitive. Unconventional living. Participants al so linked intuitive images to being unconventional. However, participants seemed to interpret the meaning of this term unusually. , (Participant 3 3) , such as in an image of a mom and dad smiling and letting their toddler water plants. Alternately, some seemed to use the word to describe activities that were outside the perceived norm. For example, Participant 9 2 pointed that it was unconventional to be reading a book out side , while Participant 6 1 commented on the unusual dress of a student studying in the library T Finally, participants defined unconventionality as something that was i llogical or 2). Most often, this lack of practicality was identified in images where parents allowed children to help with an adult activity. To many participants, involving children in an activity complicated the process and represented a less efficient way of achieving a goal. For example, Participant 3 3 felt let kids paint w And , participants in one group saw the impracticality in terms of the family painting the wall chartreuse rather than a more typical wall color. Imagination. In many cases, participants also associated intuitive images with imagination. For example, 1). Participant 5 2 also identified imagination as important to intuition because it helped participants create something , s using

PAGE 266

266 information you collected and creating something like in this research picture or imagining Other participants noted th at imagination defined the difference between intuitive images and other more clinical, or pr actical images . For example, in one image that showed a group of scientists at a table with a large number of beakers filled with brightly colored liquids, Participant 9 3 felt that the purpose of add the extra [illustration of colored fumes coming out of a test tube] was to take that practical image and try to make it into something that seems making more of a discovery. And, making discoveries, I feel like, invol ves some aspects of experimenting, cause and effect, having an imagination to try different things. virtue of the bright colors and a swirling illustration, a science oriented image was transformed from practical and, therefore, sensing to i ntuition by infusing it with a sense of imagination and fun. Still other participants tied the idea of imagination to the themes of big picture and future thinking (discussed above). For example, Participant 9 1 felt that one image, which showed a man look ing up at a chalkboard filled with doodles and figures, was imaginative because it had such an odd collection of sketches, like Similarly, some participants connected imagining with the theme of future thinking. For instance, in an image where a woman was coloring with her child, Participant 6 1 felt the mother However, participants most commonly associated intuitive images with imagination 2) and

PAGE 267

267 3). As one participant said, 2). Creativity. Participants also consistently identified intuitive images as having e lements of creativity. Intuitive images , such as a picture of a man looking up at a chalkboard filled with doodles and figures, displayed more creative thinking : 3 1), perhaps due to the random collection of both math equations and odd drawings (e.g., a bird, a fish, a martini glass) on the chalkboard. Additionally, intuitive images coul d show creative tasks : (Participant 2 2). Other participants recognized intuitive images as taking place in a creative environment or atmosphere because of numerous characteristics such as smiling, appearing to have fun, and looking relaxed while working. In addition, their clothing marked them as creative and, therefore, intuitive : 1) . Still other participants identified elements of art or creation in intuitive images. ipant 2 (Participant 2 2), but instead they chose to create it themselves. Additionally, a few participants noted the potential for cooking to be re still 2).

PAGE 268

268 Children. Children were also frequently mentioned as one of the biggest indicator s of an intuitive image. Interestingly, participants often linked children to intuition via one of the previously discussed themes. For example, Participant 10 2 linked children to intuition using the with it. A Alternately, Participant 6 2 noted the overlap between the theme of experiencing the moment id , Other participants identified images with children as intuitive because of their unconventional nature. Participants thought images where children were helping with something indicated that the task wa 3) 3). each 3). in the crea tive nature of children. Participant 3 3 sentiments: T his creative nature can also flow from adult encoura gement or inspiration : 1). Still other participants identified images with children as intuitive via the theme of imagination. Participants felt that kids, more than adults, we re particularly imaginative. For

PAGE 269

269 instance, Participant 8 Color/lighting. Another quality of intuitiv e images was having a specific color palette and distinct lighting feel. Participants often described intuitive images as having (Participant 5 2), (Participant 3 3) (Participant 3 1), in part , because 1), and warmer 2). Participant 3 1 also identified color as one of the most important attributes for intuitive images, describin g some of the nuances of the color: Other participants focused on the lighting in intuitive images. For instance, Participant 3 d off the lighting . I t gives me more of a dreamy feel and an imagi 3 3). Illustration included. Several images in the image deck contained vectorized color illustration s that were added by the researcher. Participants strongly associated these illustrations with intuition because illustrations 3) 3), 4 2) . Some participants associated the color illustration with increased imagination because somebody had to add that to this image 2). I n some cases, they specifically

PAGE 270

270 identified the imaginative nature of t he color illustrations: That little vector thing, immediately I (Participant 5 2) . Other participants felt that the color illustrations represented something more than could be seen in the image. For i nstance, participants in the fourth group speculated that the color illustration represented what would happen when scientists mixed the contents of several beakers , while Participant 10 1 thought the colorful bubbles [illustration] above his head indica w what this is going to end up! Like, something is possible. " Several other participants also commented on how a color illustration turned cooking into a more (Participant 5 1). Thought bubbles shown. Several images in the image deck also included thought bubbles, some of which were part of the orig inal stock images and some of which were added by the researcher. For some participants, t hese thought bubbles acted as 3). One reason thought bubbles were associated with intuition was that the [ entity (Participant 8 2). Less frequently mentioned themes. In addition to the common themes discussed above, two other less frequently mentioned themes also emerged for intuition . T hese less frequently mentioned themes include d : being set in nature and reflecting/daydreaming. In several cases, participants noted a relationship between nature and intuitive images. For example, Participant 8 1 commented, land. Just, like, the

PAGE 271

271 Another less frequently mentioned pattern that some participants identified was that intuitive images might include subjects who appeared to be reflecting or daydreaming. For example: mething. (Participant 5 1) Mm hm. Pensive. something. (Participant 5 ective. (Participant 5 2) Summarizing, this analysis showed that intuitive images often displayed subjects engaged in big picture or future thinking. Alternately, they could also depict subjects experiencing a moment and taking in the here and now. Many participants agreed that intuitive images were full of emotions and feelings and often showed subjects in relationships. They also portrayed aspects of unconventional living, which participants usually defined as more relaxed, outside the perceived norm, illogical or impractical. Most often, this was identified in images where adults were allowing children to help or participant in a task. Intuitive images also had a high degree of imagination and creativity. Further, participants identified a strong link between intuition and chil dren, often sorting an image as intuitive simply because a child was shown. Additionally, intuitive images had a color palette that was bright, saturated, warm color and lighting that was dreamy, heavenly and ethereal. Participants also identified images w ith illustrations or thought bubbles as intuitive. Intuitive images might also be set in nature and show subjects reflecting or daydreaming. Sensing/intuition (S/N) summary Overall, participants perceived sensing (S) and intuitive (N) images very different ly. Sensing images had a short term focus and were goal oriented, focused on a single goal, task, end result or solution to a problem. Intuitive images, on the other hand, involved big picture or future thinking, or even possibly daydreaming. Even when int uitive images portrayed subjects

PAGE 272

272 experiencing a moment in the here and now, which could be seen as short term, it was a moment with no goal or purpose in mind quite different from sensing images. Additionally, where sensing images conveyed practicality, functionality and systematic processes, intuitive images were full of emotion, feelings and relationships. Intuitive images were also linked to unconventional living, which participants defined as impractical and illogical the exact opposite of a sensing image. Further, sensing images portrayed subjects at work, specifically in methodical, practical and systematic jobs like science and medicine. Inversely, the subjects in intuitive images were engaging their imagination and creativity. In particular, chil dren were seen as intuitive, directly opposing the working adults seen in sensing images. Also unlike intuitive images, sensing images were seen as organized and filled with conservatively dressed subjects. Moreover, sensing had links to technology, while intuition was more likely to be set in a nature. The color palettes for the two images were also drastically different. Sensing images were described as clean, neutral and, especially, sterile with lots of white. Alternately, intuitive images had colors th at were saturated, bright and warm with dreamy, ethereal lighting. Thinking (T) The first part of RQ1c ask ed which visual symbolic elements were consistent with thinking (T). Therefore, the researcher examined the images that group image sort participants sorted into the thinking pile and analyzed the verbal reasons given for the placement. Based on the findings, several common themes emerged, including logical/rational, process oriented, goal driven/end result, focused subjects, working, setting , impersona l, body language and color . Logical/rational. In many cases, participants identified thinking images as exuding logic or rationality. For participants, logic and rationality often took the form of something that required thought, planning and intelligence. For example, participants identified logical decisions that are advancing them toward something 3). They also noted that

PAGE 273

273 o do 3). More specifically, some participants pointed out how subjects in thinking images were applyin g logic and rationality to a particular situation. P articipants identified studying, finance, (Participant 6 2). Related to mathematics, Participant 9 2 also identifi Process oriented. Thinking image were also process oriented. This process could take the form of a thought process, a detailed plan or a step by step method. In some ima ges, p articipants noted that a mental process, rather than verbal, appeared to be tak ing place . For instance, participants noted that , in some thinking images , the process was ( Participant 3 3 (Participant 7 1). Partic ipants also seemed to connect 1) with trying to solve a problem: He worked through his thought process. (Participant 5 trying to reason through all this to come to some kind of thing that makes se nse. (Participant 5 1) So, it was a lot of mental, l ike, this was very mental. (Participant 5 2) Participants also considered images especially thinking if the activity required subjects to follow a process 1). Participants i n the fifth group noted that even creative thinking follows a process, saying

PAGE 274

274 (Participant 5 2) . Yeah, there has to be a plan and it has to be a system ( Participant 5 3) . Furthermore, some participants projected a process onto the subject based on other (Participant 8 2). 1). Thus, participants seemed to infer that organized elements, such as labeled beakers, organized test tubes and separate containers for individual substances, indicated the subject was in the midst of working thro ugh a precise process . Other participants specifically noted that subjects in thinking images often used a process while completing a task or as part of their profession. In particular , participants discussed the step by step process es involved in medicine and science. They felt that medical personnel 10 1). Similarly, one particip especially with chemistry, there are, like, so many steps and it's so methodical, you have to do this and then this and then this 1). Par t of how participants derived this step by step process from visuals was based on the job setting shown in the image : Just the setting. Just, um, experimental. They have to follow some sort of set of guidelines or procedures, you know, not to make somethin g explode or create toxic (Participant 8 3) Goal driven/end result. In many cases, thinking image s were also driven by a goal or end result. For example, Participant 2 3 identified an image as goal an immediate objective in the near future that

PAGE 275

275 Similarly, other participants saw the goal of thinking imag es as a final decision, specific solution or end result. For example, Participant 5 3 distinguished between open ended creative x + y only equals z kinda thing , like chemicals mixing and things like that. Participant s working on something that they need to get done P articipant 4 3). Thus, not only did thinking images include visuals of subjects working, but that work seemed to be necessary rather than as a form of enjoyment or a hobby. Additionally, participants pointed out that the subjects in thinking images appeared to be t 6 3) or deciding something specific . Specifically, Participant 4 trying to make a decis ion about Focused subjects. In addition to being logical/rational and process oriented, partic ipants identified the subjects in thinking images as displaying a high degree of focus. In a statement typical of many comments, Participant 9 Like, serious focused not fun focused like a family cooking. Many times, these types of statements seemed to be in reference to images where 3). Usually, these statements referred to images whe re subjects seemed to be engaged in tasks that required a high degree of mental focus, such as business planning, scientific experiments or surgery. In several cases, participants also linked focus to other patterns found in thinking images. For example, s ome participants associated focused images with thinking via logical or step by step processes (discussed above) because the subjects were

PAGE 276

2 76 1). log (Participant 1 2), participants agreed, that the focus was on work or a task, not a relationship or a person. Alternately, some participants noted the connection between the two thinking th emes of put into this [surgery]. I would not want to be cut open by somebody who was dealing with their In fact, an i mage of a surgeon and several nurses operating on a patient elicited the most comments about focus, most likely due to the serious nature of the activity and the single minded focus participants felt was necessary to succeed. Working. Working and job relat ed images were also described as more thinking. In 3) or 2). According to participants, this work and thinking connection was strongest when an image depicted a science or medical related occupation, similar to the process orientation theme discussed above . In fact, in many cases, participants based their decision to sort an image as Participant 8 3 said, Logical again, the whole medical, scientific setting Science, math and labor jobs (Participant 7 3) . Things you do with your mind. Just working in general. Being left brained (Participant 7 1) . In particular, participants identified almost all images depicting medical jobs as thinking a logi c to the practice. Like the farming and the medicine, for me, stand out as (Participant 5 1). Participants also identified a medical practitioner

PAGE 277

277 (Participant 4 3). And, according to Participant 2 Methodical. I a flow chart. . 3). This sentiment suggested that images of medical practi tioners in the midst of a process or treatment might be particularly thinking, especially when little or none of the patient is visible. Along with medical professions, science related occupations were also closely linked to thinking because participants a 1), and being 2). Setting. setting also played an important rol e as participants identified thinking images. 3) feel that (Participant 5 1). They were also visuall 2) and (Participant 3 1) impersonal (Participant 4 1) . A few part icipants also mentioned that, in these settings, computers, notebooks and books could also indicate a thinking image. he's in a lab, they're doing science, they're doing surgery, she 's in a library, they're in a lab. [...] Also, probably a classroom or an office (Participant 3 1) .

PAGE 278

278 1) elicited thinking. Many participants felt that a library and laborator y settings were particularly conducive to thinking. Impersonal. While participants identified some thinking settings as impersonal, other aspects of thinking images also had an impersonal feel to participants. For example Participant 4 This impersonality was often manifested visually through a lack of connection with the viewer 3) ause if 1) . This impersonality tended to produce a generic, disconnected feeling with the subject : posed. (Participant 9 3) Distant. (Participant 9 1) Participants further defined impersonal images as those where the subject is an object or activity, (Participant 9 1). Thus, in thinking images, subjects were more often focused on inanimate objects or abstract ideas. For example, Participant 3 , engagin g an idea , or your occupation , or Body language. Participants also frequently discussed the body language of subjects in thinking images. In addition to not making eye contact, p articipants enumerated several other body language factors that contributed to an image being identified as thinking, including a thinking pose, looking up, staring into space and not smiling. The most commonly mentioned body language associated with think ing was the thinking pose . This pose consisted of a subject looking up and holding his/her chin and may be The Thinker

PAGE 279

279 looking down with his chin on his fist . This pose was so stereotypically thinking, that many participants felt no need to even describe it and merely mentioned the pose in passing. However, Participant 10 2 noted the pose in conjunction with Simil arly to the thinking pose and sometimes overlapping with it some participants linked a subject who was simply looking up with thinking. For example, Participant 2 2 looking up. Even if you remove all of the equations and everything in the background, I could Participants also identified not smiling as another form of body language that indica ted a thinking image. For example, Participant 9 2 specially the surgery one . T Further, if a subject was 1). Color. Participants also commented on the strong connection between certa in color s, or lack of colors, and thinking. In particular, participants identified thinking images as having a sterile, cold or clean feeling as a result of a color palette that was neutral, cool and monochromatic. Specifically, thinking images had a large amount of blue and white: A lot of white, very clean spaces. Every one has a white background (Participant 9 3) . Participants also

PAGE 280

280 2 ). Less frequently mentioned themes. In addition to the common themes discussed above, four other less frequently mentioned themes emerged for thinking. These less frequently mentioned themes include d : being objectiv e , being alone, attire and thought bubbl es shown . One less frequently mentioned pattern that emerged in thinking images was that of follow the rules and not allow emotions to play a role in the decisi on making process. For example Participant 4 In several cases, participants identified images as being mo re thinking if there was a 1). Therefore, a single person indicated that the subject could be on their own in order to reflect or think about something personal, which participants identified as a thinking behavior. A few participants also mentioned tire that indicated an (Participant 9 2). Participant 6 2 also noted that one related attire as 3). Overall, participants seemed to feel that the subjects had 2).

PAGE 281

281 Finally, s ome participants also identified images with thought bubbles as thinking : put this in thinking . I 1) . One parti cipant felt that a thought bubble visually represented thinking about future decisions: (Participant 8 1). In summary, this analysis showed that thinking images were typically logical or rational. This often took the form of tasks or activities that required thought, planning and intelligence. Thinking images were also process oriented, implying the use of intric ate mental thought processes, detailed plans or step by step methods. In many cases, thinking images were also driven by a goal or an end result where subjects were working on a task that had an clear end product or was something that needed to be accompli shed. In thinking images, subjects tended to be highly focused on a single task or activity. In many cases, images of people working were also considered to be more thinking, particularly when the job was science or medical in nature. Particular settings a lso tended to indicate thinking, including those with a sterile, clean feel, such as a library or a laboratory. Further, thinking images often felt impersonal, with no visual connection between the viewer and the subject. Body language, too, played an impo rtant role in determining thinking images. In these images, subjects were often identified as in a thinking pose (i.e., looking up with hand on chin), looking up, staring into space and/or not smiling. Additionally, cool, neutral color palettes with an abu ndance of white and blue also gave thinking images a sterile, clean feel. Thinking images might also have some connection to a single subject portraying objectivity (e.g., following rules, no emotion), being alone and wearing certain attire (e.g., business , glasses). In a few cases, thought bubbles were also connected to thinking.

PAGE 282

282 Feeling (F) The second part of RQ1c ask ed which visual symbolic elements were consistent with feeling (F). Therefore, the researcher examined the images that participants sorted i nto the feeling pile and analyzed the verbal reasons given for the placement. Based on the findings, several common themes emerged, including relationships, emotions, experiencing a moment, enjoyable pastimes, relaxation, creative endeavors, body language, nature, color and lighting. Relationships. In many cases, participants identified feeling images as those that displayed some sort of relationship. For example, participants felt it was important to see connectivity, whether with the subject and the vie wer (Participant 7 2) or that the subject was thing or a person. And then in a few , (Participant 6 2). In fact, participants from most groups id entified relationships or connections as one of the most important or 2) components of a feeling image . 3) and that the not about the thing they re doing, it s more about the camaraderie or their relationship with (Participant 1 2) positive energy unit y (Participant 1 1) 2). Furthermore, the relationships in feeling images were further divided into four categories family, romantic , plutonic and animal. Each typ e of relationship is discussed below. Group image sort participants consistently identified images displaying family relationships as feeling. In the image deck, most family images included a mom, dad, and one or more small children. Participants recognize d family relationships as feeling because they were

PAGE 283

283 1). In fact, in an image of a family painting a wall, one particip ant identified family relationships as a key difference between feeling and thinking images: The people they were doing the activity with. Because painting by yourself would seem more thinking cause your thinking about color that you want, like, the proces s of it, but because they interacted with their kids and included them in a chore , i t (Participant 2 1) Some participants even suggested that they identified images as fee ling almost wholly based on the subject s being a family. For instance, 3) I probably would say feeling. Only because it is a family , and I think that just the aspect of family is more feeling than thinking. (Participant 3 (Participant 3 3) Additionally, some participants identified family relationships as more feeling than romantic r (Participant 7 3). Participant 7 Although some participants identified family relationshi ps as stronger than romantic relationships, many participants still classified images that portrayed romantic relationships as couple and not just one perso 1) 2). Other participants focused on the emotional aspect, 3). Pa rticipants from the third group agreed, saying,

PAGE 284

284 sensual, personal. (Participant 3 1) Smiley. (Participant 3 3) Yeah, the interaction between them and it just looks very personal. And passionate. (Par ticipant 3 1) Although not as strong as family or romantic relationships, plutonic relationships could 3). These plutonic relationships could also occur between friends or coworkers as long as there was a 3). The final type of feeling relationship identified by participants was betwe en a human and an animal. For instance, participants in the third group specifically noted the relational aspect of a man washing his dog: pant 3 1) You can also go get your dog groomed if you want to. Or, you can wash him yourself, which is a 3) You know, their fur in your hands, and your talking to them and playing with them. (Participant 3 1) And, also, (Participant 3 3) Yeah, it seems more like focusing on his relationship with the dog t know, they both look happy. (Participant 3 1) 4 2) , rticipant 5 3). Another participant noted that feeling had to with 2). Some participants even indicated that simply having an animal or a pet in the photo was a strong ind icator of feeling. For example, Those pets, man. (Participant 6 owner smiling , , i [makes happy noise]. (Participant 6 1) All feeling. (Participant 6 3)

PAGE 285

285 Emotions. In addition to images that depicted relationships, participants identified feeling images as portraying emotions. One way in which emotions were portray ed was thinking about emotions or something emotional . For example, one image showed a young man looking at the camera with his arm around a young woman who was looking up at a thought bubble floating over their heads, which contained sketches of a city, a house, a shopping car, a lock and 3 noted: your feelings. T hinking about getting a house together and stuff like that, which is a very emotional kind of thing P arti cipants also noted that the image subjects seemed focused only on emotions . For instance, in an image of a couple cuddling under a tree, participants in the first group pointed out: 1). 2). Based on participant comments, the most emotional feeling images seemed to be those which showed multiple people and no real task. Ins 3). Additionally, participants pointed out that subjects often visually displayed this infatuation by looking at and touching each other. Similarly, in m any cases, participants honed in on the romantic emotions displayed in an image . In response to one image, which showed a man making a salad in the kitchen while a women held a glass of wine and kissed his cheek, Participant 4 2 said, T re . And in some cases, the romantic emotions elicited a personal connection for the participant . For example, while looking at an image of a couple cuddling u nder a tree, P articipant 3

PAGE 286

286 In other cases, participants noted that a depiction of happiness, rather than romance, was the emotion that made them classify an image as feeling. Furthermore, t his happiness was occasionally connected to other positive emotions : 2). Although all the images in the image deck conveyed positive emotions, multiple participants note d that negative emotions (e.g., grief, sadness, fear) would have also been classified as feeling images. For example, participants suggested images of people crying, a team losing a sporting event, a funeral, older people, sick people or a war could also e licit a feeling personality. Experiencing a moment. Participants also identified feeling images as those in which the subjects were experiencing a moment or making a memory with another subject. Related to the relationships theme discussed above, it was al most as if feeling images captured subjects in a particular moment within their relationship. In some images, the moment was among a family. For example Participant 2 2 stated, er. Having a In other cases, participants identified an intimate moment between romantic partners. For instance, p articipants in the eighth group sharing 1) and lo 2). Alternately, the moment could also be shared between a human and an animal where viewers got doing. Like, a little moment between her and the cat (Participant 3 1) . Enjoyable pastime. In addition to relationships and emotions, participants associated feeling images with enjoyable pastimes, hobbies and non work situations. In particular, when the

PAGE 287

287 act of volunteering appeared in an image, participants consistently identified it as feeling , seeing (Participant 9 of person that can feel strongly about those issu 1) would do because, in part, it you to do it unless you have an emotional tie to it 2). Other participants noted that read ing a book was often a relaxing pastime and usual ly constituted a feeling image. this relatively long book. (Participant 5 relaxed in this day, I feel li 1) However, most participants did stipulate that the type of book would make a difference. Only images where subjects seemed to be reading books chosen for fun or relaxation would fall into the feeling category bec 1). For example, Participant 8 e book 2) and the 2) , making it feeling. Relaxation. In addition to participating in enjoyable pastimes, participants indicated that 1). For example, participants in the first group noted the lack of urgency in feeling images: The re just sitting under a tree in the middle of the day (Participant 1 1) . (Participant 1 2) . Additionally, several participants directly compared the relaxation in the feeling images to the lack of relaxation in the thinking images. According to on e participant:

PAGE 288

288 [Feeling images] themselves. And, other things, like math and science and beakers, it looks like so much goes into that, like, so much effort. ( Participant 3 1 ) Creative endeav ors. Participants further identified creativity as a component of feeling images because they thought of that they thought creative thinking could be process oriented. In some cases, the crea tive endeavor depicted in the image was more akin to professional artist 1) . Despite the more professional nature of the subjects, the creativity made the image feeling because partic 1). In an image of two people throwing pottery, (Participant 3 2). The artwork itself was also Rather, than, like , 1). For these participants, artwork see med more about using instinct or following your feelings to create rather than following instructions or a process. In other cases, participants identified more novice forms of art and creativity as feeling. For example, Participant 10 s feeling because they have colored pencils. So S imilar to professional art, participants still identified the freedom and need to follow your feelings in any type of artistic expression: by numbers. ( Participant 4 2 ). Body Language. Participants also noted the importance of body language to feeling images. Based on participant comments, feeling body language was divided into several categories touching, loving looks, smiling and laughing. When all shown together, participants

PAGE 289

289 1) feel. More specifica lly, participants identified images in which the subjects were touching another human or an animal as strongly feeling because touching indicated an emotional or intimate connection : ant 10 2). In addition to touching, participants identified images as more feeling when there was a look of love between subjects : T (Participant 9 1). Participa nt 1 1 agreed even if Another attribute of feeling images was smiling , which visually signified emotion : B 2). Participants in the third group also indicated that the more genuine the smiles looked, the more feeling they communicated. Therefore, while any smile seemed to communicate some degree of feeling, more ge nuine smiles elicited much stronger feeling responses from participants. Finally, when subjects appeared to be laughing in an image, participants were also more likely to classify it as feeling because it conveyed warmth and connection. For example, Parti cipant 9 1 noted that feeling images could show human subjects laughing with their pets because it showed Set in nature. For many participants, images set in nature were also more feeling. The connection be tween nature and feeling was so clear to the majority of participants, that they gave little additional explanation. For example, in a typical comment , Participant 9 1 note d only,

PAGE 290

290 Those participants who were able to explain the conn ection they saw between nature and feeling voiced a variety of opinions. For instance, Participant 9 2 related nature to relaxation because are more stressors indoors. W e automatically associate the outdoors with relaxation and 3 mentioned a connection between nature and romance : Everything ju Finally, participants in the fourth group identified nature as a place people go for emotional clarity, saying, The beach is kinda like a place you go to sort out your feelin gs. (Participant 4 2) Yeah. You go sit on the beach and think about life. (Participant 4 2) Color. Participants also strongly associated feeling images with a specif ic color palette. In particular feeling images were composed of a lot of bright, saturated, warm colors with specific mentions of yellows and greens. For example, p articipants in the fifth group said : the sky got blown out (Participant 5 2) . Really bright colors (Participant 5 1) . And, then the same with the really rainbow colorful one in the kitchen (Participant 5 2) . Other participants focused more on the warm feeling of the color palette , sayin g: not (Participant 7 1), a ( Participants 5 2 ). In several cases, participants noted the presence of warm, yellow tones as particularly indicative of feeling. For instance, Participant 7

PAGE 291

291 association. Typically, w arm colors yellow, orange and red generally suggest energy and passion . While, yellow, in particular has happy, cheerful connotations. Lighting. Beyond particular, some participants describ 1). Other p articipants expanded on the idea of soft light to include low light times of day, mainly sunrise and sunset, implying that those times of day we re more evocative of feeling. For instance , Participant 9 2 felt an image was feeling b ecause of the Participant 4 2 Although less frequently mentioned, several participants also identified f eeling images as having more dramatic lighting to indicate high emotions and passion. For example, participants in the seventh group image sort 3). Less frequently mentioned themes. In addition to the common themes discussed above, two other less frequently mentioned theme emerged . These less frequently mentioned themes inclu de: candid photos and subject attire. In a few cases, participants identified feeling images as more candid and less posed. For example , Participant 10 1 caught this family in a moment. Like it Similarly, Participant 6 2 also thought one

PAGE 292

292 image looked Moreover, p articipants also mentioned that feeling ima ges seemed as if they 3): important to someone. Whereas [thin could throw them away and not feel guilty [laughter]. (Participant 9 3) [ T hinking feel special. (Participant 9 1) Additionally, this pattern also encompassed the idea that feeling images looked les s like a deck were stock photos, participants identified some images as feeling more relaxed, less posed, less professional in short, less like a stock image. For example, Participant 9 3 pointed out that regardless of the quality . Participants themes. For example, Participant 10 2). They also noted that this relaxed clothing style was in direct contrast to the cl othing in thinking images, which was more reminiscent of a uniform and made subjects look stern and rigid. To summarize, this analysis showed that feeling images most often contained relationships whether it was between family members, romantic partners, plutonic acquaintances, or a human and an animal. Feeling image were also rife with emotions,

PAGE 293

293 particularly romantic and happy feelings. Some of the subjects in feeling images were simply experiencing a moment, almost caught in a moment in time, while othe rs were engaged in an enjoyable pastime such as volunteering or reading. Participants also identified subjects engaged communicated a feeling personality through touch ing, loving looks, smiling and laughing. Natural, outdoor settings also contributed to a feeling personality in images. Moreover, specific colors and lighting also indicated an image was feeling. Specifically, bright, saturated and warm colors, including y ellow, were notably feeling. And, low, soft lighting, such as sunrise and sunset were often associated with feeling images. There might also be a connection between feeling images being more candid and less posed. Additionally, few participants mentioned a relaxed or creative attire (e.g., hat, Converse) as contributing to a feeling personality. Thinking/feeling (T/F) summary Overall, thinking (T) and feeling (F) displayed extreme visual differences. While thinking images were logical and rational , feeling images were filled with emotions like happiness and romance. Relationships of all types family, romantic, plutonic and with animals also dominated feeling image with subjects touching, smiling, laughing and looking lovingly at each other. On the other hand, thinking images were impersonal and often showed only a single person. Further, where thinking images depicted subjects following processes and step by step methods, the subjects in feeling images were experiencing a single moment in time o r making a memory. Participants also associated thinking images with goals and end results, specifically noting tasks that were necessary and had clear, immediate outcomes. But, feeling images showed subjects engaged in open ended creative endeavors and ar tistic pursuits that relied on instincts. While thinking subjects were focused and often shown working, particularly in scientific or medical occupations, feeling subjects were relaxing or participating in enjoyable pastimes such

PAGE 294

294 as volunteering or reading . The two dimensions also had different settings, with thinking images taking place in sterile, clean environments like a library or a laboratory, and feeling images set in nature and away from stress. Finally, the colors associated with each dimension wer e also opposite. Thinking images were composed of cool, neutral colors, specifically whites and blues, which gave off a sterile feel. Alternately, feeling images were filled with bright, saturated and warm colors, (i.e., yellow) and soft, low lighting (e.g ., sunrise, sunset), which implied happiness and emotion. The two dimensions did have one theme that might appear to be similar initially attire. However, although attire helped to identify each dimension, the type of clothing worn by subjects was drasti cally different with no actual similarity. While the subjects in thinking images wore business attire, tucked in shirts and lab coats, feeling subjects were dressed in sweaters, funky hats and Converse shoes. The connotation of these clothes was also oppos ite with the clothes in thinking images giving off a stern, rigid feel and the clothes in feeling images seeming relaxed, comfortable and creative. Research Question 2 Visual Symbolic Elements for Multidimensional MBTI Brand Personalities Research Questi on 2 (RQ2) asked how visual symbolic elements from individual MBTI dimensions combine d to portray the eight multidimensional MBTI personalities included in this study extraverted/sensing/thinking (EST), extraverted/sensing/feeling (ESF), extraverted/ int uitive/thinking (ENT), extraverted/intuitive/feeling (ENF), introverted/sensing/thinking (IST), introverted/sensing/feeling (ISF), introverted/intuitive/thinking (INT) and introverted/intuitive/ feeling (INF). RQ2 was answered using the qualitative data ga the red from the group image sorts. T he researcher was able to create image profiles by examining and combining the themes identified in RQ1. Although the themes noted in RQ1 can be combined in large variety of ways

PAGE 295

295 to produce images with specific multidime nsional brand personalities, in most cases, to achieve a multidimensional brand personality, some elements of each individual brand personality had to be compromised. Because of the complex and nuanced nature of visuals, some combinations might work better than others. An example of each multidimensional brand personality combination is below. The researcher created these combinations based on the data and her prior knowledge of the topic (see Table 4 15 ). Table 4 15 . Multidimensional Brand Personality Visu als . Dimensions Elements Extraversion (E) Sensing (S) Thinking (T) Multiple people (E) Collaborating (E) Other people in the background (E) Dressed conservatively (S) In a science or medical job (S/T) Working on a practical task that requires a clear pro cess (S/T) Touching the project being worked on (S) Smiling (E), but looking at task (T) People wearing bright colors (E) surrounded by a white/sterile/neutral/ clean environment (S/T) Extraversion (E) Sensing (S) Feeling (F) Multiple people (E) in a fami ly (F) Other people/families in the background (E) Dressed conservatively (S) Working outside (F) in a public space (E) to complete a practical task with a clear end goal (S) that they chose to engage in (E) Touching each other (E/S) Using their five sense s (S) Smiling a nd look happy (E), but looking at the task (T) Bright, saturated colors (E/F) but a clean, organized environment (S) Extraversion (E) Intuition (N) Thinking (T) Multiple people (E) who look like they know each other an d have a relationship of some type (E/N) Collaborating (E) Other people in the background (E) In a library or laboratory (T) Thinking pose or looking up (T) Thought bubble (N/T) filled with colored illustrations (N) of logical, rational or process oriented thoughts (T) People wearing bright colors (E) surrounded by a white/cool/neutral/ clean environment (T)

PAGE 296

296 Table 4 15 . Continued. Dimensions Elements Extraversion (E) Intuition (N) Feeling (F) Multiple people (E) in a family parents, children and a pet (N/F) Touching each o ther (E/F) Wearing trendy, cute clothing (E) Happy, smiling, laughing (E/F), talking (E) Other people in the background (E) In a public place (E) set in nature (N/F) Participating in an enjoyable activity together (E/F) to create a memorable moment (N/F) L etting kids participate in an adult activity or create some kind of art (N) Bright, vibrant, saturated and warm colors (E/N/F) Introversion (I) Sensing (S) Thinking (T) A single person (I) Immersed in their work (I) Not smiling or making eye contact, but looking at task (I/T) Isolated with no one else visible (I) Dressed conservatively (S) In a science or medical job (I/S/T) Working in a laboratory (T) Working on a practical task that requires a clear process (S/T) Focused on the short term, here and now to achieve an end result (S/T) Touching the project being worked on (S) Quiet atmosphere (I) White/blue/sterile/neutral/muted/clean color palette (I/S/T) Introversion (I) Sensing (S) Thinking (T) (continued) Introversion (I) Sensing (S) Feeling (F) A s ingle person (I) Alone, touching/interacting with a pet (I/S/F) and looking lovingly at the pet (F) Dressed conservatively (S) Immersed in a task with a clear end goal (S) At home with no one else in sight (I) In a clean, organized environment with a white /neutral color palette and simple, straight lined furniture (S) Low lighting (I/F)

PAGE 297

297 Table 4 15 . Continued . Dimensions Elements Introversion (I) Intuition (N) Thinking (T) A single person (I) Thinking pose or looking up (T) Not smiling or making eye con tact (I/T) Isolated with no one else visible (I) In a science or medical job (I/T) Working in a laboratory (T) Wide crop (I) Quiet atmosphere (I) White/blue/sterile/neutral/clean environment (I/T) head (N) depicting abstract scientific ideas (I/T) Introversion (I) Intuition (N) Feeling (F) A romantic couple (I/N/F) Isolated in nature (I/N/F) Quiet atmosphere (I) Smiling and touching (F), but not making eye contact (I) Isolated with no one else vi sible (I) Experiencing a moment (N/F) depicting their shared, imaginary future, including children (I/N/F) Low light (F) Warm, bright, saturated colors (N/F) Extraverted/Sensing/Thinking (ES T) To create an extraverted/sensing/thinking (EST) personality , an image should include multiple people in a science or medical job who are collaborating together on a task. The people should be dressed conservatively (e.g., not showing a lot of skin, not overly trendy, may be business style) but in bright colors. They should also be smiling but looking at the task they are working on. They should be physically touching the task they are working on. The task itself should be a practical task that requires a clear process. In the background, additional people should be visible, also working on tasks. While the people in the image are dressed in bright colors, the surrounding environment should be clean, sterile and neutral with a significant amount of white.

