Citation
The Influence of Repeated Exposure to Negative Political Advertising on the Evaluations of Candidates and Vote Intention

Material Information

Title:
The Influence of Repeated Exposure to Negative Political Advertising on the Evaluations of Candidates and Vote Intention
Creator:
FERNANDEZ, JULIANA DE BRUM ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Advertising research ( jstor )
Frost ( jstor )
Political advertising ( jstor )
Political candidates ( jstor )
Political elections ( jstor )
Psychological assessment ( jstor )
Questionnaires ( jstor )
Television commercials ( jstor )
Television programs ( jstor )
Voting ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Juliana De Brum Fernandez. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2012
Resource Identifier:
700510616 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

1 THE INFLUENCE OF REPEATED EXPO SURE TO NEGATIVE POLITICAL ADVERTISING ON THE EVALUATIONS OF CANDIDATES AND VOTE INTENTION By JULIANA de BRUM FERNANDES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2007

PAGE 2

2 Copyright 2007 Juliana de Brum Fernandes

PAGE 3

3 To my fianc, Juliano; and to my mother, Margarida.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank Dr. Lynda Lee Kaid fo r her support, advice and kindness throughout this entire process. In addition, I would like to thank Dr. Johanna Cleary and Dr. Jorge Villegas for their helpful and sincere comments on the execution of this master’s thesis. I would also like to thank th e Florida/Brazil Institute for granting me the opportunity to complete my master’s degree at the University of Florida. Finally, I would like to thank my fianc, Juliano Laran, and my mother, Margarida Fernandes for always believing that I could accomplish my dream.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........8 ABSTRACT....................................................................................................................... ..............9 CHAPTER 1 INTRODUCTION..................................................................................................................11 2 LITERATURE REVIEW.......................................................................................................14 Negative Political Advertising................................................................................................14 Effects of Negative Political Adve rtisement on Candidate Evaluations.........................16 Effects of Negative Political Advertising on Vote Intention and Attitude toward the Ad............................................................................................................................. ....18 Effects of Message Repetition................................................................................................20 The Two-Factor Theory..........................................................................................................22 Research Hypotheses............................................................................................................ ..28 3 STUDY 1........................................................................................................................ ........29 Introduction................................................................................................................... ..........29 Method......................................................................................................................... ...........29 Participants and Design...................................................................................................29 Stimuli........................................................................................................................ .....29 Procedure...................................................................................................................... ...30 Dependent Variable Measurement..................................................................................31 Data Analysis.................................................................................................................. .32 Analysis and Results........................................................................................................... ....32 Candidate Evaluation.......................................................................................................32 Likelihood of Voting.......................................................................................................34 Additional Measures........................................................................................................36 Discussion..................................................................................................................... ...36 4 STUDY 2........................................................................................................................ ........45 Introduction................................................................................................................... ..........45 Method......................................................................................................................... ...........45 Participants and Design...................................................................................................45 Stimuli........................................................................................................................ .....45 Procedure...................................................................................................................... ...47

PAGE 6

6 Dependent Variable Measurement..................................................................................48 Data Analysis.................................................................................................................. .49 Analysis and Results........................................................................................................... ....49 Candidate Evaluation.......................................................................................................50 Likelihood of Voting.......................................................................................................51 Cognitive Responses.......................................................................................................53 Additional Measures........................................................................................................54 Discussion..................................................................................................................... ...55 5 GENERAL DISCUSSION.....................................................................................................64 Implications................................................................................................................... .........65 Limitations and Future Research............................................................................................66 APPENDIX A DESCRIPTION AND TRANSCRIPTION OF THE NEGATIVE POLITICAL ADVERTISEMENTS.............................................................................................................69 Advertisement with Gabe Castillo and Dan Lungren.............................................................69 Description of the Advertisement....................................................................................69 Transcript of the Advertisement......................................................................................69 Advertisement with Pete Sessions and Martin Frost..............................................................69 Description of the Advertisement....................................................................................69 Transcript of the Advertisement......................................................................................70 B QUESTIONNAIRE – STUDY 1............................................................................................71 C QUESTIONNAIRE – STUDY 2............................................................................................77 LIST OF REFERENCES............................................................................................................. ..84 BIOGRAPHICAL SKETCH.........................................................................................................89

PAGE 7

7 LIST OF TABLES Table page 3-1 Descriptive statistics for candidate ev aluation, vote likelihood a nd recall. Individual analysis for sponsor and target of the advertisement.........................................................40 3-2 One-way ANOVA results for candidate ev aluation and vote likelihood. Individual analysis for sponsor and target of the advertisement.........................................................41 3-3 Descriptive statistics for candidate ev aluation and vote likelihood. Global analysis for sponsor and target of the advertisement.............................................................................42 4-1 Descriptive statistics for candidate eval uation, vote likelihood, reca ll, attitude toward the ad, and cognitive responses. Individual an alysis for sponsor and target of the advertisement.................................................................................................................. ...59 4-2 One-way ANOVA results for candidate ev aluation and vote likelihood. Individual analysis for sponsor and target of the advertisement.........................................................60 4-3 Descriptive statistics for candidate ev aluation and vote likelihood. Global analysis for sponsor and target of the advertisement.............................................................................61

PAGE 8

8 LIST OF FIGURES Figure page 3-1 Composite means for the evaluation of the sponsor and the target of the advertisement and difference between the eval uations of th e candidates.................................................43 3-2 Composite means for likelihood of voti ng for the sponsor and the target of the advertisement and difference between the likelihood of voting for the candidates...........44 4-1 Composite means for the evaluation of the sponsor and the target of the advertisement and difference between the eval uations of th e candidates.................................................62 4-2 Composite means for likelihood of voting for the sponsor and the target of the advertisement and difference between the likelihood voting for the candidates...............63

PAGE 9

9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Mass Communication THE INFLUENCE OF REPEATED EXPO SURE TO NEGATIVE POLITICAL ADVERTISEMENT ON THE EVALUATION S OF CANDIDATES AND VOTE INTENTION By Juliana de Brum Fernandes May 2007 Chair: Lynda Lee Kaid Major: Mass Communication This study investigates the influence of repeated exposure to negative political advertisements on the evaluations of and likelihood of voting for th e sponsor and the target of the ad. Two experimental studies were developed to investigate this rela tionship. Study 1 exposed participants to a negative political advertising one, three or five tim es. The entire set of advertisements, which also included filler ads, was presented in 7, 9 or 11 minutes, characterizing the presentation as massive. Study 2 sought to investigate the effects of repeated exposure to a negative political advertisemen t on a real world condition. Participants were exposed to a real 30-minute TV program with inse rtions of the negative political advertisement one, three or five times, characterizing th e presentation as distributed or spaced. The results of Study 1 suggested that mode rate repetition of a negative political advertisement presented massively enhances th e evaluation of and the likelihood of voting for the sponsor of the ad. Conversely, moderate repe tition of a negative politi cal advertisement did not affect the evaluation and likelihood of voting fo r the target of the ad as compared to low and high repetition. In addition, Study 1 revealed th at under moderate exposure to the ad, the difference between the evaluation of and the likeli hood of voting for the sponsor and the target of the ad was greater than under lo w and high repetition to the ad.

PAGE 10

10 The results of Study 2 suggested that mode rate repetition of a negative political advertisement in a spaced presentation did not af fect the evaluation of and the likelihood of voting for both candidates as compared to low and high repetition. Additionally, the results of Study 2 revealed that the difference between th e evaluation of and the likelihood of voting for the candidates increased as exposure to the adve rtisement increased. Therefore, these results indicated that when presentation of advertisements is spaced, positive affect would probably be reached under higher levels of re petition to the advertisement. The contribution of this research to the field of political a dvertising lies in understanding how repetition of a negative polit ical advertisement affects the evaluation of and the likelihood of voting for the sponsor and the target of th e ad. Moreover, findings suggest that the way advertisements are presented to viewers (i.e., massively vs. spaced) can influence how people evaluate political candidates.

PAGE 11

11 CHAPTER 1 INTRODUCTION Imagine a political candidate running for a s eat in Congress. His cam paign strategies to promote his candidacy include broadcasting some TV advertisements. Among the types of advertisements chosen to be aired, the candidate and his campaign team decide to air political advertisements that directly attack his opponent (i.e., negative advertising) . In addition to these advertisements, he will air a few ads promoting his own positions on issues such as education, taxes, and others. Now, imagine that the major ity of the candidate’s TV advertisements are negative in tone. That is, he points out the cont roversial positions his opp onent favors. Also, in the same ad, the candidate who sponsors the ad presents little information about his own positions on the same issues and about the thi ngs he will do for the country if elected. How would these advertisements influence the evalua tions of the candidates? What would be the impact of airing the same advertisement over and over again on the evaluation of each candidate? One alternative is that repetition may result in more positive evaluations toward the sponsor of the ad and more negative evaluations to ward the target of the ad. However, the more the ad is repeated, the more the viewers may b ecome tired of watching that advertisement and will evaluate both candidates negatively. Repetition of negative advertisements may also backfire and have a negative impact on the eval uation of the sponsor a nd a positive impact on the evaluation of the target of the ad. This scenario is not unlikely during electoral periods sinc e a great deal of political advertising is aired everyday. In 2002, for instan ce, U.S. candidates spent more than $1 billion on political advertising. In 2006, the spending on TV advertisement reached the $2 billion mark (TNS Media Intelligence 2006), and $160 million were spent on negative political advertisement (Kuhnhenn 2006). It is clear that candidates are spending a great deal of their campaign budget

PAGE 12

12 on political advertising. One explanation fo r the increase in the number of political advertisements is the fact that they have beco me the major means for candidates to communicate with voters in a massive manner (Jamieson 2004 cited in Crawford 2004, para. 9). As a consequence, a variety of studies have been done to verify the effects of such advertisements. Researchers have found that negative political ad vertising is considered unethical (Surlin and Gordon 1977) and untrustworthy (Garramone 1984) . However, people find these ads more informative than the positive version (Lau 1985; Surlin and Gordon 1977). Other studies have examined the role of the sponsor in political advertising and the findings indicate that if an advertisement is sponsored by an independent orga nization, people tend to ev aluate this ad more positively (Garramone and Smith 1984). However, if the ad is sponsored by a candidate or political party, people generally dislike it (Garramone 1984; Johnson-Cartee and Copeland 1987). Therefore, people’s percep tion of the candidate as the sponsor of the ad plays an important role in defining whether people will like or dislike the advertisements. Yet, other factors may affect the likeability of negative political advertisements. For instance, the number of exposures of a negative political ad could result in positive affect toward the candidate and the ad. After repeated expos ure to the advertisements, people would show increased affect toward negative ads and th e candidate. However, overexposure to the advertisement could result in tedium or reacta nce toward the ad and the candidate. Repeated exposure would lead to an increase, and then a de crease in liking toward th e candidate and the ad due to a two-step process. Initia lly, with a moderate level of e xposure, an effort to learn and comprehend the message develops. In this stage, according to Cacciopo and Petty (1979, 1980), the viewer has opportunity and ability for mo re elaborate processing of the message. As

PAGE 13

13 exposure continues, however, tedium or reacta nce toward the ad may develop. Motivation to learn about the message decreases, producing nega tive evaluations toward the message or ad. The two-factor process has been widely stud ied using noncommunicative stimuli, such as polygons, drawings and nonsense words. In addition, th e idea of an increase then a decrease in liking has produced some work that utilized co mmunicative stimuli, such as advertisements. Nevertheless, little is known about the effects of repetition on political advertising. Thus, the current study represents an attempt to investig ate the effects of repeat ed exposure to negative political advertisements on the evaluations of the sponsor and the target of the ad and on the likelihood of voting for the sponsor and the target. The remaining chapters of this research are organized as follows. The next chapter reviews the literature on negative political advertising and th e theories used to study repetition. Chapters 3 and 4 present two experimental studies that were developed to inve stigate the effects of repeated exposure to negative political advert isements. Chapter 3 presents Study 1, which was designed to test the effects of repetition on a sequentia l presentation of advertisements. Chapter 4 presents Study 2, which was designed to test the effects of repetition us ing a real TV program. Therefore, Study 2 intended to explain a real wo rld condition. A general discussion of the results of the studies is presented in chapter 5 along with the limitations of the study, implications and directions for future work.

