<%BANNER%>

JUR



Signage Preference in Grocery Stores
ALL VOLUMES CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091523/00635
 Material Information
Title: Signage Preference in Grocery Stores
Series Title: Journal of Undergraduate Research
Physical Description: Serial
Language: English
Creator: Jones, Nancey Renee
Park, Nam-Kyu
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: Fall 2012
 Subjects
Subjects / Keywords: Interior Design
Sinage
Retail Design
Grocery Store Design
Wayfinding
Kaplan's Preference Theory
Genre: serial   ( sobekcm )
 Notes
Abstract: This study aims to understand if currently available grocery store signage is adequately assisting customers in finding the items they’re looking for. Kaplan’s Preference Theory, which focuses on complexity and legibility, is used as a framework for testing the effectiveness of both existing signs and newly designed ones. Four new signs in Virtual 3D scenes were designed to reveal whether the manipulation of the two variables had any effect on their preference toward the new signage. Participants were given a task-oriented questionnaire that consisted of multiple-choice, Likert Scale, ranking, and open-ended questions. It was originally hypothesized that participants would prefer more information and the presence of a graphic (high complexity x low legibility); however, results indicated that people preferred signage with less information but more graphic appeal (low complexity x low legibility). Implications of these results suggest that designers and store owners should carefully consider limiting the amount of information they use on their signs, and potentially provide relevant graphics to enhance visual cues.
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: sobekcm - UF00091523_00602
System ID: UF00091523:00639

Downloads

This item is only available as the following downloads:

Jones_final ( PDF )


Full Text

PAGE 1

University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 1 Signage Preference in Grocery Store s Nancey Renee Jones and Dr. Nam Kyu Park College of Design, Construction and Planning University of Florida This study aims to understand if currently available grocery store signage is adequately assisting customers in finding the items which focus es on complexity and legibility is used as a framework for testing the effectiveness of both existing signs and newly designed ones. Four new signs in Virtual 3D scenes were designed to reveal whether the manipulation of the two variables had any effect on their preference toward the new signage. Participants were given a task oriented questionnaire tha t consisted of multiple choice, Liker t Scale, ranking and open ended quest ions. It was originally hypothesized that participants would prefer more information and the presence of a graphic (high complexity x low legibility) ; however r esults indicated that people preferred signage with less info rmation but more graphic appeal (l ow complexity x low legibility). Implications of these results suggest that designers and store owners should carefully consider limiting the amo unt of information they use on their signs, and potentially provide relevant graphics to enhance visual cues. INTRODUCTION One of the more recent topics in the field of Architecture and Interior Design is environmental psychology and the study of human behavior within built environments. The influence of the environment been recognized a nd studied for quite some time (Donovan, 1982) though only in the last few decades has it becoming increasingly important to apply these findings to create more effective built environments. For example, architects and designer s have made more efforts to integrate natural light into spaces now that we understand that sunlight is necessary for humans to function productively. A s the Interior Design profession grows so does research in this topic of environmental psychology The interactions, responses and preferences is not only invaluable for understanding human behavior, but can also assist designers in creating spaces that are better suited to the needs o f a targeted demographic. A n architectural interior can elicit certain reactions and emotions as well as shape behaviors both positively and negatively intentional ly or not. Much of the general public spends most of their time every day indoors whether they are working, eating, sleeping, shopping, or enjoying leisure activities. For that reason, it is imperative to study the perceptions and preferences of interior spaces. Retail stores have the potential to offer unique environments that may influence the cust Since many grocery shoppers decide which items they are going to purchase on the spot, in store elements can often have more profound effects than advertising (Baker, Grewal, & Parasuraman, 1994) For example, a shopper already made a decision about what type of cereal. In store displays, signage and other design and marketing decisions have the capacity to greatly persuade the shopper one way or another. Although retail environments provide us with the opportunity to get things that we need for our everyday lives, the shopping experience is complex and a poorly designed space has the potential to dissuade the consumer. Therefore, it is important to unders tand the variety of possibilities available for designing a successful retail space as well as what elements will make the space successful. experience with the space can begin before they even step into the store (Kopec, 2006) The experience can be broken down into three distinct processes: objective (a need/desire for a particular item or service), process (time, effort, and expense required to acquire it), and outcome (feelings evoked before, during, and after the acquisition). There are two basic types of shoppers: recreational and utilitarian (Kopec, 2006) These types of shoppers behave very differently and have different attitudes in a store environment. Recreational shoppers typically shop for the entertainment aspect, while utilitarian shoppers are on a mission to acquire a specific item or items and do not necessarily find the experience enjoyable. Different types of stores may cater to th ese different types of shoppers; for example a boutique store may draw in recreational shoppers while a drugstore more likely draws in the utilitarian. Grocery stores are somewhat unique in that they draw in all different sorts of consumers no matter which type of shopper they are. People may avoid shopping for clothes or eventually most must go shopping for food to survive.