PAGE 298

298 An example of this image might be: Multiple nurses in bright scrubs with long sleeved shirts underneath work ing ng the patient more of a task than a subject). The nurses smile faintly while looking down at a watch or device. They are not making eye contact with each other or the patient clearly focused on the task. In the background, others nurses can be seen work ing together to perform practical tasks in a hospital room that contains several beds and patients (none of whose faces are visible). No one is make eye contact, but everyone looks happy. The room is clean, sterile and white. Extraverted/Sensing/Feeling (E SF) To create an extraverted/sensing/feeling (ESF) personality , an image should include multiple members of a family working outside in a public space to complete a practical task with a clear end goal. The task should be something the family chose to enga ge in. Family members should be dressed conservatively but in bright colors. They should be touching each other and actively working, engaging the task with their five senses. The family members should be smiling and appear happy but looking at their task. The outdoor, natural environment should be filled with bright, saturated colors. However, the environment should also be clean and organized. Other people and families should be visible in the background. An example of this image might be: A family, which includes parents and children , volunteers to pick up trash at an outdoor park. The park is filled with trees, lush grass and some distant playground equipment. Other families are visible in the background also picking up trash. The family members look hap py to be there, smiling, touching each other and interacting but keeping their eyes on the task. They are dressed conservatively in pants, long sleeved work clothes and work gloves. The males have on hats, and the females have their hair tied back in ponyt ails or bandanas. Some family members are bending down to physically touch/pick up

PAGE 299

299 trash. Others hold trash bags. Everything in the environment is organized, and little trash is shown (most of it having already been picked up). Extraverted/Intuitive/Thinki ng (ENT) To create an extraverted/intuitive/thinking (ENT) personality , an image should include multiple people who look like they know each other and have a relationship of some type. The people should be collaborating together in a library or laboratory setting. They should not appear to be working on a task but instead are looking up (some might also be holding their chin in a thinking pose). A thought bubble filled with colored illustrations of logical, rational or process oriented thoughts hovers above their collective heads. While the people in the group are wearing bright colors, their environment (e.g., library, laboratory) should be cool, neutral, clean and white. Other people can also be seen in the background of the photo, although it should be cl ear that those people are not part of the main group. An example of this image might be: Multiple young people collaborate in a library setting. A group of at least 5 young people sit around a single table wearing bright, trendy clothes. Some are touching in a friendly manner (e.g., hand on shoulder, sitting closely, leaning over each other). Thick, textbooks are open on the edge of the table, but no one is looking at them. Instead, everyone is looking up as if in thought. A thought bubble floating above th eir collective heads is filled with brightly colored illustrations of math formulas and symbols. Extr averted/Intuitive/Feeling (ENF) To create an extraverted/intuitive/feeling (ENF) personality , an image should include multiple members of a family with par ents, children and a pet. The family is happy; they are smiling, laughing, talking and touching. All the family members should be wear ing trendy, cute clothing. The family should be in a public place set in nature. They should be participating in an enjoya ble activity together to create a memorable family moment. The parents should be letting

PAGE 300

300 the children create some kind of art or participate (with their help) in an adult activity. The surroundings should be filled with bright, vibrant, saturated and warm colors. An example of this image might be: A family, which includes a mom, dad and two kids, looks happy as they participate in an outdoor painting class at a local park. The park is filled with green trees, lush grass and bright flowers. A large pet dog o n a leash sits near one of the family , watching as he/she paint s . The family is clustered together and the focus of the image, but you can see other class members painting in the background. The family smiles, laughs and talks. The children have t heir own easels and paints, participating like the adults. The mom leans over one child to help. The family is wearing trendy, bright clothing, and the image is highly saturated with bright colors. Introverted/Sensing/Thinking (IST) To create an introverte d/sensing/thinking (IST) personality , an image should include a single person in a science or medical related job working alone in a laboratory. He/she should be dressed conservatively and immersed in his/her work. He/she should not be smiling or making e ye contact with any other subject (including the viewer) but looking and touching the task at hand. The task should be practical and require a clear process to achieve a short term, here and now, end result. The laboratory should feel quiet and isolated, w ith no other people visible. The setting should have a sterile, neutral, muted and clean color palette with a n abundance of white and blue. An example of this image might be: A female scientist is alone in a lab wearing a long sleeved, white lab coat. She also has on blue medical gloves and protective eye goggles. She is focusing on a scientific experiment with a clear process. She is holding a dropper filled with clear liquid, which she is dispensing into a row of test tubes that will then be transferred t o an empty machine waiting next to her. She is looking down at her work, concentrating and not

PAGE 301

301 smiling. No one else is in the room, and it seems quiet and still. The lab is clean and sterile, filled with white and blue hues. Introverted/Sensing/Feeling (IS F) To create an introverted/sensing/feeling (ISF) personality , an image should include a single person alone at home , dressed conservatively in muted colors. Although there are no other human subjects in sight, the person should be lovingly looking at and touching/interacting with a pet. While petting the animal, the person should also clearly immersed in a task with a n obvious end goal (e.g., painting a room, mowing the grass). The lighting in the image should be low , and t he home environment should be cle an and organized with simple, straight lined furniture and a white, neutral color palette. An example of this image might be: A woman sitting at a table with a cat in her lap. She is clearly loving/touching/cuddling the cat while working to complete a cro ssword or jigsaw puzzle. She is dressed conservatively in muted colors. No one else is visible, and the house is clean and modern with a white, neutral, muted color palette. The lights in the house are low with a sunset shown outside the window. Introverte d/Intuitive/Thinking (INT) To create an introverted/intuitive/thinking (INT) personality , an image should include a single person in a science or medical related profession who is isolated and alone in a laboratory with no one else visible. The person sho uld not be smiling or making eye contact with anyone (including the viewer). Instead, he/she should be looking up or striking a thinking pose (looking up with hand on chin). The laboratory environment should seem quiet and have a sterile, neutral, clean co lor palette with an abundance of white and blue. However, a brightly colored, illustrated thought bubble should

PAGE 302

302 image should be widely cropped showing a large portion of the laboratory with the subject relatively small in the frame. An example of this image might be: A single male scientist is alone in a sterile, clean, lab full of white and blue hues. No one else can be seen. He is wearing a white lab coat, striking a thinking pose (lo oking up with hand on chin) and has a serious look on his face. The image is widely cropped so the scientist takes up relatively little room in the frame and much of the lab can be seen. Above his head, floats a brightly colored, illustrated thought bubble depicting abstract scientific ideas (e.g., an idea light bulb, atoms, planets, a microscope, chemical formulas). Introverted/Intuitive/Feeling (INF) To create an introverted/intuitive/feeling (INF) personality , an image should include a romantic couple is olated in nature with no other subjects visible. The atmosphere should seem quiet and still with little noise. Both the subjects should be smiling and touching but not making eye contact with each other or the camera. They should seem to be experiencing a moment together. The lighting should be low, but the image should be filled with warm, bright and saturated colors. A colorful, illustrated thought bubble should hover above the depicting their shared, imaginary future, which includes child ren. An example of this image might be: A boy and girl lay side by side on a colorful, patchwork quilt in the middle of a field as the sun sets. The field is filled with bright, saturated grass and flowers but seems quiet and isolated with no one else in s ight. His arm is around her, and she rests her head on his chest. They are not making eye contact, but they are smiling. A colorfully, illustrated thought bubble hovers over both their heads, depicting their shared dreams for the future a house, kids and a dog. The sunset casts a warm, yellow light across the scene.

PAGE 303

303 Research Question 3 Understanding Images Research Question 3 (RQ3) asked how consumer read and negotiated personality information in images. During the group image sorts, participants were p rimarily concerned with sorting the image deck into individual MBTI personality dimensions. However, as they sorted and discussed each image, they also provided valuable insights into how consumers understand and make sense of images. Therefore, the resear cher answered RQ3 by reviewing the group image sort data for information about how participants read, negotiated and, ultimately, made decisions about each image in the image deck. Three groups of themes were identified . First, three themes emerged concern ing how the meaning negotiations, sort patterns and hatred of stock photos. Second, the researcher identified three themes related to the MBTI personality dimensions used in the sor t dimension overlap, family problems and sensing/intuition (S/N) difficulties. Finally, a number of themes developed regarding how participants derived meaning from the images themselves, including storytelling, prior n personality, self projection and intertextuality . Each set of theme s is explored in detail below . Group image sort themes As participants sorted the image deck, the researcher identified several patterns related to the group image sort process that influ enced how participants assigned meaning to each image. These themes included, meaning negotiations, sort patterns and hatred of stock photos. Meaning negotiations. Each group image sort was composed of three participants who were asked to work as a team to During this process, for most images, participants engaged in meaning negotiations to determine and -

PAGE 304

304 take. T hey also necessitated each participant voicing their opinion and the reasons for that opinion. For example, this process might involve one participant stating their opinion about an isagree and state an opposite opinion (i.e., introvert). The third participant would then either agree with one side or remain momentarily undecided. Next, each participant would make a case for his/her opinion. In the end, the dissenting parties would oft en come around to the other side, creating an agreed upon meaning. Participants in the sixth group provided a quick example of this process for an image that showed a young man holding a tablet and with illustrated ideas in bubbles floating above his head: Sensing. (Participant 6 sensing. (Participant 6 and functional. (Participant 6 3) OK. I ca n see that. (Participant 6 1) In other cases, none of the participants had strong initial opinions. In these cases, they worked together to create meaning by each weighing in and pointing out different aspects of the image. For instance, participants in th e fourth group discussed an image of a woman washing dishes with her cat from multiple angles, eventually agreeing to place the image in the sensing pile: What is she doing? (Participant 4 bigger part of th is picture is her interaction with the cat. (Participant 4 2) Mm hm. (Participant 4 first. (Participant 4 2) But she could be thinking, like, talking to my cat is what I should do every da dishes and talking to your cat. (Participant 4 1) No. (Participant 4 of a neither. (Participant 4 3) Or both. (Participant 4 1) Yeah. (Participant 4 3) (Participant 4 hing dishes. (Participant 4 1) Yeah. (Participant 4 2)

PAGE 305

305 (Participant 4 3) If we had t o force it into a category, it would probably be sensing, but I don't feel like it should be in that category. (Participant 4 2) I would say 4 1) And, she does look very c onservative. I guess, yeah. (Participant 4 2) Sort patterns. Another theme identified during group image sorts was the emergence of sort patterns. As participants worked through each of the three individual dimensional sorts, many groups began to rely on p atterns created by the earlier images to determine the personality of the later images. For example, participants in the seventh group established rules early during their sensing/intuition (S/N) sort about families with children and then stuck to those ru les with little deviation. Specifically, they determined that doing activities with children was not practical, involving children in adult tasks was unconventional and that children evoked creativity. Therefore, if an image had a child in it, they almost automatically sorted the image as intuitive. While some groups unconsciously established patterns, like in the case above, other groups were more aware that they were following patterns. For instance, Participant 8 2 noted, I feel like I am forming sort of a generalization [about where In some cases, it seemed difficult for participants to break sort patterns once they had been established. For example, while looking at an image o f a woman washing dishes with her cat, Participant 7 2 ). However, in other cases, participants acknowledged the difference more easily and simply moved on. For instance, while discussing one image that showed a group of scientists holding brightly colored beakers of liquid, Participant 3

PAGE 306

306 Hatred of stock photos. A nother common theme of stock photos. As they examined images, participants often bluntly expressed thei r dislike of (Participant 3 2). Participants also discussed the unrealistic feel of some stock photos, refer 1) and 3). Sometimes, participants even acknowledged the subjects as playing real scientist 2). (Participant 8 2) nature of the stock photos also made it harder for participants to determine an ima couple daydreaming about their future, Participant 3 However, despite many participants general dislike of stock images and their admission that the staged nature sometimes made it hard to identify a personality, it did not seem to significantly affect the perso nality sorting. In most cases, one participant mentioned disliking an image; the whole group agreed and lamented the nature of stock imagery; and then the sort moved on with the group doing their best to look beyond the fact that it was a stock image. For example, while trying to determine the personality of an image of a family painting an interior 2).

PAGE 307

307 Although the theme of hating sto ck photos was quite common during the group image study, participants were either advanced advertising or graphic design majors. Therefore, they might have bee n more familiar with stock photos, better able to identify them, and quicker to dislike them. MBTI personality dimension themes After reviewing the group image sort data, the researcher recognized three additional themes related to the MBTI personality dim ensions used in this study. These themes included, dimension overlap, family problems and sensing/intuition (S/N) difficulties. Dimension overlap. Another pattern that emerged during group image sorts was the overlap of multiple MBTI dimensions. As partici pants discussed the same images multiple times instructions to think about each sort separately. For example, during the thinking/feeling (T/F) sort of an image that sho wed a woman sitting at the beach with her dog, Participant 7 3 noted pe rsonalities aligned: I would put introverted with sensing and thinking. (Participant 1 1) These three [INF] boom, boom, boom all the way across. (Participant 1 3) I would do introverted, sensing, thinking and then extroverted, intuit ion, feeling. (Part icipant 1 2) Family Problems. In addition to some dimensions overlapping, participants also had some problems deciding where to place images of families. While most groups were able to easily place family images in feeling (F) rather than thinking (T), som e groups struggled with sorting family images as either sensing (S) or intuition (N). However, these family problems

PAGE 308

308 were, by far, most evident and consistent during the extraversion/introversion (E/I) sorts. (Participant 6 2). rticipant 10 2). Thus, they were together because they were family. And, this lack of choice made it harder for participants to decide whether a family image was extraverted or introverted. Some participants found this sentiment to be particularly true whe n young children were included as part of a family image. Participants felt that because the children were unable to be alone, they were unable to act on their extraverted or introverted tendencies. For instance, in an image of a woman sitting next to a sm all boy as he colored, participants in the second group noted, that [child], right? Be nd the world. (Participant 2 3) Additionally, participants often mentioned that families made it particularly hard to woul d be comfortable around his/her family members. For instance, while discussing an image of a family cooking in a kitchen, Participant 9 out to see other people. While discussing the problems associated with family images, several groups suggested could be a 3). Alternately, family images could more easily be labeled

PAGE 309

309 (Participant 3 3). Similarly, participants felt that families who were in their own home were more likely to be introverted, while families that were in public were more likely to be extraverted because they chose to be in public. For instance, Participant 2 doing something. They coul d be at Disneyland, anywhere, but they chose to stay home and have Sensing/intuition (S/N) difficulties. Although images of families caused some confusion during group image sorts, participants noted that, overall, the se nsing/intuition (S/N) dimension was the most difficult for them to understand and assign to images because it was hard for them (Participant 6 1). Participants also felt like the sensing 2). In some cases, the sensing and intuition dimension labels proved confusing for participants. For example, Participant 3 3 was confused by the word sensing because it made her Similarly, Participant 10 picture aspe ct] into my definition with the other sensing and intuition which are significan tly different than how the words are used in common vernacular. Image related themes Finally, a number of themes also emerged regarding how participants derived meaning from the images themselves. These themes included, storytelling, prior knowledge, subje personality, own personality, self projection and intertextuality .

PAGE 310

310 Storytelling. One of the most common patterns observed during the group image sorts was storytelling. As participants attempted to understand and negotiate the meaning of each image an d determine its personality, they constructed elaborate stories to provide the images with additional meaning. And, these stories were not isolated incidences. Dozens of stories were told in each group, and almost every image had at least one story told ab out it. 3). For example, during the third and final sort, when group five came to an image that showed a male and female scientist smil ing at the camera while working together on a task, Participant 5 2 noted how differently he felt about the image now smiles might indicate they were having an of Moreover, participants in the same group sometimes told differ ent stories for the same image. For example, two participants in one group were both looking at an image of a couple making dinner in the kitchen. Participant 9 to go out so they just quickly fixed a sal ad, while Participant 9 example, a fter the conversation above, Participant 9 3 felt the image was sensing and Participant 9 2 felt the image was intuitive. In some cases, participants created stories after they determined the personality. In these instances, the stories seemed designed to bolster the personality they had already decided on.

PAGE 311

311 While looking at the same image of a couple cooking together in their kitchen, another group unanimously decided the image was introverted, then provided a story to backup their choice: They obviously ha ve discretionary money. They have a chandelier. [group (Participant 2 house. So they just recharge their batte ries by having a quiet dinner at home. (Participant 2 2) In other cases, when they felt the image lacked a clear narrative of its own, participants attempted to build their own story. For example, while discussing an image of a dad helping a small boy cut peppers with a sharp knife, the tenth group decided the only reason that such a (Participant 10 2). Participants also told stories to help them explain and identify the personality of images 3) in an imag e of family cooking around a stove. After discussing voiced more than th 2). The other two members immediately agreed, sounding almost relieved because they now had a reference point to understand the scene. Some participants told their stories through the eyes of the image subjects. For instance, Participant 1 1 created a story through the lens of the mother in the image who was helping her to blossom into something beautiful, you know, a bush or a flower or something like that. And,

PAGE 312

312 Similarly, in other cases, participants created stories that speculated about the future of an 2). Alternately, some of the subjects. In an image of a man washing his dog, participants in the first group took into account a 1). Other times, participants based their stories on surface level visual factors. For example, P articipant 10 10 (Participant 10 2). In other cases, participant s used clothing as the basis of their story, but the building a playground, Participant 5 And flats. A Another interesting finding was that once a story was established, it often followed the group through each of the remaining sorts. For example, early on, participants of the ninth group imagined an i

PAGE 313

313 2). Then, when the same image came up again in the next personality sort, the story continued, still gu iding how they viewed and sorted the image: This is probably kind of practical because they are studying for something and getting things done. (Participant 9 ormed us ahead of time that it was actually a brainstorming for a start up company or something, I would see that as more creative. (Participant 9 3) On a few occasions, participants told stories that were peculiar and a bit far fetched. For example, parti cipants in the fifth group believed the dog that a man was washing was actually a wolf. And, despite the strangeness of seeing a dog as a wolf, the story persisted through the 5 2) and he had to wash the animal himself instead of taking it to the groomer because groomers Although most of the unusual stories were isolated, participants from multiple groups told the same unexpected story in a few instances. For example, while looking at an image of a young couple daydreaming about their future, multiple groups told similar stories about the female being a gold digger because the male was looking at the camera, while she was looking up at the th ought bubble over their heads, which contained sketches of a city, a house, a dollar the couple imagining their future together like most participants, a few participants said the because she 3). In some cases, participants told stories that were completely unexpected, running opposite to the personality dimension the researcher predicted. And yet, the story was told

PAGE 314

314 consistently. For instance, the researcher labeled one particular image of a girl sitting of the floor of a library with her laptop and books as both introverted and thinking. However, several groups decisions w ere based on stories they told of the girl probably being on Facebook instead of studying and sitting on the floor with her legs stretched out into the aisle. Participants also talked 3) and 2) that Participant 3 3). Participants then used these details to tell 3) or she had dresse 3). Prior knowledge. In addition to telling stories about images, participants also used their immediately app 2). personality. Specifically, participants used prior knowledge associated with personal expe riences, important others and general information. When negotiating image meaning, participants often relied on personal experiences to ipant 2 1 related an image of a young man and woman throwing pottery t o a recent experience with his girlfriend: and I did something similar in D.C. over winter break. It was like this couples thing where a bunch of people all did something lik e this. We created art, drank wine,

PAGE 315

315 company of others. Other participants relied on persona l experiences related to topics such as their school major, childhood, family and animals to determine image personalities. college major, which, for this study, was either advertising or graphic design. For example, while looking at an image of a female scientist working in a lab, Participant 3 1, an advertising major, little bi Similarly, those participants who were graphic design majors also made assumptions about images based on their major. For instance, Participant 6 2, a graphic design major, saw similarities between her major and an image showing a group of young adults collaborating on a project with at the same time, just knowing our major, when you c had very specific thoughts on an image depicting a young man and woman throwing pottery. Because of their prior kn owledge on the subject, which advertising majors did not have, these participants were able to make a fine distinction between pottery as a craft rather than an art icipant 5 2). Along with their school major, participants also used their childhood personal experiences to read and negotiate image personalities. Often, when an image showed a child, participants conjured up their own childhood memories to understand the image and assess its personality. For example, in an image of a dad helping a young child cut vegetables, participants

PAGE 316

316 in the second group discussed a shared personal childhood experience about kitchen knives. One 2). In reaction to another image of a family painting a wall in their house, one group expressed different childhood experiences: (Participant 3 3) Oh, my family always made that a project. And, I hated it when I was little. (Participant 3 2) [group laughter] I was not allowed to go near the walls, ever! Ev (Participant 3 3) In addition to using childhood memories, participants also relied on experiences in their current family life. For instance, while discussing an image of a family coo king together in the kitchen, two participants in the tenth group both used their own families to provide very different illustrations of how an introverted family might look, demonstrating the role personal experiences play in reading and negotiation imag e personalities. Participant 10 introverted family wa Finally, participants also used their personal experiences with animals to interact with images and determine personality. In many cases, participants related images back to positive experiences with pets. For example, participants in the sixth group identified an image of a man ause dogs (Participant 6 1). But, while dogs seemed to elicit positive memories, cats, in particular, seemed to bear the brunt of the negative associations. For instance, participant s in the fifth group factored their own dislike of

PAGE 317

317 rticipant 5 2). Participants also seemed to have strong personal opinions on horses and what they implied about the people associated with them . Participants tended to use the s e personal experiences with horse pe personality to an image of a girl standing next to her horse as it grazes in an open field. For instance, different participants identified horse people as passionate, impractical, quiet an d focused. However, although many participants seemed to have strong opinions about horse people based on their personal experience with horses or knowing someone who loved horses, it was not always similar and did not always lead to the same personality c onclusions. In addition to personal experiences, a second type of prior knowledge that participants either friends or family members. For instance, while discussing an image depicting a surgeon performing surgery, Participant 2 times, participants relied on family members to assign an image personality. For instance, while discussing the same surgical imag e, Participant 6 However, like with personal experiences, somet imes participants had prior knowledge from the same family acquaintances but came to opposite personality conclusions. For example, two participants in different groups both had family members who were farmers. Yet, when viewing an image of a farmer crouch ed in a wheat field, one participant found the image to be and now, practical, how much effort goes into making

PAGE 318

318 3). However, the other 2). Finally, in addition to using prior knowledge based on personal experience and important others, pa rticipants sometimes relied on general information that they had gathered throughout their information but seemed to rely on what they felt to be common un derstandings of how the world works. However, there was no guarantee that their common knowledge was the same as another participants. Most often, participants seemed to apply this general information to their understanding of medical professions, farmers and horse people. For example, several participants commented in general terms about the requirement for surgeons to be unemotional, detached, methodical and studious, which made them introverts. In a comment typical of many, while looking at an image of s urgeons performing surgery, Participant 10 1 noted that doctor Alternately, when participants had no prior knowledge from personal experiences, important others or general information to use as frame of reference, some participants found it hard to assign a personality to images. For example, Participant 3 1 noted, al experience with. Like the farming, I have absolutely zero g about that. (Participant 3 1) Beyond storytelling and prior knowledge, participants also relied , participants found it difficult, if not impossible, to separate the personality of an image from the

PAGE 319

319 personality of the subjects shown in the image. Despite initial instructions and numerous redirects from the moderator, participants often determined the people in it. Often participants seemed to forget or, perhaps, never understand that they were supposed to be basing their personality judgments on the whole image. For example, while discussing an image of a man washi ng his dog, Participant 2 3). However, when the moderator reminded him to concentrate on the personality of the image, not the person in the image, he quic kly identified the image as introverted. the little information cont ained in a single image. Many participants were torn between that snapshot and what the subject might be like at other times. For example, while looking at an image that showed a couple cuddling under a tree, participants in the first group image sort said , 2). Own Personality. While some participants used the subjects in images to guide their personality decisions, other part icipants used their own personalities to interact with and Although personalities probably interacted with and influenced their views of the image personalities to some degree, a few participants were self aware enough to expressly mention the connection between their own MBTI personality and that of the image. For example, Participant 5 2 noted

PAGE 320

320 Some participants seemed to use their own personality as a barometer of sorts, gaging whether an image was like them or not. Images that were like them received the same person ality as them, and those that were not were assigned the opposite dimension. For example, in a discussion about the overall sorting process at the end of the group image sort, Participant 3 1 noted, But in each category, even though I honestly tried not to , I kinda placed which one hat so it must be thinking. For instance, Participant 5 2 noted that an image of surgeons conducting surgery was thinking 5 nervous! I cou 2 took this as proof that his assessment of the Self projection. For example, while looking at an image of a man holding a tablet with thought bubbles floating over his head, Participant 7 1 stated, situations, and if I was in his situation, are doing, which makes me feel extraverted. While some participants understood their self placement, other participants seemed to inst while discussing an image of a woman and her dog sitting on the beach, Participant 10

PAGE 321

321 I was sitting on the beach with the sunset, I would only be focused o n the sand and the sun Intertextuality. Participants also read and negotiated image personality using intertextuality, or the connecting of one text to another (Chandler, 2013). Specifically, during the group image sort, pa rticipants used other familiar texts including popular books, TV shows and movies to understand images in the image deck and assign them personalities. For example, while looking at an image of scientists holding brightly colored beakers, Participant 2 2 said, 3 thought that sons included references such as: The Brady Bunch (TV show): Participants in the second group thought an image showing a family (i.e., mom, dad and two children one girl and one boy) painting a wall in their house looked like a Brady Bunch ticipant 2 3). Ghost (movie): While looking at an image of a young man and a young woman throwing Ghost (Participan t 2 2). And, because of the connection with the movie, Participant 6 1 classified the images as feeling. Twilight (book and movie series): In response to one image, where a young woman is sitting under a tree in a field of grass reading a book, Participant 1 Taylor Swift (musical artist): Immediately upon seeing an image of a young, female scientist working al one in a laboratory, Participant 10 In addition to references for individual images, participants from the first group also used intertextuality to explain the entire thinking/feeling (T/F) dimension in terms of characters from a popular TV show, The Big Bang Theory

PAGE 322

322 Research Question 3 (RQ3) summary In summary, this analysis identified multiple themes regarding how participants read and negotiated personality information in images. These themes were divided into themes that related to the group image sort process, themes that focused on the MBTI perso nality dimensions and themes regarding how participant derived meaning from the images themselves. Some of the themes were related to how the group image sort process influenced terns and hatred of stock photos. During the sorting process, the group dynamic encouraged meaning negotiations where the three participants would give and take until they came to a single acceptable personality for an image. Participants also established sort patterns, in which they identified certain characteristics that they felt embodied a personality dimension and then sorted according to those qualities rather than examining the details of each image separately. Finally, while viewing the image deck, many participants expressed hatred of the stock photos they saw. In particular, participants disliked the unrealistic and posed nature of stock photos. In part, this aversion to stock photography may have stemmed from their majors in school, graphic design and advertising, which are more visually literate and aware. Three other themes found during the analysis were related to the MBTI personality dimensions used in the sorts, including dimension overlap, family problems and sensing/intuition (S/N) difficult ies. During the sorts, some participants had difficulty separating each MBTI dimension from the others, and referred back to how an image was sorted in a previous dimension or commented about how multiple dimensions seemed similar. Participants also had pa rticular trouble sorting images of families into either extraversion (E) or introversion (I), noting the complications of having multiple people in an image and determining personality based on how people acted around their immediate family members. Lastly , the sensing/intuition