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW Political campaigns are periods of time during which candidates and their parties transmit their points of view about issues that are facing the nation, and they are also the time to persuade people to vote in a certain way on election day. In this regard, in every election in the United States, candidates and their parties spend millions of dollars on political advertising to influence and obtain voters’ support. Political advertising, in turn, is one of the most preferred and most powerful campaign strategies that candidate s utilize to promote their candidacies. According to Kaid, Chanslor and Hovind ( 1992), the reasons why political candidates spend a great deal of money on TV advertisem ents are well-know: “television advertising reaches large numbers of voters, requires limited information-seeking behavior on the part of the voter, and provides the candidate with comple te control of the message. [] television advertising also allows the pol itical candidates to choose the specific television programming environment in which political commercials wi ll appear” (Kaid et al . 1992, p. 303). This huge financial investment, however, has been producing other effects on the democratic process, such as negativism toward the campaign process, cy nicism toward the political system, apathy, and lower voter turnout (Ansolabehere and Iyengar 1995; McLeod and McDonald 1985; Cohen and Davis 1991; Johnson-Cartee and Copeland 1991). C oupled with these symptoms of political disaffection, one element that has been receiving attention from candidates and researchers is the use of negative political advertisements. Th is topic is discussed in the next section. Negative Political Advertising Although an exact definition ha s not been reached yet, ge nerally, a negative political advertisement directly attacks its target, pointing out the broken promises of the target, voting record inadequacies, or denigra ting the target’s pers onal characteristics or traits (Merritt 1984;

PAGE 15

15 Johnson-Cartee and Copeland 1989; Pinkleton 1997). Garramone (1984) and Stewart (1975) refer to this type of advertis ing as “mudslinging.” Based on this definiti on, researchers have identified three types of negative advertising : negative issue advertising: usually containi ng information about an opponent’s political record, voting record, criminal record, and positions on issues or items of public policy (Johnson-Cartee and Copeland 1989) negative image advertising: us ually containing information about an opponent’s medical history, personal life, religion, sex life, or family member. Such advertisements focus on the opponent’s personal characteristics and us ually they do not address issue positions. (Johnson-Cartee and Copeland 1989) comparative advertising: a more complex fo rm of political advertising in which both candidates are directly compared. The sponsor candidate claims superi ority over the target candidate based on both candida tes’ issue positions, experi ence, or voting records. (Johnson-Cartee and Copela nd 1991; Merritt 1984). Considering these definitions and the amount of negative campaign tactics that have been presented in past elections, many studies have been conducted to i nvestigate the impact that such advertisements might have on the democra tic process (Ansolabeh ere and Iyengar 1996; Ansolabehere, Iyengar, and Simon 1999; Ka hn and Kenney 1999). They argue that negative campaign tactics, including negative political adve rtising, turn voters off and keep people away from the polls, therefore, contributing to a uniformed, uninvolved electorate. In a meta-analysis study, Lau et al. (1999) f ound that people like negative ads less than positive ads.1 However, research has shown that negativ e political advertising can be useful. For instance, Surlin and Gordon (1977) found that negative political a dvertisements were considered unethical, but also were perceived to be inform ative. Not only is negative political advertising informative, but it may also be considered more informative than the positive version of the advertisements. Lau (1982) found th at negative information is more influential than comparable 1 According to Lau et al. (1999), this finding seems to be the only point of agreement in the literature regarding the effects of negative ads.

PAGE 16

16 positive information in candidate evaluations . A more recent study conducted by Lau and Pomper (2002) found that negative campaigning is relatively effective for challengers, while positive campaigning is more effective for incumb ents. In addition, some studies have found that negative information in general is percei ved as more credible than positive information (Hamilton and Zanna 1972, Leventhal and Singer 1964). Many researchers have examined the impact of political adve rtising on political candidates’ evaluations (Ha ddock and Zanna 1997; Kaid and Boydston 1987; Pinkleton 1997; Shapiro and Rieger 1992), vote intention, and attitude toward the ad (Garramone 1985; Garramone and Smith 1984; Meirick 2002; Pi nkleton 1997; Roddy and Garramone 1988). Of particular importance to this st udy are the studies which investigat ed candidate evaluations, vote intention, attitude toward the ad and that have exposed part icipants to negative political advertising. These studie s are examined next. Effects of Negative Political Advertisement on Candidate Evaluations It is well known that exposur e to political advertising a ffects candidates’ evaluation (Garramone 1984; Garramone and Smith 1984; Haddock and Zanna 1997, Kaid and Boydston 1987; Kaid and Chanslor 1995; Pinkle ton 1997; Shapiro and Rieger 1992). According to Kaid and Boydston (1987), audi ences have rated candi dates as being less qualified, less honest, less serious, le ss sincere, less successful, and le ss fiscally responsible after exposure to negative political ads. In addition, resear ch results show that ev aluations of the target candidate worsen in greater proportion than eval uations of the sponsor candidate among persons exposed to negative issue advertising (Shapiro and Rieger 1992). On the other hand, Hill (1989) found that the target of the ad was liked more if the ad was negative rather than positive, while the sponsor of the ad was liked less if th e ad was negative rather than positive.

PAGE 17

17 In a study that exposed participants to negativ e comparative political advertising, Pinkleton (1997) investigated the effects of such advertisements on candidate evaluations. The results show that the comparative stimuli containing the most negative information produced significantly greater negative changes in the ta rget candidate evaluations than any other stimuli (i.e., low and moderate negative information). Although candi dates receive negative evaluations when featuring negative ads, negativ e advertising makes a greater impression on the audience than positive or neutral advertising because of the greater weight given to negative information over positive information (Pinkleton 1997). Research in psychology has shown that individuals weigh negative information more heavily than positive information in individual impression formation, assessments of likeability, and judgmental d ecision making (Fiske 1980; Hamilton and Huffman 1971; Hamilton and Zanna 1972). Furthermore, negative information is weighted more heavily than positive information in political perception and behavi or (Lau 1982; 1985). Lau (1982) found that negative information receives more weight than comparable positive information in the formation of evaluations of presidential candidates. The results of these studies are divergent. Some results indica te that exposure to negative political advertising either lowers evaluations of the sponsor of th e ad or it lowers evaluation of the target of the ad. With respec t to lowering evaluations of the sponsor of the ad, research has identified this effect as a “backlash effect or boomerang effect” (Garramone 1984; Merritt 1984). In other words, unintended effects are generated to ward the sponsor of the ad instead of intended effects toward the target of the ad. Although so me of these studies have shown that negative political advertising produces backlash effects toward the sponsor of the ad, researchers have identified that third-party or independent s ponsorship of political ads can alleviate these unintended effects (Garramone 1984; 1985; Garr amone and Smith 1984). In addition, negative

PAGE 18

18 issue advertising, that is, ads that attack the ta rget’s issue positions, tend to be more effective than attacks made to the character or imag e of the opponent (Kahn and Geer 1994; Roddy and Garramone 1988). As a result, because the sponsor of the advertising is issue-attacking his opponent, his evaluations will be higher than the target of the ad. On the other hand, for the candidate who is the target of th e advertisement, rebuttals can be considered a strategy to lessen the effects of the sponsor’s attack (Garramone 1985; Roddy and Garramone 1988). In short, while some researchers have found that negative ads are effective and produce the intended effects toward the target of the ad, others have found th at negative ads fail to deliver their main purpose, causing backlash effects towa rd the sponsor of the ad. Consequently, the evaluation of the target of the ad is improved. Moreover, these studies did not investigate the effects of repeated exposure to the advertisements on candidate evaluations. In fact, repeated exposure to negative ads is a real condition duri ng election periods, and until now, little attention was given to this topic. Effects of Negative Political Advertising on Vo te Intention and Attitude toward the Ad Primarily, when a candidate decides to air ne gative political advertisements, the desired goals are to denigrate the image of the opponent and increase the evaluations of the sponsor candidate. Ultimately, a negative political advertisem ent aims to persuade people to vote for the sponsor candidate on election day. Research on this issue has found that political advertisements do produce behavioral effects. In a study that explored the in fluence of sponsor and rebuttal2 on perceptions of the candidates and vote intentions, Garramone ( 1985) exposed participants to negative 2 According to Garramone (1985, p.150), “a rebuttal politi cal commercial may be defined as one which charges as false the claims of another commercial. Effective rebuttal of an attack should attenuate the attack’s negative impact on the targeted candidate.”

PAGE 19

19 advertisements embedded in a newscast. The resu lts show that viewers of an independent ad opposing a candidate demonstrated a decreased likelihood of voting fo r the target as compared to viewers of an advertisement that was sponsored by a candidate. Furthermore, viewers in the rebuttal condition demonstrated a lesser likeliho od of voting for the target’s opponent than did viewers in the no-rebuttal conditi on. These findings are in line w ith an earlier study developed by Garramone and Smith (1984). In this study, the au thors aimed to investigate the process of sponsor effects, focusing on the mediating role s of sponsor trustworthiness and advertisement evaluation. As a result, Garramone and Smith ( 1984) found that an ad that is independently sponsored is perceived as more trustwort hy than an ad sponsored by a candidate, and consequently, the ad itself is evaluated more positively. Also, the result s show that for voters who rely on political advertisements, an independen t sponsor is more trustw orthy than a party or a candidate sponsor. In a study that attempted to co mpare negative and comparative political advertisements in terms of the cognitive responses their viewers ha ve, Meirick (2002) expose d participants to two political advertisements embedded in a sequence of consumer ads. Meir ick (2002) selected two political ads from each candidate, one negative and one comparative. The results show that participants who saw comparative ads were ab le to produce more counterarguments than did those who saw negative ads. S ource derogations, in contrast, were more common for negative ads than for comparative ads. Although negative ads can generate more s ource derogations, Kaid (1997) found that participants were significantly more likely to say they intended to vote for the candidate in the condition with a distorted version3 of the advertisement than in the condition with the undistorted 3 A distorted political advertising includes computerized alterations, slow motion, superimpositions, misleading editing or sequence, and other techniques that might lead viewers to false conclusions (Kaid, 1997).

PAGE 20

20 version of the ad. Conversely, positive image ad s produced a greater likelihood of voting for the sponsor than negative ads in a study that sought to determine wh ether the type of TV program surrounding the political advertising influences th e effectiveness of thes e ads (Kaid, Chanslor and Hovind 1992). Overall, the results of these studies suggest that negative pol itical advertisements influence the way people would vote for a candidate and the way people absorb the ad itself. Yet, the findings of these studies do not lead to a one -way direction since th ey utilize different methodologies and different kinds of stimuli. On the one hand, negative political ads may be very beneficial to the sponsor of the ad, as it was found in the st udies described earlier. On the other hand, the unintended effects may emerge, a nd the target of the ad may benefit from it. Therefore, it is difficult to affirm that negative po litical advertisements are harmful or beneficial to either the sponsor or the ta rget of the ad. For example, Ansolabehere and Iyengar (1995) suggest that the negativity associat ed with attack ads may turn voters off from the polls or worse than that, they may turn voters off from voting at all. In their study, exposure to negative ads reduced voter turnout by 5%. It is interesting to note, however , that other studies have found little relationship between exposure to the ads and intention to vote (Garramone et al. 1990; Martinez and Delegal 1990). In sum, how would a candidate be evaluate d if he decided to air the same negative advertisement over and over again? What effects would repeated exposure to negative political advertisements have on candidates’ evaluations and vote intention? Examined next is another important issue discussed in the literature: the effect s of message repetition. Effects of Message Repetition Generally speaking, research has shown that repetition of political ads produces higher evaluations of the candidates spons oring the ads (Becker and Doolitt le 1975). Research in social

PAGE 21

21 psychology provides evidence that as exposure to a novel stimulus increases, so does the affective evaluation of that stimulus (Zajonc 1968). This phenomenon is known as the mere exposure effect. According to Zajonc (1968), mere exposure is a condition which “just makes the given stimulus accessible to the individual’s perception” (Zajonc 1968, p. 1) and the mere exposure effect is defined as the increase in pos itive affect that resu lts from the repeated presentation of previously unfamiliar stimuli (Seamon et al. 1995). Zajonc (1968) was the first one to study the exposure-affect relationship. In Experiment 1, Zajonc exposed participants to a series of nonsense words at frequencies ranging from 0 to 25 exposures. After the exposure pha se, participants ra ted each stimulus word for goodness of meaning, that is, the extent to which the word co nnotes “good” versus “bad” affect. As a result, Zajonc found a positive relations hip between the number of e xposures and the average of goodness rating for a word. Becker a nd Doolittle (1975) utilized pa radigms of the mere exposure effect to study how repetition affects evalua tion of and information seeking about political candidates. In this study, the authors utilized 34 advertisements of six fictitious candidates recorded by a professional broadcaster as stimuli. Participants had to hear the advertisements and not see them. The results of this study indicate that moderate exposure (5 exposures) produces higher affective evaluations of the candidates than does either low (2 exposures) or high exposure (10 exposures). In a ddition, likelihood of informati on seeking is highest at the moderate exposure level (Becker and Doolittle 1975). Moreover, findings in Becker and Doolittle’s (1975) study suggest that affect increases with exposure only to a point, after which it decreases as exposure continues. This finding can be explained by the two-factor theories. A number of theoretical accounts have been pr oposed to explain th e relationship between exposure frequency and affect. Among them ar e classical conditioning, response competition,