PAGE 2

NANCEY RENEE JONES A ND DR NAM KYU PARK University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 2 always a leisurely trip : people often arrive with a list of things to purchase and are interested in getting home quickly to put away cold items. The population that visits grocery stores is vast and varied from young to old, singles to families, etc This large demographic implies a large array of differen t needs including but not limited to varying levels of physical abilities (mobility, visual acuity, etc.), range in proficiency with cooking/pairing ingredients, and available time (in a hurry vs. not in a hurry). The other issue that grocery stores face is that they stock thousands of items in many different categories, some items even being very similar or indistinguishable but produced by different brand s While organization and for, not everyo ne personally categorizes foods in the same way and categorical cues may not trigger the same associations for everyone. For example, if a customer is aisle, someone who does not typically use it for baked goods might not pick up on that association and have trouble finding it An effective tactic for dealing with all of these issues related to grocery shopping is to provide effective signage. If the system of signage is designed properly, the majori ty of people will be able to read it, find what they are looking for, and leave the store in a timely manner. Signage should be a whole system that helps customers navigate and make decisions, and if the signage is done well it should effectively disappear (Mellgren, 2005) Research Purpose & Questions The purpose of this study was to investi gate alternate ways of designing signage for grocery store aisles. Given the large demographic, vast number of items to choose from, and th possible, it is important that signage be clear and informative for the largest number of shoppers. This study was based on the Kaplan and Kaplan Preference Framework and tested a combination of two variables: leg ibility and complexity. preferences at varying levels of complexity and legibility in order to conclude which type of signage would be best for grocery store aisles. At this point it is unknown whether more complex (itemized) signage would be more helpful or more confusing fo r people trying to locate items, as well as how the level of legibility (graphic design) will affect this. For example, would it be better to have a very detailed sign that listed every item in the aisle, or would it be better to have a simple sign with just a few words as cues ? Within those two options, would it be better to provide a decorative graphic for a visual cue or a sign with no embellishment ? This study aims to pinpoint people preference for these different options and to find out their effectiveness T his study will contribute to the body of knowledge concerning signage issues in grocery store environments and provide designers with knowledge for creating a more effective re tail space. This in turn will provide a better sho pping experience for consumers as they t ry to locate items in the store, and perhaps boost sales and create repeat customers. In this study, some of the questions being asked are as follows : How do people feel about grocery store aisle signage that is currently available? What are the perceived problems/advantages concerning existing signage? How will people to respond to more complex, itemized signage? How will people respond to signage with varying level s of graphic design? Will people prefer a newly proposed signage design over any of the existing designs? Results from this study responded to the significance of this topic and demonstrate d whether these issues are important in the world of grocery store design. It is anticipated that participants will prefer a signage design with high comp lexity and low legibility based on the hypothesis that more information and visual stimulation will be the most appealing. RESEARCH METHOD S Initial Investigations Pre de sign research began with visiting different existing grocery stores and documenting signage types that were already being implemented. The discovery was that four main types of signage were present in all six grocery stores (some stores used more than one type) and each type had their own pros and cons. These four types are indicated in Figure 1 (shown in aerial vie w, signs have arrows indicating the direction of information):