PAGE 323

323 (S/N) dimension was, by far, the hardest dimension for participants to understand and assign to images. For some, the dimension labels of sensing and intuition were confusing because the MBTI definitions were different than the comm on definitions they knew. Others understood the definitions, but found it harder to apply to visual characteristics. Finally, a number of themes developed regarding how participants derived meaning from the images themselves, including storytelling, prior personality, self projection and intertextuality . While assigning personality to images, participants often told elaborate stories to bolster the personality they had already decided on or to provide context when they participants told similar stories about an image, while, other times, the stories were very different. Stories could also be based on the past, present or future surround ing the image. Participant s also used their prior knowledge associated with personal experiences (e.g., school major, childhood, family, animals), important others (e.g., family members, friends, coworkers) and general information (i.e., information they felt was common knowledge) to help determine rsonality they assigned the overall then determining the image pe rsonality based on those self projections. Lastly, participants also connected the images to other familiar texts, including popular books, TV shows and movies, to

PAGE 324

324 Research Question 4 Most Repres entative Images Research Question 4 (RQ4) asked which images group image sort participants selected as the most representative for each multidimensional MBTI personality included in the current study extraverted/sensing/thinking (EST), extraverted/sensin g/feeling (ESF), extraverted/intuitive/thinking (ENT), extraverted/intuitive/feeling (ENF), introverted/sensing/thinking (IST), introverted/sensing/feeling (ISF), introverted/intuitive/thinking (INT) and i ntroverted/intuitive/feeling (INF) . After each sort , the researcher recorded the individual personality dimensions extraverted (E) or introverted (I), sensing (S) or intuition (N), and thinking (T) or feeling (F) participants assigned to each image. Next, for each image, the three individual dimensions one from each sort example, if Image 1 was sorted into the extraverted (E) pile during the first sort, the sensing (S) pile during the second sort and the thinking (T) pile dur ing the third sort, Image 1 would be assigned an extraverted/sensing/thinking (EST) multidimensional personality. Then, to answer RQ4, after all 10 group image sorts were conducted, the researcher analyzed the number of times a group assigned each image a particular multidimensional personality. For each of the eight multidimensional personalities, the researcher selected the image that the most groups had identified to use in the phase three survey. Secondarily, the researcher also checked to see if the to choices, two images were selected for the survey: 1) the top participant choice overall and 2) the top particip ant choice that overlapped with the images in these cases, the researcher was able to compare the results of images selected through the group image sort process and images selected by the researcher to be tter understand the

PAGE 325

325 effectiveness of the group image sorts process in selecting effective advertising images with specific MBTI personaliti es. If a tie arose, the researcher then looked at whether any of the images were selected by participants as the most representative of any single dimension. The images selected for each multidimensional MBTI personality are discussed in detail below. Images, sorted by their unique identification number, can be viewed in Appendix H . Extraverted/Sensing/Thinking (EST) Res earch Question 4a (RQ4a) asked which image participants selected as most representative for the extraverted/sensing/thinking (EST) personality. According to group image sort results, Image 10, which pictured a male and female scientist working in a lab and smiling at the camera, was assigned an EST personality by more groups (60%) than any other image (see Table 4 16 ). EST. Therefore, Image 10 was selected to represent the EST personality in the phase three survey. Table 4 16 . Images sorted as extraverted/sensing/thinking (EST). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a group Extraverted/Sensing/Thinking (EST) 10 1 * 6 11 1 2 Note: 1 Image overlaps with the resea *Image selected for use in the phase three survey. Extraverted/Sensing/Feeling (ESF) Research Question 4b (RQ4b) asked which image participants selected as most representative for the extraverted/sensing/feeling (ESF) personality. According to group image sort results, five different images were assigned an ESF personality by the same number of groups (20%) (see Table 4 1 7 ). Because one of those images, Image 9, which pictured a family

PAGE 326

326 personality assignment of ESF, that image was chosen to represent the ESF personality in the phase three survey. Table 4 17 . Images sorted as extraverted/sensing/feeling (ESF). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a group Extraverted/Sensing/Feeling (ESF) 1 2 5 2 9 1 * 2 12 2 20 2 Note: 1 Image overlaps with the resea *Image selected for use in the pha se three survey. Extraverted/Intuitive/Thinking (ENT) Research Question 4c (RQ4c) asked which image participants selected as most representative for the extraverted/intuitive/thinking (ENT) personality. According to group image sort results, Image 6, whi ch pictured a group of young adults collaborating on a project, was the only image that groups assigned an ENT personality. Results showed that four different groups (40%) assigned the image an ENT personality. Image 6 also overlapped with the original personality assignment of ENT. Therefore, Image 6 was selected to represent the ENT personality in the phase three survey. Extraverted/Intuitive/Feeling (ENF) Research Question 4d (RQ4d) asked which image participants selected as most representat ive for the extraverted/intuitive/feeling (ENF) personality. According to group image sort results, Image 8, which pictured a family (mom, dad, small daughter) outside watering a garden, was assigned an ENF personality by more groups (70%) than any other i mage, but did (see Table 4 18 ). Therefore, in order to compare the effectiveness of images chosen by participants with images chosen by the researcher, two images were selecte d to represent the ENF personality.

PAGE 327

327 Because participants ranked it as the top overall image, Image 8 was selected as the first ENF image (i.e., ENF participants). For comparative purposes, a second image was chosen that was both ranked highly by participan Image 1 and Image 3, were each selected by five groups (50%) and both overlapped with the archer examined whether participants selected either image as the most representative of a single dimension. Image 1 was selected as most representative of the sensing (S) dimension one time and the intuitive (N) dimension two times. Alternately, Image 3 w as selected as most representative of the intuitive (N) dimension three times and the feeling (F) dimension one time. Image 1 was selected as a most representative image less total times than Image 3, including once for opposite dimension of those desired sensing (S). Therefore, Image 3, which pictured a woman and small boy drawing sitting at a table drawing with colored pencils, was selected as the second ENF image (i.e., ENF researcher) because it represented the top participant choice that also overlap Table 4 18 . Images sorted as extraverted/intuitive/feeling (ENF). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a group Extraverted/Intuitive/Feeling (ENF) 1 1 5 3 1 * 5 4 4 5 6 7 3 8* 7 9 6 12 6 Note: 1 Image overlaps with the resea *Images selected for use in the phase three survey.

PAGE 328

328 Introverted/Sensing/Thinking (IST) Research Question 4e (RQ4e) asked which image participan ts selected as most representative for the introverted/sensing/thinking (IST) personality. According to group image sort results, Image 22, which pictured a female scientist working alone in a lab while looking at her task, was assigned an IST personality by more groups (90%) than any other image (see Table 4 19 Therefore, Image 22 was selected to represent the IST personality in the phase three survey. Table 4 19 . Imag es sorted as introverted/sensing/thinking (IST). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a group Introverted/Sensing/Thinking (IST) 2 2 11 5 18 4 22 1 * 9 23 1 4 24 1 4 Note: 1 Image overlaps with the resea *Image selected for use in the p hase three survey. Introverted/Sensing/Feeling (ISF) Research Question 4f (RQ4f) asked which image participants selected as most representative for the introverted/sensing/feeling (ISF) personality. According to group image sort results, Image 21, which pictured a woman washing dishes in her kitchen while looking at her cat, was assigned an ISF personality by more groups (50%) than any other image (see Table 4 20 Ther efore, Image 21 was selected to represent the ISF personality in the phase three survey.

PAGE 329

329 Table 4 20 . Images sorted as introverted/sensing/feeling (ISF). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a g roup Introverted/Sensing/Feeling (ISF) 19 1 3 20 1 4 21 1 * 5 Note: 1 Image overlaps with the resea *Image selected for use in the phase three survey. Introverted/Intuitive/Thinking (INT) Research Question 4g (RQ4g) asked which image participants selected as most representative for the introverted/intuitive/thinking (INT) personality. According to group image sort results, Image 16, which pictured a man looking up at a chalkboard filled with drawings/ideas, was assigned an INT personality by more groups (70%) than any other image (see Table 4 21 of INT. Therefore, Image 16 was selected to represent the INT personality in the phase three survey. Table 4 21 . Images sorted as introverted/intuitive/thinking (INT). Multidimensional MBTI Personality Unique image identification number Number of times sorted b y a group Introverted/Intuitive/Thinking (INT) 2 2 15 2 16 1 * 7 18 1 2 Note: 1 * Image selected for use in the phase three survey. Introverted/Intuitive/Feeling (INF) Research Question 4h (RQ4h) asked which image participants selected as most representative for the introverted/intuitive/feeling (INF) personality. According to group i mage sort results, Image 17, which pictured a girl reading a book alone while sitting outside under a tree, was assigned an INF personality by more groups (80%) than any other image, but did not , which was INT (see Table 4 22 ).

PAGE 330

330 Therefore, in order to compare the effectiveness of images chosen by participants with images chosen by the researcher, two images were selected to represent the INF personality. Because participants ranked it as the top o verall image, Image 17 was selected as the first INF image (i.e., INF participants). For comparative purposes, a second image was chosen that was both ranked highly by mages, Image 13 and Image 15, were each selected by six groups (60%) and overlapped with the whether participants selected either image as the most representat ive of a single dimension. Image 13 was selected as most representative of the intuitive (N) dimension two time s and the feeling (F) dimension nine times. Alternately, Image 15 was selected as most representative of the extraverted (E) dimension one times, the introverted (I) dimension four times, the intuitive (N) dimension three times and the feeling (F) dimension one time. Image 15 was selected as a most representative image less total times than Image 13, including once for opposite dimension of those d esired extraversion (E). Therefore, Image 13, which pictured a romantic couple (male and female) cuddling alone under a tree, was selected as the second INF image (i.e., INF researcher) because it represented the top participant choice that also overlapp ed with the

PAGE 331

331 Table 4 22 . Images sorted as introverted/intuitive/feeling (INF). Multidimensional MBTI Personality Unique image identification number Number of times sorted by a group Introverted/Intuitive/Feeling (INF ) 2 2 13 1 * 6 14 1 4 15 1 6 17* 8 19 3 21 2 23 2 Note: 1 Image overlaps with the resea *Images selected for use in the phase three survey. Research Question 4 (RQ4) summary In summary, based o n the results of the group image sorts, 10 images were sel ected for use in the phase three survey. Table 4 23 provides a brief description along with the unique image number for each selected images.

PAGE 332

332 Table 4 23 . Images selected for phase three survey . Unique Number Personality Dimension Image description 1 0 E xtroverted/sensing/thinking (EST) Photo of a male and female scientist working in a lab and smiling at the camera 9 Extroverted/sensing/feeling (ESF) Photo of a family (mom, dad and two kids) painting a wall in their home 6 Extroverted/intuitive/thinkin g (ENT) Photo of a group of young adults collaborating on a project 8 Extroverted/intuitive/feeling (ENF) participant s Photo of a family (mom, dad, small daughter) outside watering a garden 3 Extroverted/intuitive/feeling (ENF) researcher Photo of a woman and small boy drawing with colored pencils at a table 22 Introverted/sensing/thinking (IST) Photo of a female scientist working alone in a lab while looking at her task 21 Introverted/sensing/feeling (ISF) Photo of a woman washing dishes in her k itchen while looking at her cat 16 Introverted/intuitive/thinking (INT) Photo of a man looking up at a chalkboard filled with drawings/ideas 17 Introverted/intuitive/feeling (INF) participant s Photo a girl reading a book alone while sitting outside un der a tree 13 Introverted/intuitive/feeling (INF) researcher Photo of a romantic couple (male and female) cuddling alone under a tree Note: Images can be seen in Appendix H . Phase Three Survey After completing the scale reduction and the qualitati ve group image sorts, the final phase of this study was to conduct a quantitative survey. Using the BHPI scale, the online survey asked respondents to evaluate both their own personality and the brand personality of 10 advertisements. For each advertisemen t, respondents were also asked to answer questions about

PAGE 333

333 attitude toward the ad ( A Ad ), attitude toward the brand ( A Brand ) and purchase intent ( PI ). Last, the survey asked several demographic questions (e.g., gender, age, race, major). The 10 survey adverti sements contained images that were identified during the group (see RQ4) . Combining the data from the three individual MBTI personality dimension sorts, the researcher selected the image that were most often identified for of the eight multidimensional MBTI personalities included in this study, including extraverted/sensing/thinking (EST), extraverted/sensing/feeling (ESF), extraverted/intuitive/thinking (ENT), extraverted/intuitive/feeling (ENF), introverted/ sensing/thinking (IST), introverted/sensing/feeling (ISF), introverted/intuitive/thinking (INT) and introverted/intuitive/feeling (INF). For six of the eight personalities, the image most often identified for each multidimensional personality by the partic dimensions extraverted/intuitive/feeling (ENF) and introverted/intuitive/feeling (INF) the participant identified ima ge differed from the researcher identified i mage. Therefore, in these two cases, the researcher opted to test both images in order to compare group image sort choices , resulting in a total of 10 images. The researcher then created 10 survey advertisements using the 10 ide ntified images. See Table 4 24 for the full name and brief description of each advertisement along with an abbreviation, which will be used during the remainder of this chapter.

PAGE 334

334 Table 4 24 . Survey advertisements. Advertisement full name Advertisement abbreviation Advertisement description 1 Extroverted/sensing/thinking advertisement EST ad Photo of a male and female scientist working in a lab and smiling at the camera 2 Extroverted/sensing/feeling advertisement ESF ad Photo of a family (mom, dad an d two kids) painting a wall in their home 3 Extroverted/intuitive/thinking advertisement ENT ad Photo of a group of young adults collaborating on a project 4 Extroverted/intuitive/feeling participant advertisement ENF participants ad Photo of a family (m om, dad, small daughter) outside watering a garden 5 Extroverted/intuitive/feeling researcher advertisement ENF researcher ad Photo of a woman and small boy drawing with colored pencils at a table 6 Introverted/sensing/thinking advertisement IST ad Photo of a female scientist working alone in a lab while looking at her task 7 Introverted/sensing/feeling advertisement ISF ad Photo of a woman washing dishes in her kitchen while looking at her cat 8 Introverted/intuitive/thinking advertisement INT ad Photo of a man looking up at a chalkboard filled with drawings/ideas 9 Introverted/intuitive/feeling participant advertisement INF participants ad Photo a girl reading a book alone while sitting outside under a tree 10 Introverted/intuitive/feeling researcher advertisement INF researcher ad Photo of a romantic couple (male and female) cuddling alone under a tree Note: Advertisements can be seen in Appendix E.

PAGE 335

335 After collecting the survey data, the researcher then ran a series of statistical tests to examine multidimensional MBTI personality dimensions are described. Potentially confounding variables are then examined. Finally, an analysis of each individual hypothesis and research question is provided. Survey Respondents The survey was completed by 526 respondents and had a completion rate of 91.0% . A series of descriptive stati breakdown (see Table 4 25 ). The researcher used the minimum 95% confidence value of 4.23%, rounded up to 5%, to collapse small categories. Therefore, categories with less than 26 respondents , which is 5% of the total number of respondents ( n = 526 ) , were grouped and labeled other . The large difference in the number of female and male respondents was expected given the gender proportions in communication related majors. The majority of partici pants were non Hispanic white females who were seniors in college and majoring in t elecommunication. Participants were an average of about 21 years old ( M = 20.54, SD = 1.68), and t he most common 3 dimension MBTI personality type was ENF.

PAGE 336

336 Table 4 25 . Ove rall Demographics for Survey Respondents . Variable N % Gender 1 Female 384 73.0 Male 142 27.0 Major 1 Advertising 104 19.8 Journalism 52 9.9 Public Relations 97 18.4 Telecommunication 149 28.3 Other 124 23.6 Race 1 African American/Black 44 8.4 Hispanic/Latino/Latina 106 20.2 White (Non Hispanic) 346 65.8 Other 30 5.7 Year in school 1 Freshman 85 16.2 Sophomore 132 25.1 Junior 171 32.5 Senior 133 25.3 O ther 5 1.0 3 Dimension Personality Type 1 EST 63 12.0 ESF 34 6.5 ENT 78 14.8 ENF 184 35.0 IST 44 8.4 ISF 26 4.9 INT 38 7.2 INF 59 11.2 Note: 1 n = 526 Scale Reliability The current study involved a scale reduction in phase one, which resulted in the Brand and Human Personality Indicator (BHPI) , and then the use of the reduced scale in the phase three survey Each respon dent took the BHPI to measure their own human personality and the brand

PAGE 337

337 personality of 10 advertisements. Because the BHPI is composed of four independent subscales extraversion/introversion (E/I), sensing/intuition (S/N), thinking/feeling (T/F) and judg ing/perceiving (J/P), the alpha values were calculated separately for each subscale and are presented below. extraversion/introversion (E/I) subscale of the BHPI (see Table 4 26 ). The alpha values ranged personality. The average alpha value for the extraversion/introversion (E/I) subscale across all 11 applications was .586. All alpha values were value for research. Table 4 26 . Stimuli Human Personality 1 .741 INT ad 1 .636 INF participant ad 1 .636 ISF ad 1 .601 ESF ad 1 .592 ENF researcher ad 1 .587 ENT ad 1 .580 INF researcher ad 1 .524 ENF participant ad 1 .522 EST ad 1 .518 IST ad 1 .513 Note: 1 n = 526. sensing/intuition (S/N) subsc ale of the BHPI (see Table 4 27 ). The alpha values ranged from a low of .417 for the IST advertisement to a high of .663 for the INT advertisement. The average alpha value for the sensing/intuition (S/N) subscale across all 11 applications was .515. More t research.

PAGE 338

338 Table 4 27 . Stimuli INT ad 1 .663 ENF researcher ad 1 .560 ISF ad 1 .560 EN T ad 1 .551 INF researcher ad 1 .540 Human Personality 1 .524 ESF ad 1 .492 EST ad 1 .491 INF participant ad 1 .444 ENF participant ad 1 .427 IST ad 1 .417 Note: 1 n = 526. thinking/feeling (T/F) subscale of the BHPI (see Table 4 28 ). The alpha values ranged from a low of .430 for the ENF 1 advertisement to a high of .528 for the INT advertisement. The average alpha value for the thinking/feeling (T/F) subscale across all 11 applications was .490. research. Table 4 28 . Stimuli INT ad 1 .528 ESF ad 1 .519 ISF ad 1 .519 ENF researcher ad 1 .509 INF participant ad 1 .506 EST ad 1 .502 Human Personality 1 .485 ENT ad 1 .481 INF researcher ad 1 .465 IST ad 1 .441 ENF participant ad 1 .430 Note: 1 n = 526.

PAGE 339

339 Lastly, although the judging/perceiving (J/P) dimension was not included in the current study, it was retained in the scale administration in order to keep the full scale intact. However, the judging/perceiving (J/P) dimension was not used for any analysis in the current study. ated for each of the 11 applications of the judging/perceiving (J/P) subscale of the BHPI (see Table 4 29 ). The alpha values ranged from a low of .329 for the EST advertisement to a high of .522 for the INF 1 advertisement. The average alpha value for the judging/perceiving (J/P) subscale across all 11 applications was .436. research. Alpha values show the judging/perceiving (J/P) subscale to be substantially weaker than th e other three subscales and in need of further development and testing. Table 4 29 . Stimuli INF participant ad 1 .522 INT ad 1 .521 Human Personality 1 .495 INF researcher ad 1 .463 ENT ad 1 .442 ISF ad 1 .437 ESF ad 1 .435 ENF researcher ad 1 .422 IST ad 1 .368 ENF participant ad 1 .357 EST ad 1 .329 Note: 1 n = 526. three were lower than those calculated during the scale reduction in phase one (see Table 4 30 ). This could be due to several factors including the small number of scale items and the exploratory nature of this study which will be discussed in more detail in the next chapt er.

PAGE 340

340 Table 4 30 . Subscale Reduction Survey 2 Survey 2 Extraversion/Introversion (E/I) .758 .513 .741 .586 Sensing/Intuition (S/N) .624 .417 .663 .515 Thinking/Feeling (T/F) .67 9 .430 .528 .490 Judging/Perceiving (J/P) .641 .329 .522 .436 Note: 1 Data from scale reduction (phase one). 2 Data from survey (phase three). Individual and Multidimensional MBTI Personality Dimensions In light of lower than expected scale reliability , and in order to better understand the data, some basic statistical analyses were run on the individual and multidimensional MBTI personality data. First, examining individual dimensions, descriptive statistics showed that more than half of the respondent s correctly assigned advertisement s the appropriate single dimension (see Table 4 31 ). A chi square test was run to determine if significantly more than 50% of respondents make the correct individual personality dimension assignment for the advertisements. Therefore, the chi square compared observed values to expected values of 50%. The chi square was significant ( X 2 = 29.19, df = 5, p < .001), confirming that, for each individual MBTI dimension, significantly more than 50% of the respondents correctly iden tified functioned largely as intended. Table 4 31 . Correct respondent choices over 50% for individual MBTI personality dimensions. Observed Expected Adver tisement # of ads n % n % Extraverted (E) advertisements 5 1,972 75.0 1,315 50 Introverted (I) advertisements 5 1,796 68.3 1,315 50 Sensing (S) advertisements 4 1,330 63.2 1,052 50 Intuition (N) advertisements 6 2,088 66.2 1,578 50 Thinking (T) ad vertisements 4 1,625 77.2 1,052 50 Feeling (F) advertisements 6 2,467 78.2 1,578 50 Note: X 2 = 29.19, df = 5, p < .001.

PAGE 341

341 Next, a similar chi square was run using the eight multidimensional MBTI brand personalities included in this study. The chi square was significant ( X 2 = 275.8, df = 9, p < .001), indicating significant differences between the observed and expected values. Therefore, the descriptive statistics were examined for more detail (see Table 4 32 ). Results showed that when dimensions were com bined, not all percentages were above 50%. More than half of the respondents correctly assigned advertisements the appropriate multidimensional personality in only a few cases IST ad (61.6%), ENF participants ad (57.2%) and ENF respondents ad (51.5%). Du ring the phase two group image sorts, these two multidimensional combinations introverted/sensing/thinking (IST) and extraverted/intuitive/feeling (ENF) were found to have similar aesthetic profiles. Therefore, these findings have face validity and sug gest that certain multidimensional MBTI personalities are easier to achieve visually. Further, 30 40% of respondents correctly assigned advertisements the appropriate multidimensional personality in three additional cases INT participants ad (47.3%), EST ad (38.6%) and INT researcher ad (31.9%). These two multidimensional personalities EST and INT parallel the profiles of the top three ads ENT and IST with only the extraversion/introversion (E/I) dimension altered. This finding suggests that the e xtraversion/introversion (E/I) dimension may cross aesthetic barriers the easiest.

PAGE 342

342 Table 4 32 . Correct respondent choices over 50% for multidimensional MBTI personalities. Observed Expected Advertisement n % n % IST ad 324 61.6 263 50 ENF participan ts ad 301 57.2 263 50 ENF researcher ad 271 51.5 263 50 INF participants ad 249 47.3 263 50 EST ad 203 38.6 263 50 INF researcher ad 168 31.9 263 50 ISF ad 117 22.2 263 50 ENT ad 111 21.1 263 50 ESF ad 74 14.1 263 50 INT ad 57 10.8 263 50 Note: X 2 = 275.8, df = 9, p < .001. Confounding Variables for Hypotheses Because of their potential to influence brand personality results, both human personality and gender were tested as possible confounding variables. Results for each potential confounding var iable are presented below. Human personality. personality, a series of chi squares were run one for each of the 10 advertisements tested. Each chi to whether the respondent correctly square tests returned significant results (see Table 4 33 ). Therefore, the researcher determined that, overall, human personality was not a confounding variable and did not influence brand personality results.

PAGE 343

343 Table 4 33 . Human personality as a confounding variable. EST ad 1 ESF ad 2 ENT ad 3 ENF participant ad 4 ENF researcher ad 5 Human personality n Correct Incorrect Correct Incorrect Cor rect Incorrect Correct Incorrect Correct Incorrect n % n % n % n % n % n % n % n % n % n % EST 63 17 27.0 46 73.0 8 12.7 55 87.3 19 30.2 44 69.8 31 49.2 32 50.8 38 60.3 25 39.7 ESF 34 14 41.2 20 58.8 8 23.5 26 76.5 8 23.5 26 76.5 19 55.9 15 44.1 23 67 .6 11 32.4 ENT 78 30 38.5 48 61.5 9 11.5 69 88.5 18 23.1 60 76.9 45 57.7 33 42.3 44 56.4 34 43.6 ENF 184 75 40.8 109 59.2 27 14.7 157 85.3 40 21.7 144 78.3 111 60.3 73 39.7 87 47.3 97 52.7 IST 44 17 38.6 27 61.4 4 9.1 40 90.9 7 15.9 37 84.1 25 56.8 19 4 3.2 21 47.7 23 52.3 ISF 26 13 50.0 13 50.0 5 19.2 21 80.8 6 23.1 20 76.9 17 65.4 9 34.6 16 61.5 10 38.5 INT 38 5 13.2 33 86.8 6 15.8 32 84.2 5 13.2 33 86.8 20 52.6 18 47.4 19 50.0 19 50.0 INF 59 22 37.3 37 62.7 7 11.9 52 88.1 8 13.6 51 86.4 33 55.9 26 4 4.1 23 39.0 36 61.0 Note: 1 X 2 = 15.39, df = 7, p = .031* 2 X 2 = 4.89, df = 7, p = .674 3 X 2 = 7.68, df = 7, p = .362 4 X 2 = 3.49, df = 7, p = .837 5 X 2 = 12.61, df = 7, p = .082

PAGE 344

344 Table 4 33 . Continued. IST ad 6 ISF ad 7 INT ad 8 INF participant ad 9 INF researcher ad 10 Human personality Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect n % n % n % n % n % n % n % n % n % n % EST 39 61.9 24 38.1 15 23.8 48 76.2 11 17.5 52 82.5 34 54.0 29 46.0 23 36.5 40 63. 5 ESF 22 64.7 12 35.3 5 14.7 29 85.3 4 11.8 30 88.2 24 70.6 10 29.4 12 35.3 22 64.7 ENT 50 64.1 28 35.9 19 24.4 59 75.6 8 10.3 70 89.7 39 50.0 39 50.0 24 30.8 54 69.2 ENF 114 62.0 70 38.0 43 23.4 141 76.6 11 6.0 173 94.0 79 42.9 105 57.1 57 31.0 127 69. 0 IST 28 63.6 16 36.4 11 25.0 33 75.0 7 15.9 37 84.1 19 43.2 25 56.8 14 31.8 30 68.2 ISF 14 53.8 12 46.2 4 15.4 22 84.6 2 7.7 24 92.3 9 34.6 17 65.4 9 34.6 17 65.4 INT 24 63.2 14 36.8 8 21.1 30 78.9 5 13.2 33 86.8 14 36.8 24 63.2 11 28.9 27 71.1 INF 33 55.9 26 44.1 12 20.3 47 79.7 9 15.3 50 84.7 31 52.5 28 47.5 18 30.5 41 69.5 Note: 6 X 2 = 1.94, df = 7, p = .963 7 X 2 = 2.60, df = 7, p = .920 8 X 2 = 10.25, df = 7, p = .175 9 X 2 = 14.45, df = 7, p = .044* 10 X 2 = 1.21, df = 7, p = .991

PAGE 345

345 Gender. To squares were run one for each of the 10 advertisements tested. Each chi square compared the tended brand personality. Only one of the 10 chi square tests returned significant results (see Table 4 34 ). Therefore, the researcher determined that, overall, gender was not a confounding variable and did not influence brand personality results.

PAGE 346

346 Table 4 34 . Gender as a confounding variable. EST ad 1 ESF ad 2 ENT ad 3 ENF participant ad 4 ENF researcher ad 5 Gender n Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect n % n % n % n % n % n % n % n % n % n % Male 1 42 52 36.6 90 63.4 17 12.0 125 88.0 34 23.9 108 76.1 64 45.1 78 54.9 66 46.5 76 53.5 Female 384 151 39.3 233 60.7 57 14.8 327 85.2 77 20.1 307 79.9 237 61.7 147 38.3 205 53.4 179 46.6 Note: 1 X 2 = 0.32, df = 1, p = .572 2 X 2 = 0.71, df = 1, p = .400 3 X 2 = 0.94, df = 1, p = .332 4 X 2 = 11.74, df = 1, p = .001* 5 X 2 = 1.9861, df = 1, p = .160 Table 4 34 . Continued. IST ad 6 ISF ad 7 INT ad 8 INF participant ad 9 INF researcher ad 10 Gender Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect Correct Incorrect n % n % n % n % n % n % n % n % n % n % Male 82 57.7 60 42.3 27 19.0 115 81.0 17 12.0 125 88.0 64 45.1 78 54.9 43 30.3 99 69.7 Female 242 63.0 142 37.0 90 23.4 294 76.6 40 10.4 344 89.6 185 48.2 199 51.8 125 32.6 259 67.4 N ote: 6 X 2 = 1.219, df = 1, p = .271 7 X 2 = 1.17, df = 1, p = .280 8 X 2 = 0.26, df = 1, p = .612 9 X 2 = 0.40, df = 1, p = .526 10 X 2 = 0.25, df = 1, p = .620

PAGE 347

347 Hypothesis 1 Personality Within Advertisements Hypothesis one (H1) stated that within eac h advertisement, ads using an image with visual symbolic elements specific to a particular MBTI personality would be assigned that particular brand personality more often than any other brand personality, including: a) extraverted/sensing/thinking (EST) b) extraverted/sensing/feeling (ESF) c) extraverted/ intuitive/thinking (ENT) d) extraverted/intuitive/feeling (ENF) e) introverted/sensing/thinking (IST) f) introverted/sensing/feeling (ISF) g) introverted/intuitive/thinking (INT) and h) introverted/intuiti ve/ feeling (INF). In other words, H1 hypothesized that within each advertisement, the highest percentage of respondents would assign the advertisement the intended personality. For example, of the eight possible personality choices, the highest percentage of respondents should have assigned the EST advertisement an EST personality. To test H1, a series differences between percentages tests were run one for each of the 10 advertisements. Below, results are presented first for those hypotheses that were su pported H1d and H1e. Next are results for hypothesis that was partially supported H1h. Finally, results are presented for those hypotheses that were not supported H1a, H1b, H1c, H1f, H1g. Supported h ypotheses Extraverted/Intuitive/Feeling (ENF). Hypo thesis 1d (H1d) predicted that, compared to all other MBTI personality types, the highest percentage of respondents would assign the ENF advertisements an ENF personality. Based on group image sort results, two ENF advertisements

PAGE 348

348 were included in the surve y one with the highest overall participant ranking and one with the assessment. Results for each advertisement ENF participants and ENF researcher are present ed below. Based on descriptive statistics, the highest percentage of respondents did assign the ENF participants ad an ENF personality (see Table 4 35 ). A difference between percentages test was run, which showed that, compared to all other MBTI personalit y types, a significantly higher percent of respondents (57.2%) assigned the ENF participants ad an ENF personality (p < .001). Table 4 35 . MBTI personalities within the extraverted/intuitive/feeling participant advertisement (ENF participant ad). Responde nt chosen brand personality n % ENF 301 57.2 1 ESF 84 16.0 INF 34 6.5 IST 31 5.9 EST 30 5.7 ENT 25 4.8 IST 14 2.7 INT 7 1.3 Note: n = 526 1 52.7% (301) is significantly greater than all others ( p < .001). Also, based on descriptive statistics, the highest percentage of respondents assigned the ENF researcher advertisement an ENF personality (see Table 4 36 ). A differen ce between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (51.5%) assigned the ENF researcher ad an ENF personality (p < .001). Because both ENF ads ENF participant s and ENF researcher were assigned an ENF personality by a significantly higher percentage of respondents, H1d was supported.