PAGE 22

22 optimal arousal, demand artifact s, and two-factor theories (S awyer 1981). Two-factor theories have been considered the most robust explanation to the exposure -affect relationship and are the focus of the following discussion. The Two-Factor Theory The two-factor theory described by Berlyne (1970) proposes an inve rted-U relationship between two opposing factors, ha bituation (reduction of uncertainty or conflict) and tedium. These two factors operate simultaneously, and the st rength of each one is a function of exposure to the stimulus. Initially, the ha bituation factor has greater impact on liki ng than does tedium. Therefore, primarily, repeated exposure to a stim ulus leads to liking but eventually, it leads to disliking. Stang (1973, 1974) offered an extension of Be rlyne’s (1970) model: re peated exposure to a stimulus provides more opportunity to learn about it. At first, this learning is rewarding and leads to increased liking for the stimulus. However, as exposure to the stimulus continues, tedium or satiation develops. Consequently, repeated exposure to a stimulus leads to a negative affect toward the stimulus. In addition, learning reduces uncertainty, that is, as exposure to a stimulus increases, fewer new elements will be learned. As learning decreases, satiation or tedium increases and negative affect toward the stimulus is perceive d (Sawyer, 1981). Berlyne (1970) theorized that novelty and comp lexity stimuli are the elements that would drive the amount of learning (habitu ation) and satiation or tedium at different exposure levels. In other words, the more novel/complex the stimulus is, the greater the opportunity for a positive response. In contrast, familiar/simple stimuli allow people to learn about it quicker and have positive responses toward it with fewer exposures. Consequently, familiar/simple stimuli lead to satiation or tedium more quick ly (Janiszewski and Meyvis 2001).

PAGE 23

23 To determine if the effect of ad exposure a nd ad liking is moderated by the complexity of the advertisement, Cox and Cox (1988) exposed partic ipants to eight print ads of a fictional soft drink product. Four of the ads we re designed to be complex (e.g. a variety of recr eational scenes in which cola might be consumed) and four of them were designed to be simple (e.g. a large glass of cola in the middle of the ad). The authors found that exposur e has a strong positive effect on evaluations of complex advertisements, and only a slight (and n onsignificant) effect on the evaluations of simple ads. The explanation fo r this finding is that “exposure allows subjects to become more familiar with these stimuli, which may reduce excess uncertainty and increase evaluation. Simple advertisements may have less inherent uncertainty, so their evaluations tend to benefit less exposure” (Cox and Cox 1988, p. 115). Although political advertisements were used as stimuli to study the exposure-affect relationship (Becker and Doolittle 1975), there is no empirical evidence of which effects repeated exposure to negative political advertisements can pr ovoke on candidate evaluations and likelihood of voting. The use of negative stimuli to study the exposure-affect relationship was encountered in the social ps ychology literature, and the resu lts presented are mixed. A study conducted by Brickman et al. (1972) exposed participants to abstract paintings that were prerated as either positive (appealing), neutral or negative (unappealing). Each painting was exposed to a participant from 1 to 10 times and each exposure lasted three seconds. The results show that positive affect toward appealing paintings increased with exposure and unappealing paintings became less liked with exposure. Another study that presents negative stimuli was conducted by Klinger and Greenwald (1994) and mixed ev idence was found. In Experiment 1 and 2, participants were exposed to irregular octagons five times each. In order to make the octagons subliminal, stimuli were presented for 16.7 msec to the participant’s left eye while a bright

PAGE 24

24 pattern of black and white blocks were presented to the participant’s righ t eye. Octagons were first rated on attractiveness by pilot participants. After the ratings, octagons were categorized as low (negative), medium (neutral), and high (positive) in attractiv eness. The results show that repeated exposure decreased liking for unattra ctive objects (Klinger a nd Greenwald 1994). In Experiment 3, an additional condition was incl uded in which octagons were presented for 16.7 msec but were not masked, so participants coul d visibly see the stimulus and remember them. The results of Experiment 3 show that expos ure increased liking for both attractive and unattractive octagons. The difference between Brickman et al. ( 1972) study and Klinger and Greenwald’s (1994) first two experiments is that in the former one , the stimuli were always highly detectable. In other words, the authors did not use subliminal st rategies to mask the stimulus. It seems clear that exposure decreases affect toward negative (unattractive or unappealing) stimuli. However, the conditions that were presented in the stud ies were not the same. Klinger and Greenwald (1994) did not mask the stimuli in Experiment 3 and, as a result, exposure increased liking to negative stimuli. Therefore, considering these findings, negative political advertisements may receive positive evaluations as exposure increases . Also, as it was previously stated, negative information is considered more credible th an positive information (Hamilton and Zanna 1972; Leventhal and Singer 1964), which could infl uence the likeability of negative political advertisements and candidates. The two-factor cognitive response model of expos ure effects is similar to the two-factor theory developed by Berlyne (1970). The same i nverted-U curve is pr edicted, however, this version of the two-factor theo ry assumes that the attitudes toward a repeated message are mediated by the elaborations (i.e., counterar guments, support arguments) that viewers produce

PAGE 25

25 when exposed to them. Cacciopo and Petty (1979, 1980) suggest that repeated exposure to a message acts primarily as an opportunity for at tending to, thinking about, and elaborating about the message arguments, that is, learning about th e message. At high levels of exposure to the message, reactance or tedium becomes more salient, which focus viewers’ attention on counterarguing. Thus, Cacciopo and Petty (1979) linke d the notion of learning or habituation to an increase in support arguments the message evoke s. They also linked the notion of tedium or reactance to a decrease of s upport arguments and the increase of counterarguments that the message evokes. Of importance to the current research, st udies using communicative stimuli have demonstrated that repeated exposure lead first to increasing, then to decr easing affect toward the advocacy (Cacciopo and Petty 1979; Miller 1976). M iller (1976) used a one-foot-by-two-foot poster of a campaign to reduce foreign aid and also postcards that were attached to the poster as an invitation to volunteer for that campaign. The participants were rando mly assigned to one of four treatment conditions: no exposure (pretest), moderate exposure, overexposure, and exposure removal (delayed post-test). M iller (1976) found that positive attitudes towa rd reducing foreign aid were significantly enhanced by moderate exposu re to the stimulus poster as compared to the pretest evaluation. In addition, he found that ove rexposure reduced positive attitudes toward the campaign compared to moderate exposure. In the delayed post-test condi tion, attitudes were not significantly different from those in the mode rate and overexposure condition, but were more significant that those in the pr etest condition. Cacciopo and Petty (1979, Experiment 1) exposed participants to a communication message either ze ro, one, three or five times. Participants had to hear the message, rate their degree of agreement with the message, and list the message arguments they would recall. In Experiment 2, part icipants heard the message either one, three or

PAGE 26

26 five times, rated their degree of agreement w ith the message, listed their thoughts about the message, and listed the arguments that they could r ecall. The results indicate that agreement first increased, then decreased as exposure increas ed. In regard to the cognitive responses, counterargumentation first decreased, and then increased as exposure increased. Considering these findings, Cacciopo and Petty (1980, p. 100) suggested that “individuals change the way they think about the advocacy and its arguments as exposure frequency increases. It seems entirely possible that, up to a moderate number of presentations, the attitudinal effects of an advert isement are affected largely by limitations in the extent to which the content of the message is processed.” Ther efore, the effects of repeated exposure to a communicative stimulus might be explai ned by a cognitive response approach. Although these studies have shown that expos ure to persuasive communications first increases, and then decreases with increased expo sure, it is interesting to note how the stimuli were presented. For instance, Cacciopo and Petty (1979, 1980) have exposed participants to the advertising or persuasive message in a massive ma nner (i.e., a sequential presentation of stimulus can be characterized as massive). As a cons equence, positive affect was observed under moderate levels of exposure, a nd negative affect was observed unde r higher levels of exposure. Belch (1982) exposed participants to a onehour TV program (adver tisements, therefore, were spaced) with one, three or five insertions of a toothpaste ad. Participants were asked to answer questions about message acceptance (purch ase intention and attitude toward using the toothpaste) and the results did not present an inverted-U curv e as it was predicted. Neither purchase intention nor attitudes were affected by repetition. In addition, the results found for the cognitive response measures were not in agreement with the two-factor cognitive response

PAGE 27

27 model. In this study, the negativ e thoughts increased across all levels of repetition and the positive thoughts remained constant. In a study using a new product advertising co ntext, Rethans, Swasy, and Marks (1986) exposed participants to a 40-minute TV program with a never-before-shown professionally produced TV advertisement. Participants were exposed one, three or five times to the advertisement. The results indi cate that repeated e xposure to the ad increased participant’s familiarity and recall of the advertisement and the new product. Overall, these findings are in agreement with the first factor in the two-fact or theory: learning. The results also presented evidence of the second factor in the two-factor theory: reactance/tedi um. Counterargumentation was more salient as exposure in creased and participant’s attitude s toward the ad also became more negative. However, the re sults of this study did not find support for the hypothesized curvilinear relationship since th e repetition effect on attitudes was not observed. Rethans, Swasy, and Marks (1986) did find an increase in learni ng under moderate levels of exposure, however, this increase in learning did not turn into increased affect. Although a number of studies using product advertising have been conducted in this area and have found distinct results in regard to repetition of messages (Anand and Sternthal 1990; Belch 1982; Nordhielm 2002; Rethans, Swasy, a nd Marks 1986), the effects of repeated exposure to negative political advertising have received little atten tion. It is not known how a repeated negative TV ad affects the evaluation of the sponsor and the target of the ad and the likelihood of voting for the sponsor and the target . This research inves tigated the effects of repeated exposure to negative political advertis ements on candidate eval uations and likelihood of voting for sponsor and target of the ad.

PAGE 28

28 Research Hypotheses Based on the literature review, this study explores the following hypotheses: H1: Under moderate repeated exposure to a negative political advertisement, H1a: the difference between the evaluation of the sponsor and the target of the ad will be greater than low and high exposure to the ad. H1b: the evaluation of the sponsor candidate of a negative ad will be significantly higher than under low or high exposure to the ad. H1c: the evaluation of the target candidate of a negative ad will be significantly lower than under low or high exposure to the ad. H2: Under moderate repeated exposure to a negative political advertisement, H2a: the difference between the likelihood of voting for the sponsor and the target of the ad will be greater than low and high exposure to the ad. H2b: the likelihood of voting for th e sponsor of a negative ad will be significantly greater than under low or high exposure to the ad. H2c: the likelihood of voting for the target of a negative ad will be significantly lesser than under low or high exposure to the ad. H3: Under moderate repeated exposure to a negative political advertisement, H3a: the occurrence of positive thoughts elaborated while watching the ad will increase more than low and high levels of exposure to the advertisement. H3b: the occurrence of negative thoughts elaborated while watching the ad will decrease more than low and high levels of exposure to the advertisement.

PAGE 29

29 CHAPTER 3 STUDY 1 Introduction This experiment sought to investigate the eff ects of repeated exposur e to negative political advertisements on the evaluation of the sponsor a nd the target of the ad and the likelihood of voting for the sponsor and the ta rget. The method employed in this experiment is described below. Method Participants and Design Participants were 27 undergraduat e and graduate students at the University of Florida. The undergraduate students participated in the expe riment in exchange for extra credit and the graduate students participated in the experiment voluntarily. The design of the experiment consisted of a 3 (level of repetition: one, th ree, or five exposures – manipulated betweensubjects) by 2 (candidate type: s ponsor and target of the ad, mani pulated within-subjects) mixed design. Stimuli The target ad for this experiment was chosen after reviewing a number of negative political advertisements used by U.S. candidates in the 2004 U.S. elections. The advertisement chosen for the experiment was a 30-second television ad for a candidate for the U.S. Congress in California, Gabe Castillo.4 The ad does not disclose the stat e or district loca tion in which the candidate was running for office, nor does it di sclose the political party affiliation of the candidate or his opponent, Dan Lungren. The omi ssion of the party aff iliation and the state location of the candidates was an important criterion in selecti ng the ad. Using candidates that 4 Refer to Appendix A for a complete transcription of the advertisement.