PAGE 3

SIGNAGE PREFERENCE I N GROCERY STORES University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 3 Figure 1 Four Existing Signage Types, View ed From Above Participants The participant group for this study consisted of both male and female adult undergraduate students and was recruited for the study through the Behavioral Research Lab in the Marketing Department at the University of Florida. Par ticipants could volunteer to partake in the study and were awarded extra credit for classes in which they were currently enrolled. The study required participants to perceive colors in the scenes that they were shown, so they were asked whether they had a ny visual impairment or inability to see colors that could not be corrected by contact lenses or eyeglasses. Participants gave their consent to be surveyed, and Institutional Review board to engage in r esearch with human subjects This study focused specifically on legibility and complexity and how signage can be designed to test varying levels of these two characteristics. Each category had two levels of intensity (low and high); this c reates four different combinations of virtual store environments demonstrated in Table 1 Table 1 Kaplan's Theory Applied to Store Scenes Characteristic Complexity (low) Complexity (high) Legibility (low) Scene A Graphic Few words Scene B Graphic Many words Legibility (high) Scene C No graphic Few words Scene D No graphic Many words Differences in legibility and complexity were manipulated by the use of a graphic design and the amount of information, respectively. Low legib ility signage displayed a large graphic meant to enhance the aesthetics of the sign at the potential cost of being more legible, while high legibility signage contained no graphic. The graphic was also indicative of what the aisle contained. Low complexity signage listed just a few items indicating the types of foods in each aisle, and high complexity signage listed every type of item in the aisle. The different scenes were created by computer rendering through the use of a combination of AutoCAD, Revit Ar chitecture and Adobe Photoshop. Each scene contained the same types of products, colors, and lighting methods and only the signage itself in the space was manipulated. Research Setting In order with the di fferent combinations of characteristics, participants needed to be presented with realistic scenes that would give them enough context to make an informed decision. By providing d igitally rendered scenes it was much easier to control potential unwanted va riables that could appear if testing in a real life setting. The alternative would have been to construct an actual life sized grocery store model, which was not feasible given the constraints of this study The layout for the grocery store that appears i n the scenes was taken from an actual grocery store provided through a design competition presented by the Retail Design Institute. This floor plan used a generic gro cery store layout that fit the needs of this study.

PAGE 4

NANCEY RENEE JONES A ND DR NAM KYU PARK University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 4 A field investigation was done in o rder to find out what type of modular shelving was being used in grocery stores and if there was consistent siz ing among shelving units. of shelving, depending on the size of the items being displ ayed. Shape and form were more or less the same, though there was slight color variation from store to store. Stores consistently employed rows of these shelving units with endcaps at either ends of the rows. The first set of four scenes in Figure 2 was cr eated to reflect the four existing signage types found in the initial investigations Neutral colors were used in the scenes in order to avoid triggering any recognition of an existing grocery store, and shelving was populated with staple items that most participants could recognize. The lighting used in the scenes was a series of cool fluorescent fixtures, which are extremely prevalent in stores of this type. Figure 2 Scenes for existing s ignage The second set of four scenes was created using the same method ; however they contained slight differences. First they employed the new signage, which only differed graphically from scene to scene and not by shape, form or location. The point of reference was shifted slightly to focus more on one individual sign, whereas the first set of images used a wider cone of visibility in order to provide a comprehensive view of the signage as a system. The new scenes also displayed different foods on the shelv es, drawing select ions based on research done about commonly recognized S ignage was designed to be visually appealing while still being functional and not lending itself to mimic any particular Complexity and legibility were manipu lated to create four distinct new signs. In the low legibility scenes, the aisles contained pasta products and sauces, so the graphic was a close up image of a prepared pasta dish. Hypothetically, other aisles in the grocery store would have similar signag e with a different graphic for each aisle. The logic behind this decision was that the graphic may interfere with the legibility by complicating the background, but would be aesthetically appealing and indicate the contents of the aisle. The complexity of the signage was manipulated by changing the amount of information on each sign. Two of the scenes contained contained a detailed, alphabetized list of e ver y item in the aisle. These 4 combinations can be seen in Figure 3 and 4