PAGE 349

349 Table 4 36 . MBTI personalities within the extraverted/intuitive/feeling researcher advertisement (ENF researcher ad). Responde nt chosen brand personality n % ENF 271 51.5 1 ESF 74 14.1 EST 45 8.6 INF 44 8.4 ISF 36 6.8 ENT 27 5.1 IST 22 4.2 INT 7 1.3 Note: n = 526 1 51.5% (271) is significantly greater than all others ( p < .001). Introverted/Sensing/Thinking (IST). Hypothesis 1e (H1e) predicted that, compared to all other MBTI personality types, the highest percentage of respondents would assi gn the IST ad an IST personality. Based on descriptive statistics, the highest percentage of respondents did assign the IST ad an IST personality (see Table 4 37 ). A difference between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (61.6%) assigned the IST ad an IST personality (p < .001). Thus, H1e was supported. Table 4 37 . MBTI personalities within the introverted/sensing/thinking advertisement (IST ad). Respondent chosen brand personality n % IST 324 61.6 1 EST 81 15.4 INF 35 6.7 INT 34 6.5 ISF 21 4.0 ENF 12 2.3 ENT 12 2.3 ESF 7 1.3 Note: n = 526 1 61.6% (324) is significantly greater than all others ( p < .001).

PAGE 350

350 Partially supported h ypothesis Introverted/Intuitive/Feeling (INF). Hypothesis 1h (H1h) predicted that, compared to all other MBTI personality types, the highest percen tage of respondents would assign an INF advertisements an INF personality. Based on group image sort results, two I NF advertisements were included in the survey one with the highest overall participant ranking and one with the highest participant ranking assessment. Results for each ad INF participants and INF researcher are presented below. Based on descriptive statistics, the highest percentage of respondents did assign the INF particip ants ad an INF personality (see Table 4 38 ). A difference between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (47.3%) assigned the INF participants ad an INF perso nality (p < .001). Table 4 38 . MBTI personalities within the introverted/intuitive/feeling participant advertisement (INF participant ad). Respondent chosen brand personality n % INF 249 47.3 1 IST 82 15.6 ISF 7 7 14.6 INT 36 6.8 ENF 35 6.7 EST 24 4.6 ENT 14 2.7 ESF 9 1.7 Note: n = 526 1 47.3% (249) is significantly greater than all others ( p < .001). However, also based on descriptive statistics, the highest percentage of respondents assigned the INF researcher advertisement an ENF personality (see Table 4 39 ). A difference between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (38.0%) assigned the INF researcher ad an ENF personality (p = .003). Because the INF participants ad was assigned an INF personality by

PAGE 351

351 the highest percentage of respondents, but the INF researcher ad was not assigned an INF personality by the highest percentage of respondents, H1h was only partially supported. Table 4 39 . MBTI personalities within the introverted/intuitive/feeling researcher advertisement (INF researcher ad). Respondent chosen brand personal ity n % ENF 200 38.0 1 INF 168 31.9 ISF 81 15.4 ESF 40 7.6 IST 13 2.5 INT 10 1.9 EST 8 1.5 ENT 6 1.1 Note: n = 526 1 38 .0% (200) is significantly greater than all others ( p = .003). Unsupported hypotheses Extraverted/Sensing/Thinking (EST). Hypothesis 1a (H1a) predicted that, compared to all other MBTI personality types, the highest percentage of respondents would assign the EST advertisements an EST personality. However, although descriptive statistics showed that the highest percentage of respondents did assigned the EST advertisement an EST personality, a difference between percentages test showed that the difference b etween the top two assigned personalities EST (38.6%) and IST (35.9%) was not significant (p = .198) (see Table 4 40 ). Thus, H1a was not supported.

PAGE 352

352 Table 4 40 . MBTI personalities within the extraverted/sensing/thinking advertisement (EST ad). Respon dent chosen brand personality n % EST 203 38.6 1 IST 189 35.9 ENT 36 6.8 INT 26 4.9 INF 21 4.0 ENF 20 3.8 ISF 17 3.2 ESF 14 2.7 Note: n = 526 1 38.6% (203) is not significantly greater than 35.9% (189) ( p = .198) but is greater than all others ( p < .001). Extraverted/Sensing/Feeling (ESF). Hypothesis 1b (H1b) predicted that, compared to all other MBTI personality types , the highest percentage of respondents would assign the ESF advertisement an ESF personality. However, based on descriptive statistics, the highest percentage of respondents assigned the ESF advertisement an ENF personality (see Table 4 41 ). A difference between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (45.6%) assigned the ESF ad an ENF personality (p < .001). Thus, H1b was not supported. Table 4 41 . MBTI person alities within the extraverted/sensing/feeling advertisement (ESF ad). Respondent chosen brand personality n % ENF 240 45.6 1 ESF 74 14.1 EST 74 14.1 ENT 42 8.0 INF 31 5.9 ISF 25 4.8 IST 24 4.6 INT 16 3.0 Note: n = 526 1 45.6% (240) is significantly greater than all others ( p < .001).

PAGE 353

353 Extraverted/Intuitive/Thinking (ENT). Hypothesis 1c (H1c) predicted that, compared to all other MBTI personality types, the highest percentage of respondents would assign the ENT advertisement an ENT personality. However, based on descriptive statistics, the highest percentage of respondents assigned the ENT advertisement an ENF personality (see Table 4 42 ). A difference between percentages test was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (30.4%) assigned the ENT ad an ENF personality (p < .001). Thus, H1c was not su pported. Table 4 42 . MBTI personalities within the extraverted/intuitive/thinking advertisement (ENT ad). Respondent chosen brand personality n % ENF 160 30.4 1 EST 119 22.6 ENT 111 21.1 IST 43 8.2 INF 41 7.8 ESF 22 4.2 ISF 17 3.2 INT 13 2.5 Note: n = 526 1 30.4% (160) is significantly greater than all others ( p < .001). Introverted/Sensing/Feeling (ISF). Hypothesis 1f (H1f) predicted that, compared to all other MBTI personality types, the highest percentage of respondents would assign the ISF advertisement an ISF personality. However, based on descriptive statistics, the highest percentage of respondents assigned the IS F ad an INF personality (see Table 4 43 ). Although, when a difference between percentages test was run, it showed that there was no significant difference between the percentage of participants who assigned the ad an INF personality and those who assigned it an ISF personality (p = .657). Despite this finding, because the two personalities shared significance, H1f was not supported.

PAGE 354

354 Table 4 43 . MBTI personalities within the introverted/sensing/feeling advertisement (ISF ad). Respondent chosen brand perso nality n % INF 121 23.0 1 ISF 117 22.2 ENF 89 16.9 IST 77 14.6 EST 47 8.9 ESF 36 6.8 ENT 20 3.8 INT 19 3.6 Note: n = 526 1 23.0% (121) is not significantly greater than 22.2% (117) ( p = .657) but is greater than all others ( p < .001). Introverted/Intuitive/Thinking (INT). Hypothesis 1g (H1g) predicted that, compared to all other MBTI personality types, the highest percenta ge of respondents would assign the INT advertisement an INT personality. However, based on descriptive statistics, the highest percentage of respondents assigned the INT advertisement an IST personality (see Table 4 44 ). A difference between percentages te st was run, which showed that, compared to all other MBTI personality types, a significantly higher percent of respondents (42.2%) assigned the INT ad an IST personality (p < .001). Thus, H1g was not supported. Table 4 44 . MBTI personalities within the int roverted/intuitive/thinking advertisement (INT ad). Respondent chosen brand personality n % IST 222 42.2 1 EST 93 17.7 ENT 62 11.8 INT 57 10.8 INF 41 7.8 ENF 37 7.0 ISF 12 2.3 ESF 2 0.4 Note: n = 526 1 42.2% (222) is significantly greater than all others ( p < .001).

PAGE 355

355 Hypothesis 1 s ummary Therefore, based on individual advertisement results described above, H1 is partially supported. In four of the 10 advertisements ENF participants, ENF researcher, IST and INF participants the highest percentage of respondents recognized the personality specific visual symbolic elements in each advertisement, and assigned it the intende d MBTI personality. This means that, in some cases, respondents were able to use predetermined visual symbolic elements to arrive at the researcher intended brand personality. However, in other cases, respondents either responded differently than the resea with the intended personality. Hypothesis 2 Personality Across Advertisements Hypothesis two (H2) stated that across all advertisements, ads using an image with visual symbolic elements specific to a particular MBTI personality would be assigned that particular brand personality by the highest percentage of respondents, including: a) extraverted/sensing/thinking (EST) b ) extraverted/sensing/feeling (ESF) c) extraverted/ intuitive/thinking (ENT) d) extraverted/intuitive/feeling (ENF) e) introverted/sensing/thinking (IST) f) introverted/sensing/feeling (ISF) g) introverted/intuitive/thinking (INT) and h) introverted/intuit ive/ feeling (INF). In other words, H2 hypothesized that the highest percentage of respondents would assign each advertisement the intended personality when compared to the other nine advertisements tested.

PAGE 356

356 For example, a higher percentage of respondents s hould have assigned the EST ad an EST personality when compared to all the other advertisements tested. To test H2, a series differences between percentages tests were run one for each of the eight multidimensional MBTI personalities. Below, results are presented first for those hypotheses that were supported H2a, H2c, H2d, H2e, H2f, H2g and H1h. Then, results are presented for the hypothesis that was partially supported H2b. Supported h ypotheses Extraverted/Sensing/Thinking (EST). Hypothesis 2a (H2a) pred icted that the EST ad would be assigned an EST personality by a higher percentage of respondents than any other advertisement tested. Based on descriptive statistics, the EST ad was assigned an EST personality by a higher percentage of respondents tha n any other advertisement (see Table 4 45 ). A difference between percentages test was run, which showed that, compared to all other advertisements, a significantly higher percent of respondents (3.86%) assigned the EST ad an EST personality ( p < .001) . Thu s, H2a was supported. Table 4 45 . Extraverted/sensing/thinking personality (EST) across advertisements. Advertisement n % EST ad 203 3.86 1 ENT ad 119 2.26 INT ad 93 1.77 ESF ad 74 1.4 1 IST ad 47 .89 ENF researcher ad 45 .86 ENF participants ad 30 .57 INF participants ad 24 .46 ISF ad 20 .38 INF researcher ad 8 .15 Note: n = 5,260 1 3.86% (20 3) is significantly greater than all others ( p < .001).

PAGE 357

357 Extraverted/Intuitive/Thinking (ENT). Hypothesis 2c (H2c) pred icted that the ENT ad would be assigned an ENT personality by a higher percentage of respondents than any other advertisement tested. Ba sed on descriptive s tatistics, the ENT ad was assigned an ENT personality by a higher percentage of respondents than any other advertisement (see Table 4 46 ). A difference between percentages test was run, which showed that, compared to all other advertise ments, a significantly higher percent of respondents (2.11%) assigned the ENT ad an ENT personality ( p < .001) . Thus, H2c was supported. Table 4 46 . Extraverted/intuitive/thinking personality (ENT) across advertisements. Advertisement n % ENT ad 111 2.11 1 INT ad 62 1.18 ESF ad 42 .80 EST ad 36 .68 ENF researcher ad 27 .51 ENF participants ad 25 .48 ISF ad 20 .38 INF participants ad 1 4 .27 IST ad 12 .23 INF researcher ad 6 .11 Note: n = 5,260 1 2.11% (111) is significantly greater than all others ( p < .001). Extraverted/Intuitive/Feeling (ENF). Hypothesis 2d (H2d) pred icted that the ENF ad would be assigned an ENF personality by a higher percentage of respondents than any other advertisement tested. Based on group image sort results, two ENF advertisements were included in the survey one with the highest overall participant ranking and one with the highest . Results for each advertisement ENF participants ad and ENF researcher ad are presented below.

PAGE 358

358 Based on descriptive statistics, both the ENF advertisements tested ENF participants ad and ENF researcher ad were assigned an ENF personality by a higher percentage of respondents than any other advertisement (see Table 4 47 ). The ENF participants ad was ranked first compared to all other ad s, followed by the ENF researcher ad. A difference between percentages test was run, which showed no significant difference between the two ENF ads ( p = .069). However, compared to all other advertisements, a significantly higher percent of respondents ass igned both ENF ads ENF participants (5.72%, p < .001) and ENF researcher (5.15%, p = .04) an ENF personality . Thus, H2d was supported. Table 4 47 . Extraverted/intuitive/feeling personality (ENF) across advertisements. Advertisement n % ENF participants ad 301 5.72 1 ENF researcher ad 271 5.15 2 ESF ad 240 4.56 INF researcher ad 200 3.80 ENT ad 160 3.04 ISF ad 89 1.69 INT ad 37 .70 INF part icipants ad 35 .67 EST ad 20 .38 IST ad 12 .23 Note: n = 5,260 1 5.72% (301) is not significantly greater than 5.15% (271) ( p = .069) but is greater than all others (p < .001). 2 5.15% (271) is significantly gr eater all others (p = .040). Introverted/Sensing/Thinking (IST). Hypothesis 2e (H2e) predicted that the IST ad would be assigned an IST personality by a higher percentage of than any other advertisement tested. Based on descriptive s tatistics, the IST ad was assigned an IST personality by a higher percentage of respondents than any other advertisement (see Table 4 48 ). A difference between percentages test was run, which showed that, compared to all other advertisements, a

PAGE 359

359 significantly higher percent of respondents (6.16%) assigned the IST ad an IST personality ( p < .001) . Thus, H2e was supported. Table 4 48 . Introverted/sensing/thinking personality (IST) across advertisements. Advertisement n % IST ad 324 6.16 1 INT ad 222 4.22 EST ad 189 3.59 INF participants ad 82 1.56 ISF ad 77 1.46 ENT ad 43 .82 ENF participants ad 31 .59 ESF ad 24 .46 ENF researcher ad 22 .42 INF researcher ad 13 .25 Note: n = 5,260 1 6.16% (324) is significantly greater than all others ( p < .001). Introverted/Sensing/Feeling (ISF). Hypothesis 2f (H2f) predicted that the ISF advertisement would be assigned an ISF person ality by a higher percentage of respondents than any other advertisement tested. Based on descriptive s tatistics, the ISF ad was assigned an ISF personality by a higher percentage of respondents than any other advertisement (see Table 4 49 ). A difference b etween percentages test was run, which showed that, compared to all other advertisements, a significantly higher percent of respondents (2.22%) assigned the ISF ad an ISF personality ( p < .001) . Thus, H2f was supported.

PAGE 360

360 Table 4 49 . Introverted/sensing/fe eling personality (ISF) across advertisements. Advertisement n % ISF ad 117 2.22 1 INF researcher ad 81 1.54 INF participants ad 77 1.46 ENF researcher ad 36 .68 ENF participants ad 34 .65 ESF ad 25 .48 IST ad 21 .40 EST ad 17 .32 ENT ad 17 .32 INT ad 12 .23 Note: n = 5,260 1 2.22% (117) is significantly greater than all others ( p < .001). Introverted/Intuitive/Thinking (INT). Hypothesis 2g (H2g) pred icted that the INT ad would be assigned an INT personality by a higher percentage of respondents than any other advertisement tested. Based on descriptive s tatistics, the INT ad was assigned a n INT personality by a higher percentage of respondents than any other advertisement (see Table 4 50 ). A difference between percentages test was run, which showed that, compared to all other advertisements, a significantly higher percent of respondents (1. 08%) assigned the INT ad an INT personality ( p < .001) . Thus, H2g was supported. Table 4 50 . Introverted/intuitive/thinking personality (INT) across advertisements. Advertisement n % INT ad 57 1.08 1 INF participants ad 36 .68 IST ad 34 .65 EST ad 26 .49 ISF ad 19 .36 ESF ad 16 .30 ENT ad 13 .25 INF researcher ad 10 .19 ENF participants ad 7 .13 ENF r esearcher ad 7 .13 Note: n = 5,260 1 1.08% (57) is significantly greater than all others ( p < .001).

PAGE 361

36 1 Introverted/Intuitive/Feeling (INF). Hypothesis 2h (H2h) predi cted that the INF ad would be assigned an INF personality by a higher percent age of respondents than any other advertisement tested. Based on group image sort results, two I NF advertisements were included in the survey one with the highest overall participant ranking and one with the highest participant ranking that also overlapp . Results for each advertisement INF participants and INF researcher are presented below. Based on descriptive statistics, both the INF advertisements tested INF participants and INF researcher were assigned an INF personality by a higher percentage of respondents than any other advertisement (see Table 4 51 ). The INF participants ad was ranked first compared to all other ads, followed by the INF researcher ad. A difference between percentages test was run, which showed that, compared to all other advertisements, a significantly higher percent of respondents (4.73%) assigned the INF participants ad an INF personality ( p < .001). Additionally, compared to all other advertisements except the INF p articipants ad, a significantly higher percent of respondents (3.19%) assigned the INF researcher ad an INF personality ( p < .001). Thus, H2h was supported. Table 4 51 . Introverted/intuitive/feeling personality (INF) across advertisements. Advertisement n % INF participants ad 249 4.73 1 INF researcher ad 168 3.19 2 ISF ad 121 2.30 ENF researcher ad 44 .84 ENT ad 41 .78 INT ad 41 .78 IST ad 35 .67 ENF participants ad 34 .65 ESF ad 31 .59 EST ad 21 .40 Note: n = 5,260 1 4.73% (249) is significantly greater than all others ( p < .001). 2 3.19% (168) is significantly greater than all others ( p < . 001).

PAGE 362

362 Partially supported h ypotheses Extraverted/Sensing/Feeling (ESF). Hypothesis 2b (H2b) pred icted that the ESF ad would be assigned an ESF personality by a higher percentage of respondents than any other advertisement tested. Based on descriptive st at istics, the ESF ad was assigned an ENF personality by a higher percentage of respondents than any other advertisement (see Table 4 52 ). However, while the ENF participants ad was ranked first compared to all other ads, the ESF ad and the ENF researcher ad tied for second. A difference between percentages test was run, which showed no significant difference between the ENF participants ad and the ESF ad ( p = .25) or the ENF researcher ad ( p = .25). Yet, when compared to all other advertisements, a significan tly higher percent of respondents assigned both the ENF participants ad (1.60%), the ESF ad (1.41%) and the ENF researcher ad (1.41%) an ESF personality ( p < .001). Thus, t here was no significant difference between the top three ads, which included the ESF ad. But there was a significant difference between the top three ads, which included the ESF ad, and the rest of the advertisements. Therefore, H2b was partially supported. Table 4 52 . Extraverted/sensing/feeling personality (ESF) across advertisements. Advertisement n % ENF participants ad 84 1.60 1 ESF ad 74 1.41 2 ENF researcher ad 74 1.41 2 INF researcher ad 40 .76 ISF ad 36 .68 ENT ad 22 .42 EST ad 14 .27 INF participants ad 9 .17 IST ad 7 .13 INT ad 2 .04 Note: n = 5,260 1 1.60% (84) is not significantly greater than 1.41% (74) ( p = .250) but is greater than all others ( p < .001). 2 1.41 % (74) is significantly greater than all others ( p < .001).

PAGE 363

363 Hypothesis 2 s ummary Therefore, based on individual personality results described above, H 2 is supported. With the exception of the ESF ad, which tied with two other advertisements, a higher per centage of respondents assigned the intended advertisement the intended personality when compared with all advertisements. Thus, even though the intended personality may not have been the most prevalent within the advertisement (see H1 above), H2 showed th at all but one advertisement received the highest percentage of intended personality when compared across advertisements. In other words, each advertisement demonstrated significantly more of the intended personality than any other advertisement tested. Th is means that the predetermined visual symbolic elements did have the intended effect, although not always to the degree desired by the researcher. However, H2 suggests that strengthening the instances of the predetermined visual symbolic shown in the test images could increase results. Research Question 5 Attitude Toward the Ad personality influenced his/her attitude toward the ad (A Ad ). To answer RQ5, the researcher ran a three wa y between groups analysis of variance (ANOVA) to explore the impact of the confounding variables Variance was significant ( p = .00 1 ), indicating that the data violated the homogeneity of variance assumption. Therefore, a more stringent significance level ( p = .01) was used to test for significance (Pallant, 2010). For attitude toward the ad, there was a statistically significant main effect for brand personality, F (7, 5134) = 5.58, p < .001; however, the effect size was small (partial eta squared

PAGE 364

364 = .008). This indicates that brand personality does influence attitude toward the ad, although the influen ce is very small. There was also a statistically significant main effect for human personality, F (7, 5134) = 2.89, p = .005; however, the effect size was also very small (partial eta squared = .004). This means that human personality also influences attit ude toward the ad, although the degree of influence is very small. There were no other statistically significant main or interaction effects (see Table 4 53 ). Therefore, to answer RQ5, MBTI brand personality had a very small influence on attitude toward th e ad. Table 4 53 . Between subjects effects for attitude toward the ad. Effect df 1 df 2 F p Partial Eta Squared Brand Personality 7 5134 5.58 .000* .008 Human Personality 7 5134 2.89 .00 5 * .004 Gender 1 5134 .131 .718 .000 Brand Personali ty * Human Personality 49 5134 1.25 .111 .012 Brand Personality * Gender 7 5134 .73 .646 .001 Human Personality * Gender 7 5134 1.34 .229 .002 Brand Personality * Human Personality * Gender 47 5134 1.31 .074 .012 Note: n = 5,260 * p < .01 Res earch Question 6 Attitude Toward the Brand personality influenced his/her attitude toward the brand (A Brand ). To answer RQ6, the researcher ran a three way between groups analysis of variance (ANOVA) to explore the impact of the confounding variables Variance was significan t ( p < .001), indicating that the data violated the homogeneity of variance assumption. Therefore, a more stringent significance level ( p = .01) was used to test for significance (Pallant, 2010).

PAGE 365

365 For attitude toward the brand, there was a statistically sig nificant main effect for brand personality, F (7, 5134) = 6.31, p < .001; however, the effect size was small (partial eta squared = .009). This indicates that brand personality does influence attitude toward the brand, although the influence is very small. There was also a statistically significant main effect for human personality, F (7, 5134) = 2.91, p = .005; however, the effect size was also very small (partial eta squared = .004). This means that human personality also influences attitude toward the br and, although the degree of influence is very small. Using the more stringent significance level ( p = .01) required, there were no other statistically significant main or interaction effects (see Table 4 54 ). Therefore, to answer RQ6, MBTI brand personalit y had a very small influence on attitude toward the brand. Table 4 54 . Between subjects effects for attitude toward the brand. Effect df 1 df 2 F p Partial Eta Squared Brand Personality 7 5134 6.31 .000* .00 9 Human Personality 7 5134 2.91 .00 5* .004 Gender 1 5134 .24 .628 .000 Brand Personality * Human Personality 49 5134 1.38 .042 .01 3 Brand Personality * Gender 7 5134 .64 .720 .001 Human Personality * Gender 7 5134 2.38 .020 .00 3 Brand Personality * Human Personality * Gender 47 5134 1.29 .091 .012 Note: n = 5.260 * p < .01 Research Question 7 Purchase Intent personality influenced his/her purchase intent (PI) . To answer RQ7, the researcher ran a thr ee way between choice of brand personality on purchase intent and to include potential confounding variables riance was significant ( p =

PAGE 366

366 .012), indicating that the data violated the homogeneity of variance assumption. Therefore, a more stringent significance level ( p = .01) was used to test for significance (Pallant, 2010). For purchase intent, there was a statis tically significant main effect for brand personality, F (7, 5134) = 4.63, p < .001; however, the effect size was small (partial eta squared = .006). This indicates that brand personality does influence purchase intent, although the influence is very small . There were no other statistically significant main or interaction effects (see Table 4 55 ). Therefore, to answer RQ7, MBTI brand personality had a very small influence on purchase intent. Table 4 55 . Between subjects effects for purchase intent. Effect df 1 df 2 F p Partial Eta Squared Brand Personality 7 5134 4.63 .000* .006 Human Personality 7 5134 1.69 .108 .002 Gender 1 5134 .92 .337 .000 Brand Personality * Human Personality 49 5134 1.37 .045 .013 Brand Personality * Gender 7 513 4 .46 .862 .001 Human Personality * Gender 7 5134 1.29 .253 .002 Brand Personality * Human Personality * Gender 47 5134 1.24 .128 .011 Note: n = 5,260 * p < .01

PAGE 367

367 CHAPTER 5 DISCUSSION The current study explored the connection between specific sym bolic visual elements and particular MBTI brand personality dimensions. Rooted in meaning transfer theory (McCracken, 1986; 1988; 1990, 2005), the researcher in vestigated advertising image s abil ity to manipulate and transfer mult idimensional brand persona lities . Specifically, the current study used a mixed method composed of three phases a scale reduction, qualitative group image sorts and a quantitative survey. In the first phase, the researcher created the Brand and Human Personality Indicator (BHPI), a reduced MBTI personality scale useful for determining both human and brand personalities. The second phase consisted of qualitative group image sorts, which were used to answer research questions about which specific visual symbolic elements in images we re associated with particular MBTI personalities, how those visual symbolic elements could be combined to represent multidimensional brand personalities, and how participants read and negotiated brand personality meaning in images. Additionally, the group image sorts served to select the best examples of images with multidimensional personalities, which were used as stimuli during the third phase, a quantitative survey. Through the survey, the researcher sought to better understand whether images that conta ined specific symbolic visual elements associated with particular multidimensional brand personalities could transfer that personality to a brand through an advertisement . Additionally, the survey asked questions regarding the relationship between multidim ensional MBTI brand personalities and attitude toward the ad, attitude toward the brand and purchase intent. This chapter provides a survey reduction, qualitative and quantitative. However, the disc ussion within each section is informed

PAGE 368

368 by the totality of the findings, referencing the data from other phases when appropriate. Findings are also discussed in terms of pas t literature and theory. Phase One Discussion Scale Reduction The first phase of t personality scale from 24 items to 12 items using a series of criteria established by the researcher, which includ ed reducing the number of scale items, avoiding equal personality results, correlatin g above a .6, staying within the original alpha range, limiting evaluative adjectives and considering judgmental item qualities. Based on the application of these criteria, a new, reduced scale was created, which consists of three items for each of the fou r MBTI dimensions (see Table 5 1 ) . Table 5 1 . Brand and Human Personality Indicator (BHPI) . Extraversion/ Introversion (E/I) Sensing/ Intuition (S/N) Thinking/ Feeling (T/F) Judging/ Perceiving (J/P) 1 Outspoken/ Quiet Systematic/ Imaginative Firm/ Soft Hearted Decided/ Flexible 2 Active/ Reserved Practical, Functional/ Creative, Theoretical Rational, Reasonable/ Passionate, Perceptive Deliberate/ Adaptable 3 Sociable/ Shy Conservative/ Unconventional Logical/ Emotional Clear Cut, Definite/ Undecided, Variable The BHPI, created during phase one of the current study, was then used to conduct the reliability statistics were then run on the data gathered from the s urvey, resulting in average alphas of .59 (E/I), .52 (S/N), .49 (T/F) and .44 (J/P). Although the scale reliabilities calculated from the phase three survey data were lower than might be liked, the scale might still have value for research. According to Nu nnally (1967), saves time and energy by working with instruments that have only modest reliability, for which

PAGE 369

369 (p. 226). Additionally, other researchers agreed .52 (S/N), .49 (T/F) and .44 (J/P) were above or approaching minimum reliability standards for research . low to moderate reliability results during its use in the phase three survey. One reason for the less than ideal reliability might stem from the low number of scale items per subscale. According to Pallant (2010), scales with a small number of items (less than 10) could produce smaller alpha values (e.g. .50), despite their usefulness. Additionally, in an exploration of how to calculate the reliability of small, two item scales, Eisinga, Grotenhuis and Pelzer (2013) noted ore, it was reasonable to expect that the small number of scale items per subscale in the BHPI might have negatively affected the reliability scores. applied during explor atory research in the current study. The current study attempted to examine a previously unexplored area of visual communication research regarding visuals and multidimensional MBTI brand personalities. As survey results showed, for some advertising images , the majority of participants were able to accurately identify the intended MBTI personality. However, in other cases, the images returned mixed personality results, indicating that they were not good examples of any particular multidimensional MBTI perso nality. These mixed personality images might have confused participants as they attempted to rate the visual

PAGE 370

370 should be tested on stronger, purer visual examp les of multidimensional brand personality to ability to measure brand perso nality with non visual stimuli. y research that might have influenced the elements that represent each MBTI dimension. Group image sort participants were able to identify a number of visual symbolic elements associated with each pole of the three dimensions included in this study extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). For example, aesthetically, participants noted that bright colors often indica soft hearted, passionate/perceptive and emotional might not have fully captured the visual symbolic element of bright colors. Therefore, future refinement of the BHPI s hould take visuals into consideration w hen evaluating adjective pairs. In a ddition , more research should be conducted to better understand how the visual symbolic elements that represent each MBTI brand personality are captured, or fail to be captured, by the current BHPI scale items. Such additional research could be used to improve the BHPI for visual use or to create a new scale focused solely on measuring visual brand personality. For example, researchers could use qualitative methods, including group i mage sorts, to identify images that are rated highly on each individual dimension. Then, working backwards, use the images to create and test polar adjective pairs that capture the visual aspects of each MBTI dimension. To more accurately understand and me asure visuals, other researchers have set a precedent for creating separate measurement tools for visual specific research. For example, rather than use an established taxonomy of rhetorical figures in advertising, Phillips

PAGE 371

371 and McQuarrie (2004) created a t ypology specific to visual rhetoric in advertising. Because of the differences in text and images, the researchers felt that verbal based taxonomies did not differentiations within the visual domain 200 4, p. 114). The same may also hold true for the differences in measuring verbal and visual brand personality, necessitating a brand personality scale designed specifically for measuring visual brand personality. Despite these possible justifications for th than ideal reliability, the scale still has room for improvement. Some of the adjective pairs could be fine tuned to more accurately represent each MBTI dimension. This process might be even more valuable now that each subscale contains fewer items. In particular, examining the individual adjectives from , which were removed during the reduction process , may serve as a good starting point for replacing potentially problematic single adjectives in the current BHPI. However, in its current state, the reduced BHPI has several advantages for researchers. These advantages include a more manageable length, the elimination of problems associated with five factor models, and the ability to directly measure and compare both human and brand personality with a single tool. Regarding manageable length, o ne of the main scale reduction goals was to shorten item length was prohibitive for scholars who w ished to examine the brand personality of multiple stimuli in a item length makes the scale suitable for multiple applications in a single study. For example, the current study was able to administ er the scale 11 times on a single online questionnaire with no difficulties and very few survey

PAGE 372

372 respondents who did not complete the questionnaire after progressing beyond the informed consent (4.8%). A second advantage for researche rs is the new B H P ility to overcomes problems that other scholars (e.g., Austin et al., 2003; Avis, 2012; Azoulay & Kapferer, 2003; Lee, 2013; factor models. The BHPI is different from many other brand personality scales because it is based on the MBTI, which is both a human personality measure and is composed of four polar dimensions, rather than five factors. In particular, the use of the BHPI in the current study demonstrated the advantages of using this scale to align brand personality with human personality in addition to the brand personality of multiple advertisements. Because the survey us ed the same scale to measure both personalities, the human and brand personalities could be directly compared. For example, the current study used the BHPI results to test whether human personality influenced brand personality. These advantages suggest tha t the BHPI could be valuable for researchers as they expand their knowledge of brand personality. Although the scale could be improved, it still provides a helpful tool for researcher looking for a shorter, more manageable alternative to problematic five f actor models. Phase Two Discussion Group Image Sorts The second phase of the current study, qualitative group image sorts, sought to better understand which specific visual symbolic elements within images were associated with particular MBTI personalitie s both at an individual dimension and multidimensional level. Additionally, the researcher wanted to explore how participants read and negotiated the images in order to assign them personality. Finally, the group image sorts also served to identify the