PAGE 30

30 are unknown to the participants help to eliminate th e influence of previous beliefs about political parties and candidates’ image. Three other 30second negative political advertisements were included in the set to avoid pa rticipants’ suspicion about the purpose of the study. All negative political advertisements were from candidates running for Congress and li kely to be unknown to participants in Florida. A set with the target ad, the three other nega tive political advertisements, and other product advertisements, used to avoid suspicion, were placed in a PowerPoint presentation. The product advertisements consisted of the following produc ts: car (BMW, Lexus and Lincoln), tennis shoes (New Balance), anti-drug campaign, investment company (Charles Schwab), newspaper (Wall Street Journal), internet provide r (Microsoft, Nortel), professi onal staffing (K-force.com). Three PowerPoint presentations were created with the set of advertisements selected. The first presentation, which corresponds to condition 1, had one insertion of the ta rget political ad along with the other political advertis ements and the product advertis ements. The second presentation, which corresponds to condition 2, had three inserti ons of the target poli tical ad along with the other political ads and the product advertisements . The third presentation, which corresponds to condition 3, had five insertions of the target political ad along with the other political advertisements and the product advertisements. Th ere is no control conditi on in this experiment due to the nature of the study. Since the politi cal candidates are likely to be unknown to the participants, it was not practical to incl ude a condition that has no treatment. Procedure Participants were not randomly assigned to one of the three conditions since the experiment took place during normal class hours. That is, there were three classes available and each class was exposed to one condition. After bei ng seated in the room, participants read and signed the informed consent. Af ter that, the experimenter announced the following cover story:

PAGE 31

31 “For the next 7 (9 or 11) minutes, you will be watching a set of advertisements. After watching the advertisements, you will be asked to fill ou t a questionnaire”. Part icipants watched a 7minute PowerPoint presentation in condition 1, a 9-minute PowerPoint presentation in condition 2, and an 11-minute PowerPoint presentation in c ondition 3. In conditions 2 and 3, the ones that involved repetition, other negativ e political advertisements a nd product advertisements were repeated to avoid participants ’ suspicion about the purpose of the study. Participants in all conditions watched the same negative politic al advertisements and the same product advertisements. After watching the advertisements, participants were asked to fill out a questionnaire. The questionnaire consisted of ques tions involving evaluation of the candidates, likelihood of voting for the candidates, and recall of the advertisement. Several fi ller questions were inserted throughout the questionnaire to avoid participants’ suspicion about the purpose of the study. After completing the questionnaire, participants were thanked by their participation and the experimenter left the room. Dependent Variable Measurement Twelve bipolar adjectives were used to assess participants ’ evaluation of the sponsor and the target of ad. The bipolar adjectives were presented as seven-point scales: unqualifiedqualified, bad-good, dishonest-honest, believabl e-unbelievable, unsuccessful-successful, unpleasant-pleasant, attractive-unatt ractive, unlikable-likable, insi ncere-sincere, calm-excitable, aggressive-unaggressive, and strong -weak. This semantic differential scale is similar to that developed and described by Kaid (2 004) and has generally produced high levels of reliability. That was also true in this study. The Cronbach ’s alpha for the scale was .86 when used to measure responses to the sponsor candidate and .91 when used to measure responses to the target candidate.

PAGE 32

32 Three items used in previous political adve rtising research were selected to measure likelihood of voting for the sponsor and the target of the ad. Specifically, Kaid (1997) items were employed, asking participants to indicate their agreement or disa greement with the statements. Scales ranged from 1-totally agr ee to 7-totally disagree. The Cr onbach’s alpha for the scale was .86 when used to measure responses to the sponsor candidate and .83 when used to measure responses to the target candidate. Recall was assessed by asking participants to list the specific points they remembered about the ad with Gabe Castillo. Data Analysis The Statistical Package for Social Sciences (SPSS 11 for Windows) was employed in this study for statistical analysis. To test the resear ch hypotheses, repeated measures analysis of variance and univariate analysis of variance were performed. Analysis and Results Study 1 investigated whether or not under moderate repeated exposure to a negative political advertisement: the difference between the eval uation of the sponsor and the target of the ad and the difference between the likelihood of voting for th e sponsor and the target of the ad would be greater than under low a nd high exposure to the ad the evaluation and the likelihood of voting for th e sponsor of the ad would be greater than low and high exposure to the ad the evaluation and the likelihood of voting for the target of the ad woul d be lower than low and high exposure to the ad. Candidate Evaluation The means for each cell on the candidate evalua tion measure are displayed in Table 3-3. A 3 x 2 repeated measures ANOVA was conducted wher e level of repetition of the advertisement (1, 3, or 5 times) was the between-subjects factor a nd candidate type (sponsor vs. target of the

PAGE 33

33 ad) was the within-subjects factor . Evaluation of the sponsor and the target of the ad was the dependent variable. The analys is did not yield a significant interaction between level of repetition and candidate type (F (2, 19) = 2.968, p < .08). However, a significant main effect of candidate type on candidate evaluation was observed (F (1, 19) = 6.825, p < .02). Pairwise comparisons indicated that the diffe rence between the evaluation of the sponsor and the target of the ad under moderate e xposure (three exposure s) was significant (Mmod-diff = 1.30, p < .01). Conversely, this difference under low (one exposure) and high (five exposures) was not significant (Mlow-diff = .10, p > .80 and Mhigh-diff = .36, p > .30, respectively) (see Table 33). Therefore, the difference between the evaluati on of the sponsor and the target of the ad was greater when participants were exposed three tim es to the advertisement than when they were exposed one or five times to the a d. These results support Hypothesis 1a. Hypothesis 1b predicted that unde r moderate repeated exposu re to a negative political advertisement, the evaluation of the sponsor candidate of a nega tive ad would be significantly higher than under low or high exposure to the ad. To test this hypot hesis, a one-way ANOVA was performed. As predicted, level of repetiti on produced a significant main effect on the evaluation of the sponsor candidate (F (2, 24) = 4.404, p < .03) (see Table 3-2). Moderate exposure to the negative ad produ ced the highest evaluation for th e sponsor candidate. As Table 3-1 shows, when exposed to the negative ad only once, participants gave the sponsor candidate a composite mean rating of 3.84. This rating rose significantly to 4.89 under moderate exposure (t = -2.576, df = 16, p < .03). However, when participan ts were exposed to the ad five times in the high exposure condition, the evaluation of th e sponsor candidate declined to 4.46. This difference, however, was not si gnificantly different from the moderate exposure level (t = 1.211, df = 18, p > .20). Therefore, Hypot hesis 1b is partially supported.

PAGE 34

34 Hypothesis 1c predicted that under moderate repeated exposure to a negative political advertisement, evaluation of the target candidate would be significantly lower than under low or high exposure to the ad. To test this hypothesis, a one-way ANOVA was performed. Level of repetition did not produce a significant main effe ct on target candidate evaluation (F (2, 19) = .404, p > .60) (see Table 3-2). As Table 3-1 shows, when exposed to the ne gative ad only once, participants gave the target candidate a compos ite mean rating of 3.76. Th is rating declined to 3.74 in the moderate exposure condition, but the di fference was not significant (t = .052, df = 13, p > .90). Similarly, when participants were expo sed to the ad five times in the high exposure condition, the evaluation of the target candidate rose to 4.07, but this difference was not significantly different from the moderate exposur e level (t = .829, df = 15, p > .40). Hence, these results do not support Hypothesis 1c. Likelihood of Voting The means for each cell on the likelihood of vo ting measure are presented in Table 3-3. A 3 x 2 repeated measures ANOVA was conducted wher e level of repetition of the advertisement (1, 3, or 5 times) was the between-subjects factor a nd candidate type (sponsor vs. target of the ad) was the within-subjects factor . Likelihood of voting for the spons or and the target of the ad was the dependent variable. The analysis did not yield a significant intera ction between level of repetition and candidate type (F (2, 22) = 2.537, p > .10). No significant main effect of candidate type on likelihood of voting was obser ved (F (1, 22) = 2.365, p > .10). Pairwise comparisons indicated that the diffe rence between the likelihood of voting for the sponsor and the target of the ad unde r three exposures was significant (Mmod-diff = 1.12, p < .01). Conversely, this difference under one exposure and five exposures was not significant (Mlow-diff = .16, p > .70 and Mhigh-diff = .20, p > .60, respectively) (see Tabl e 3-3). Therefore, the difference between the likelihood of voting for the sponsor and the target of the ad was greater when

PAGE 35

35 participants were exposed three times to the adve rtisement than when they were exposed one and five times to the ad. Hence, thes e results support Hypothesis 2a. Hypothesis 2b predicted that unde r moderate repeated exposu re to a negative political advertisement, likelihood of voting for the spons or would be significantly greater than under low or high exposure to the ad. To test this hypothesis, a one-way ANOVA was conducted. As predicted, level of repetition pr oduced a significant main effect on likelihood of voting for the sponsor of the ad (F (2, 24) = 3.446, p < .05) (see Table 3-2). Moderate exposure to the negative ad produced the highest likelihood of voting for the sponsor candi date. As Table 3-1 shows, when exposed to the negative ad only once, part icipants gave the sponsor candidate a composite mean rating of 3.47. This rating rose significan tly to 4.63 under moderate exposure (t = 2.086, df = 16, p = .053). However, when participants were exposed to the ad five times in the high exposure condition, the likelihood of voting for the sponsor candidate declined to 3.40, and this difference was significantly different from the moderate exposure level (t = 2.156, df = 18, p < .05). Therefore, these resu lts support Hypothesis 2b. Hypothesis 2c predicted that under moderate repeated exposure to a negative political advertisement, likelihood of voti ng for the target would be signif icantly lesser than under low or high exposure to the ad. To te st this hypothesis, a one-way ANOVA was performed. Level of repetition did not produced a signif icant main effect on likelihood of voting for the target of the ad (F (2, 22) = .165, p > .80) (see Table 3-2). As Table 3-1 shows, when exposed to the negative ad only once, participants gave the target candi date a composite mean ra ting of 3.72. This rating declined to 3.51 in the moderate exposure, but this difference did not yield a signif icant result (t = .648, df = 15, p > .50). When participants were exposed to the ad five times in the high exposure condition, the likelihood of voting for the target candidate was 3.50, and this difference

PAGE 36

36 was not significantly different from the mode rate exposure level (t = .038, df = 17, p > .90). Therefore, Hypothesis 2c was not supported. Additional Measures An additional measure of advertising recall was investigated in Study 1 and no hypotheses were elaborated for this measure. An ANO VA was conducted to explore the relationship between level of repetition and advertising recall. A significant main effect of repetition on recall was observed (F (2, 24) = 4.043, p < .05). As Table 3-1 shows, when exposed to the negative ad only once, participants recalled a mean of 0.71 it ems. This mean rose significantly to 3.72 under moderate exposure (t = 2.905, df = 16, p < .02). Ho wever, when participants were exposed to the ad five times in the high exposure condition, the mean of items recal led declined to 3.22, and this difference was not significan tly different from the moderate exposure level (t = .473, df = 18, p > .60). Thus, under moderate repeated exposu re to a negative political advertisement, advertising recall increases signi ficantly. However, it is not possi ble to affirm that under higher levels of exposure, adver tising recall decreases. Discussion The results of Study 1 are in pa rtial agreement with the twofactor theory anticipated by Cacciopo and Petty (1979). Cacciopo and Petty (1 979; Experiment 1) found evidence that a persuasive appeal repeated in a massive manner leads to an increase and then a decrease in agreement with the message. In Study 1, the adve rtisements were shown in a massive manner as well. However, the dependent variable was not message acceptance, but candidate evaluation and likelihood of voting for the candidates. The results of the global analysis lead to the conclusion that under moderate repeated exposure to a negative political advertisement, the difference between the evaluation of the sponsor and the target of the ad is greater than under low and high levels of exposures. The same

PAGE 37

37 conclusion can be applied to the likelihood of vo ting for the candidate variable. In addition, the curvilinear relationship presente d by the results of Study 1 is in harmony with the two-factor theory designed by Berlyne (1970) and extende d by Stang (1973, 1975). At first, repeated exposure to a stimulus provides more opportunity fo r learning about it, and consequently it leads to positive affect toward the stimulus. However, under higher levels of exposure, learning stops and tedium or satiation de velops, causing negative affect toward the stimulus. Although some of the results were not signifi cant, the means in the candidate evaluation measure (see Table 3-3) reveal an interesting pe rspective. The literature on negative political advertisement predicts that the sponsor of a negative ad might suffer backlash effects when airing this type of ad. Even though the sponsor of th e ad is denigrating the image of the target of the ad, which may not be well regarded by viewers, his evaluation is high er than the evaluation of the target of the ad acro ss all levels of re petition. In addition, although the sponsor’s evaluation decreases and the targ et’s evaluation increases with high levels of exposure, the evaluation of the sponsor of the a dvertisement is still higher than the evaluation of the target of the ad (see Figure 3-1). In other words, the main purpose of a negative political advertisement is being accomplished: lowering the target’s evaluati on without putting at risk the evaluation of the sponsor of the ad. These evaluations, in a political race perspec tive, could be translated into likelihood of voting for the candidates. The means of the like lihood of voting measure (s ee Figure 3-2) reveal that the sponsor of the ad received greater vote in tention with three and fi ve exposures to the ad. The target of the ad, on the contrary, received a lower vote intention with three exposures and this tendency was maintained with five exposures to the ad. Conversely, with one exposure, the sponsor of the ad received lower vote intention than the target of th e ad. Although only under the