PAGE 6

SIGNAGE PREFERENCE I N GROCERY STORES University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 5 Figure 3 New signage designs Figure 4 Final rendered s cenes Instrument After creating all eight rend ered scenes, a questionnaire was created that could preferences toward the signage as well as gather qualitative responses about their preferences and w hy they made certain decisions. The questionnaire used a combination of Li kert Scale, ranking open ended and multiple choice questions. Seven point Likert Scale questions were used to find o ut how participants viewed characteristics for each scene. Each participant was first presented with one random scene and then later with all four scenes and was asked to rank them. Participants were also asked to simply describe what they did or did not like a bout the scene and later their reasoning for which they ranked first and last. Previous research studies using similar methods to te st referenced in order to develop an effective questionnaire. (1998) about perceived danger and environmental preference uses a very similar approach by showing participants dif ferent scenes and asking them a

PAGE 7

NANCEY RENEE JONES A ND DR NAM KYU PARK University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 6 series of these types of questions Results were statistically analyzed and revealed the desired representations of preference for each scene. Data Collection A survey tool called Qualtrics was used to create the questionn aire and distribute it to participants. This software allowed for the creation of a questionnaire that could distribute a random scene to each participant and ensure that a similar amount of participants received each of the scenes. After the data collecti on was complete, each of the four scenes collected responses from about 25% of the participants making the results very even ly distributed RESULT S Analysis of variances was used on the questionnaire data in order to find out if the different levels of co mplexity and scenes The first section addresses demographic information of the participant group and their overall preference towards shopping The next section contains both Likert Scale open e nded questions concerning as well as rankings where participants order all four scenes from mos t preferred to least preferred and give a rationale. Participant Demographics As shown in Table 2, males represen ted a higher percentage of the participant group than females Only 3 participants failed to provide their gender information. The largest percentage of the participants was between the ages of 21 and 30, and only 3 participants failed to provide their age. Table 2 Participant Demographics Source Frequency % Frequency % Gender Age N/A 3 2.6 N/A 2 1.8 Male 60 52.6 41+ 8 7.0 Female 51 44.7 31 40 23 20.2 Total 114 100.0 21 30 53 46.5 18 20 28 24.6 Total 114 100.0 Participants were also asked if they had any type of visual deficiencies that could not be overcome using corrective lenses. Any data from participants who study relie d on visual capabilities. One of the demographic questions asked to participants was how they felt about grocery shopping in general, and they responded on a scale of one to seven with one being The reason for this was to find out any of the other variables in the study. Most people felt neutral or better towards grocery shopping, with only 21.1% responding below neutral (see Table 3) Further analysis of this data showed no significan t difference between genders and it did not largely affect preference for any of the scenes. Table 3 Participant Shopping Preference Source Frequency % 1 (Hate it) 6 5.3 2 8 7.0 3 10 8.8 4 (Neutral) 18 15.8 5 36 31.6 6 25 21.9 7 (Love it) 11 9.6 Total 114 100.0 *7 point Bipolar Semantic Scale: 1 = Hate it; 7 = Love it Scene Rankings When first presented with one random scene, responses were positive negative and neutral towards the signage: participants were given n o other scenes to compare with and therefore responded abstractly about what came to mind. Once all four scenes were presented and participants were asked to rank them from highest to lowest preference, there was a very strong, consistent ranking pattern a nd responses began to produce similar themes. Upon seeing the first scene, participants were asked if they liked the signage and if they thought it provided adequate information. People were relatively neutral about liking the signage, and their decision was not significantly affected by either complexity or the legibility. They were slightly more positive about feeling that the signs provided adequate information, but again were not highly affected by either the complexity or legibility variables (see Ta bl es 4 and 5 )