PAGE 373

373 m ost representative images for each of the eight multidimensional MBTI brand personalities included in this study. During the group image sorts, participants sorted an image deck of 24 images three times, once for each of the MBTI personality dimensions inc luded in the current study extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F). As they sorted, participants were asked to explain why they chose to place each image in a particular personality pile. Combined, the sorted i ch, detailed data for analysis. Although one purpose of the group image sorts was to identify the visual symbolic elements associated with individual and multidimensional MBTI personality dimensions, some of the most interesting findings came from the second goal, which was to examining how participants read and negotiated personality information in images. Because the lenses through e role in the ultimate personality outcome, those lenses are discussed first, including, participants use of prior experiences, their own personality, stories and subjects to determine image personality. Then, there is a discussion about the particular vis ual symbolic elements associated with each individual MBTI personality and how they can be combined to create multidimensional MBTI personalities. In reviewing group image sort results the degree with which personal history and prior knowledge influenced e image personality decisions was clear. In many cases, participants could look at the same image and assign it opposite personalities based on an event as if almost all personality decisions were influenced

PAGE 374

374 and discussed each image, some nality meaning came through personality. Therefore, while participants wer e often able to identify and understand the dominant meaning in the image, the final image personality was most commonly dete rmined by a negotiated reading, which involv ed the mixing of their own personal experiences with the researcher intended personalit y messages. In order to increase the number of customers who engage in a dominant reading of their advertisements companies may choose to add text to their advertisements, which Barthes (1977) suggested could anchor the preferred meaning and point consumer s toward a dominant reading. Additionally, research from Perception Research Services (PRS) suggested that participants who are exposed to advertisements for an extended amount of time (i.e., a minute or more) recognize s many more meanings than the typical consumer, who, on average, spends only between 3 to 20 seconds looking at a print advertisement (Young, n.d.). In light of the fact that group image sort participants spent several minutes looking at each image, this data suggests that ipants may have relied more heavily on their own personal background, experiences and knowledge to create negotiated readings than could be expected of the average consumer. Therefore, when tested in a more real world environment, consumers may rely more h eavily on the visual symbolic elements contained in the ad, rather than their own experiences, increasing the likelihood of a dominant reading. Besides using their own personal background, experiences and knowledge, participants used their own individual p ersonality to created negotiated readings (Hall, 1980) as they sorted

PAGE 375

375 infl uenced sometimes consciously and sometimes unconsciously by their own MBTI personality. Th is finding was similar to other past brand personality literature , which also had personality (e.g., Johar et al., 2005; Puzakova et al., 2013). Participants identifi ed images that looked like something they would do or enjoy as having the same personality as their own. Alternately, they also identified images that they were not attracted to as not like them and assigned them a personality opposite their own. For examp le, if a feeling participant saw an image of an activity that they would enjoy (e.g., volunteering, reading a book), they labeled it like them feeling. Alternately , if a feeling participant viewed an image of an activity they were adverse to (e.g., chemi stry, surgery), they labeled the images as the opposite dimension pole thinking. This finding is particularly interesting in light of a study by Puzakova et al. (2013) , which showed that, not only do consumers use their own personality to determine obser vab le brand personality traits (i.e ., those that are explicitly communicated in advertising), but they also use their own personality to determine unobservable brand personality traits ( i.e., those not shown in advertising). Specifically, Puzakova et al. ( 2013) found that consumers first determined whether the observable brand personality traits in an ad were like them or unlike them. Then, based on that decision, they assumed that all the unobservable brand personality traits must follow the same pattern a nd be either like them or unlike them. When applied to the current study, the findings from Puzakova et a l. (2013) suggest that consumers could use the visual symbolic elements they recognize in an image to determine the observable brand personality traits , then use their own multidimensional MBTI personality to assign the remaining unobservable brand

PAGE 376

376 personality traits. Because consumers could potentially use a single, observable MBTI ality, it becomes even more important for brands to understand what specific symbolic visual elements align with each particular MBTI personality. Armed with this knowledge, brands could use images to either 1) strongly align themselves with a single perso nality dimension that fits the majority of their target market, or 2) present a more robust, multidimensional MBTI brand personality that leaves no unobservable traits for consumers to extrapolate. Although this may be difficult with visuals only, using th e findings from the current study to better transfer multidimensional brand personalities and adding text to anchor the meaning (Barthes, 1977) could increase the likelihood that consumers would recognize the full multidimensional personality. In addition to using their previous experience and own personality to read and negotiate After looking at only a single, still image, participants would frequently const ruct a complex story that add ed depth and detail to the scene . Often, the stories were unique to the teller, with different participants telling different stories based on the same image . A lthough, in some cases, multiple participants from different groups told similar stories. These narratives could take the form of stories about a n image his/her future, or the current moment he/she was engaged in. Participants also told stories from different perspectives. Some stories were told through th e eyes of one subject in the image. Others were holistic, established around the entire scene. Still others were based on surface level visual symbolic elements such as clothes, hairstyles or pets. The choice of which elements to base their stories on seem ed to have no discernable pattern among participants.

PAGE 377

377 Further, once participants established a story, it often followed the image through the remainder of the s orts, growing as they assigned t he image additional personality traits. While certain stories we re somewhat predictable and based on shared cultural norms, others stories were unusual and unexpected. And, when participants told these unusual stories, they sometimes arrived at personality decisions opposite to those the researcher intended. Although s tories were often born out of the visual symbolic elements in an image, in these cases, the story itself, rather than the visual symbolic elements, seemed to ultimately determine the i Viewing this storytelling pattern through the lens o framework provides some additional insight. According to Hall (1980), meaning cannot be ; producers use audio/visual components to understood by the audience (p. 164). Similarly, advertisers must use symbolic visual elements to encode an image with meaning that can then be decoded and understood by the target market. However, in the case of the still images in the current study, it seemed that participants sometimes decoded more than was encoded. In the absence of information they felt necessary to erive from the visual symbolic element with additional meaning of their own in the form of stories. So, like stories. In order for advertisers to have more control over their brand meaning, companies may consider adding text (e.g., headlines, body copy) to carefully chosen images in order to anchor the text and guide consumers toward the desired meaning (Barthes, 1977). enco ding/decoding framework pro vided some insight for this storytelling pattern, to the

PAGE 378

378 from images has not yet been addressed in previous brand personality or visual literature and should, therefore, be investigated mor e thoroughly in future studies. inherent meaning of their own, but are dependent on culture time, place, experiences and associations to provide meaning (e.g., McCracken, 1990 ; Postrel, 2003; Zuckerman, 1990 ). And, like Zuckerman (1990) noted, in order for an image to successfully communicate meaning, such as brand personality, it must first have meaning to the viewer thorough some syste m of real world familiarity. Therefore, it makes sense that participants would be better able to assign personality to images when they had some prior knowledge or experience with the subject matter or felt had some connection to their own personality. Add also have been an attempt to make the image more familiar, allowing them to assign it meaning or personality. Finally, participants seemed largely unable to separate the perceived personality of the people in the i mage from the image as a whole. Although the moderator frequently reminded participants to consider the whole image when deciding its personality, they often treated the participants felt that the subject looked extraverted (e.g., smiling, dressed trendy) but was in an introverted setting (e.g., alone, neutral colors), they would most often rate the image as extraverted based on how they felt about the human that was port rayed. to assign image personality based on the subject also led to problems when multiple people were they felt that the subjects had op posite personalities. Even when directed to consider the whole image (e.g., colors, setting, task), participants still felt conflicted when the people seemed

PAGE 379

379 different from each other. For instance, despite other visual symbolic elements, participants did not know how to rate an image if they felt that one person in an image was thinking and the other was feeling. Based on how strongly participants associated the personality of the subjects within an image with the overall personality of the image , it is im portant for advertisers to strongly consider the models they use in their ad vertising image s. findings, while Postrel (2003) assumption that we make judgments about people, places and things based on how they look is true, it is most true for people i n the case of brand personality . Based on the group image sort findings, participants often judged almost all other places and things in images based on how they judged the people. Because the current study was investigating per sonality, a concept most commonly applied to humans rather than inanimate objects (e.g., products , brands), this finding makes sense. Although Ramaseshan and Tsao ( 2007 ) noted that consumers have no trouble attaching personality to brands, it seems that, g iven a choice, they may still find it easier to connect personality to humans. In light of this finding, future research should endeavor to understand whether simply changing the visual symbolic Knowing that participants used a variety of lenses or techniques to arrive at negotiated helps set the stage to understand the specific visual symbolic elements they associated with p articular MBTI personalities. In some cases, p articipants were able to identify more co ncrete visual symbolic elements, such as multiple people being extraverted and a person alone being introverted , or bright colors i ndicating feeling and neutral colors i ndicating thinking. However, a lthough participants were able to identify some concrete visual symbolic elements for each dimension, other themes were more abstract and less tied to

PAGE 380

380 specific aesthetics. Participants would often agree on the words and concep ts that they used to describe a personality, but their application s were quite different. For example, participants consistently said that practical images were sensing and big picture images were intuitive. However, when looking at an image of parents hel ping their child cut vegetables in the kitchen, some participant said that cooking and cutting were practical tasks, while others said that the parents were teaching the child a big picture life lesson to prepare him for the future. Therefore, in most case s, it seemed to be a combination of specific visual symbolic elements and the spirit of an image that together determined its personality. As previous scholars have noted, visuals are complex and it is not only the content that communicates, but also the w ay the content is presented ( Messaris, 1997 ). While participants were able to point out some of the visual symbolic elements that influenced their personality assignment, it can be assumed that Messaris (1997) was correct in that there were many more visua l devices (e.g., point of view, angle, metaphor) that participants were less aware of, yet still subconsciously influenced their personality assignments. As such, it is important for advertisers to consider not only a few specific visual symbolic elements as they choose an image, but also spirit of an image, which is In addition, scholars have lamented the serious literature gap exploring the visual microlanguage in advertisi ng, particularly concerning visuals and brand personality ( Batra et al. , 1993 ew other studies that have focused sp ecifically on the link between specific visual symbolic elements and brand personality (e.g., Akay, 2001; Boudreaux & Palmer, 2007; Cunningham et al., 2007; Jewett & Sutherland, 2013; Park et al., 2005).

PAGE 381

381 Although images are understudied, complex combinatio ns of visual symbolic elements, spirit and meaning, participants in the group image sorts were able to identify a visual profile for each of the six individual MBTI personality dimensions extraversion (E), introversion (I), sensing (S), intuition (N), th inking (T) and feeling (F). For each dimension below, a summary of elements in light of previous research. Extraversion (E). According to participants, extrave rted images contained multiple happy people who were smiling, talking, communicating, laughing and making eye contact. The subjects could be making eye contact with either the viewer (i.e., looking at the camera) or another person in the image. Participant s also indicated that there was some sort of activeness in extraverted images. Subjects seemed to be in motion or taking part in something dynamic. Collaboration between subjects was particularly noted as indicative of an extraverted image. According to pa rticipants, subjects in extraverted images also appeared to have made a choice to participant in an activity , to be social and/or to be in public. Finally, extraverted images were filled with lively, loud and/or vibrant colors such as orange and green. Par ticipants also less frequently mentioned that the subjects in extraverted images were in close proximity to each other, appeared to feel comfortable and wore trendy/cute/revealing clothing. For some, including dogs in the images or tight cropping also made the image more extraverted. Several of these findings overlapped with aesthetic elements noted in the few other past studies that have explored visuals and brand personality. For example, both Cunningham et al. ( 2007 ) and Jewett and Sutherland (2013) iden tified the link between extraverted images and smiling, a public setting and bright/saturated colors as representative of an extraverted image.

PAGE 382

382 Akay (2001) al so pointed out that extraverted images had a sense of fun or humor, similar Additionally, Akay (2001) drew attention to the link between sexual elements and extraverte d images. The current study also found a connection between extraverted images and subjects wearing trendy, cute or revealing clothes, which could be considered sexual, particularly when viewed in light of Umiker at linked small amounts of clothing (e.g., revealing clothing, hardly any clothing, no clothing at all) with the sexualization of women. Like the current study, Jewett and Sutherland (2013) also found that extraverted images often showed the subject making eye contact with the viewer. Messaris (1997) suggested that an image o f someone looking directly at the viewer could innately cause him/her to return the gaze . Alternately, Jewitt and Oyama (2004), identified images with subjects making direct eye contact as demand images because they were symbolically demanding something from the viewer. The authors suggested that exactly what the subjects were demanding depended on additional facial expressions and gestures. It is possible that participants in the curren t study could have read this direct eye contact as a returned gaze (Messaris, 1997) or demand (Jewitt & Oyama, 2004), thereby increasing the interaction between the image and the viewer. In turn, this increased interaction might have given the image a more social or collaborative feel and increased its extroverted qualities. Finally, similar to the current study, Jewett and Sutherland (2013) also identified tightly cropped images as more extraverted. According to Jewitt and Oyama (2004), these close up shot s mimic the physical closeness of intimate rela tionships, which might also have added to the

PAGE 383

383 images social, collaborative and, ultimately, extraverted feel noted by participants in the current study. Introversion (I). Participants in the current study iden tified introverted images as having fewer people, usually only one. However, an image with two people could be considered introverted, especially if those two people were an intimate couple. A single person alone with an animal was also indicative of intro version. It was also important that the subject was not gaze was focused on an inanimate object. According to participants, the most introverted settings were in the subject seemed to have chosen isolation over being around other people. The image should also give off a feeling of quietness or peacefulness, with a lack of noi se or talking. Often, the subject in an introverted image was immersed in thought or in activity, unaware of their surroundings. Further, certain professions, including doctors, scientists and farmers, seemed to indicate introversion in an image. Finally, introverted images were composed of soft, neutral and muted colors, such as blues and whites, which gave the image a sterile feel. Participants also less frequently mentioned that the subjects in introverted images could be relaxing and that the image migh t be more widely cropped. Other past studies have also noted some similar visual symbolic elements in introverted images. For instance, like the current study, Jewett and Sutherland (2013) identified that introverted image s often portrayed only one person , although, they also noted that the single person in the image had their back to the camera. While participants in the current study did not specifically mention the subject turning his/her back on the viewer, they did note that the subject was not making

PAGE 384

384 was more likely focused on an inanimate object or gazing out of the frame. Jewitt and Oyama (2004) referred to images where the subjects are not engaging the viewer as offe rs because they simply offer information. The authors also noted that viewers tend to observe the subject in offer images in a more detached, impersonal way, which fits with the introverted MBTI personality. Similarly, Cunningham et al. (2007) picked up on the feeling of solitude shown in introverted images. Cunningham et al. (2007) also linked that solitude to nature, which participants in the current study identified as well. Specifically, participants noted that being away from manmade structures and cha os increased the feeling of introversion. they related to gender, women sh own in the domestic sphere suggested a lack of purposeful activity (e.g., self absorbed grooming, no action) . However, when viewed more holistically, comfortabl e place where introverted people would be happy and able to recharge. Unlike Goffman (1979), they found no negative connotation for either men or women shown in their own homes. Additionally, like participants in the current study, several other authors no ted the presence of low and/or romantic lighting in introverted images (Akay, 2001; Jewett & introverted images to mood lighting and seclusion, which incre Finally, along with participants, Jewett and Sutherland (2013) also identified widely cropped images as more introverted. According to Jewitt and Oyama (2004), far away shots reflected the way an individual normally saw a str anger impersonal and in outline . This

PAGE 385

385 understanding of widely cropped images fits with the less personal qualities of introverted images. Sensing (S). During the group image sorts, participants described sensing images as goal oriented. Often, this goal orientation took the form of the subject(s) focusing on a single goal, task, end result, or solution to a problem. Additionally, this focus was often short term and related to a task that would soon be completed, such as painting a wall or washing a dog. S ensing images were also deemed to be more practical and functional, which was usually shown in the form of being down to earth, conventional, normal, day to day, or almost boring, with no imagination or spontaneity. For example, participants often identifi ed financial or numerical tasks as practical or functional and, therefore, sensing. Participants also assigned a sensing personality to images that showed systematic processes taking place, such as technical medical procedures or scientific lab work. In fa ct, any image depicting either a science or medical related job was regularly identified as sensing just based on the occupation. This may be, in part, because participants identified images where subjects were using one or more of their five senses as sen sing. Images that showed subjects using their sense of touch were considered especially sensing. Sensing images were also highly organized, with little clutter or mess. In particular, participants pointed out the organization in images where individual sub stances were placed in their own containers and labeled. More concretely, participants identified images where subjects were dressed conservatively (e.g., long sleeved shirt, little skin showing) as sensing. Sensing images were also associated with a clean , neutral color palette with lots of white, creating a sterile feel to the image. Participants also less frequently mentioned that the sensing images showed technology (e.g., laptops, tablets).

PAGE 386

386 While a few other studies also identified aspects of sensing i mages, the sensing/intuition (S/N) dimension has been less studied than either extraverted/introverted (E/I) or thinking/feeling (T/F). However, Akay (2001) did identify sensing images as simple and factual with direct, rational messages. This finding over laps with participants from the current study images is also similar to the organized nature of sensing images and their clean, white color palette. Additionall y, Cunningham et al. (2007) recommended using real photos or timelines to create sensing images. Because all the images in the image deck were composed of mostly real photos, participants did not specifically identify sensing images as real. However, a few images dimensional pole (discussed in more detail below). Therefore, it is likely that, if illustration is identified with intuition, then realistic images are more likely to be identified with sensing. did equate sensing with systematic or step by step processes. In many ways, such a process seems to be the photogra phic equivalent of the timeline mentioned by Cunningham et al. (2007). Intuition ( N ). When describing intuitive images, participants noted that the images had elements of big picture or future thinking. Subjects seemed to be trying to piece together many d isparate ideas or thinking about something long term. Alternately, instead of thinking about the big picture or the future, the subject in an intuitive image could also be just experiencing the moment living wholly in the present, almost frozen in time. Participants also identified images that were full of emotions or feelings as intuitive. Almost any strong emotion seemed to elicit an intuitive personality; however, romance was especially noted. Similarly, images that showed

PAGE 387

387 relationships, whether romant ic, plutonic, family oriented or with an animal, were considered more intuitive. Unconventional living, which was often defined as less staged, systematic, practical or logical, was also connected to intuition. In many cases, this unconventional living was demonstrated by parents allowing their children to help with adult activities. Further, intuitive images often had elements of imagination and creativity. For example, the subject in an image might be using their imagination as they read a book or creatin g some sort of artwork. Children also prompted participants to label an image intuitive. In many cases, children were linked to other themes, such as future thinking, unconventionality, imagination or creativity. Regardless, participants were often quick t o label images that included children as intuitive. The color palette in intuitive images was identified as bright, warm and dreamy. Intuitive images also included elements of illustration. Participants identified the illustrated elements as creative, arts y, and representative of more than could be seen in the image. Similarly, including thought bubbles in images also elicited an intuitive personality. Participants also less frequently mentioned that the intuitive images were set in nature and could show su bjects reflecting or daydreaming. As mentioned above, few studies have investigated the sensing/intuition (S/N) dimension. One study did suggest that intuitive images would include artistic representations and illustrations (Cunningham et al., 2007). The c ed this suggestion by Cunningham et al. (2007) . Images that include d illustration were often identified as intuitive. Additionally, participants pointed out the illustrations as a large part of why they label ed the image as i ntuitive, in part because they added elements of fun and imagination. Akay (2001) also suggested that intuitive images would use product personification and combine multiple images. While none of the image in the image deck utilized any sort of product per sonification, some of the images with illustration could be considered to combine multiple

PAGE 388

388 images. In some cases, the illustrations were separate, small pictures housed in a thought bubble. Although, it may be the illustration that participants associated with intuition, this finding does ntuitive may also have validity, although more research should be conducted. Thinking (T) . According to participants, thinking images were logical and rational, often indicated by a task or action having been thoroughly thought through. For example, participants identified studying, finance, science, medical and business situations as logical activities. Similarly, thinking images were process oriented a nd goal driven, with a detailed plan, step by step strategy or end result in mind. Subjects in thinking images were also very focused on a single task or trying to figure something out. Additionally, participants identified thinking images as impersonal, c reating a feeling of disconnect with the viewer. Sometimes this than another human. Body language also played a large role in identifying images as thinking. Part up with hand on chin), looking up, staring into space, not smiling, and/or not making eye contact as particularly thinking. Portraying certain jobs in an image wa s also consistently considered more thinking, specifically science and medical related occupations. Like particular jobs, images with particular settings were also more thinking. Participants often assigned images set in a library or laboratory a thinking personality. Finally, thinking images had cool, neutral color palettes with high amounts of white, giving off a sterile feel. Participants also less frequently mentioned that the thinking images portrayed objectivity (i.e., following the rules with no emo tion), a single subject alone , particular attire (e.g., business, glasses), or thought bubbles.

PAGE 389

389 The visual profile participants associated with the thinking dimension overlapped with the visual symbolic elements noted in other brand personality studies. Ac cording to Jewett and Sutherland (2013), thinking images show the product with no people and use lines/text to identify product features. Although the group image sort phase of the current study used only images with no product or text, there were some sma ll similarities between the two sets of thinking image, participants in the current study indicated that thinking images often had a single person alone. Additional ly, participants noted the impersonal nature of the thinking images. images will portray fewer rather than more people. Similarly, although people were included in th e images, participants in the current study noted the subjects did not make eye contact with either the viewer or another human in thinking images. As discussed above regarding introversion, Jewitt and Oyama (2004) referred to images where the subject does not engage the viewer through eye contact as offer images because they simply offer information. The authors further noted that viewers tend to observe the subjects in offer images in a more detached, impersonal way. This explanation fits perfectly with p being impersonal with few subjects and no eye contact. Also, Akay (2001), found thinking images to be logical, direct and unemotional. Participants in the current study noted that thinking images were more logic al and rational. Additionally, they described thinking images as objective, with the participants seeming to follow the rules with no emotion. Further, participants in the current study identified feeling images, the direct opposite personality pole to thi nking images (discussed below) as highly

PAGE 390

390 current study. In another study, Cunningham et al. (2007) suggested that thinking images could contain tabulated data and no data, there might be some overlap between the tabulated data and graphs suggested by Cunningham et al. rational, process oriented and driven by goals and end results. Like the strict order and along with specifics like calculators, money, medicine and science s uggest similarities between the two. Cunningham et al. (2007) also described thinking images as monochromatic or black and color images, participants still identified thinking images as hav ing monochromatic color palettes. Additionally, they pointed out the large amount of whites, blues and neutrals in thinking images. When mixed with white and neutral, blue tends to appear clean and sterile, reminiscent of medical scrubs and hospitals. Blue s can also be associated with large corporations, which often have a reputation as more impersonal, high pressure and thought oriented. Therefore, without exposing participants to black and white photos, they still chose the least colorful, most monochroma tic images as thinking. Finally, participants described both intuitive and thinking images as showing thought bubbles. However, when discussing the thought bubbles, participants described the thinking thought bubbles displayed future plans or creative ideas. Therefore, although participants saw the

PAGE 391

391 same visual symbolic element a thought bubble the content inside helped determine the Feeling (F). some type of relationship, whether it was a family relationship, romantic relationship, or a relationship with an animal. Feeling images were also filled with emotions, particul arly romantic or happy emotions. Additionally, participants noted that the subjects in feeling images were often participating in enjoyable activities (e.g., volunteering, reading), relaxing, experiencing a moment or participating in some sort of creative endeavor (e.g., painting, pottery, art). A participants identified an image as feeling when subjects were touching, looking lovingly at another subject, smili ng or laughing. Images set in nature were also more likely to be classified as feeling. Further, feeling images were filled with bright, saturated and warm colors, particularly yellow. And, the lighting in feeling images was often low or soft. Specifically , participants identified images taken at sunrise or sunset as feeling. Participants less frequently mentioned that the feeling images seemed to be more candid and less posed , and contain subjects wearing more relaxed attire . Several other studies identifi ed similar visual symbolic elements associated with the feeling dimension. For example, some scholars noted the relational aspect of feeling images, particularly when an image portrayed a person and an animal (Jewett & Sutherland, 2013) or a family (Akay, 2001). The same scholars also noted several overlapping types of body language more feeling.

PAGE 392

392 , smiling and touch as a sign of feeling in certain cases . Moreover, similar to the curr subject(s) relaxing with no sense of urgency, Jewett and Sutherland (2013) underscored the relaxed atmosphere of feeling images. Akay (2001) also pointed out the presence of emotion in feeling imag es. Likewise, the current study identified emotions as a component of feeling images, citing romantic and happy emotions as particularly indicative of feeling. However, A kay (2001) also noted a sense of accomplishment in feeling images, and the current stu dy found no such pattern. Although, this could be because none of the images in the current study contained content that would indicate accomplishment. Finally, Cunningham et al. (2007) suggested that feeling images were colorful and harmon ious. The curren t study supported the notion that feeling images were colorful, with participants more specifically identifying a bright, saturated color palette with warm tones and lots of yellow. However, participant did not necessarily identify the colors as harmonious . Instead, they were more often referred to as bright and saturated, combining many different colors. Similar to color, participants in the current study identified feeling images as having low or romantic lighting. However, both Jewett and Sutherland (201 3) and the current study participants also paired low or romantic lighting with introversion. Therefore, depending on the intended personality, advertisers should exercise caution when using low or romantically lit images, being careful not to confuse view ers. After examining the visual profiles for each of the six individual MBTI personality dimensions extraversion (E), introversion (I), sensing (S), intuition (N), thinking (T) and

PAGE 393

393 feeling (F) the researcher then combined elements from several individu al dimensions to construct visual profiles for multidimensional MBTI personalities. While doing this, t wo visual patterns emerged in how participants described the visual symbolic elements associated with personality dimensions. Aesthetically, extraversion (E), intuition (N) and feeling (F) were very similar. Alternately, introversion (I), sensing (S) and thinking (T) were also very similar visually . In the case of both ENF and IST personalities, aspects such as color palette, number of people, relationship s, and goals all had similarities . These overlaps made it easier to create visual profiles for these two multidimensional personality types ENF and IST because the re were no conflicting visuals. I n other cases, creating multidimensional personality pro files was more difficult . For the remaining six multidimensional MBTI personalities extraverted/sensing/thinking (EST), extraverted/sensing/feeling (ESF), extraverted/intuitive/thinking (ENT), introverted/sensing/feeling (ISF), introverted/intuitive/thin king (INT) and introverted/intuitive/feeling (INF) creating visual profiles required sacrificing certai n elements from each dimension. In each case, two of the individual personality dimensions were relatively similar, while the third was quite different . Dimensions that paired well together included, extraversion/intuition (EN), extraversion/feeling (EF), introversion/sensing (IS), introverted/thinking (IT) and sensing/thinking (ST). Take, as a single example of this finding, the visual profile for the i ntroverted/intuitive/thinking (INT) personality. In this multidimensional personality profile, introversion and thinking (IT) were aesthetically similar. However, intuition (N) was visually quite different. Both introversion (I) and thinking (T) typical de picted a single person alone, isolated and immersed in their thoughts or task, creating an impersonal feel where the subject

PAGE 394

394 was not making eye contact with any other subject or the viewer. However, intuitive (N) images were distinguished by the relationsh ips portrayed in the image. They were full of emotions and feelings, often happy or romantic in nature. Intuitive (N) images also showed subjects who were reflecting or daydreaming and engaging their imagination and creativity. Moreover, introverted (I) an d thinking (T) images often showed subjects at work, particularly as scientists or medical personnel. Alternately, intuitive (N) images portrayed children and families experiencing a moment together or making a memory. The dimensions also showed extreme di fferences in their color palettes. Introverted (I) and thinking (T) images had neutral, muted and cool colors, particularly whites and blues, which produced a sterile feel. Conversely, intuitive (N) images were filled with bright, warm and dreamy colors. T here were a few visual similarities between introverted (I) and intuitive (N) images, such as being set in nature or showing intimate couples. However, both of these aesthetic elements were directly opposite to thinking (T) images, which were set in academ ic or scientific settings such as a library or laboratory. Thinking (T) images also typically portrayed one person alone and were more impersonal in nature, which is in direct opposition to a romantic or intimate relationship. And, although both intuitive (N) and thinking (T) images could show thought bubbles, the content within those thought bubbles was drastically different, with intuitive (N) thought bubbles portraying brightly colored creative or future thought and thinking (T) thought bubbles depicting logical or rational mental processes. Given the combination of two similar dimensions (IT) and one disparate dimension (N), it becomes obvious that compromises must be made across all three dimensions in order to create a single visual profile for introve rted/intuitive/thinking (INT) images. Below is one example,

PAGE 395

395 crafted by the researcher, of how the individual visual profiles for introversion (I), intuition (N) and thinking (T) could be combined. An introverted/intuitive/feeling (INF) image, c ould include a single person in a science or medical related profession isolated and alone in a laboratory with no one else visible. The person is not smiling or making eye contact with anyone (including the viewer). Instead, he/she is looking up or striking a thinki ng pose (looking up with hand on chin). The laboratory environment seems quiet and has a sterile, neutral, clean color palette with an abundance of head depi cting abstract scientific ideas. The image is widely cropped showing a large portion of the laboratory with the subject relatively small in the frame. The above example, which details the compromises necessary to create the visual profile for the introvert ed/intuitive/thinking (INT) personality, is representative of the careful give and take process that is required to create visual profiles for the majority of the multidimensional MBTI personality. It seems clear that the extraverted/intuitive/feeling (ENF ) and introverted/sensing/thinking (IST) are the simplest multidimensional MBTI personalities for advertisers to create with visuals alone. Although the qualitative group image sorts showed that it is possible to create the other six multidimensional MBTI personalities extraverted/sensing/thinking (EST), extraverted/sensing/feeling (ESF), extraverted/intuitive/thinking (ENT), introverted/sensing/feeling (ISF), introverted/intuitive/thinking (INT) and introverted/intuitive/feeling (INF) they will require calculated visual compromises, and, therefore, more careful attention as creatives design their advertisements. According to Ma (2008) poorly chosen or weakly executed advertising visuals can cause consumers to lose the

PAGE 396

396 preferred meaning, underscoring the importance of paying close attention to visual symbolic elements . Additionally, because each of these remaining six three dimensional MBTI personalities includes one personality dimension that is visually opposite the other two, creating multidimensional quantitative results have shown. Therefore, in addition to carefully selected visuals, advertisers may need to rely on added text (Barthes, 1977) to consistently transfer all th ree dimensions. attempted to examine visuals and multidimensional MBTI brand personalities. In their 2007 study, Cunningham et al. explored the overlap between MBTI brand per sonality and website design, specifically in regards to e commerce websites. Like the current study, Cunningham et al. (2007) first identified visuals they predicted would be associated with each individual MBTI dimension, and then combined those visuals t o create multidimensional MBTI brand personality profiles. However, unlike the current study, Cunningham et al. (2007) did not test any aspect of their visual profiles. Rather, the aesthetic elements were simply suggested by the researcher based on their o wn knowledge. Additionally, the visual suggestions for each dimension were simple with no depth, nuance or explanation. And, the combinations for each multidimensional visual profile were made without consideration of potential conflicts or need for compro mise. Therefore, although some suggestions by Cunningham et al. (2007) overlapped with the current associated with multidimensional MBTI personality in images, as opposed to focusing solely on e depth, nuance and explanation than those offered by Cunningham et al. (2007). Finally, and