PAGE 38

38 moderate exposure condition the difference between the likelihood of voting for the sponsor and the target was significant, the results of this study may indicate that repetition of a negative political advertisement might enhance the likeli hood of voting for the sponsor of the ad more than the likelihood of voting for the target. The individual analysis of the sponsor of the ad showed that overall , repetition did affect the evaluation of the sponsor candidate. More specifically, under moderate exposure to a negative political advertisement, the evaluation of the sponsor of the ad was higher compared to a single exposure. However, under high levels of exposure, the evaluation of the sponsor of the ad did decrease as compared to moderate le vels of exposure, but this difference was not significant. Therefore, it is po ssible to conclude that as e xposure to a negative political advertisement increases, the evaluation of the spons or candidate increases, but only up to a point. After five exposures, for instance, repetition di d not cause a significant effect on the evaluation of the sponsor of the ad. Hence, the results of th is study suggest that high le vels of repetition of a negative political ad will not increase the evaluation of the sponsor candidate. Likewise, repetition of a nega tive political advertisement affe cted the likelihood of voting for the sponsor of the ad. The individual analysis revealed that under moderate repeated exposure to the ad, the likelihood of voting for the sponsor of the ad wa s significantly different from the low and high levels of exposure. As a result, it is possible to conclude that as exposure to a negative political advertisement increases, so does the likelihood of voting for the sponsor of the ad. However, under high levels of exposure, for instance, the likelihood of voting for the sponsor decreased. These results are in agreement with th e two-factor theory. Cons equently, repetition of an ad can be beneficial only up to a point. After overexposure to the ad, people start losing interest in it, and tedium or reactance develops.

PAGE 39

39 Repetition of a negative political advertisement di d not affect the evaluation of the target of the ad. The individual analysis of the target of the ad showed th at under moderate exposure to a negative political advertisement, the evaluation of the target of the ad was lower compared to a single exposure and five exposur es. However, the differences between low – moderate and moderate – high exposures were not significantly different from each other. The same pattern was observed in the voting likeli hood variable. Therefore, the results of this study cannot confirm that repeated exposure to a negative political advertisem ent lowers target’s evaluation and likelihood of voting for him. Further rese arch should be conducte d to investigate the relationship between evaluation of the target of a negative political ad and repetition. In sum, the results of Study 1 encouraged th e development of a study with more external validity. Study 1 was more exploratory in natu re, serving as evidence that repetition of communicative stimuli, especially negative politi cal advertisements, might influence candidate evaluation and likelihood of voting for the candidates. Study 2 is an attempt to investigate the same variables as those of Study 1, but under cond itions that are more aligned with real life situations.

PAGE 40

40 Table 3-1. Descriptive statis tics for candidate evaluation, vote likelihood and recall. Individual analysis for sponsor and target of the advertisement Level of Repetition Low Moderate High Dependent Variable Object M (SE) M (SE) M (SE) Evaluation Sponsor 3.84*(.191)4.89*(.298)4.46 (.143) Target 3.76 (.216)3.74 (.316)4.07 (.131) Vote Likelihood Sponsor 3.47*(.298)4.63*(.398)3.40*(.399) Target 3.72 (.218)3.51 (.202)3.50 (.377) Recall Ad .71*(.856)3.72*(.683)3.22 (.755) Note: * indicates significance at the .05 level. N = 27

PAGE 41

41 Table 3-2. One-way ANOVA results fo r candidate evaluation and vote likelihood. Individual analysis fo r sponsor and target of the advertisement Dependent Variable Candidate df MS F p Evaluation Sponsor 2 2.354 4.404 .023 Target 2 .248 .440 .651 Vote Likelihood Sponsor 2 4.694 3.446 .048 Target 2 .104 .165 .849 Note: N = 27

PAGE 42

42 Table 3-3. Descriptive sta tistics for candidate eval uation and vote likelihood. Global analysis for sponsor and target of the advertisement Level of Repetition Low Moderate High Dependent Variable Object M (SE) M (SE) M (SE) Evaluation Sponsor 3.86 (.314)5.04*(.222)4.44 (.265) Target 3.76 (.336)3.74*(.237)4.07 (.284) Difference .10 (.455)1.30 (.322).36 (.385) Vote Likelihood Sponsor 3.55 (.443)4.63*(.327)3.70 (.384) Target 3.72 (.325)3.51*(.240)3.50 (.282) Difference .16 (.499)1.12 (.369).20 (.432) Note: * indicates that difference between the sponsor and the target is significant at the .05 level. The means in the global analysis are slig htly different from those of the individual analysis due to missing values. N = 27

PAGE 43

43 Figure 3-1.Composite means for the evaluation of th e sponsor and the target of the advertisement and difference between the ev aluations of the candidates 3.86 5.04 4.44 3.76 3.74 4.07 3 3.5 4 4.5 5 5.5 135 Level of Repetition Sponsor Target 0.1 1.3 0.37 0 0.2 0.4 0.6 0.8 1 1.2 1.4 135 Level of Repetition Difference

PAGE 44

44 Figure 3-2.Composite means for likelihood of voti ng for the sponsor and the target of the advertisement and difference between the likelihood of voting for the candidates 3.55 4.63 3.7 3.72 3.51 3.5 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 135 Level of Repetition Sponsor Target -0.17 1.12 0.2 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 135 Level of Repetition Difference

PAGE 45

45 CHAPTER 4 STUDY 2 Introduction This experiment sought to investigate the eff ects of repeated exposur e to negative political advertisements on candidate evaluations and likelihood of voting for the candidates. The objective of this experiment was to provide evid ence that frequent exposure to negative political advertisements affects the way people evaluate po litical candidates and al so vote intentions. The method employed in this experiment is described below. Method Participants and Design Participants were 81 male and female underg raduate students who pa rticipated in the experiment in exchange for extra credit. Par ticipants were recruite d from introductory mass communication courses at the Univer sity of Florida. The design of the experiment consisted of a 3 (level of repetition: one, three, or five exposures – manipulated between-subjects) by 2 (candidate type: sponsor and target of the a d, manipulated within-subjects) mixed design. Stimuli The target ad for this experiment was chosen after reviewing a number of negative political advertisements used by U.S. candidates in the 2004 U.S. elections. The advertisement chosen for the experiment was a 30-second television ad fo r a candidate for the U.S. Congress in Texas, Pete Sessions5. The ad does not disclose the state or di strict location in which the candidate is running for office, nor does it disc lose the political party affili ation of the candidate or his opponent, Martin Frost. The omission of the pa rty affiliation and the state location of the candidates was an important crite rion in selecting the ad. Usi ng candidates that are unknown to 5 Refer to Appendix A for a complete transcription of the advertisement.

PAGE 46

46 the participants help to elimin ate the influence of previous be liefs about political parties and candidates’ image. The prime-time TV show “W ill & Grace” was utilized as the TV program watched by the participants. The target political advertisement was embedd ed within the regular commercial breaks of the TV program. Three similar DVDs were created with the “Will & Grace” TV show. The first DVD, which corresponds to condition 1, had one insertion of the target ad within the first commercial break. More specifically, the target ad was the first advertisement to appear within the first commercial break. The second DVD, which corresponds to condition 2, had th ree insertions of the target ad throughout the commercial breaks. More specifically, the first insertion of the target ad was the first advertisement to appear within the first commercial break; the second insertion was the fourth advertisement within the second co mmercial break; and the third insertion was the first advertisement to appear w ithin the third and last commer cial break. The third DVD, which corresponds to condition 3, had five insertions of the target ad throughout the commercial breaks. More specifically, the first insertion of the target ad was the first advertisement to appear within the first commercial break; the second inserti on was the fourth advertisement within the second commercial break; the third inser tion was the seventh advertisemen t to appear within the second commercial break; the fourth inser tion was the twelfth advertisement to appear within the second commercial break; and the fifth and last insertion wa s the first advertisement to appear within the third and last commercial break. With the inserti ons of the target ad, all participants in all conditions have watched the political ad as the fi rst ad within the first commercial break. There is no control condition in this experiment due to the nature of the study. Since the political candidates are unknown to the participants, it is no t justifiable to include a condition that has no

PAGE 47

47 treatment, as participan ts in the control condition would not be able to rate either sponsor or target at all. This experiment was elaborated to simulate a real world condition. The use of a real TV program and political advertisement provide a high level of external validity to the experiment. Even though the setting in which the experiment took place does not correspond to a “real living room”, the TV program and political advertisem ent definitely emulate a real TV session. Procedure Participants were randomly assigned to one of the three experimental conditions. After being seated in the experimental room, participan ts read and signed the in formed consent. After that, the experimenter announced the following cove r story: “For the next 30 minutes, you will be watching a TV program. The TV program wa s recorded from a normal weekday prime-time hour. After watching the program, you will be asked to fill out a questionnaire”. Participants watched a 30-minute TV program (“Will and Grace”). The target ad was embedded in the TV program in the regular commercials breaks. A ll conditions watched the same TV program and the same negative political advertisement. Th e only difference between conditions was the number of times the target ad was repeated. After watching the TV program, participants were asked to complete a cognitive response task. Participants were asked to list whatever thoughts, ideas, or reactions they experienced while watching the target adver tisement. Participants had 5 minutes to complete the cognitive response task. After completing the cognitive response task, participants were asked to fill out a questionnaire. The questionnaire consisted of que stions involving evaluati on of the candidates, likelihood of voting for the candidate s, attitude toward the ad, a nd recall of the advertisement. Several filler questions were inserted thr oughout the questionnaire to avoid participants’

PAGE 48

48 suspicion about the purp ose of the study. After completing th e questionnaire, the experimenter thanked everyone and the participants were dismissed. Dependent Variable Measurement Twelve bipolar adjectives were used to assess participants ’ evaluation of the sponsor and the target of ad. The bipolar adjectives were presented as seven-point scales: unqualifiedqualified, bad-good, dishonest-honest, believabl e-unbelievable, unsuccessful-successful, unpleasant-pleasant, attractive-unatt ractive, unlikable-likable, in sincere-sincere, trustworthyuntrustworthy, strong-weak, unreas onable-reasonable. Like the scale used in Study 1, this semantic differential scale is similar to that developed and described by Kaid (2004) and has generally produced high levels of reliability. That was also true in this study. The Cronbach’s alpha for the scale was .91 when used to meas ure responses to the s ponsor candidate and .86 when used to measure respons es to the target candidate. Five items used in previous political adve rtising research were selected to measure likelihood of voting for the sponsor and the target of the ad). Specifically, Kaid (1997) items were employed, asking participants to indicat e their agreement or disagreement with the statements. Scales ranged from 1-totally agree to 7-totally disagree. The Cronbach’s alpha for the scale was .92 when used to measure responses to the sponsor candidate and .91 when used to measure responses to the target candidate. A task asking participants to list all the thoughts, ideas or re actions they experienced while watching the Pete Sessions/Martin Frost advertisement was used to assess participants’ cognitive elaborations about the advertisement and candida tes. Thoughts counted as positive thoughts were statements in favor of the advertisement or the candidates and statements that supported the message of the advertisement. Thoughts counted as negative thoughts were statements directed against the advertisement or the candidates and st atements that mentioned specific unfavorable

PAGE 49

49 consequences of the advertisement or candidate s. All other items were counted as neutral thoughts or irrelevant, such as a description of the advertisement or candidate without attribution of valence. Similar thoughts were counted as one thought. Six items used in consumer research were se lected to measure attitude toward the ad (Bruner and Hensel 1996; Zinkhan, William and Leigh 1986).The question asked how well the words believable, interesting, irritating, liked the ad, enjoye d the ad, worth remembering described the advertisement. Recall was assessed by asking participants to list the specific points they remembered about the ad with Pete Sessions/ Martin Frost. Data Analysis The Statistical Package for Social Sciences (SPSS 11 for Windows) was employed in this study for statistical analysis. To test the resear ch hypotheses, repeated measures analysis of variance and univariate analysis of variance were performed. Analysis and Results Study 2 investigated the same relationship as that of Study 1 with a few differences. Instead of using a massive presentation of adve rtisements, Study 2 utilized a real TV Program with insertions of the political advertisement, wh ich characterized the presentation of the ad as spaced. The purpose of this experiment, therefore, was to verify whether or not under moderate repeated exposure to a negative political advertisement: the difference between the eval uation of the sponsor and the target of the ad and the difference between the likelihood of voting for th e sponsor and the target of the ad would be greater than low and high exposure to the ad the evaluation and the likelihood of voting for th e sponsor of the ad would be greater than low and high exposure to the ad the evaluation and the likelihood of voting for the target of the ad woul d be lower than low and high exposure to the ad.