PAGE 8

SIGNAGE PREFERENCE I N GROCERY STORES University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 7 Table 4 Mean and S tandard deviation (SD) S cores for S E valuation of P reference Source n Mean 2 SD p value Complexity .994 Simple 57 3.79 2.27 Complex 57 3.81 2.12 Legibility .221 Easy to read 56 4.05 2.19 Difficult to read 58 3.55 2.17 Complexity by Legibility 114 Scene A (Simple x Easy) 28 3.71 2.39 Scene B (Complex x Easy) 28 4.39 1.97 Scene C (Simple x Difficult) 29 3.86 2.20 Scene D (Co mplex x Difficult) 29 3.24 2.13 *7 point Bipolar Semantic Scale: 1 = Not at all; 7 = Very much Table 5 Mean and S tandard D eviation (SD) S cores for S E valuation of A dequate I nformation Source n Mean 2 SD p value Comple xity 0.50 Simple 56 5.61 1.80 Complex 57 5.37 1.88 Legibility 0.40 Easy to read 55 5.64 1.70 Difficult to read 58 5.34 1.95 Complexity by Legibility 0. 48 Scene A (Simple x Easy) 27 5.63 1.74 Scene B (Complex x Easy) 28 5.64 1.70 Scene C (Simple x Difficult) 29 5.59 1.88 Scene D (Complex x Difficult) 29 5.10 2.02 *7 point Bipolar Semantic Scale: 1 = Strongly disagree; 7 = Strongly agree When presented with all four scenes at once and asked to rank them, the themes that the responses began to create were very clear and able to be organized into three categories: complexity, legibility, and overall appearance (see Table 7) Most participants used words to describe the y to read, clear busy that matched the definitions of the variables. This was a very positive outcome since it reinforced that the study achiev ed what it set out to discover. Overall, there was an overwhelming preference f or scene A (low complexity, low legibility) and an overwhelming dislike for scene D (high complexity, high legibility). People found scene A to be simple and attractive, and scene D to be boring and clutt ered (see Table 6) Table 6 Ranking F requencies for each S cene Scene Ranking 1 2 3 4 A 66 33 9 6 B 18 20 46 30 C 25 50 24 15 D 5 10 37 62 Table 7 presents the frequency of open ended answer s in each category. Comments in the It is clear that scene A was thought to be clear and legible, while scene D was too busy and aesthetically unappealing.

PAGE 9

NANCEY RENEE JONES A ND DR NAM KYU PARK University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 8 Table 7 Qualitative Themes Themes (Most Liked) Themes (Least Liked) Complexity Legibility Appearance Total Complexity Legibility Appearance Total n % n % Scene A 35 31 44 110 59 .46 5 1 2 8 4.88 Scene B 12 1 14 27 14.59 22 10 10 42 25.61 Scene C 15 19 4 38 20.54 8 1 10 19 11.59 Scene D 5 4 1 10 5.41 49 13 33 95 57.93 Total 67 55 63 185 100.00 84 25 55 164 100.00 Statistical analysis revealed t hat the scenes did indeed vary in complexity and legibility as they were designed; however the factors affecting them were not as expected (see Tables 8 and 9) For complexity, the amount of text in the sign was manipulated and created either a simpler or more complex sign as intended For legibility, the presence of a graphic element was supposed to create a sign that was either easier to read or harder to read. However, the graphic actually did little to affect the legibility and it was the amount of text on the sign that affected it instead. As a signage with interesting graphics, but in this case it did not affect how easy or difficult it was for someone to read it. Table 8 Mean and S tandard D eviation (SD) S cores for S ubj Evaluations of Complexity Source N Mean 2 SD p value Complexity .000 Simple 57 6.04 1.60 Complex 57 4.32 2.22 Legibility .835 Easy to read 56 5.21 2.21 Difficult to read 58 5.14 2.04 Complexit y by Legibility 712 Scene A (Simple x Easy) 28 6.14 1.82 Scene B (Complex x Easy) 28 4.29 2.19 Scene C (Simple x Difficult) 29 5.93 1.39 Scene D (Complex x Difficult) 29 4.34 2.29 *7 point Bipolar Semantic Scale: 1 = Complex ; 7 = Simple Table 9 Mean and S tandard D eviation (SD) S cores for E valuations of L egibility Source N Mean 2 SD p value Complexity .001 Simple 57 5.67 2.0 Complex 57 4.39 2.03 Legibility 331 Easy to read 56 5.21 2.18 Difficult to read 58 4.84 2.03 Complexity by Legibility .570 Scene A (Simple x Easy) 28 5.96 2.12 Scene B (Complex x Easy) 28 4.46 2.01 Scene C (Simple x Difficult) 29 5.38 1.86 Scene D (Complex x Difficult) 29 4.31 2.09 *7 point Bipolar Semantic Scale: 1 = Difficult to read ; 7 = Easy to read CONCLUSIONS AND DISC USSIONS This study was designed to investigate preference for signage based on varying levels of complexity and legibility. The results indicated that the original hypothesis was part ial ly incorrect, and that participants did not find the signage more appealing when it was densely populated with information. Despite the fact that the more complex signs offered more specific and detailed information about they were more cumbersome to read. However, this finding was consistent with the hypothesis that participants were more drawn to the signage with the graphic detail and in general thought it was a positive aspect. The combination of aesthetics and a moderate amount of information (low complexity, low legibility) was the most appealing.