PAGE 397

397 perhaps most importantly, the curren t study used a scientific, mixed method study to obtain its results, rather than exclusively on the judgment of the researcher. Phase Three Discussion Survey The third and final phase of the current study was to conduct a quantitative survey. Using image s identified in the phase two group image sorts, the researcher created 10 advertisements, which were used as stimuli. Then, using the BHPI scale, the online survey asked respondents to evaluate both their own personality and the brand personality of the 1 0 advertisements. For each advertisement, respondents also answer questions about their attitude toward the ad ( A Ad ), attitude toward the brand ( A Brand ) and purchase intent ( PI ). Last, the survey asked several demographic questions (e.g., gender, age, race , major). Examining the survey results by individual dimension extraversion/introversion (E/I), sensing/intuition (S/N) and thinking/feeling (T/F) showed that significantly more than 50% of the respondents correctly assigned advertisements the appropri ate single dimension. In fact, the percentages were all well above 50%, anywhere from 63.2% for the sensing (S) dimension to 78.2% for the feeling (F) dimension. This data suggests that, despite lower than expected scale ividual subscales functioned largely as intended as significantly more than 50% of survey respondents were able to correctly identify images based on a single dimension. Further, this data suggests that it is possible, and perhaps quite easy, for advertise rs to transfer a single MBTI personality dimension to a brand through an advertising image. None of the images used to create the advertising stimuli for the current study were chosen for their ability to best represent a single dimension. Instead, they we re chosen as compromising images that visually represented parts of three separate MBTI personality dimensions. Yet, most

PAGE 398

398 respondents 63.2% to 78.2% were still able to correctly identify single dimensions. This finding aligns with the Jewett and Suther images were able to successfully transfer a single MBTI personality dimension. Using images that were selected as ideal images for four MBTI dimensions extraversion (E), introversion (I), thinking ( T) and feeling (F) Jewett and Sutherland were able to achieve more than 80% agreement for each dimension. In light of the difference between Jewett and Sutherland (2013) using ideal images for a single personality and the current study using compromising images for % to 78.2% seem appropriate and even high. After looking at the individual MBTI personality dimensions, the researcher then examined the results by multidimensional advertisement . When the individual MBTI personality dimensions were combined into multidimensional personalities, a chi square showed a significant difference ( X 2 = 275.8, df = 9, p < .001 ) between the observed values and the expected value of 50%, with some ads above 50% agreement and some below. When the data was examined closer, two findings emerged regarding 1) ENF and IST brand personalities, and 2) the extraversion/introversion (E/I) dimension. First, the advertisements with more than 50% agreement were the IST ad (61.6%), the ENF participants ad (57.2%) and the ENF researcher ad (51.5%) . Yet, the three ads had only two multidimensional personalities extraverted/intuitive/feeling (ENF) and in troverted/sensing/thinking (IST). These findings parallel findings from the group image sorts in the second phase of the current study. The group image sort participants identified extreme similarities in the visual profiles of extraversion (E), intuition (N) and feeling (F) and in the visual profiles of introversion (I), sens ing (S) and thinking (T). For example, extraversion (E),

PAGE 399

399 intuition (N) and feeling (F) were all visually identified by showing relationships, emotions (e.g., happy, romantic), smiling, laughing, touching and bright colors (e.g., vibrant, warm, saturated). Alternately, introversion (I), sensing (S) and thinking (T) were all visually identified by showing subjects focused or immersed in a thought/activity/goal, science or medical professions, and sterile colors (e.g., white, neutral, cool). Combined, the grou p image sort and survey data suggest that, because of their visual similarities, extraverted/intuitive/feeling (ENF) and in troverted/sensing/thinking (IST) brand personalities may be the easiest to achieve through images and visual symbolic elements. After looking at the advertisements where more than 50% of the respondents correctly identified the multidimensional MBTI personality, the researcher examined the advertisements with correct identifications in the 30% and 40% range. Between 30 % and 49% of the s urvey respondents assigned the correct multidimensional MBTI personality to three advertisements the IN F participants ad (47.3%), the EST ad (38.6%) and the IN F researcher ad (31.9%). These three ads also had only two multidimensional personalities int roverted/intuitive/feeling (INF) and extraverted/sensing/thinking (EST) which, except for the extraversion/introversion (E/I) dimension, paralleled the two top dimensions extraverted/intuitive/feeling (ENF) and in troverted/sensing/thinking (IST). This finding suggests that the extraversion/introversion (E/I) dimension might cross the aesthetics barrier the easiest. Likely, the extraversion/introversion (E/I) dimension of an image could be swapped by showing more (E) or less (I) people and/or adjusting t object (I). Put another way, the top six advertisements had either an intuitive/feeling (NF) combination or a sensing/thinking (ST) combination, suggesting that visually pai ring

PAGE 400

400 intuitive/thinking (NT) or sensing/feeling (SF) could be quite difficult, which again parallels group image sort findings. Results from the partially supported Hypothesis 1 (H1) lent further support to extraverted/intuitive/feeling (ENF) and in trovert ed/sensing/thinking (IST) personalities being the easiest to communicate visually. According to survey results, the ENF participants ad ( 57.2%, p < .001), ENF researcher ad ( 5 1 . 5 %, p < .001), IST ad (61 . 6 %, p < .001) and INF participants ad (47.3%, p = .00 3) were the only four advertisements where a statistically significantly higher percentage of respondents assigned an advertisement the correct personality. These results again underscored the relative ease of achieving an extraverted/intuitive/feeling (EN F) or introverted/sensing/thinking (IST) personality using only advertising images. Additionally, other aspects of the results from H1 support the idea that INF and EST personalities may be more difficult but still possible to communicate using advertising images. While one of the two introverted/intuitive/feeling (INF) advertisements tested INF participants had the highest percentage of participants assign it an introverted/intuitive/feeling (INF) personality, t he other introverted/intuitive/feeling (I NF) advertisements tested INF researcher had highest percentage of participants assign it an extraverted/intuitive/feeling (ENF) personality . This result may again suggest that extraversion/introversion (E/I) may be the dimension that crosses the aesth etic barriers most easily. This conclusion may be explained by the fact that the INF researcher ad showed two people, a romantic couple, in the image, while the INF participants ad showed only one. While group image sort participants felt that a romantic c ouple was often identified as a single unit and could be perceived as introverted, not all survey respondents seemed to respond in the same way. Survey results showed that a larger percentage of respondents (38.0%) identified the INF researcher ad as extra verted, presumably because of

PAGE 401

401 the two people, instead of introverted (31.9%). This difference in response could also be explained, in part, by the difference in the amount of time group image sort participants (i.e., several minutes) and survey respondents (i.e., a few seconds) spent looking at the images. Combined, however, more than two thirds (69.9%) of the survey respondents recognized the ad as intuitive (N) and feeling (F). Therefore, it appears that advertisers may be able to move an advertisement fr om extraverted (E) to introverted (I) by adding or subtracting the number of people shown in the image. Future research should further test this assumption by using the same image with a varying number of people to see if this hypothesis holds true in mult iple images. It is also possible that the visual profiles of the extraverted/intuitive/feeling (ENF) personality and the introverted/sensing/thinking (IST) personality are so strong that they function as default personalities. That is, the visuals are so c losely tied together, that an image containing two of the three personality dimensions will be assigned the third almost instinctually. For example, if an image is strongly extraverted (E) and intuitive (N), it might, by its visual nature, be almost automa tically identified as feeling (F) as well. Therefore, the visual nature of multidimensional MBTI personalities may be more bipolar in nature, rather than having eight nuanced personality options like the current study hypothesized. This idea may also be re inforced by some of the unsupported parts of H1. For example, in addition to the two ENF ads, the highest percentage of survey respondents assigned the ESF ad (45.6%, p < .001 ), the ENT ad (30.4%, p < .001 ) and the INF researcher ad (38.0%, p = .003 ) an EN F personality . Additionally, along with the IST ad, the highest percentage of survey respondents assigned the INT ad (42.2%) an IST personality ( p < .001). And, while the highest percentage of respondents assigned the EST ad an EST personality (38.1%), it was not significantly higher than the percentage of respondents who assigned the same advertisement an

PAGE 402

402 IST personality (35.9%) ( p = .198). Therefore, out of the ten total advertisements, respondents assigned five advertisements the ENF participant ad, th e ENF researcher ad, the ESF ad, the EST ad and the INF researcher ad an ENF personality and two advertisements the IST ad and the INT ad an IST personality, with a third advertisement the EST ad showing no statistical personality and an IST personality. Thus, in total, survey respondents assigned eight of the 10 advertisement either an ENF or IST personality, which supports the idea that these two multidimensional MBTI personalities may function as default visual perso nalities. As discussed above, the notion that extraverted/intuitive/feeling (ENF) and in troverted/sensing/thinking (IST) personalities may function as a set of bipolar, default personalities may help to partially explain the unsupported parts of H1. Additi onally, the significant number of unsupported hypotheses those where the highest percentage of survey respondents did not identify the predetermined brand personality may also have been caused by using images in the survey that did not align with all o f the visual symbolic elements identified by group image sort participants. In this study, the researcher set out to test a single set of images using multiple methods. Therefore, despite the fact that group image sort participants did not strongly associa te each multidimensional MBTI personalities with a specific image, the researcher chose to use the best option available rather than adding or replacing images in the image deck. While this allowed for continuity and comparisons within and between the qual itative and quantitative data, it may have weakened some findings and resulted in more unsupported hypotheses. A future survey should be conducted using new advertising images that completely conform to the group image sort findings to test whether a great er number of

PAGE 403

403 multidimensional MBTI personalities are recognized by the majority of survey respondents, thus lowering the number of unsupported hypotheses. For advertisers, the implications of these findings mean that extraverted/intuitive/feeling (ENF) and in troverted/sensing/thinking (IST) may be the easiest multidimensional personalities to achieve solely through images. With careful attention to the extraverted/introverted (E/I) dimension, it may also be possible to use images to transfer an introverted/ intuitive/feeling (INF) or an extraverted/sensing/thinking (EST) personality by adjusting the number of people in the extraverted, sensing/thinking (EST), extrav erted/intuitive/thinking (ENT), introverted/sensing/feeling (ISF) and introverted/intuitive/thinking (INT) may be significantly qualitative group image sorts, adv ertisers may be able to increase the percentage of consumers who assign an advertisement the desired multidimensional MBTI personality. However, this possibility needs further testing before it c an be said with any confidence. Alternately, advertisers migh t choose to use text to anchor their desired meaning as suggested by Barthes (1977). Rather than placing the full brand personality burden on the advertising visual alone, adding text could reinforce the intended brand personality and guides the reader tow research should further test the possibility of transferring personalities through images alone. However, more importantly, additional research is needed to explore the role text plays in partnering with images to anchor visual personality meaning. The H1 results also showed a difference in how the qualitative group image sort participants and the researcher picked representative images for some personalities. For six of the

PAGE 404

404 eight multidimensional personalities included in this study, the participants top choice aligned ENF and INF participants chose an image that was different from those the resea rcher chose. In these cases, two images were tested in the phase three survey percentage of survey respondents assigne d the desired personality to the advertisement with the participant chosen image rather than the advertisement with the researcher chosen image. This finding indicates that researchers and practitioners should have more confidence in images chosen through group image sort processes rather than by a single researcher or practitioner. After reflecting on the differences between herself and the participants, the researcher recognized several differences that might explain why group image sort participants more effectively chose some images. First, the researcher assumed that participants would make brand personality decisions based mostly on certain, more obvious visual symbolic elements discussed in literature and design theory, such as color, body language an d number of people. But, while participants did pick up on these specific visual symbols as indicators of particular personalities, they also took into account a wider range of visual symbolic elements than the researcher expected, such as trendy attire, h ow posed an image seemed and subtle setting cues (e.g., chandeliers indicated discretionary income, extra crayons and paper on the table indicated the possibility of more children around, failing to cover nice parquet floors while painting indicated imprac ticality). In this way, participants were more visually literate and sophisticated than the researcher expected. Overall, participants also seemed to draw more personality meaning from the human subjects in the images. While the researcher surveyed the ent ire image holistically, equally

PAGE 405

405 examining all aspects, participants seemed to more heavily weight their impressions of the human subjects in the image. As qualitative results suggested, many participants seemed unable to separate the perceived personality of the human subject from the personality of the whole image. Therefore, when choosing the images, the researcher did not foresee the increased importance of the humans in the image versus the other elements. Additionally, participants mixed these visual i ndicators with stories, personal experiences and prior knowledge, to a degree which the researcher did not anticipant, often relying on the spirit of the image, rather than just the visual symbols, to determine personality. Because the researcher is older and in a different phase of life than the participants, she did not share some of the same cultural experiences and meanings as the participants. However, because participants in the group image sorts were more similar to the survey respondents in age, lif e stage, degree, etc., they may have relied on more shared cultural meanings to come to similar personality conclusions. Moreover, having been trained to conduct unbiased research, the researcher tried to detach herself and objectively choose images based on visual symbolic elements such as those in past literature and design theory. However, participants seemed to incorporate more subjective meanings as they read and negotiated personality. Finally, differences may have also occurred because the researcher was much more invested in the study process and results than participants. After spending a great deal of time seeking to understand the intricacies of brand personality, MBTI dimensions, and how images house meaning, the researcher then carefully and tho roughly examined hundreds of potential images on stock image sites. After comparing such a large number of images, the researcher chose those images that represented each multidimensional MBTI personality better than the rest. However, participants did not have the same background understanding, comparative basis

PAGE 406

406 or investment in the image personalities. Therefore, participants made their judgments based solely on what was in front of them for a relatively short amount of time. Further, these quicker more d In addition to examining the survey data within individual test advertisement, the current study also investigated how each of the eight MBTI personalities included in this study were evaluated across advertisements. Analysis for Hypothesis 2 (H2), which was mostly supported, compared each multidimensional MBTI personality across all 10 adverti sements. Results showed that for all but one of the eight multidimensional MBTI personalities each personality was significantly higher in the predicted advertisement (s) than in any other advertisement. For example, a higher percentage of respondents assig ned the EST ad an EST personality when compared to all the other advertisements tested . For the tenth personality extraverted/sensing/feeling (ESF) the hypothesis was only partially supported. Although the highest percentage of participants assigned an ESF personality to the ENF participants ad (1.60%), it was not significantly higher than the percentage of participants that assigned an ESF personality to the ESF ad (1.41%, p = .250), resulting in the statistical equivalent of a tie between the personal ities. Specifically, in addition to identifying symbolic visual elements associated with sensing (S), during group image sorts, participants also noted several intuitive (N) visual symbolic elements in the ESF image, which showed a family (mom, dad and you ng boy and girl) painting an interior wall. In particular, participants commented on the fac t that the family had chosen to paint the wall a bright, chartreuse green color. The family was also painting a wall with what participants deemed a variety of oddl y shaped paintbrushes. However, perhaps the most common intuitive visual symbolic element

PAGE 407

407 participants pointed out was that the family had not taped the walls or covered their nice, parquet floors prior to painting. Combined, some participants felt that th ese aspects of the image made it unconventional , which was one of the BHPI scale items for intuition. Thus, it seems that this image was clearly extraverted (E) and clearly feeling (F), but that the sensing/intuition (S/N) dimension was not well defined vi sually. However in most cases, this finding indicates that, while the intended multidimensional personality may not have been the most assigned personality within the ad, it was assigned more for the intended ad than for any other ad. Therefore, the resear cher was successful at choosing images that were more aligned with the visual profile of the intended personality than any other image. In other word, no other advertisement was more visually aligned with the intended personality than the hypothesized adve rtisement. For example, no other advertisement was more visually aligned with an EST personality than the EST ad. which were based on previous research ( e.g., Akay, 2001; Boudreaux & Palmer, 2007; Cunningham et al., 2007; Jewett & Sutherland, 2013; Park et al., 2005 ) for each multidimensional MBTI personality had some validity. Therefore, by combining past literature for both individual and multidimensional MBTI brand personalities, advertisers may be able to fine tune the visual profile for each multidimensional MBTI personality and more effectively use visual symbolic elements in images to transfer a specific person ality. Beyond examining the influence of visual symbolic elements on brand personality, the current study also investigated the link between brand personality and attitude toward the ad ( A Ad ), attitude toward the brand ( A Brand ) and purchase intent ( PI ) . Bo th human personality and

PAGE 408

408 gender were also included in a three way analysis of variance (ANOVA) to determine if they acted as confounding variables. For attitude toward the ad, results showed a statistically significant main effect for brand personality, F (7, 5134) = 5.58, p < .001 , and human personality, F (7, 5134) = 2.89, p = .005 . However, in both cases, the effect size was so small (partial eta squared = .008 and .004 , respectively) that neither brand nor human personality was determined to be a factor in determining attitude toward the ad. Similarly, a statistically significant main effect for brand personality, F (7, 5134) = 6.31, p < .001 , and human personality, F (7, 5134) = 2.91, p = .005 , were also found for attitude toward the brand. But, again, in both cases, the effect size was so small (partial eta squared = .008 and .004 , respectively) that neither brand nor human personality was determined to be a factor in determining attitude toward the brand. Finally, in the case of purchase intent, a stat istically significant main effect for brand personality, F (7, 5134) = 5.58, p < .001 , was found, but with such a small effect size (partial eta squared = .00 6), that it was determined not to be a factor. Therefore, although statistically significant main effects were observed for attitude toward the ad, attitude toward the brand and purchase intent, the practical significance as determined by the effect sizes, suggested that brand and human personality had no discernable effect on behaviors. These findings are particularly interesting in light of the group image sort findings that a esults might seem to suggest somewhat contradictory conclusions since individual personality appeared to be influential in group image sorts, but not during surveys. However, this difference may be due to the difference in the amount of time participants s pent looking at each image during the different methods. As

PAGE 409

409 mentioned above, research from Perception Research Services (PRS) suggested that participants who are exposed to advertisements for an extended amount of time (i.e., a minute or more) recognized m any more meanings than the typical consumer, who, on average, spends only between 3 to 20 seconds looking at a print advertisement (Young, n.d.). Because group image sort participants spent several minutes looking at each image while survey participants we re more likely to quickly view the advertisements, this data suggests that participants may have relied more heavily on their own personalities during group image sorts. Alternately, survey participants, who more quickly assessed the images, may have engag ed in a more cursory, surface reading that did not include relating the image to their own personalities. Therefore, the differences in time spent viewing images may influence the likelihood that a consumer allows their own personality to influence their b ehavior and attitudes. Moreover, Young (n.d.) suggested that print ads were more accurately judged through shorter exposures (i.e., 5 10 seconds) that more closely mimic the way consumers would actually interact with an advertisement. Thus, while the curre influences how participants read and negotiate image meaning, are valuable for general visual research, the survey findings, which show little to no influence, may be more accura te for advertising researchers and practitioners. More research is needed to understand the effect brand personality has on consumer behaviors and attitudes. Although other past studies showed that brand personality could create more positive, unique and c ongruent brand association (Freling & Forbes, 2005), increase brand trust, brand affect and brand loyalty (Sung & Kim, 2010) and increase purchase intent toward th e ad, attitude toward the brand or purchase intent . However, in each of these previous

PAGE 410

410 studies where a link was found between brand personality and consumer behavior and attitudes (Boudreaux & Palmer, 2007; Freling & Forbes, 2005; Sung & Kim, 2010), resear chers were e.g., Avis, 2012; Azoulay & Kapferer, 2003 ) brand image, which is more inclusive than just brand personali ty. Brand image measures, such as related to intelligence, gender and social class , which personality scholars do not consider to be personality facets . However, the current study used the BHPI, which is based on a tr ue personality measure, the MBTI. brand personality and consumer behavior and attitudes, may more accurately reflect the relationship, or lack thereof, between true bran d personality and consumer behavior and attitudes. Because there seems to be some connection between brand image and consumer behavior and attitudes (Boudreaux & Palmer, 2007; Freling & Forbes, 2005; Sung & Kim, 2010), it is possible that one of the variab les that differentiate the two concepts, such as intelligence, gender or social class, is actually responsible for these effects. The current study tested gender as a potential confounding variable and found no effect; however, neither intelligence nor soc ial class was measured. Future studies should further investigate the connection between these non personality aspects of brand image and consumer behavior and attitudes, such as attitude toward the ad, attitude toward the brand and purchase intent. It is also possible that brand personality has some effect on another variable not related to brand image, which was not measured in this study, that more directly affects consumer attitudes and behaviors. For example, other scholars have found that one of brand valuable contributions to a brand is its ability to create brand differentiation (e.g., Biel, 1993; D.

PAGE 411

411 Aaker, 1996; Park, Jaworski & Maclnnis, 1986; Ramaseshan & Tsao, 2007 ) . Perhaps brand personality aids brands in differentiating them selves from the competition, and this differentiation, in turn, shapes consumer behaviors and attitudes. Future research should attempt to better understand what variables brand personality most directly influences, thereby determining if brand personality has an indirect effect on measures such as attitude toward the ad, attitude toward the brand and purchase intent. Theoretical Implications The current study also made important contributions to meaning transfer theory (McCracken, 1986; 1988; 1990; 1993; 2 005 moves from culture to brand and on to the consumer. According to meaning transfer theory, cultural meaning is transferred to consumers via a two step process : 1) from culture to the consumer good or brand via advertising or fashion and then 2) from the good or brand to the individual consumer through purchase and consumption ritual s (McCracken, 1986; 1988; 1990; 1993; 2005 ). McCracken ( 1986; 1988; 1990; 2005 ) also noted that advertising was one of the primary vehicles for transferring meaning from the culturally constituted world to a consumer good. Therefore, by focusing on how brand personality as meaning is transferred from culture to brand through advertising, the current study was able to expand meaning tr ansfer theory (see Figure 5 1) and increase its usefulness for advertisers. In particular, the researcher suggested that advertisements were composed of two distinct parts words and images. Further, the researcher asserted that advertising images were co mposed of visual symbolic elements that house and visual symbolic elements within images to contain and transfer meaning, making them a viable addition to McC 1986; 1988; 1990; 1993; 2005 ) meaning transfer theory. However, this

PAGE 412

412 study found that visuals alone might only be able to consistently transfer ENF and IST brand personalities. ( 2012 ), Therefore, like Campbell and Warren (2012), this study also adds nuance to meaning transfer theory. Additionally, the researcher speculates, and future research shou ld test, that visual symbolic elements are most valuable when paired with carefully written text to serve as an anchor that more obviously guides consumers toward the dominant reading (Barthes, 1977; Hall, 1980). Figure 5 1 . M odifications to mean ing tran sfer theory . The new model is s (1993) updated model, and the current study .

PAGE 413

413 In addition to findings on the abilities and limitations of visuals symbolic elements to house and transfer meanings within advertising images, the current study also added to According to McCracken ( 1986; 1988; 1990 ), advertising creatives must exercise extreme care to s elect visual symbolic elements that hold the desired meanings and suppress unwanted ones. reaction (McCracken 1986). By shedding some light on the science behind these instincts, the current and multidimensional MBTI brand personality provides both scholars and practitioners with a more detailed understanding of advertisin (Batra et al., 1993) . Although the individual and multidimensional MBTI brand personalities provide a starting place for further investigation . However, while study participants were able to provide some specifics for individual personality dimensions, such as the number of people in extroverted/introverted (E/I) images and the body language of thinking/feeling (T/F) images, in several cases, mo re abstract themes emerged with less clearly defined visual symbolic elements. For example, participants described images in the sensing (S) dimension as goal oriented , but they had trouble pointing to the exact visual symbols that represented goal orienta tion. Or, in another example, participants identified thinking (T) images as logical/rational contributed to this feel. Therefore, by conducting additional studies with more images, researchers can attempt to more clearly define some of these abstract categories, adding to the

PAGE 414

414 list of specific visual symbolic elements that consumers associate with particular MBTI personalities. Finally, the current study opens the door for scholars to begin considering the role of the image in meaning transfer theory. Although several scholars have identified visuals as perhaps the most powerful and meaning filled elements of advertising (McCracken, 1986; Postrel, 2003; Williamson, 1978/2010), thus far, the role of images in mea ning transfer theory ha s been largely ignored and remains woefully untested . Hopefully, the current study will be the first of many to begin exploring how the specific visual symbolic elements within images wor k to convey particular meanings with future st udies both continuing to add to our knowledge of the role visuals play in transferring brand personality, as well as investigating other types of cultural meaning. Limitations After concluding this mixed method study, several limitations were identified th at had the potential to a ffect results and should be considered in future studies. Because this current study included three different phases a scale reduction, group image sorts and a survey limitations for each method are discussed below. Scale Reduc tion First, the BHPI scale reduction was based on (1998) original brand personality scale. Therefore, any limitations inherent in the original scale may have been passed on to the BHPI . Additionally, any anomalies in t he sample may have affected reliability. Further, the BHPI retained (1998) original scale. Although the BHPI provid es an adequately reduced MBTI personality scale for both humans and

PAGE 415

415 brands , more work should be done to fine pair should be examined to discover any weak links , outdated words or misunderstood meanings . Replacing single adjectives in one or more pair s , could make the scale more representative of the true MBTI dimensions and increase reliability. Also, original scale did not specifically consider the visual aspect s of brand personality. There fore, the BHPI may more accurately measure text rather than visuals. Future studies should attempt to determine if there are any differences when the BHPI is applied to text as opposed to visuals. And, as individual adjectives are potentially refined and r eplaced, their ability to measure visual symbolic elements should be considered. Alternately, researchers may want to consider creating a visual specific version of the BHPI scale to better measure visual brand personality. Group Image Sorts Second, the gr oup image sorts provided i nteresting findings that had high face validity, fit with results from other phases (i.e., the survey) of the current study and reached saturation. However, the qualitative results could still be nefit from conducting additional gr oup image sorts, which would provide added support. Additional group image sorts should be conducted with the same images to ensure that findings remain stable and no new themes emerge. Moreover, conducting a second round of group image sorts using a new s et of images, which are based on symbolic elements associated with particular MBTI brand personalities. Also , the group image sorts only included advanced adver tising and graphic design majors who had high knowledge of brand personality and visual theory. While this provided a level of detail and understanding that the researcher was looking for, it may have also skewed results. More studies should be conducted w ith other majors to understand the differences in

PAGE 416

416 participants who are less visually literate . F uture research should also include students who are more acquainted with the subject matter in the images. For example, several images showed scientists in labs and doctors. C onduct ing group image sorts with students in scie nce and medical related majors could provide new insights. It is also possible that the images themselves could have been a limitation. Results indicated that some images were better fits for each dimension, which could have affected results. And, although the researcher chose the 24 images in the image deck with care, because images are so complex and so filled with meaning, there is always the potential, and even likelihood, for unidentified factors within the images to influenced results, as was the case in the Additionally, the current study asked participants to sort 24 images three times once for each MBTI dimension. However, b y the third sort, some participants were showing signs of fatigue , indicating that 24 images were too many for a single group image sort session . In the future, researchers should consider either using an image deck with less than 24 images or conducting less than three sorts in a single session. Based on group image sort results, the terms sensing and intuition also caused confusion among participants. Because theses labels are part of the official Myers Briggs Type Indicator, the researcher had little control over the words use d. However, future studies may chose to use alternate labels that are less likely to confuse participants. Alternately, future researchers could also provide participants with definitions, but no labels. Instead, each dimension extraversion (E), introver sion (I), sensing (S), intuition (N), thinking (T) and feeling (F) could be labeled as a unique number or letter (e.g., Dimension 1/Dimension 2 or Dimension A/Dimension B) . This process would simultaneously eliminate confusion and prevent participants fr om resorting to the

PAGE 417

417 common understandings of all the labels rather than particular MBTI definitions. In addition , the MBTI may also want to consider officially renaming the sensing/intuition (S/N) dimension to more effectively and easily communicate their meaning beyond scholars and academics. Moreover, advertisers are most likely to use the group image sort technique to find the best image for their predetermined brand personality. Therefore, they would only be interested in discovering one multidimensiona l personality, rather than eight , like the researcher. Because this might cause problems unidentified by the current study, additional research should be conducted using the group image sort technique more like an ad vertising agency might use it. Survey Fi nally, the survey phase of the current study used a student sample. While this sample was appropriate for the current study , additional research should be conducted with non student samples to increase the external validity and determin e the transferability of the results. Also, inherently, surveys do not provide information on why respondents choose specific answers. Therefore, the researcher was unable to determine if the survey respondents based their personality decisions on the same visual symbolic elements as group image sort participants. To address this issue, future researchers may want to conduct additional in depth interviews with some survey respondents after completing the survey to better understand how they made brand perso nality decisions. Moreover, a s exploratory researcher, the current study chose to use a survey, rather than an experiment, which limited the ability to understand which specific parts of an image most cisions. However, with the current now

PAGE 418

418 conduct experiments to manipulate particular visual symbolic elements in order to achieve a desired brand personality.