PAGE 50

50 In addition, Study 2 investigated an additiona l hypothesis: whether or not under moderate repeated exposure to a negative political adve rtisement the occurrence of (1) positive thoughts elaborated while watching the advertisement would increase more than under low and high levels of exposure and (2) negative thoughts elab orated while watching the ad would decrease more than under low and high le vels of exposure. Thus, this study focused on the effects that repeated exposure to a negative political adve rtisement generates in a real world condition. Candidate Evaluation The means for each cell on the candidate evalua tion measure are displayed in Table 4-3. A 3 x 2 repeated measures ANOVA was conducted wher e level of repetition of the advertisement (1, 3, or 5 times) was the between-subjects factor a nd candidate type (sponsor vs. target of the ad) was the within-subjects factor . Evaluation of the sponsor and the target of the ad was the dependent variable. The analys is did not yield a significant interaction between level of repetition and candidate type (F (2, 76) = .254, p >.70). However, a significant main effect of candidate type on candidate evaluation was observed (F (1, 76) = 17.205, p = .001). Pairwise comparisons indicated that the diffe rence between the evaluation of the sponsor and the target of the ad under one exposure was not significant (Mlow-diff = .58, p < .09). Conversely, this difference under three exposures and five expos ures approached significance (Mmod-diff = .83, p < .02 and Mhigh-diff = .87, p < .01, respectively) (see Table 4-3). Therefore, the results presented above do not support Hypothesi s 1a. The difference between the evaluation of the sponsor and the target of the ad increased as repetition increased. Hypothesis 1b predicted that unde r moderate repeated exposu re to a negative political advertisement, evaluation of the sponsor candida te would be significantly higher than under low or high exposure to the ad. To test this hypot hesis, a one-way ANOVA wa s performed. Level of repetition did not produce a significant main effe ct on sponsor candidate evaluation (F (2, 78) =

PAGE 51

51 .521, p > .50) (see Table 4-2). As Table 4-1 shows, when exposed to the ne gative ad only once, participants gave the sponsor candidate a compos ite mean rating of 4.07. Th is rating declined to 3.77 under moderate exposure, but the difference was not significant (t = .916, df = 46, p > .30). On the other hand, when participants were exposed to the ad five times in the high exposure condition, the evaluation of the ta rget candidate rose to 4.00. This difference, however, was not significantly different from the moderate exposur e level (t = .755, df = 55, p > .40). Hence, these results do not support Hypothesis 1b. Hypothesis 1c predicted that under moderate repeated exposure to a negative political advertisement, evaluation of the target candidate would be significantly lower than under low or high exposure to the ad. To test this hypothesis, a one-way ANOVA was performed. Level of repetition did not produced a signif icant main effect on target ca ndidate evaluation (F (2, 76) = 2.649, p >.07) (see Table 4-2). Moderate exposure (three times) to the ne gative ad produced the lowest evaluation for the target candidate. As Ta ble 4-1 shows, when exposed to the negative ad only once, participants gave the target candida te a composite mean rating of 3.49. This rating decreased significantly to 2.96 under moderate exposure (t = 2.350, df = 45, p < .03). However, when participants were exposed to the ad five times in the high exposure condition, the evaluations of the target candida te increased to 3.12, but this difference was not significantly different from the moderate exposure level (t = .625, df = 53, p < .50). Therefore, Hypothesis 1c is partially supported. Likelihood of Voting The means for each cell on the likelihood of vo ting measure are displayed in Table 4-3. A 3 x 2 repeated measures ANOVA was conducted wher e level of repetition of the advertisement (1, 3 or 5 times) was the between-subjects factor and candidate type (sponsor vs. target of the ad) was the within-subjects factor. Li kelihood of voting for the sponsor and the target of the ad was

PAGE 52

52 the dependent variable. The analysis did not yi eld a significant interac tion between level of repetition and candidate type (F (2, 78) = .027, p >.90). However, a significant main effect of candidate type on likelihood of voting was observed (F (1, 78) = 17.046, p = .001). Pairwise comparisons indicated that the diffe rence between the likelihood of voting for the sponsor and the target of the ad under one exposure was significant (Mlow-diff = .66, p = .05). Similarly, this difference under three exposures and five exposures approached significance (Mmod-diff = .73, p < .05 and Mhigh-diff = .77, p < .01, respectively) (see Table 4-3). Therefore, the results presented above do not support Hypothesi s 2a. The difference between the likelihood of voting for the sponsor and th e target of the ad increas ed as repetition increased. Hypothesis 2b predicted that unde r moderate repeated exposu re to a negative political advertisement, likelihood of voting for the spons or would be significantly greater than under low or high exposure to the ad. To test this hypot hesis, a one-way ANOVA was conducted. Level of repetition did not produce a significant effect on sponsor candidate eval uation (F (2, 78) = .458, p > .60) (see Table 4-2). As Table 4-1 shows, when exposed to the negative ad only once, participants gave the sponsor candidate a com posite mean rating of 3.64. This rating decreased to 3.30 under moderate exposure, but the differe nce was not significant (t = .892, df = 46, p > .30). On the other hand, when partic ipants were exposed to the ad five times in the high exposure condition, the likelihood of voting for the sponsor candidate rose to 3.53, but this difference was not significantly different from the moderate exposure level (t = .702, df = 55, p > .40). Hence, these results do not support Hypothesis 2b. Hypothesis 2c predicted that under moderate repeated exposure to a negative political advertisement, likelihood of voti ng for the target would be signif icantly lesser than under low or high exposure to the ad. To te st this hypothesis, a one-way ANOVA was performed. Level of

PAGE 53

53 repetition did not produced a signif icant main effect on likelihood of voting for the target of the ad (F (2, 78) = .976, p >.30) (see Table 4-2). As Table 4-1 shows, when exposed to the negative ad only once, participants gave the target candi date a composite mean ra ting of 2.97. This rating decreased to 2.56 under moderate exposure, but the difference wa s not significant (t = 1.364, df = 46, p > .10). Similarly, when participants were exposed to the ad five times in the high exposure condition, the likelihood of voting for th e target candidate rose to 2.76, but this difference was not significantly different from th e moderate exposure level (t = .714, df = 55, p > .40). Hence, these results do not support Hypothesis 2c. Cognitive Responses A total of 495 thoughts were generated in the cognitive response task. Among these thoughts were the positive thoughts (64 favorab le arguments), negative thoughts (284 unfavorable arguments), and neutral thoughts (147 irrelevant arguments). Hypothesis 3a predicted that under moderate repeated exposure to a negative political advertisement the occurrence of positive thoughts while watching the ad would increase more than under low and high levels of exposure. A uni variate analysis of variance was conducted to test this hypothesis. The anal ysis reveled that the minimal increase in positive thoughts across the three levels of repetition was not significant (F (2, 116) = .302, p > .70). As Table 4-1 shows, when exposed to the negative ad only once, part icipants generated a composite mean of 0.44 of positive thoughts. This mean increased to 0.57 under moderate exposure, but the difference was not significant (t = .613, df = 76, p < .50). Similarl y, when participants were exposed to the ad five times in the high exposure condition, a mean of 0.58 of positive thoughts was generated, and this difference was not significan tly different from the moderate exposure level (t = .059, df = 81, p > .90). Therefore, these resu lts do not support Hypothesis 3a.

PAGE 54

54 Hypothesis 3b predicted that unde r moderate repeated exposu re to a negative political advertisement, the occurrence of negative though ts while watching the ad would decrease more than under low and high levels of exposure. To test this hypothesis, a univariate analysis of variance was conducted. The result s indicate a significant main e ffect on negative thoughts (F (2, 116) = 9.736, p < .01). As Table 4-1 shows, when exposed to the negative ad only once, participants generated a compos ite mean of 1.47 of negative t houghts. This mean increased significantly to 2.78 under moderate exposure (t = 3.657, df = 76, p < .001). However, when participants were exposed to the ad five times in the high exposure condition, the mean was maintained, Mhigh = 2.78, and of course, this difference was not significantly different from the moderate exposure level (t = .015, df = 81, p > .90). Hence, these results do not support Hypothesis 3b. As can be seen in Table 4-1, negative t houghts about the advertis ement increased after three exposures. This increase in negative thoughts is not in agreement with previous studies. Cacciopo and Petty (1979; Experiment 2) tested the hypothesis that the effects of message repetition on message acceptance were mediated by a two-factor process. In this process, the repetition of the message arguments would provide more opportunity to think about the message, to understand it clearer and ultimately, to realiz e its favorable implications. On the other hand, counterarguments or negative thoug hts would decrease at moderate levels of repetition (three exposures). However, at higher levels of repeti tion, the favorable thoughts would decrease due to tedium or reactance to the message a nd the counterarguments would increase. Additional Measures Two additional measures, advertising recall and attitude toward the ad, were investigated in Study 2. No hypotheses were elab orated for these measures. A uni variate analysis of variance was conducted to explore the relationship betw een repetition and advertising recall. No

PAGE 55

55 significant main effect was ve rified (F (2, 78) = .865, p > .40). As Table 4-1 shows, when exposed to the negative ad only once, participan ts recalled a mean of 1.22 items. This mean decreased to 1.20 in the moderate exposure, but this difference did not yield a signif icant result (t = .111, df = 46, p > .90). When participants were exposed to the ad five times in the high exposure condition, the mean of items recalled increased to 1.40, but this difference was not significantly different from the moderate exposur e level (t = -1.040, df = 55, p > .20). Thus, there is no evidence that advertising re call increases as exposure to a negative political advertising increases. The means for each cell on the attitude toward the ad measure are displayed in Table 4-1. A univariate analysis of variance was performed to verify the relationship between attitude toward the ad (Aad) and repetiti on. No significant main effect wa s verified (F (2, 78) = .637, p > .50). As Table 4-1 shows, when exposed to the ne gative ad only once, partic ipants gave the ad a composite mean of 2.54. This mean decrea sed to 2.47 under moderate exposure, but this difference was not significant (t = .383, df = 46, p > .70). When part icipants were exposed to the ad five times in the high exposure condition, th e mean decreased to 2.37, and this difference was not significantly different from the moderate exposure level (t = .623, df = 55, p > .50). As a result, there is no evidence that Aad increases under moderate repeated exposure to a negative political advertisement and decreases under high levels of exposure. Discussion The results of Study 2 did not replicate the fi ndings of Study 1. Furtherm ore, the results of this study are not in agreement with the two-factor theory an ticipated by Cacciopo and Petty (1979). The global analysis lead to the conclusion th at under moderate repeated exposure to a negative political advertisement, the difference between the evaluation of the sponsor and the

PAGE 56

56 target of the ad increased as exposure to the advertisement increased. The same conclusion can be applied to the likelihood of voting for the can didate variable. The individual analysis of the sponsor of the ad showed that repetition did not affect the eval uation and the likelihood of voting for the sponsor. Under moderate exposure to a negative political advertisement, the evaluation of the sponsor was lower as compared to a singl e exposure and five exposures. However, the differences between low –moderate and modera te – high exposures were not significantly different from each other. The same pattern was observed in the voting likelihood variable. Therefore, the results of this study cannot confir m that repeated exposure to a negative political advertisement affects the ratings of the sponsor of the advertisement. Similarly, repetition of the negative political advertisement did not affect the evaluation of the target of the ad. The individua l analysis of the target showed that under moderate exposure to a negative political advertisement, the evaluation of the target of the ad was lower as compared to a single exposure and five exposures. However, the only significant result was verified in the difference between the low and moderate exposures. Therefore, if the spon sor attacks the target of the ad and this ad is repeated three times, it is very likely that the ev aluation of the target will decrease. In regard to the voting likelihood va riable, under moderate exposure to a negative political advertisement, the like lihood of voting of the target wa s lower as compared to a single exposure and five exposures. However, the differe nces between low – moderate and moderate – high exposures were not significantly different from each other. Therefore, the results of this study cannot confirm that repeated exposure to a negative political advertisement lowers the likelihood of voting for the target. Even though the results for the ta rget were not significant, the means reveal a tendency in the direction predic ted (see Figure 4-1 and 4-2). In other words,