PAGE 10

NANCEY RENEE JONES A ND DR NAM KYU PARK University of Florida | Journal of Undergraduate Research | Volume 14, Issue 1 | Fall 2012 9 Some possible limitations and/or opportunities for further investigation may include the effects of reducin g the area for product display on the endcap, varying the to read the signage at different distances. There are many opportunities for this study as there are several other variables and environm ental characteristics that can be manipulated. To further this study, the next logical step would be to redesign the new signs so that legibility is being measured in the way that it was intended. As previously stated, legibility was successfully manipulat ed but as a result of complexity and not the influence of the graphic. This was a minor setback and it does indicate that a pleasing graphic does not interfere with how easy or difficult it i s for people to read the signs. These findings suggest that inter ior/graphic designers employing these types of signs. Placing signs at eye level, including visually appealing elements and limiting the amount of information on each sign are positive factors for th is type of setting. Signage that uses these design elements could lead to higher sales and less customer confusion. This information could potentially be a useful addition to the current body of knowledge, and will continue to expand as more is discovere d about interior environmental preference. BIBLIOGRAPHY Ang, S., & Leong, S. (1997). The mediating influence of pleasure and arousal on layout and signage effects. Journal of Retailing and Consumer Services 13 24. Baker, J., Grewal, D., & Parasuraman, A. (1994). The influence of store environment on quality inferences and store image. Journal of the Academy of Marketing Science 22 (4), 328 339. Buffalo, School of Architecture and Planning University at Buffalo. (2012 ). Buffalo, NY. University of Bu ffalo. Carmel Gilfilen, C. (2011). Advancing retail security design: uncovering shoplifter perceptions of the physical environment. Journal of Interior Design 36 (2), 21 38. Donovan, R. J. (1982). Store atmosphere: An environmental psychology approach. Journal of Retailing 58 (1), 34 58. Doyle, J. R., & Bottomley, P. A. (2008). The massage in the medium: Transfer of connotative meaning from typeface to names and products. Applied Cognitive Psychology 23 (3), 396 409. Herzog, T. R., & Leverich, O. L. (2003). Searching for legibility. Environment and Behavior 35 (4), 459 477. Herzog, T., & Miller, E. (1998). The role of mystery in perceived danger and environmental preference. Environment and Behavior 30 (4), 429 449. Kaplan, S., & Kaplan, R. (1989) The experience of nature: A psychological perspective. New York: Cambridge University Press. Kaplan, S. (1987). Aesthetics, affect, and cognition: environmental preference from an evolutionary perspective. Environment & Behavior 19 (1), 3 32. Kopec, D (2006). Environmental Psychology for Design. New York: Fairchild Publications, Inc. Mellgren, J. (2005, July 16). Signs of success: the silent sales force. pp. 30 33. Newman, E. (2007). Retail Design for 2008: Thinking Outside the Big Box. p. 26. Olsen, S. O., & Kare, S. (2011). Journal of Consumer Marketing 28 (7), 532 539. Retail Design Institute. (2011). Tarrytown, NY. R etail Design Institute. Reutterer, T., & Teller, C. (2009). Store format choice and shopping trip types. International Journal of Retail & Distribution Management 37 (8), 695 710. Stamps, A. E. (2004). Mystery, complexity, legibility and coherence: A met a analysis. Journal of Environmental Psychology 24 (1), 1 16. Wilkinson, C., Mickle, S. J., & Goldman, J. D. (2003). Trends in food and (Kaplan & Kaplan, 1989) nutrient Intakes by adolescents in the United States. Family Econo mics and Nutrition Review 15 (2), 15 27.