PAGE 419

419 CHAPTER 6 CONCLUSION AND PRACTICAL APPLICATION In conclusion, this study found that, through advertising, companies do have some control over their brand and its meaning, including brand personality. Although some scholars ( e.g., Davis, 2000; Elliot, 2006; Neumei er, 2006, Plummer, 1984/2000 ) have asserted that the ng. Although companies do have a signif individual personalit y and gender did not have any significant influence on how respondents assigned personality to brands, participants in the group image sorts used their own prior knowledge, past experiences and individual personality as they assigned personality to images. This difference between the group image sorts and the survey may be due to the fact that participants in the group image sort spent several minutes viewing and interacting with each image, while survey respondents viewed the advertisements that contained the images much more quickly. In fact, a report from Perception Research Services (PRS) found that the 5 seconds to a high of 15 20 seconds (Young, n.d.), which is similar to the amount of time survey respondents spent on each advertisement. Moreover, PRS also found that when consumers have extended exposure to an ad (i.e., many of which will not come across when consumers interact with ads for a more typical 5 10 seconds (Young, n.d., p. 2). Therefore, participants in the group image sorts may have

PAGE 420

420 incorporated more of their own prior knowledge, experiences and personality because of the extended viewing time. According to Hall (1980), by mixing their own experiences with the participated in a negotiated reading of the personality. Yet, even wh en participating in a negotiated reading, which is tempered by their own personal experiences and interests, Hall (1980) asserted that consumers still understood and generally agreed with the preferred meaning. However, not all group image sort participants who engaged in a negotiated reading ideal image choices by the researcher. For some images, participants pointed out additional visual symbolic el ements or identified a spirit to the image that the researcher had not taken into the preferred meaning may have to do with the group image sort process. Because participants were allowed to review and discuss the image for much longer than a consumer typically views an advertisement, participants often talked themselves out of their initial gut reaction to an image. For example, a participant might initially agree with the preferred personality meaning, identifying an image as introverted (I). But, after looking closely at the image for another minute, he/she would start telling a story or relying on personal experience or prior knowledge to identify reasons why th e image could be extraverted (E). And, in some cases, these tactics resulted in the participants actually talking himself/herself into the opposite personality from their initial gut reaction . Group dynamic may also have played a part in convincing some pa rticipants, who initially agreed with the preferred meaning, to assign the image an alternate meaning instead. In

PAGE 421

421 some groups, it almost seemed as if participants felt the need to figure out how an image could fit both personalities, like it was too simple to just place it in their gut reaction category. advertising images might be more true to life. However, the strong influence of personal experiences, prior knowledge and individual experiences expressed by group image sort participants cannot be ignored. Combined, these findings suggest that, while present, the influence of personal experiences, prior knowledge and individual personality may be tempered by the short amoun t of time spent viewing an ad. Therefore, with careful attention to the specific visual symbolic elements contained in an image, it seems that advertisers can still influence Furthermore, this study has provided advertisers with a list of visual symbolic elements that represent individual and multidimensional MBTI brand personalities. Although this list does not include all possible visual symbolic elements that could be present in an image a potentiall y im possible task given the complexity of visual symbolism it does provide advertisers with a starting point. With additional research and practical application, this list can grow to encompass even more visual symbolic elements, providing advertisers wi th an increasing resource that links specific aesthetics to particular personalities. While these findings may not surprise most advertising creatives who have spent their careers choosing specific images in order to elicit particular consumer responses, i t does provide more concrete language and scientific backing for their decisions. Although creatives often know instinctually which image is appropriate for an advertisement (McCracken, 1986), the c urrent some objective corroborati on for those decisions. In addition, these

PAGE 422

422 findings can help less experienced practitioners, who may not be as visually savvy, choose images that are more likely to produce the desired brand personality outcomes. Beyond providing a list of visual symbolic elements associated with particular brand personalities, the current study has also shown that implementing specific visual symbolic elements can lead consumers to identify the desired brand personality outcome. While not all of the images tested worked eq ually well, the study did show that, in some cases, the majority of consumers were able to correctly associate a brand with the intended personality. Further, if the findings from the current study were used to select a new set of images, which were tested in an additional study, future results could be even stronger. In particular, because of extreme similarities in their visual profiles, the current study found that some brand personalities extraverted/intuitive/feeling (ENF) and introverted/sensing/thi nking (IST) might be easier to elicit through advertising visuals alone. Additionally, one dimension extraversion/introversion (E/I) might cross those aesthetic boundaries more easily than others. In fact, exchanging extraversion and introversion mig ht be as simple as adjusting the number of people in an image (i.e., adding multiple people for contact (i.e., looking at another subject or the viewer for ex traversion or looking at an inanimate object or out of the frame for introversion). Therefore, with careful attention to the visual symbolic elements associated with the extraversion/introversion dimension, advertisers may also be able to use visuals to pr ovide brands with an introverted/intuitive/feeling (INF) or extraverted/sensing/thinking ( EST) personality fairly easily. might not be as easy to achieve visually. In these cases, advertisers might need to 1) pay extreme

PAGE 423

423 attention to visual symbolic elements and the compromises necessary to account for differences in visual profiles, or 2) use headlines and/or body copy to anchor their intended meaning (Barthes, 1977). According to Barthes (1977), there is no doubt that advertising images have intentional meanings imbedded in them a priori by the advertiser through carefully chosen visual symbolic elements. However, Barthes (1977) suggested that text, when paired with a n image, could serve and anchoring function that tells viewers which aspects of the image are most important and guide s them toward a specific meaning . Like a remote control, Barthes (1977) suggested that text could serve to steer readers toward the advert which, in this case, is a particular brand personality. Thus, adding guiding text to advertising visuals could increase the likelihood that consumers identify the advertisement with the intended brand personality. In fact, the only advertising (e.g., Marchand, 1985 ; McQuarrie, 2008 ; Pracejus, 2003 ) may be negatively , in the case of brand personality, eschewing words in favor of only images seems to leave too much room for interpretation, personal input and misunderstanding. So, while an advertisement that is dominated by a beautiful image with only a small representa tion of the company product and logo may be attractive and modern, it may also be ineffective. In actuality, previous generations of advertisers may have had it right by using text to anchor visual meaning specifying and guiding consumers toward the adve the bandwagon of image only advertisements and perpetuating an ineffective trend, practitioners should consider reinstating copy as a vital component of successful advertising.

PAGE 424

424 For advertisement s where text would be helpful in guiding consumers to the predetermined brand personality, group image sorts may provide advertisers with an effective source for gathering headline and body copy ideas. During group image sorts, participants often told elab orate stories about the images they viewed. From only one, still image participants based on a n image were engaged in. They were also told from different perspectives. Some stories were told through the eyes of one subject in the image. Others were established around the entire scene. Still others were based on surface level visual symbolic elements such a s clothes, hairstyles or pets. No matter what prompted the narrative, many stories would work as an excellent base for work together like a story (Felton, 2006), it makes sense for copy to start as a narrative. Through group image sorts, advertisers can listen to the stories participants build around an image. They can then morph these stories into copy that speaks directly back to their audience in a voice they r meaning and can be transmitted to the consumer (Hall, 1980, p. 164). T ake, for example, the stories participants told about an image that showed a man washing his dog (see Figure 6 1). Several participants told stories about how much this man loved his dog, about how taking the time to bathe the dog was a sign of care and at tention, and about the bond of love and companionship shown in the image. One participant even mentioned dollars. If stories such as these were told during a group image sorts conducted for a dog food

PAGE 425

425 company, it would require no great leap to translated them into copy that maintained the spirit of the narrative. For instance, an advertiser might write, We know you love your dog. You have an unbreakable bond. You kn ow where he your companion deserves the freshest, most wholesome ingredients at an affordable pri cost that much. Figure 6 1. Image of a man washing his dog, which was used during the group image sorts. However, for this process to work, companies must ensure that their group image sorts are composed of participants who mirror the target audience. Because the current study showed that group image sort participants did participant in negotiated readings mixing their own personal experiences and identity with the adverti sers meaning (Hall, 1980) it is important that

PAGE 426

426 as possible. In addition to providing stories for advertising copy, group image sorts composed of participants advertisers to select appropriate advertising images with specific multidimensional brand participants in group image s orts were capable of selecting images with particular multidimensional MBTI brand personalities. For six tions, indicating that they agreed with the ENF and INF the participants disagreed, choosing an image that the researcher had originally assigned a different personality. Therefore, in these two cases, the researcher chose to include both images in the survey 1) the personality assignment. Results showed that, in both cases, the images chosen by the participants were assigned the desired personality by a higher percentage of survey respondents than the images chosen by the researcher. These results indicate that the g roup image sort process can be a valid and valuable option for both advertisers and researchers who want to identi brand personality, perhaps even more so than a single researcher or practitioner. Overall, this study suggested that companies s hould not buy into the notion that consumers are the sole owners of their brand ( e.g., Davis, 2000; Elliot, 2006; Neumeier, 2006, Plummer, 1984/2000 ). While this study suggested that consumers might participant in negotiated meanings by mixing their own pe rsonal experiences, background knowledge and individual

PAGE 427

427 personality with the companies preferred meaning, it still appeared as if the desired meaning was often recognized and understood, particularly when viewed quickly as most advertisements are (Young, n .d.). And, by reversing the trend of image only advertising and adding copy back into advertisements, practitioners may be able to elicit even more dominant readings from consumers. Therefore, although they may not be in total control, companies should emb race their role as brand guiders with the ability to escort consumer toward the desired brand outcome by embedding and transferring brand personality through advertising. It also appears that advertising images, in particular, may play an important role in the transfer of particular meanings. While the input/out take (Plummer, 1984/2000) or encoding/decoding (Hall, 1980) acting as a guiding force to lead the consum er toward the brand personality established by a

PAGE 428

428 APPENDIX A SAMPLE ADS FOR SCALE REDUCTIO N DATA Extraverted (E) Ad Introverted (I) Ad Thinking (T) Ad Feeling (F) Ad

PAGE 429

429 APPENDIX B IMAGE DECK IMAGES EST Multiple scientists in a lab (possibly looking at the camera), lots of instruments, crowded around with a test in progress or measuring something (but not touching each), close crop Group of coworkers (possibly looking at the camera, may be wearing round glasses), in action, bright/saturated colors, working on project (building a structure, conducting surgery, etc.), hands busy/dirty, people not touching, close crop Group of professors/stude nts (possibly looking at the camera), wearing bright colors, working on project, hands busy/dirty, in action, design classroom, room relatively clean, people not touching, close crop ESF Family (parents and kids) in a kitchen (possibly looking at camer a), lots of cooking tools, crowded around a bowl touching each other, parent helping to exactly measure/ chop something, close crop Group (possibly looking at camera), bright/saturated colors, in action, hands busy/ dirty, worki ng on project (gardening/etc.), people touching/showing emotion, smiling/ enjoying the outdoor environment, close crop Group of people (possibly looking at camera), bright/ saturated colors, working on project, hands busy/dirty, in action, painting studi o, walls colorful, messy painting tools, people touching and showing emotion, close crop

PAGE 430

430 ENT Group of scientists in a lab (may be wearing square or thick rimmed glasses, possibly looking at the camera), colorful swirly illustrated fumes coming out of vi als, not touching, close crop [Note: Researcher added swirls] Group of coworkers (may be wearing square or thick rimmed glasses, possibly looking at camera), bright/saturated colors, people not touching, daydreaming about project possibilities (looking o ut over large garden/imagining a playground), hands idle, brightly colored illustrated looking at building plans, close crop [Note: Researcher added playground bubble.] Group of professors/students (may be wearing square or thick rimmed glasses, possibly looking at the camera), wearing bright colors, daydreaming about project possibilities, hands idle, classroom, walls white, computer areas clean, art on walls framed if any, people not touching, c lose crop

PAGE 431

431 ENF Family (parents and kids) (possibly looking at the camera), in a kitchen, pots simmering with colorful crowded together, touching each other, close crop Couple/family (possibly looking at the camera), daydrea ming about project possibilities (looking at large garden/imagining a playground), bright/saturated colors, hands idle (not actively building/constructing), brightly colored illustrated touching and showing emotion, close crop T eacher and student(s) (possibly looking at camera), wearing bright colors, daydreaming about project possibilities, hands idle, painting studio, walls colorful, messy painting tools, people touching and showing emotion, close crop IST One scientist alone in a darkish lab, muted colors, lots of instruments, concentrating on a test in progress, possibly measuring something or reading instruments wide crop One person (may be wearing round ish glasses), muted colors, sunrise/s unset/low light, working on project (gardening/building/etc.), hands busy/dirty, in action, looking at building plans, wide crop One student, wearing muted colors, low light, working on project, hands busy/dirty, in action/studying, walls white, computer areas clean, art on walls framed if any, wide crop

PAGE 432

432 ISF Couple alone in low lit kitchen (e.g., candles on small table set for two), possibly wearing apron, muted/neutral colors, lots of cooking tools, man or woman concentrating on measuring something, people touching and showing emotion, wide crop One person, muted/neutral colors (if possible), low light (if possible), interacting with pet (washing/petting/etc.), hands busy/dirty, in action, enjoying/soaking in the outdoor environment/beauty, pet tou ching and showing emotion, wide crop One person, muted/neutral colors (if possible), low light (if possible), working on project, hands busy/dirty, in action, pet touching and showing emotion, wide crop INT One person (may be wearing square or thick f ramed glasses), not looking at the camera, neutral/muted colors, One person (may be wearing square or thick framed glasses), not looking at the camera, muted colors, sunrise/sunset/low light, hands idl e, reading a book, wide crop One student (may be wearing square/thick framed glasses), not looking at the camera, muted colors, low light (if possible), daydreaming about project possibilities, illustrated computer, hands idle, comp uter areas clean, art on walls framed if any, wide crop

PAGE 433

433 INF Couple, not looking at the camera, no other people/animals, low light, muted/neutral colors, touching and showing emotion, wide crop One person, muted colors, sunrise/sunset/low light, daydr eaming about possibilities, hands idle, enjoying/soaking in the outdoor environment/beauty, pet touching and showing emotion, wide crop One person, wearing muted/neutral colors, sunrise/sunset/low light, daydreaming about project possibilities, hands idl e, pet touching and showing emotion, wide crop

PAGE 434

434 APPENDIX C (NAME OF ASSISTANT) . Thank you for participating in our group. to please turn off our cell phones. you choose to put them in a certain group. This discussion is part of a project to learn more about images and personality. During the group, you will have to work as a team. In some cases, you may disagree with the other group members. If this happens, I ask that you respectfully listen and then share your view. to hear from all of you equally. o know what you think. Although we will be audio taping our discussion this afternoon, your identity will be kept confidential. I will not associate your real name with any comments you make here. The recordings will be transcribed and your real name wil l not be associated with any of the comments. The transcriber and myself will be the only person with access to the recordings. Once transcribed, I will destroy all audio recordings. I am asking that you do not share what is said here with others outsid e of this room. look back at all of the comments and use them for additional research. One last thing, because I am taping it is important for only one person to speak at a time. I want to avoid anything that may affect the quality of my tape recording. (NAME OF ASSISTANT) will also be taking notes while we talk and helping me organize everything. This group discussion will last anywhere from an hour to two h ours. In order to gather the most helpful information, it is important that you actively participate the whole time. If you need to step out to use the restroom, the closest bathrooms are located (fill in with closest bathroom location) . Now, before we b egin our group discussion, I need everyone to read and sign a consent form. (CONSENT FORM NEXT PAGE)

PAGE 435

435 Consent to Participate in MBTI Group Image Sort The purpose of the group discussion and the nature of the questions have been explained to you. You co nsent to take part in a group interview where you will provide your opinion on images and personality dimensions. This group interview will last from 1 to 2 hours. You also consent to be tape recorded during this group interview. Your participation is v oluntary. You understand that you are free to leave the group at any time. You also understand that there are no personal risks or benefits associated with this research, and that you will not be compensated for your participation. If you decide not to p articipate at any time during the discussion, there will be no negative repercussions. Your identity will be kept confidential to the extent provided by law. Confidentiality cannot be guaranteed from group members. However, none of your thoughts or opinio ns will be shared with anyone outside of the researcher group unless all your identifying information is removed first. Further, the information that you provide during the focus group will be grouped with answers from other people so that you cannot be id entified. I have read the above information. I voluntarily agree to participate in the procedure and I have received a copy of this description. __________________________ ___________________________ Please Print Your Name Please Sign Your Name ____ ______________________ ___________________________ Date Principal Investigator Signature If you have questions about this study, please contact: Adriane (Jewett) Grumbein, Ph.D. Student Department of Advertising, Weimer Hall University of Florida, College of JM/COM PO Box 118400, Gainesville, FL 32611 8400 352 273 1643 adriane@ufl.edu Dr. J. Robyn Goodman, Associate Professor Department of Advertising, Weimer Hall University of Florida, College of JM/COM PO Box 118400, Gainesville, FL 3 2611 8400 352 392 2704 rgoodman@jou.ufl.edu If you wish to speak with someone about your rights as a research participant in this study, please contact: IRB02 Office University of Florida Box 112250, Gainesville, FL 32611 2250 352 392 0433

PAGE 436

436 Thanks again for participating today. Is everyone ready to begin? Please take a second to fold the paper in front of you like this (DEMONSTRATE) and write your ICEBREAKER To get us started and help us major and what year you are. Good! Thanks! DISCUSSION STARTER Great! Thanks for sharing e veryone. It may seem silly to think of yourself as a color, but different colors represent different computer they use also make us think different things a personality. People often think things like: Apple computers are creative, Mountain Dew is exciting and Jaguars are sophisticated . going to ask you to sort these pictures into several different piles based on different personality ular parts of each image make you put it into one of the piles. Remember, there is only one set of images, so you will have to work as a team to decide which pile the image belongs in. It will be easiest if you talk about each image out loud and discuss w hy you think it should be in a certain pile. MBTI DIMENSION #1 DIMENSION A & DIMENSION B . descriptio on the table top signs. MODERATOR READS DIMENSION DESCRIPTIONS LOCATED ON THE LAST PAGE We will have one pile for DIMENSION A and another pile for DIMENSION B . We will have a pile.

PAGE 437

437 Also, if you think an image may fit in both personality piles, please discuss why with your group. If the rest of the group agrees, we Does anyone have any questions? OK, here are the images. You may start sorting. THE MODERATOR HANDS AN IMAGE DECK TO THE GROUP THEN LISTENS, WATCHES AND TAKES NOTES WHILE THE GROUP SORTS. THE MODERATOR MAY ALSO PROBE FOR MORE DETAIL ON WHY THE GROUP DECIDED TO PLACE AN IMAGE IN A CERTAIN PILE IF THE DECISION IS NOT DISCUSSED IN OUT LOUD IN DETAIL (e.g., What about that particular im age made you decide to put it in the _________ pile?). THE MODERATOR WILL ALSO MONITOR THE GROUP DYNAMIC TO ENSURE THAT ALL MEMBERS ARE ABLE TO PARTICIPATE AND GIVE THEIR OPINION (e.g., _________, do you agree with the decision to place that image in the _________ pile? _________ what do you think about that image? _________, what stands out to you in this image?) Great! Does everyone agree with the piles the way they are? Now, can you tell me what aspects of the images were most important when you were sorting? What were some of the things that made you sure an image was DIMENSION A ? How about DIMENSION B ? Looking back though the DIMENSION A pile, which image do you think is most DIMENSION A ? How about DIMENSION B ? Can you think of any particular thi ngs that were missing from the images? Or something that you would have liked to see? If you could describe the ideal DIMENSION A image, what would it be? How about DIMENSION B ? Thanks for you feedback, that was really helpful! MODERATOR WILL CLIP EACH TABLE TOP SIGN TO THE IMAGES IN THAT PILE AND PLACE THEM TO THE SIDE. MBTI DIMENSION #2 dimensions.

PAGE 438

438 DIMENSION C & DIMENSION D . are also printed on the table top signs. MODERATOR READS DIMENSION DESCRIPTIONS LOCATED ON THE LAST PAGE Just like last time, we will have one pile for DIMENSION C and another pile for DIMENSION D do And again, if you think an image may fit in both personality piles, please discuss why with your one of the person Does anyone have any questions? OK, here are the images. You may start sorting. THE MODERATOR HANDS A SECOND IMAGE DECK TO THE GROUP THEN LISTENS, WATCHES AND TAKES NOTES WHILE THE GROUP SORTS. THE MODE RATOR MAY ALSO PROBE FOR MORE DETAIL ON WHY THE GROUP DECIDED TO PLACE AN IMAGE IN A CERTAIN PILE IF THE DECISION IS NOT DISCUSSED IN OUT LOUD IN DETAIL (e.g., What about that particular image made you decide to put it in the _________ pile?). THE MODERAT OR WILL ALSO MONITOR THE GROUP DYNAMIC TO ENSURE THAT ALL MEMBERS ARE ABLE TO PARTICIPATE AND GIVE THEIR OPINION (e.g., _________, do you agree with the decision to place that image in the _________ pile? _________ what do you think about that image? _____ ____, what stands out to you in this image?) Great! Does everyone agree with the piles the way they are? Now, can you tell me what aspects of the images were most important when you were sorting? What were some of the things that made you sure an image was DIMENSION C ? How about DIMENSION D ? Looking back though the DIMENSION C pile, which image do you think is most DIMENSION C ? How about DIMENSION D ?

PAGE 439

439 Can you think of any particular things that were missing from the images? Or something that you would have liked to see? If you could describe the ideal DIMENSION C image, what would it be? How about DIMENSION D ? Thanks, everyone! That gives me a much better understanding of DIMENSION C and DIMENSION D . MBTI DIMENSION #3 with the same set of photos. This time, I have another set of personality dimensions for you to use. DIMENSION E & DIMENSION F . and give you a copy of the are also printed on the tabletop signs. MODERATOR READS DIMENSION DESCRIPTIONS LOCATED ON THE LAST PAGE Just like the first tw o time, we will have one pile for DIMENSION E and another pile for DIMENSION F And again, if you think an image may fit in both personality piles, please discuss why with your Does anyone have any questions? OK, here are the images. You may start sorting. THE MODERATOR HANDS A THIRD IMAGE DECK TO THE GROUP THEN LISTENS, WATCHES AND TAKES NOTES WHILE THE GROUP SORTS. THE MODERATOR MAY ALSO PROBE FOR MORE DETAIL ON WHY THE GROUP DECIDED TO PLACE AN IMAGE I N A CERTAIN PILE IF THE DECISION IS NOT DISCUSSED IN OUT LOUD IN DETAIL (e.g., What about that particular image made you decide to put it in the _________ pile?). THE MODERATOR WILL ALSO MONITOR THE GROUP DYNAMIC TO ENSURE THAT ALL MEMBERS ARE ABLE TO PAR TICIPATE AND GIVE THEIR OPINION (e.g., _________, do

PAGE 440

440 you agree with the decision to place that image in the _________ pile? _________ what do you think about that image? _________, what stands out to you in this image?) Great! Does everyone agree with the piles the way they are? Now, can you tell me what aspects of the images were most important when you were sorting? What were some of the things that made you sure an image was DIMENSION E ? How about DIMENSION F ? Looking back though the DIMENSION E pil e, which image do you think is most DIMENSION E ? How about DIMENSION F ? Can you think of any particular things that were missing from the images? Or something that you would have liked to see? If you could describe the ideal DIMENSION E image, what would it be? How about DIMENSION F ? Thanks, for all your help! I think I see the differences between DIMENSION E and DIMENSION F . THANK YOU & ENDING QUESTION anyo ne have anything else they would like to add? That concludes our group. Thanks again for all your help!

PAGE 441

441 APPENDIX D DIMENSION DESCRIPTIONS FOR GROUP IMAGE SORTS Extroversion/Introversion (E/I) where a person gets and directs his/her energy. Extraverted (E): Extraverts relate easily to the outer world of people and things. They focus on the world around them and gain energy from other people. Extraverts often know a little about a lot, preferrin g to explore a variety of subjects and ideas. Extraverts often need to think out their ideas aloud. Adjectives that describe extroverts include: outspoken, active and sociable. Introverted (I): Introverts relate easily to the inner world of ideas. They dr aw energy from their inner world and prefer to spend time alone. Introverts tend to have fewer but deeper interests, immersing themselves in the topics that appeal to them. Introverts prefer to quietly think through an idea in their head before speaking it . Adjectives that describe introverts in clude: quiet, reserved and shy. -------------------Sensing/Intuition (S/N) The sensing/intuition (S/N) dimension analyzes how an individual takes in information. Sensing (S): Sensors prefer to work with known fact s. They concentrate on concrete information from their five senses. Sensors appreciate simple, matter of fact reality. Sensors accurately remember details and excel at here and now practical matters. Adjectives that describe sensors include: systematic, pr actical/functional and conservative. Intuition (N): Intuitives prefer to look for possibilities and relationships. They interpret and add meaning to information, looking at big picture patterns. Intuitives value big ideas and see obscure connections other s may miss. Intuitives tend to be creative and good at imagining long term possibilities for the future. Adjectives that describe intuitives include: imaginative, creative/ theoretical and unconventional. -------------------Thinking/Feeling (T/F) The thin king/feeling (T/F) dimension represents how individuals make decisions. Thinking (T): Thinkers base their judgments on impersonal analysis and logic. They put more weight on objective principles, impersonal facts and logic. Thinking is closely tied to in tellect. Thinkers are more driven to be fair and consistent. Thinkers tend to take things less personally, are harder to offend and are uncomfortable with emotions. Adjectives that describe thinkers include: firm, rational/reasonable and logical. Feeling (F): Feelers base their judgments on personal values. They place more emphasis on people and personal concerns. Feeling places a high premium on the development and maintenance of personal relationships. Feelers are guided by their personal values and exte nuating circumstances. Feelers often wear their heart on their sleeves, are comfortable with emotions and may get their feelings hurt more easily. Adjectives that describe feelers include: soft hearted, passionate/perceptive and emotional.

PAGE 442

442 APPENDIX E SURV EY ADVERTISEMENTS 1. EST ad 2. ESF ad 3. ENT ad 4. ENF participants ad 5. ENF researcher ad 6. IST ad

PAGE 443

443 7. ISF ad 8. INT ad 9. INF participants ad 10. INF researcher ad

PAGE 444

444 APPENDIX F HAND SOAP LOGO AND BOTTLE

PAGE 445

445 APPENDIX G Q UESTIONNAIRE

PAGE 446

446

PAGE 447

447

PAGE 448

448

PAGE 449

449

PAGE 450

450

PAGE 451

451

PAGE 452

452

PAGE 453

453

PAGE 454

454

PAGE 455

455

PAGE 456

456

PAGE 457

457

PAGE 458

458

PAGE 459

459

PAGE 460

460

PAGE 461

461

PAGE 462

462

PAGE 463

463 APPENDIX H NUMBERED IMAGE DECK IMAGES Image 1 Image 2 Image 3 Image 4 Image 5 Image 6

PAGE 464

464 Image 7 Image 8 Image 9 Image 10 Image 11 Image 12

PAGE 465

465 Image 13 Image 14 I mage 15 Image 16 Image 17 Image 18

PAGE 466

466 Image 19 Image 20 Image 21 Image 22 Image 23 Image 24

PAGE 467

467 LIST OF REFERENCES Aaker, D. (1996). Building strong brands . New York, NY: Free Press. Aaker, D. (1991). Managing brand equity: Capitalizing on the value of a brand name . New York, NY: Free Press. Aaker, J. (1997). Dimensions of brand personality. Journal of Marketing Research, 34 (3), 347 356. Aaker, J. (1999). The malleable self: The role of self expression in persuasion. Journal of Marketing Research, 36 (1), 45 57. Aaker, J., Benet Martinez, V., & Garolera, J. (2001). Consumption symbols as carriers of culture: A study of Japanese and Spanish brand personality constructs. Journal of Personality and Social Psychology , 81 (3), 492 508. Abreu, G. D., & Cline, T. (2003). Schooled mathematics and cultural knowledge. Pedagogy, Culture and Society , 11 (1), 11 30. Akay, E. (2001). The effects of executional elements of advertisements on the perception of brand personality. (Unpublished masters thesis). U niversity of Florida, Gainesville, FL. Ang, S.H., & Lim, E.A.C. (2006). The influence of metaphors and product type on brand personality perceptions and attitudes. Journal of Advertising , 35 (2), 39 53. Arnheim, R. (1969). Visual thinking . Berkeley, CA: Uni versity of California Press. Arora, R., & Stoner, C. (2009). A mixed method approach to understanding brand personality. Journal of Product & Brand Management , 18 (4), 272 283. Austin, J.R., Siguaw, J.A., & Mattila, A.S. (2003). A re examination of the gene ralizability of the Aaker brand personality measurement framework. Journal of Strategic Marketing , 11 , 77 92. Avis, M. (2012). Brand personality factor based models: A critical review. Australasian Marketing Journal , 20 , 89 96. Azoulay, A., & Kapferer, J. (2003). Do brand personality scales really measure brand personality? Journal of Brand Management , 11 (2), 143 155. Azzadina, I., Huda, A.N. & Sianipar, C.P.M. (2012). Understanding relationship between personality types, marketing mix factors, and purchasi ng Decisions. Procedia Social and Behavioral Sciences , 65 , 352 357. Babbie, E. (2010). The practice of social research (12th ed.). Belmont, CA: Wadsworth. Bakalash, T., & Riemer, H. (2013). Exploring ad elicited emotional arousal and memory for the ad u sing fMRI. Journal of Advertising , 42 (4), 275 291.

PAGE 468

468 Balnaves, M., & Caputi, P. (2001). Introduction to quantitative research methods: An investigative approach . Thousand Oaks, CA: Sage Publications. Baran, S.J., & Davis, D.K. (2009). Mass communication theo ry: Foundations, ferment and future (5 th ed.). Boston: Wadsworth Cengage Learning. Bartlett, J.E., Kotrlik, J.W., & Higgins, C.C. (2001). Organizational research: Determining appropriate sample size in survey research. Information Technology, Learning and Performance Journal , 19 (1), 43 50. Batra, R., Lehmann, D.R., & Sinh, D. (1993). The brand personality component of brand goodwill: Some antecedents and consequences. In D.A. Aaker & A.L. Biel (Eds.), Brand ng strong brands (83 96). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Berger, A.A. (1998). Media research techniques (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Berger, A.A. (2011). rican character and society (4th ed.). Lanham, MD: Rowman & Littlefield Publishers, Inc. Bergkvist, L., Eiderbäck, D., & Palombo, M. (2012). The brand communication effects of using a headline to prompt the key benefit in ads with pictorial metaphors. Jour nal of Advertising , 41 (2), 67 76. Bergkvist, L., & Rossiter, J.R. (2007). The predictive validity of multiple item versus single item measures of the same constructs. Journal of Marketing Research , 44 , 175 184. Bergkvist, L., & Rossiter, J.R. (2008). The r performance . Journal of Advertising , 37 , 85 97. Bergkvist, L., & Rossiter, J.R. (2009a). Tailor made single item measures of doubly concrete constructs. International Journal of Advertising , 28 (4), 607 62 1. Bergkvist, L., & Rossiter, J.R. (2009b). The importance of choosing one good item for single item measures of attitude towards the ad and attitude towards the brand and its generalization to all measures. Transfer Werbeforschung & Praxis , 55(2), 8 18. B ernard, H.R., & Ryan, G.W. (2010). Analyzing qualitative data: Systematic approaches . Los Angeles, CA: Sage Publications, Inc. Biel, A. L. (1993). Converting image into equity. In D. A. Aaker & A. L. Biel (Eds.), Brand equity and advertising: Advertising's role in building strong brands (pp. 67 82). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Boudreaux, C.A., & Palmer, S.E. (2007). A charming little Cabernet: Effects of wine label design on purchase intent and brand personality. International Journal o f Wine Business Research , 19 (3), 170 186.

PAGE 469

469 Brewer, J. (2010, October 25). Consumers want to run your campaign. Will you let them? Retrieved November 19, 2013 from h ttp://brolik.com/blog/consumers want to run your campaign will you let them/ Briggs, S.R., & Cheek, J.M. (1986). The role of factor analysis in the development and evaluation of personality scales. Journal of Personality , 54 , 106 148. Brown, J.D. (2001). Point biserial correlation coefficients. Shiken: JALT Testing & Evaluation SIG Newsletter , 5 (3), 12 16. Retrieved October, 19 2013 from http://jalt.org/test/PDF/Brown12.pdf Bruning, J.L., & Kintz, B.L. ( 1968). Computational handbook of statistics . Glenview, IL: Scott, Foresman and Company. Byrne, A., Whitehead, M., & Breen, S. (2003). The naked truth of celebrity endorsement. British Food Journal , 105 (4/5), 288 296. Callister, M.A., & Stern, L.A. (2007). The role of visual hyperbole in advertising effectiveness. Journal of Current Issues and Research in Advertising , 29 (2), 1 14. Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi Experimental Designs for Research. Chicago, IL: Rand McNally Col lege Publishing Company. Campbell, M.C., & Warren, C. (2012). A risk of meaning transfer: Are negative associations more likely to transfer than positive associations? Social Influence , 7 (3), 172 192. Caplan, R.D., Naidu, R.K., & Tripathi, R.C. (1984). Cop ing and defense : Constellations vs components. Journal of Health and Social Behavior , 25 , 303 320. Carlyn, M. (1977). An assessment of the Myers Briggs Type Indicator. Journal of Personality Assessment , 41 (5), 461 473. Chandler, D. (2013). Semiotics for B eginners . Retrieved May 16, 2014 from http://users.aber.ac.uk/dgc/Documents/S4B/sem09.html Chandler, D. (2007). Semiotics: The basics (2 nd ed.). New York: Routledge. Childers, T.L., & Hou ston, M.J. (1984). Conditions for a picture superiority effect on customer memory. Journal of Consumer Research , 11 (2), 643 654. Cochran, W. G. (1977). Sampling techniques (3rd ed.). New York, NY: John Wiley & Sons. Cooper, T.B. (2009). Appropriating visua representative form. Visual Communication Quarterly , 16 , 90 107. Costa, P. T., & McCrae, R. R. (1992). Four ways five factors are basic. Personality and Individual Differences, 13 (6), 653 665.