PAGE 57

57 under moderate exposure to the negative politi cal advertisement, th e evaluation and the likelihood of voting for the target of ad was lo wer than under low and high exposure to the ad. One important difference between Study 2 and Study 1 is that, in Study 2, the advertisements were presented embedded in a TV program. Thus, the presentation is characterized as distributed or spaced. In Study 1, on the contrary, th e presentation of the advertisements was massive. One possible explan ation for the results of Study 2, and especially for the sponsor of the ad, is that with a spaced presentation of advertisements, the liking and satiation toward the ads might ha ve been delayed when compared to a massive presentation of advertisements. Two-factor theory posits th at when the interval between the stimulus presentations is increased, sati ation is reduced (see Janiszewsk i and Meyvis 2001 for evidence of this assertion). In Study 1, the st imulus was presented in a massive manner and the optimal point of the evaluation of the sponsor of the ad was ve rified with three exposur es and decreased after five exposures. On the other hand, Study 2 utilized a spaced presentation of advertisements and a decrease on the evaluations of both candidates was not observed with five exposures. Although the results of Study 2 were not significant, they point to some directions. The increase in the evaluations of the candidates at fi ve exposures to a negative polit ical advertisement as compared to three exposures might be a cue of the optim al point of liking within a spaced presentation. Under higher level of exposure, therefore, tedium or satiation could have been developed. The results of the likelihood of voting variable followed the same pattern as that of the candidate evaluation variable. Hence, it can be assumed that five exposures within a spaced presentation would result in a higher likelihood of voting for the candidates. Massive vs. spaced presentation of advertisements might be an important modera tor when analyzing the in fluence of repetition on candidate evaluation and likelihood of voting. This point will be further discussed on the general

PAGE 58

58 discussion section. In addition, the length of the TV program might have caused the feelings and reactions toward the can didates to fade away. Even though th e political ad was repeated three and five times, the amount of additional information that was being presented might have weakened the effects of the political ad. One possible explanation for the increase in negative thoughts across the three levels of exposure is that multiple exposures within a 30-mi nute program might have resulted in satiation and development of reactance by the participants toward the advertisement. Belch (1982) has attributed the increase in the number of negative thoughts acro ss all levels of exposure to the participant’s negative reactions to message repeti tion, rather than negati ve evaluations of the content of the message.

PAGE 59

59 Table 4-1. Descriptive sta tistics for candidate evalua tion, vote likelihood, recall, attitude toward the ad, and cognitive responses. Individual analysis for sponsor and target of the advertisement Level of Repetition Low Moderate High Dependent Variable Object M (SE) M (SE) M (SE) Evaluation Sponsor 4.07 (.192)3.77 (.263)4.00 (.168) Target 3.49 (.108)2.96*(.198)3.12 (.154) Vote Likelihood Sponsor 3.64 (.263)3.30 (.277)3.53 (.208) Target 2.97 (.181)2.56 (.237)2.76 (.169) Attitude toward the Ad Ad 2.54 (.101)2.47 (.150)2.37 (.079) Recall Ad 1.22 (.124)1.20 (.140)1.40 (.109) Cognitive Responses Positive .44 (.122).57 (.160).58 (.120) Negative 1.47*(.216)2.78*(1.78)2.78 (.202) Note: * indicates significance at the .05 level. N = 81

PAGE 60

60 Table 4-2. One-way ANOVA results fo r candidate evaluation and vote likelihood. Individual analysis fo r sponsor and target of the advertisement Dependent Variable Candidate df MS F p Evaluation Sponsor 2 .592 .521 .596 Target 2 1.74 2.649 .077 Vote Likelihood Sponsor 2 .746 .458 .634 Target 2 1.000 .976 .381 Note: N = 81

PAGE 61

61 Table 4-3. Descriptive sta tistics for candidate eval uation and vote likelihood. Global analysis for sponsor and target of the advertisement Level of Repetition Low Moderate High Dependent Variable Object M (SE) M (SE) M (SE) Evaluation Sponsor 4.07 (.220)3.80*(.225)4.00*(.190) Target 3.49 (.166)2.96*(.169)3.12*(.143) Difference .58 (.330).83 (.338).87 (.286) Vote Likelihood Sponsor 3.64*(.261)3.30*(.261)3.53*(.222) Target 2.97*(.207)2.56*(.207)2.77*(.176) Difference .66 (.339).73 (.339).77 (.289) Note: * indicates that the difference between the sponsor and target is significant at the .05 level. The means in the global analysis are s lightly different from the individual analysis due to missing values. N = 81

PAGE 62

62 Figure 4-1.Composite means for the evaluation of th e sponsor and the target of the advertisement and difference between the ev aluations of the candidates 4.07 3.8 4 3.49 2.96 3.1 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 135 Level of Repetition Sponsor Target 0.58 0.84 0.9 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 135 Level of Repetition Difference

PAGE 63

63 Figure 4-2.Composite means for likelihood of voting for the sponsor and the target of the advertisement and difference between the likelihood voting for the candidates 3.64 3.3 3.53 2.97 2.56 2.77 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 135 Level of Repetition Sponsor Target 0.67 0.74 0.76 0.65 0.67 0.69 0.71 0.73 0.75 0.77 135 Level of Repetition Difference

PAGE 64

64 CHAPTER 5 GENERAL DISCUSSION The goal of this research was to investigate th e effects of repeated exposure to a negative political advertisement (1) on the evaluations of the sponsor and the target of the ad, and (2) on the likelihood of voting for the sponsor and the ta rget of the ad. Two experimental studies were developed to study the effects of repeated e xposure to negative pol itical advertising. The results of Study 1 showed that under m oderate exposure to a negative political advertisement, the difference between the evaluati on of the sponsor and the target of the ad is greater than under low and high exposure to th e ad. Study 1 also showed that under moderate exposure to a negative political advertisement, the difference between the likelihood of voting for the sponsor and the target of the ad is greater than under low and high exposure to the ad. Individual analysis of the sponsor of the ad reve aled that repetition did affect the evaluation and the likelihood of voting for the sponsor candidate . Conversely, repetition of a negative political advertisement did not affect the evaluation and likelihood of voting for th e target candidate. The results of Study 2, on the other hand, do not replicate the results of Study 1. When participants were exposed to a negative political advertisement at a moderate level of exposure within a regular TV program, the difference betw een the evaluation of the sponsor and the target of the ad increased across all levels of exposure. In addition, the difference between the likelihood of voting for the candidates increased acro ss all levels of exposure. Individual analysis for the sponsor and the target of the ad revealed that repetition did not affect the evaluation and likelihood of voting for both candidates. A possible explanation for these distinct re sults might lay on the context in which the advertisements were presented. In Study 1, the advertisements were presented in a massive manner. That is, participants watched all adver tisements in sequence without any TV program as

PAGE 65

65 a background. In Study 2, conversely, participants watched the poli tical ad embedded in a real TV program. The exposure to the political ad was spaced, which might have delayed the achievement of the optimal point at three exposures. These results , therefore, indicate a tendency that the optimal point could have been reached at five exposures. According to D’Souza and Rao (1995), when exposures are forced, massed, and wit hout competition, it is likely that wearout of the stimulus occurs with few exposures. Conve rsely, when exposures ar e voluntary, distributed and include competition, it is likely that w earout occurs after many more exposures. In addition, the length of th e TV program might have cause d the feelings and reactions toward the candidates to fade away. Even though the political ad was re peated three and five times, the amount of additional information that was being presented might have weakened the effects of the political ad. In regard to the increase in negative thought s across the three levels of exposure, one possible explanation is that multiple exposures within a 30-minute program might have resulted in satiation and development of reactance by the participants toward the advertisement. Belch (1982) has attributed the incr ease in the number of negativ e thoughts across a ll levels of exposure to the participant’s ne gative reactions to message re petition, rather than negative evaluations of the content of the message. Implications Although the results of this study failed to suppo rt some of the research hypotheses, they do have implications for polit ical candidates and for the development of negative political advertisements. A particular inte resting finding is that in both studies, the evaluations of the target of the advertisement were always lower than the evaluations of the sponsor of the ad. This finding provides directional evid ence that even though the sponso r of the advertisement attacks the target of the ad, he still receives higher evaluations than his opponent. Therefore, the main

PAGE 66

66 purpose of a negative political advertisement is being accomplished: lower evaluations of the target. The results of this study s uggest that repetition of a ne gative ad within a massive presentation results in more positive evaluations toward the sponsor of the ad than toward the target of ad. However, massively presenting an ad can result in backlash effects much faster than presenting an ad within a TV program (i.e. sp aced presentation). Therefore, candidates should carefully consider how many times they would like to air a specific ad and in which type of presentation they would like to have the ad inserted. Limitations and Future Research This study has several limitations. First, the questionnaire did not have a randomized set of questions. The questions asking about the sponsor of the advertisements were always asked first, followed by the questions about the target of the a d. Therefore, the evaluati ons of the target of the ad might have been influenced by the eval uations given to the s ponsor. Future research should involve two types of questionnaires cont aining the same questions. The only difference between the questionnaires would be the order of the questions, which would help to rule out order effects on the dependent measures. In addi tion, the questionnaire lacks a measure of tedium or boredom. Although in Study 1 an inverted-U curv e was verified, it is no t possible to affirm that the decline on the evaluations and on the like lihood of voting for the candidates is due to the development of tedium or boredom. A pre a nd posttest measurement of tedium would be appropriate to verify how the st imulus affected participants. Second, the experimental groups should be sm aller in size. Having more than twenty participants in a single session propitiates lateral c onversations and consequently, participants tend to pay less attention to the TV program. In addition, the length of the TV program might

PAGE 67

67 have satiated participants. Sitting through a 30-minute TV program without changing the channel, without leaving the room, might ha ve induced boredom fo r the participants. Third, to eliminate the possible influence of pr evious beliefs, advertisements that omitted the candidate’s party affiliation and state which they belong to were selected. This strategy, however, could have contributed to the lack of empathy toward the candidates. Neutral information about the candidates, such as career in politics, ac hievements as a politician, could have been provided to the partic ipants. Thus, after watching th e advertisements, participants would have more information to evaluate the candidates. Moreover, th e use of two different advertisements in the studies might have pr opitiated the differences in the results. The advertisement selected for Study 2 was more comple x in content compared to the ad selected for Study 16. The ad in Study 2 was about taxes, while the ad in Study 1 had a variety of topics mentioned, however none of them was discusse d in depth. In addition, the ad in Study 1 presented the sponsor candidate sp eaking directly to the audience, while the ad in Study 2 did not have this characteristic. This di fference in the presentation of th e sponsor of the ad might have caused less empathy toward the candidate that did not speak directly to the audience, and consequently it might have influenced the results. Additionally, due to the complexity in content, three exposures to an ad might not be the optim al point of liking of an advertisement, and consequently, liking toward the candidates. In Study 2, for instance, the target ad was more complex as compared to the target ad in Study 1. Therefore, complexity of the advertisement might be an important moderator when inve stigating liking of po litical candidates. Additionally, for future researc h, the effects of repeated expo sure to negative and positive political advertisements on the evaluation of th e candidates should be ex plored. A design similar 6 Refer to Appendix A and B for a complete description of the advertisements.

PAGE 68

68 to this study could be employed. Moreover, a valence variation of the advertisements could result in interesting insights to the use of political a dvertisements within a TV program. For instance, what would happen if there were both positive and negative ads about the same candidate within a TV program? Or, what would happen if ther e were both positive and negative ads about two candidates running for the same office? Primacy and recency effects could also provide interesting insights to the use of political advertisements within a TV program. For instance, if the advertisement is exposed earlier or later wi thin the program, how would the evaluations of the candidates be affected? Attitude toward the ad was not investigated in depth in this study. Thus, future research should investigate which effects Aad has on candi date evaluation and on likelihood of voting for the sponsor and the target of the ad. Furthermore, future research should investigate the role of massive and spaced presentations of negative political advertisements. An experiment manipulating these two variable s should be conducted to furt her investigate this topic.