PAGE 470

470 Creswell, J.W. (2007) . Qualitative inquiry & research design: Choosing among five approaches (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Creswell, J.W. (2012). Educational research: Planning, conducting and evaluating quantitative and qualitative research (4th ed.). Boston, MA: Pearson. Cunningham, D., Thach, L., & Thompson, K. (2007). Innovative e commerce site design: A conceptual model to match consumer MBTI dimensions to website design. Journal of Internet Commerce , 6 (3), 1 27. Danesi, M. (2008). Why it sells: Dec oding the meanings of brand names, logos, ads, and other marketing and advertising ploys. Lanham, MD: Rowman & Littlefield Publishers, Inc. Davis, S. (2000). The power of the brand. Strategy & leadership, 28 (4), 4 9. Delbaere, M., McQuarrie, E.F., & Philli ps, B.J. (2011). Personification in advertising: Using a visual metaphor to trigger anthropomorphism. Journal of Advertising , 40 (1), 121 130. Denzin, N.K. (1978). The research act (2nd ed.). New York: McGraw Hill. Depositphotos ® . (2014). Retrieved from www.depositphotos.com DeVellis, R.F. (2003). Scale development: Theory and applications (2nd ed.). Thousand Oaks, CA: Sage. Doyle, P. (1989). Building successful brands: The strategic options. Journal of Marketing Management , 5 (1), 77 9 5. Dreamstime ® . (2014). Re trieved from www.dreamstime.com Edinger, E.R. (n.d.). C.G. Jung: An outline of analytical psychology. Retrieved October, 2013 from http://www.capt.org/using type/c g jung.htm August, 2013 from http ://www.businessinsider.com/the 10 best ads of 2012 2012 11?op=1 Eisinga, R., Grotenhuis, M., & Pelzer, B. (2013). The reliability of a two item scale: Pearson, Chronbach or Spearman Brown? International Journal of Public Health , 58 , 637 642. Elliot, S. (2 006, October 9). Letting consumers control marketing: Priceless. The New York Times . Retrieved November 19, 2013 from www.nytimes.com/2006/10/09/business/media/09adcol.html?_ r=0 eMarketer. (2012, September). Advertising spending in the United States from 2010 to 2016 (in billion U.S. dollars). Retrieved October 8, 2013 from ht tp://www.statista.com/statistics/182389/forecast of advertising spending in the us/ Erdogan, B.Z. (1999). Celebrity endorsement: A literature review. Journal of Marketing Management , 15 , 291 314.

PAGE 471

471 Escalas, J.E., & Bettman, J.R. (2005). Self construal, refe rence groups, and brand meaning. Journal of Consumer Research , 32 (3), 378 389. Estimated frequencies of the types in the United States population. (n.d.). Retrieved October, 2013 from the Center for Applications of Psychological Type at http://www.capt.org/mbti assessment/estimated frequencies.htm Experian Information Solution. (2013). Usage of liquid hand soap in the U.S. 2013. Retrieved October 8, 2013 from Statista database, http://www.statista.com/statistics/276368/us households usage of liquid hand soap/ Felton, G. (2006). Advertising concepts and copy (2nd ed.). New York, NY: W. W. Norton & Company. Fennis, B.M., Das, E., & Fransen, M.L. (2012). Print advertising: Vivid content. Journal of Business Research , 65 , 861 864. Freling, T., & Forbes, L. (2005). An empirical analysis of the brand personality effect. Journal of Product & B rand Management, 14 (7), 404 413. Fetscherin, M., & Toncar, M. (2009). Visual puffery in advertising. International Journal of Market Research , 51 (2), 147 148. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to th eory and research reading . Massachusetts: Addison Wesley. Gardner, M.P. (1985). Does attitude toward the ad affect brand attitude under a brand evaluation set? Journal of Marketing Research , 22 , 192 98. Gaskell, G. (2000). Individual and group interviewing . In M.W. Bauer and G. Gaskell (Eds.), Qualitative researching with text, image and sound: A practical handbook (pp. 38 56). Thousands Oak, CA: Sage Publications, Inc. Geuens, M., Weijters, B., De Wulf, K. (2009). A new measure of brand personality. Intern ational Journal of Research in Marketing , 26 , 97 107. Gibbs, A. (1997). Focus groups. Social Research Update , 19 . Retrieved October 12, 2013 from http://sru.soc.surrey.ac.uk/SRU19.html Gkiouzepas, L., & Hogg, M.K. (2011). Articulating a new framework for visual metaphors in advertising: A structural, conceptual, and pragmatic investigation. Journal of Advertising , 40 (1), 103 120. Goffman, E. (1979). Gender advertisements . New York, NY: Harper Colophon B ooks. Gould, S.J. (1991). Jungian analysis and psychological types: An interpretive approach to consumer choice behavior. Advances in Consumer Research , 18 , 743 748.

PAGE 472

472 Govers, P., & Schoormans, J. (2005). Product personality and its influence on consumer pre ference. Journal of Consumer Marketing , 22 (4), 189 197. Greene, J. (2007). Mixed methods in social inquiry. San Francisco, CA: Jossey Bass. Grumbein, A. (2013). Brand Who?: Teaching students the value of brand personality. Journal of Advertising Education , 17 (2), 43 50. Gwinner, K.P., & Eaton, J. (1999). Building brand image through event sponsorship: The role of image transfer. Journal of Advertising , 28 (4), 47 57. Hall, S. (1980). Encoding/decoding. In S. Hall, D. Hobson, A. Love, & P. Willis (Eds.), Cult ure, media, language (pp. 128 138). London: Hutchinson. Halonen Knight, E., & Hurmerinta, L. (2010). Who endorses whom? Meanings transfer in celebrity endorsement. Journal of Product & Brand Management , 19 (6), 452 460. Hampp, A. (2010, October 11). Gap to scrap new logo, return to old design. Advertising Age. Retrieved November 19, 2013 from adage.com/article/news/gap scrap logo return design/146417/ Harman, R.C. (2001). Activities of contemporary Mayan elder. Journal of Cross Cultural Gerontoloy , 16 , 57 77 . Hamilton, M., & Sun, X. (2005). Actual self and ideal brand image: An application of self congruity to brand image positioning . Paper presented at International Communication Association, New York, NY. Hartmann, P., & Apaolaza Ibáñez, V. (2012). Consume r attitude and purchase intention toward green energy brands: The roles of psychological benefits and environmental concern. Journal of Business Research , 65 , 1254 1263. Henderson, P.W., & Cote, J.A. (1998). Guidelines for selecting or modifying logos. Jou rnal of Marketing , 62 ,14 30. Hirschman, E.C. (2007). Metaphor in the marketplace. Marketing Theory , I7 (3), 227 248. Hirsh, S.K., & Kummerow, J. (1989). Life types: Understand yourself and make the most of who you are . New York, NY: Warner Books. Huang, H., Mitchell, V.W., & Rosenaum Elliot, R. (2012). Are consumer and brand personalities the same? Psychology & Marketing , 29 (5), 334 349. Huang, J.H., & Yang, T.K. (2011). The impacts of homepage screen density on website evaluations: The moderating role of pe rsonality type. Social Behavior and Personality , 39 (3), 381 390. iStock ® . (2014). Retrieved from www.istockphoto.com

PAGE 473

473 Janiszewski, C. (2008). Rethinking visual communication research: Updating old constructs and co nsidering new metaphors. In Wedel, M., & Pieters, R. (Eds.), Visual marketing: From attention to action (pp. 277 293). New York, NY: Psychology Press. Jeong, S.E. (2008). Visual metaphor in advertising: Is the persuasive effect attributable to visual argum entation or metaphorical rhetoric? Journal of Marketing Communications , 14 (1), 59 73. Jewett, A. & Sutherland, J. (2013). Picture perfect personality: Examining the influence of advertising images on brand personality. Association for Education in Journali sm and Mass Communication (AEJMC) Annual Conference , Washington DC, August 8 11. Jewitt, C., & Oyama, R. (2004). Visual meaning: A social semiotic approach. In Van Leeuwen, T. & Jewitt, C. (Eds.), Handbook of visual analysis (pp. 134 156). Thousand Oaks, C A: Sage Publications Ltd. Johar, G.V., Sengupta, J., & Aaker, J.L. (2005). Two roads to updating brand personality impressions: Trait versus evaluative inferencing. Journal of Marketing , 42 , 458 469. Johnson, B., & Christensen, L. (2004). Educational resea rch: Quantitative, qualitative, and mixed approaches (2nd ed.). Boston, MA: Pearson Education, Inc. Johnson, C.A. (2009, February 11). Cutting through advertising clutter. CBS News. Retrieved November 10, 2012 from http://www.cbsnews.com/8301 3445_162 2015684.html Johnson, R.B., & Onwuegbuzie, A.J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher , 33 (7), 14 26. Jung, C.G. (1971). Psychological T ypes. In J. Campbell (Eds.), The portable Jung (pp.178 269). New York, NY: Penguin Books. (Original work published in 1923) gender analysis revisited. Sex Roles, 37, 9 79 997. Kapoor, H., & Heslop, L.A. (2009). Brand positivity and competitive effects on the evaluation of brand extensions. International Journal of Research in Marketing , 26 , 228 237. Kroeber, A.L., & Kluckhohn, C. (1952). Culture: A critical review of con cepts and definitions . New York, NY: Vintage Books. Krueger, R.A. (1988). Focus groups: A practical guide for applied research . Newbury Park, CA: Sage Publications. Krueger, R.A. (1998). Developing questions for focus groups. Thousand Oaks. CA: Sage Public ations.

PAGE 474

474 LaBarbera, P.A., Weingard, P. & Yorkston, E.A. (1998). Matching the message to the mind: Advertising imagery and consumer processing styles. Journal of Advertising Research , 38 , 29 43. Labrecque, L.I., Patrick, V.M., & Milne, G.R. (2013). The mark review of color research and future direction. Psychology and Marketing , 30 (2), 187 202. Lehmann, D., & Shemwell, D.J. (2011). A field test of the effectiveness of different print layouts: A mixed model field experiment in alter native advertising. Journal of Promotion Management , 17 , 61 75. Lee, E.J. (2013). A prototype of multicomponent brand personality structure: A consumption synmbolism approach. Psychology and Marketing , 30 (2), 173 186. Lindner, K. (2004). Images of women in general interest and fashion magazine advertisements from 1955 to 2002. Sex Roles , 51 (7/8), 409 421. Lowry, R. (2013). Point biserial correlation coefficient. Retrieved October 19, 2013 from http://www .vassarstats.net/pbcorr.html Lutz, R.J., MacKenzie, S.B., & Belch, G.E. (1983). Attitude toward the ad as a mediator of advertising effectiveness: Determinants and consequences. In R.P. Bagozzi, & A.M. Tybout (Eds.), Advances in Consumer Research (Vol. 10 , pp. 532 539). Ann Arbor, MI: Association for Consumer Research. Ma, L. (2008). Pictorial metaphor in advertising and consumer interpretation of its cultural meaning. China Media Research , 4 (3), 9 17. e: The increasing global importance of image banks in corporate media. Visual Communication , 3 (3), 316 336 . MacKenzie, S.B., & Lutz, R.J. (1989). An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. Journal of Marketing , 53 , 48 65. Maehle, N., & Supphellen, M. (2011). In search of the sources of brand personality. International Journal of Market Research , 53 (1), 95 114. Malar, L., Krohmer, H, Hoyer, W.D., & Nyffenegger, B. (2011). Emotional b rand attachment and brand personality: The relative importance of the actual and the ideal self. Journal of Marketing , 75 , 35 52. Marchland, R. (1985). Advertising the American dream: Making way for modernity 1920 1940. Berkeley, CA: University of Californ ia Press. Marketing Charts. (2013). College student spending power flat this year. Retrieved October 8, 2013 from http://www.marketingch arts.com/wp/topics/automotive/college student spending power flat this year 30388/

PAGE 475

475 Martin, C. (n.d.). The sixteen types at a glance. Retrieved October, 2013 from http://www.capt.org/ mbti assessment/type descriptions.htm Martin, L., & Gourley Instructional Science , 1 20. Maxwell, J.A. (2013). Qualitative research design: An interactive approach (3rd ed.). Los Angeles, CA: Sage Publications, Inc. MBTI ® Instrument Comparison with Other Psychological Tests. (n.d.). Retrieved October, 2013 from the Center for Applications of Psychological Type at http: //www.capt.org/mbti assessment/mbti test comparison.htm McCarley, N.G., & Carskadon, T.G. (1986). The perceived accuracy of elements of the 16 type descriptions of Myers and Keirsey among men and women: Which elements are most accurate, should type descri ptions be different for men and women, and do the type descriptions stereotype sensing types? Journal of Psychological Type , 11 , 2 29. McCracken, G. (1986). Culture and consumption: a theoretical account of the structure and movement of the cultural meanin g of consumer goods. Journal of Consumer Research , 13 (1), 71. McCracken, G. (1988). Culture and consumption . Bloomington, IN: Indiana University Press. McCracken, G. (1989). Who is the celebrity endorser? Cultural foundations of the endorsement process. J ournal of Consumer research, 16 (3), 310 321. McCracken, G. (1990). Cultural and consumer behaviour: An anthropological perspective. Journal of the Market Research Society , 32 (1), 3 11. McCracken, G. (1993). The value of the brand: An anthropological persp ective. In D.A. Aaker & A.L. Biel (Eds.), Brand equity and advertising: Advertising's role in building strong brands (pp. 125 139). Hillsdale, NJ: Lawrence Erlbaum Associates. McCracken, G. (2005). Culture and consumption II: Markets, meaning, and brand ma nagement . Bloomington, IN: Indiana University Press. McCrae, R.R. (2002). Cross cultural research on the five Factor model of personality. Online Readings in Psychology and Culture , 4 (4). http://dx.doi.org/10.9707/2307 0919.1038 McQuarrie, E.F. (2008). Dif ferentiating the pictorial element in advertising. In Wedel, M., & Pieters, R. (Eds.), Visual marketing: From attention to action (pp. 91 112). New York, NY: Psychology Press. McQuarrie, E.F., & Mick, D.G. (1992). On resonance: A critical pluralistic inqui ry into advertising rhetoric. Journal of Consumer Research, 19 , 180 197.

PAGE 476

476 McQuarrie, E. F., & Mick, D. G. (1996). Figures of rhetoric in advertising language. Journal of Consumer Research, 22 , 424 461. McQuarrie, E. F., & Mick, D. G. (1999). Visual rhetoric in advertising: Text interpretive, experimental and reader response analyses. Journal of Consumer Research 26 , 37 53. McQuarrie, E.F., & Phillips, B.J. (2005). Indirect persuasion in advertising: How consumers process metaphors presented in pictures and w ords. Journal of Advertising , 34 (2), 7 20. Meenaghan, T. (1995). The role of advertising in brand image development. Journal of Product & Brand management , 4 (4), 23 34. Merskin, D. (2001). Winnebagos, Cherokees, Apaches, and Dakotas: The persistence of ste reotyping of American Indians in American advertising brands. The Howard Journal of Communications , 12 , 159 169. Messaris, P. (1997). Visual persuasion: The role of images in advertising . Thousand Oaks, CA: Sage Publications. Messaris, P. (1998). Visual as pects of media literacy. Journal of Communication , 48 (1), 70 80. Mitchell, A.A., & Olson, J.C. (1981). Are product attribute beliefs the only mediator of advertising effects on brand attitude? Journal of Marketing Research , 18 (3), 318 332. Morgan, D.L. (19 96). Focus groups. Annual Review of Sociology , 22 , 129 152. Morgan, D.L. (1997). Focus groups as qualitative research (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Morgan, S.J. (2005). More than pictures? An exploration of visually dominant magazin e ads as arguments. Journal of Visual Literacy , 25 (2), 145 166. Morgan, S.E., & Reichert, T. (1999). The message is in the metaphor: assessing the comprehension of metaphors in advertisements. Journal of Advertising , 28 (4), 1 12, Moriarty, S., & Rohe, L. ( 2005). Cultural palettes in print advertising: Formative research design method. In K.L. Smith, S. Moriart, G. Barbatsis, & K.Kenny (Eds.), Handbook of visual communication: Theory, Methods, and media (pp. 117 124). Hillsdale, NJ: Lawrence Erlbaum Associat es. Morrison, M.A., Haley, E., Sheehan, K.B., & Taylor, R.E. (2002). Using qualitative research in advertising: Strategies, techniques, and applications . Thousand Oaks, CA: Sage Publications. Multicultural use of the MBTI instrument (n.d.). Retrieved Octob er, 2013 from http://www.myersbriggs.org/more about personality type/international use/multicultural use of the mbti.asp

PAGE 477

477 Myers, I.B. (1976a). Characteristic frequently associated with each type. Palo Alto, CA: Consulting Psychologists Press, Inc. Myers, I.B. (1976b). Report form for Myers Briggs Type indicator. Palo Alto, CA: Consulting Psychologists Press, Inc. Myers, I.B. (1995). Gift s differing: Understanding personality type . Palo Alto, CA: Davies Black Publishing. NAA (2009). How America uses advertising. Retrieved November 10, 2012 from http://www.naa.org/Topics and Tools/Advertising/Research/~/media/NAACorp/Public%20Files/TopicsAndTools/Advert ising/Research/HowAmericaUsesAdvertising.ashx National Student Clearingh ouse (2013). College enrollments decrease for second straight year; State by state figures included. Retrieved January 6, 2014 from http://www.studentclearinghouse.org/about/media_center/press_releases/files/releas e_20 13 12 12c.pdf Nielsen, J., & Landauer T.K. (1993). A mathematical model of the finding of usability problems. Proceedings of INTERCHI , 206 213. Nielsen, J. (2000, March 19). Why you only need to test with 5 users. Nielsen Norman Group . Retrieved Octob er 12, 2013 from http://www.nngroup.com/articles/why you only need to test with 5 users/ Nielsen, J. (2004, July 19). Card sorting: How many users to test. Nielsen Norm an Group . Retrieved October 15, 2013 from http://www.nngroup.com/articles/card sorting how many users to test/ Neumeier, M. (2006). The brand gap: How to bridge the distan ce between business strategy and design . Berkeley, CA: New Riders. Nienstedt, H.W., Huber, F., & Seelmann, C. (2012). The influence of the congruence between brand and consumer personality on the loyalty to print and online issues of magazine brands. The I nternational Journal on Media Management , 14 , 3 26. entertaining, intriguing, powerful and beautiful spots. Retrieved August, 2013 from http://www.adweek.com/news/advertising branding/10 best commercials 2012 145324?page=1 Nunnally, J. C. (1967). Psychometric theory. New York, NY: McGraw Hill. Nunnally, J.C. (1978). Psychomet ric theory (2nd ed.). New York, NY: McGraw Hill. Ouwersloot, H., & Tudorica, A. (2001). Brand personality creation through advertising. MAXX Working Paper Series , 2001 01, 1 25.

PAGE 478

478 Pallant, J. (2010). SPSS survival manual: A step by step guide to data analysi s using SPSS (4th ed.). New York, NY: Open University Press. Park, C., Jaworski, B., & Maclnnis, D. (1986). Strategic brand concept image management. The Journal of Marketing , 135 145. Park, S., Choi, D., & Kim, J. (2005). Visualizing e brand personality: Exploratory studies on visual attributes and e brand personalities in Korea. International Journal of Human Computer Interaction , 19 (1), 7 34. Journalism Monographs , 123 , 1 29. Pedhazur, E.J., & Schmelkin, L.P. (1991). Measurement, design, and analysis: An integrated approach . Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Phillips, B.J., & McQuarrie, E.F. (2004). Beyond visual metaphor: A new typology of visual rh etoric in advertising. Marketing Theory . 4 (1 2), 111 134. Plummer, J. (2000). How personality makes a difference. Journal of Advertising Research, 40 (6), 79 84. (Original work published 1984) Pomering, A., Johnson, L.W., & Noble, G. (2013). Advertising co rporate social responsibility: Results from an experimental manipulation of key message variables. Corporate Communications: An International Journal , 18 (2), 249 263. Porter, B.J., & Roll, S. (1992). Personality and perception: Rorschache and Luescher corr elates of Jungian types as measured by the Myers Briggs Type Indicator. In C.D. Spielberger & J.N. Butcher (Eds.), Advances in personality assessment volume 9 (pp. 117 124). New York, NY: Routledge. Portland State University (2011). Personal care product s urvey: Final results report . Retrieved October 8, 2013 from Statista Database. Postrel, V. (2003) The substance of style: How the rise of aesthetic value is remaking commerce, culture and consciousness . New York, NY: Harper Perennial. Pracejus, J.W. (2003) . Overview. In J.W. Pracejus (Chair), Visuals are information: How meaning is transferred to consumers through executional elements in advertising (Special Session Summary) . Valdosta, GA: Association for Consumer Research. Retrieved December 9, 2013 from www.acrwebsite.org/search/view conference proceedings.aspx?Id=8761 recept ion of advertising meaning through the use of white space. In J.W. Pracejus (Chair), Visuals are information: How meaning is transferred to consumers through executional elements in advertising (Special Session Summary) . Valdosta, GA:

PAGE 479

479 Association for Consu mer Research. Retrieved December 9, 2013 from www.acrwebsite.org/search/view conference proceedings.aspx?Id=8761 Puzakova, M., Kwak, H., & Taylor, C.R. (2013). The ro brand personality traits: Consumer inference of unobservable attributes. Journal of Advertising , 42 (1), 16 29. Qualtrics ® . (2014). Version 56686. Retrieved from www.qualtrics.c om Quintiliani, L.M., Campbell, M.K., Haines, P.S., & Webber, K.H. (2008). The use of the pile sot method in identifying groups of healthful lifestyle behaviors among female community college students. Journal of the American Dietetic Association , 108 (9), 1503 1507. Ramaseshan, B., & Tsao, H. (2007). Moderating effects of the brand concept on the relationship between brand personality and perceived quality. The Journal of Brand Management, 14 (6), 458 466. Rayner, K., & Castelhano, M.S. (2008). Eye movement s during reading, scene perception, visual search, and while looking at print advertisements. In Wedel, M., & Pieters, R. (Eds.), Visual marketing: From attention to action (pp. 9 42). New York, NY: Psychology Press. Rayner, K., Miller, B., & Rotello, C.M. (2008). Eye movements when looking at print advertisements: The goal of the viewer matters. Applied cognitive psychology , 22 , 697 707. Rayner, K., Rotello, C.M., Stewart, A.J., Keir, J., & Duffy, S.A. (2001). Integrating text and pictorial information: E ye movements when looking at print advertisements. Journal of Experimental Psychology: Applied , 7 (3), 219 226. Reid, L.N., & Buchanan, L. (1979). A shopping list experiment of the impact of advertising on brand images. Journal of Advertising , 8 (2), 26 28. Richek. H.G. (1969) Note on intercorrelations of scales of the Myers Briggs Type Indicator. Psychological Reports, 25 , 28 30. Ritson, M., & Elliot, R. (1999). The social uses of advertising: An ethnographic study of adolescent advertising audiences. Journ al of Consumer Research , 26 , 260 277. Rogers, E.M. (1994). A history of communication study . New York, NY: The Free Press. Romaniuk, J. (2008). Comparing methods of measuring brand personality traits. Journal of Marketing Theory and Practice , 16 (2), 153 16 1. Rossiter, J.R. (2002). The C OAR SE procedure for scale development in marketing. International Journal of Research in Marketing , 19 , 305 335. Rubin, R.B., Rubin, A.M., & Piele, L.J. (1990). Communication research: Strategies and sources (2nd ed.). Belm ont, CA: Wadsworth.

PAGE 480

480 Ryan, G., and H. R. Bernard. (2004). Techniques to identify themes in qualitative data . Retrieved October 12, 2013 from http://w ww.analytictech.com/mb870/Readings/ryan bernard_techniques_to_identify_themes_in.htm Sabri, O., & Obermiller, C. (2012). Consumer perception of taboo in ads. Journal of Business Research , 65 , 869 873. Schmidt, K., & Ludlow, C. (2002). Inclusive Branding . New York, NY: Palgrave Macmillan. Schmidt, L.D., & Fretz, B.R. (1968). Teacher student similarity in an educational skills course. The Journal of Experimental Education , 36 (4), 78 83. Schuldt, J.P., Konrath, S.H., & Schwarz, N. (2012). The right angle: Vis ual portrayal of products Psychology and Marketing , 29 (10), 705 711. Shadish, W.R., Cook, T.D., & Campbell, D.T. (1963). Experimental and quasi experimental designs for generalized causal inference. Boston, MA: Hou ghton Mifflin Company. Shank, M.D., & Langmeyer, L. (1994). Does personality influence brand image? Journal of Psychology , 128 (2), 157 167. Shutterstock ® . (2014). Retrieved from www.shutterstock.com Siguaw, J., M attila, A., & Austin, J. (1999). The brand personality scale: An application for restaurants. Cornell Hotel and Restaurant Administration Quarterly , 40 (3), 48. Simon, J.L., (1969). Basic research methods in social science: The art of empirical investigatio n. New York, NY: Random House. Sirgy, M. (1982). Self concept in consumer behavior: A critical review. Journal of Consumer Research , 9 (3), 287. Sorrell, M. (2000). Branding the new era. Foreign Policy , 119 , 61 63. Stanton, J.M., Sinar, E.F., Balzer, W.K. & Smith, P.C. (2002). Issues and strategies for reducing the length of self report scales. Personnel Psychology , 55 (1), 167 194. Strausbaugh, K.L. (1998). assessing brand personality and buildi ng brand personality profiles . Retrieved from ProQuest Dissertations and Theses. Stricker. L.L., & Ross, J. (1963). Intercorrelations and reliability of the Myers Briggs Type Indicator scales. Psychological Reports, 12 , 287 293. Stricker, L.J., Schiffman, H., & Ross, J. (1965). Predictions of college performance with the Myers Briggs Type Indicator. Educational and Psychological Measurement, 25 , 1081 1095.

PAGE 481

481 Suh, J. (2009). The role of consideration sets in brand choice: The moderating role of product charact eristics. Psychology & Marketing , 26 (6), 534 550. Sukamolson, S. (2005). Fundamentals of quantitative research. E Journal for Researching Teachers , 3 (1). Retrieved February 27, 2014 from http://www.culi.chula.ac.th/e journal/bod/suphat%20sukamolson.pdf Sung, Y., & Kim, J. (2010). Effects of brand personality on brand trust and brand affect. Psychology and Marketing , 27 (7), 639 661. Sung, Y. & Tinkham, S.F. (2005). Brand person ality structures in the United States and Korea: Common and culture specific factors. Journal of Consumer Psychology , 15 (4), 334 350. Sutherland, J., Marshall, S. & Parker, B. (2004). The Real, Ideal, and Undesired Self Concepts and Their Effects on Viewer Preferences: Who Do You Love? American Academy of Advertising Conference , Baton Rouge, LA. Swaminathan, V., Stilley, K.M., & Ahluwalia, R. (2009). When brand personality matters: The moderating role of attachment styles. Journal of Personality and Social Psychology , 92 (3), 542 556. The Myers & Briggs Foundation. (n.d.). Retrieved October, 2013 from http://www.myersbriggs.org The reliability and validity of the Myers Briggs Type Indicator ® instrument. (n.d.). Retrieved October, 2013 from the Center for Appl ications of Psychological Type at http://www.capt.org/mbti assessment/reliability validity.htm . Children and Society, 22 , 418 428. Thorne, A., & Gough, H. (1991). Portraits of type: An MBTI research compendium . Palo Alto, CA: Consulting Psychologists Press, Inc. Tieger, P.T., & Barron Tieger, B. (1995). Do what you are: Discover the perfect career fo r you through the secrets of personality type . New York, NY: Little, Brown & Company. Tieger, P.T., & Barron Tieger, B. (2000). wanted using the secrets of personality type . New York, NY: Little, Brown & Company. Till, B.D., Zdravkovic, S., & Morison, B. (2010). M B indicator and advertising careers. Journal for Advancement of Marketing Education , 17 , 28 38. Toncar, M., & Fetscherin, M. (2012). A study of visual puffery in fragrance advertising: is the m essage sent stronger than the actual scent? European Journal of Marketing , 46 (1/2), 52 72.

PAGE 482

482 Trotter, R.T. III, & Potter, J.M. (1993). Pile sorts, a cognitive anthropological model of drug and AIDS risks for Navajo teenagers: Assessment of a new evaluation t ool. Drugs and Society , 7 , 23 29. Tullis, T., & Wood, L.E. (2004). How many users are enough for a card sorting study? Annual Meeting of the Usability Professionals Association , Minneapolis, MN, June 10 12. Umiker Sebeok, J. (1996). Power and construction of gendered spaces. International Review of Sociology, 6 , 389 404. Van Mulken, M. (2003). Analyzing rhetorical devices in print advertisements. Document Design , 4 (2), 114 128. Van Mulken, M., Le Pair, R., & Forceville, C. (2010). The impac t of perceived complexity, deviation and comprehension on the appreciation of visual metaphor in advertising across three European countries. Journal of Pragmatics , 42 , 3418 3430. Van Riel, C.B.M., Stroeker, N.E., & Maathuis, J.M. (1998). Measuring corpora te images. Corporate Reputation Review , 1 (4), 313 326. Vaughn, R. (1980). How advertising works: A planning model. Journal of Advertising Research , 20 (5), 27 33. Vaughn, R. (1986). How advertising works: A planning model revisited. Journal of Advertising R esearch , 26 (1), 57 66. Veblen, T. (1994). The Theory of the Leisure Class . New York: Dover Publications, Inc. (Original work published 1899) Webb, S.C. (1964). An analysis of the scoring system of the Myers Brigs Type Indicator. Educational and Psychologic al Measurement , 24 , 765 781. Wedel, M., & Pieters, R. (Eds.) (2008). Visual marketing: From attention to action . New York, NY: Psychology Press. Weller, S.C., & Romney, A.K. (1988). Systematic data collection: Qualitative researcher methods . Newbery Park, CA: Sage Publications. Williams, F. (1992). Reasoning with statistics: How to read quantitative research (4th ed.). Fort Worth, TX: Harcourt Brace Jovanovich College Publishers. Williamson, J. (1978/2010). Decoding advertising: Ideology and meaning in adv ertising . New York, NY: Marion Boyars. Wimmer, R.D., & Dominick, J.R. (2011). Mass media research: An introduction (9th ed.). Boston, MA: Wadsworth.

PAGE 483

483 Wood, J.R., & Wood, L.E. (2008). Card sorting: Current practices and beyond. Journal of Usability Studies , 4 (1), 1 6. Yoo, C., Ban, H.K., & Kim, Y. (2009). The effects of a consistent ad series on consumer evaluations: A test of the repetition variation hypothesis in a South Korean context. International Journal of Advertising , 28 (1), 105 123. Young, S. (n.d.). A c ampaign for i mprovement: Getting m ore f rom p rint a dvertising . Perception Research Services . Retrieved May 28, 2014 from http: //www.prsresearch.com/prs insights/article/a campaign for improvement getting more from print advertising/ Zmuda, N. (2013, October 9). J.C. Penney scuttles logo introduced by former CEO. Advertising Age . Retrieved November 19, 2013 from www.adage.com/article/cmo strategy/j c penney scuttles logo introduced ceo/244660/ Zuckerman, C.I. (1990). Rugged cigarettes and sexy soap: Brand images and the acquis ition of meaning through the associational juxtaposition of visual imagery thesis). University of Pennsylvania, Philadelphia, PA.

PAGE 484

484 BIOGRAPHICAL SKETCH commu nication with a minor in graphic design from Abilene Christian University in spring of 2005. After working for several years as a marketing specialist for a small software company, she left the industry to pursue a graduate career. In the fall of 2010, she received her degree in journalism with an emphasis in strategic communication from the University of Kansas . She then went on to complete her doctoral degree in m ass communication with an emphasis in advertising from the University of Florida in the summer of 2014.