PAGE 69

69 APPENDIX A DESCRIPTION AND TRANSCRIPTION OF THE NEGATIVE POLITICAL ADVERTISEMENTS Advertisement with Gabe Castillo and Dan Lungren Description of the Advertisement The advertisement is sponsored by candidate Ga be Castillo and it had the claim message at the beginning of the ad: “I’m Gabe Castillo and I approve this me ssage”. After the claim, a black screen with a picture of Gabe Castillo’s opponent, Dan Lungren, appeared concomitantly with the attacks to Dan Lungren. The attacks were made by a narrator and after approximately 10 seconds, candidate Gabe Castillo started to talk about the bad things th at “career politicians” have done in Congress. After this , a white screen with Gabe Cas tillo’s picture and the slogan: “Gabe Castillo, New Leadership for a Be tter America” appeared and the ad ended. Transcript of the Advertisement Gabe Castillo: “I’m Gabe Casti llo and I approve this message” Narrator: “Voted against women’s rights, voted against the clean air act, voted against the clean water bill, voted against the safe dri nking water act, voted against public education funding” Gabe Castillo: “My opponent, Dan Lungren, w ho’s a career politician, has called me a rookie. When I take a look how the career politicians in Congre ss have driven up our national debt and wasted our tax dollars, I say, we need more rookies in Washington” Narrator: “Gabe Castillo, new leadership for a better America” Advertisement with Pete Sessions and Martin Frost Description of the Advertisement The ad starts describing Martin Frost as desperate since he is wrongly accusing Pete Sessions (the sponsor of the ad) of voting to increas e taxes. After that, the narrator says that the

PAGE 70

70 truth is Pete Sessions never voted to raise taxes and that the ac cusations and ads backing Martin Frost are false. The narrator also says that these ads were pulled off the air by some TV managers due to their false content. After this, an attack saying that Mart in Frost is the biggest spender in Congress and that he voted for higher taxes 87% of the time is made. Right after this attack, the narrator says that Pete Sessions works to cut taxes. The ad ends with the claim “I’m Pete Sessions and I approve this message”. Transcript of the Advertisement Narrator: “Martin Frost is desperate. Now Fros t is accusing Pete Sessions of raising taxes. Who’s he kidding? Truth is Pete Sessions never voted to raise taxes. Not once. And those ads backing Frost, totally false. So inaccurate some TV station managers pulled them off the air. The true difference: Martin Frost, biggest spender in Congress. For higher taxes 87% of the time. Pete Sessions, he cuts taxes across the board” Candidate: “I’m Pete Sessions and I approve this message”

PAGE 71

71 APPENDIX B QUESTIONNAIRE – STUDY 1 Questionnaire Thank you for your participation! 1. How many times did you see/hear the last name SULLIVAN in the advertisements? one two three four five 2. How many times did you see a car in the advertisements? one two five ten more than eleven 3. Please react to Candidate PETE SESSIONS on each of the scales below: For example, if you think the candidate is very pleasant you would check the UNPLEASANTPLEASANT scale as follows: UNPLEASANT:____:____:____:__ __:____:____:_X_:PLEASANT On the other hand, if you think he is very unpleasant , you would rate him as follows: UNPLEASANT:_X_:____:____:____: ____:____:____:PLEASANT If you think he is some where between these two ex tremes, then you would ch eck the space that best represents your reaction on that sc ale. If you feel that you have no reaction to him on any one scale , please check the middle space num ber to indicate your neutrality. Candidate PETE SESSIONS UNQUALIFIED:____:__ __:____:____:____: ____:____:QUALIFIED BAD:____:____:____:____: ____:____:____:GOOD DISHONEST:____:____:____: ____:____:____: ____:HONEST BELIEVABLE:____:____:____:____ :____:____:____:UNBELIEVABLE UNSUCCESSFUL:____:____:____:__ __:____:____:____:SUCCESSFUL UNPLEASANT:____:____:____:__ __:____:____:___:PLEASANT ATTRACTIVE:____:__ __:____:____:____:__ __:____:UNATTRACTIVE UNLIKABLE:____:____:____:__ __:____:____:____:LIKABLE

PAGE 72

72 INSINCERE:____:____:____:__ __:____:____:____:SINCERE CALM:____:____:____:____:__ __:____:____:EXCITABLE AGGRESSIVE:____:____: ____:____:____:__ __:____:UNAGGRESSIVE STRONG:____:____:____:__ __:____:____:____:WEAK 4. Using the same scales, please react to Candidate GABE CASTILLO : Candidate GABE CASTILLO UNQUALIFIED:____:__ __:____:____:____: ____:____:QUALIFIED BAD:____:____:____:____: ____:____:____:GOOD DISHONEST:____:____:____: ____:____:____: ____:HONEST BELIEVABLE:____:____:____:____ :____:____:____:UNBELIEVABLE UNSUCCESSFUL:____:____:____:__ __:____:____:____:SUCCESSFUL UNPLEASANT:____:____:____:__ __:____:____:___:PLEASANT ATTRACTIVE:____:__ __:____:____:____:__ __:____:UNATTRACTIVE UNLIKABLE:____:____:____:__ __:____:____:____:LIKABLE INSINCERE:____:____:____:__ __:____:____:____:SINCERE CALM:____:____:____:____:__ __:____:____:EXCITABLE AGGRESSIVE:____:____: ____:____:____:__ __:____:UNAGGRESSIVE STRONG:____:____:____:__ __:____:____:____:WEAK 5. Using the same scales, please react to Candidate DAN LUNGREN : Candidate DAN LUNGREN UNQUALIFIED:____:__ __:____:____:____: ____:____:QUALIFIED

PAGE 73

73 BAD:____:____:____:____: ____:____:____:GOOD DISHONEST:____:____:____: ____:____:____: ____:HONEST BELIEVABLE:____:____:____:____ :____:____:____:UNBELIEVABLE UNSUCCESSFUL:____:____:____:__ __:____:____:____:SUCCESSFUL UNPLEASANT:____:____:____:__ __:____:____:___:PLEASANT ATTRACTIVE:____:__ __:____:____:____:__ __:____:UNATTRACTIVE UNLIKABLE:____:____:____:__ __:____:____:____:LIKABLE INSINCERE:____:____:____:__ __:____:____:____:SINCERE CALM:____:____:____:____:__ __:____:____:EXCITABLE AGGRESSIVE:____:____: ____:____:____:__ __:____:UNAGGRESSIVE STRONG:____:____:____:__ __:____:____:____:WEAK 6. Please indicate your agreement or disagreement with the following statements. (1)The ads I saw made me interest ed in buying the product advertised. totally disagree 1 2 3 4 5 6 7 totally agree (2)Buying a LEXUS car woul d be a great acquisition. totally disagree 1 2 3 4 5 6 7 totally agree (3)I would not like to have a LEXUS car. totally disagree 1 2 3 4 5 6 7 totally agree (4)Having a LEXUS car would make me happier. totally disagree 1 2 3 4 5 6 7 totally agree (5)I do not think that a LEXUS car is very expensive: totally disagree 1 2 3 4 5 6 7 totally agree (6)Buying a BMW car would be a great acquisition.

PAGE 74

74 totally disagree 1 2 3 4 5 6 7 totally agree (7)I would not like to have a BMW car. totally disagree 1 2 3 4 5 6 7 totally agree (8)Having a BMW car would make me happier. totally disagree 1 2 3 4 5 6 7 totally agree (9)I do not think that a BMW car is very expensive: totally disagree 1 2 3 4 5 6 7 totally agree (10)The message that I heard from Ca ndidate GABE CASTIL LO was compelling. totally disagree 1 2 3 4 5 6 7 totally agree (11)People should make a strong effort to vote for Candidate GABE CASTILLO. totally disagree 1 2 3 4 5 6 7 totally agree (12)Voting for Candidate GABE CA STILLO is a smart choice. totally disagree 1 2 3 4 5 6 7 totally agree (13)I like Candidate GABE CASTILLO. totally disagree 1 2 3 4 5 6 7 totally agree (14)I think GABE CASTILLO wo uld make a great politician. totally disagree 1 2 3 4 5 6 7 totally agree (15)If it were time to do so, I woul d vote for Candidate GABE CASTILLO. totally disagree 1 2 3 4 5 6 7 totally agree (16)The message from GA BE CASTILLO was dumb. totally disagree 1 2 3 4 5 6 7 totally agree (17)The message from Candida te GABE CASTILLO was weak. totally disagree 1 2 3 4 5 6 7 totally agree

PAGE 75

75 (18)The message that I heard from Candidate DAN LUNGREN was compelling. totally disagree 1 2 3 4 5 6 7 totally agree (19)People should make a strong effort to vote for Candidate DAN LUNGREN. totally disagree 1 2 3 4 5 6 7 totally agree (20)Voting for Candidate DAN LUNGREN is a smart choice. totally disagree 1 2 3 4 5 6 7 totally agree (21)I like Candidate DAN LUNGREN. totally disagree 1 2 3 4 5 6 7 totally agree (22)I think DAN LUNGREN woul d make a great politician. totally disagree 1 2 3 4 5 6 7 totally agree (23)If it were time to do so, I would vote for Candidate DAN LUNGREN. totally disagree 1 2 3 4 5 6 7 totally agree (24)The message from DAN LUNGREN was dumb. totally disagree 1 2 3 4 5 6 7 totally agree (25)The message from Candi date DAN LUNGREN was weak. totally disagree 1 2 3 4 5 6 7 totally agree 7. Please list the specific points you remember about the ad with candidate GABE CASTILLO: 1. 7. 2. 8. 3. 9. 4. 10. 5. 11. 6. 12. 8. Please list the specific points you remember about the ad with candidate PETE SESSIONS: 1. 7. 2. 8.

PAGE 76

76 3. 9. 4. 10. 5. 11. 6. 12. 9. Please list the specific points you remember about the car LEXUS: 1. 7. 2. 8. 3. 9. 4. 10. 5. 11. 6. 12. 10. What do you think was the purpose of this study? 11. My age is _______ years. 2. I am MALE FEMALE 12. Please indicate your ethnicity: African American Asian Caucasian Latin American Middle Eastern Native American Pacific Islander Other_______________________ 13. Which of the following best represents your politi cal beliefs? (circle one) (1) Democrat (2) Republican (3) Independent or ot her __________________

PAGE 77

77 APPENDIX C QUESTIONNAIRE – STUDY 2 Questionnaire Thank you for your participation! 1. How much of the each of the following emotions are you curre ntly experiencing? SAD none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling DREARY none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling DISMAL none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling GLOOMY none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling HAPPY none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling JOYFUL none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling CONTENT none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling CHEERFUL none of this feeling 1 2 3 4 5 6 7 a great deal of this feeling 2. Please react to Candidate PETE SESSIONS on each of the scales below: For example, if you think the candidate is very pleasant you would check the UNPLEASANTPLEASANT scale as follows: UNPLEASANT:____:____:____:__ __:____:____:_X_:PLEASANT On the other hand, if you think he is very unpleasant , you would rate him as follows: UNPLEASANT:_X_:____:____:____: ____:____:____:PLEASANT

PAGE 78

78 If you think he is some where between these two ex tremes, then you would ch eck the space that best represents your reaction on that sc ale. If you feel that you have no reaction to him on any one scale , please check the middle space num ber to indicate your neutrality. Candidate Pete Sessions UNQUALIFIED:____:__ __:____:____:____: ____:____:QUALIFIED BAD:____:____:____:____: ____:____:____:GOOD DISHONEST:____:____:____: ____:____:____: ____:HONEST BELIEVABLE:____:____:____:____ :____:____:____:UNBELIEVABLE UNSUCCESSFUL:____:____:____:__ __:____:____:____:SUCCESSFUL UNPLEASANT:____:____:____:__ __:____:____:___:PLEASANT ATTRACTIVE:____:__ __:____:____:____:__ __:____:UNATTRACTIVE UNLIKABLE:____:____:____:__ __:____:____:____:LIKABLE INSINCERE:____:____:____:__ __:____:____:____:SINCERE TRUSTWORTHY:____:__ __:____:____:____:____:_ ___: UNTRUSTWORTHY STRONG:____:____:____:__ __:____:____:____:WEAK UNREASONABLE: ____:____:____:_ ___:____:____:__ __: REASONABLE 3. Using the same scales, please react to Candidate MARTIN FROST : Candidate Martin Frost UNQUALIFIED:____:__ __:____:____:____: ____:____:QUALIFIED BAD:____:____:____:____: ____:____:____:GOOD DISHONEST:____:____:____: ____:____:____: ____:HONEST BELIEVABLE:____:____:____:____ :____:____:____:UNBELIEVABLE UNSUCCESSFUL:____:____:____:__ __:____:____:____:SUCCESSFUL UNPLEASANT:____:____:____:__ __:____:____:___:PLEASANT

PAGE 79

79 ATTRACTIVE:____:__ __:____:____:____:__ __:____:UNATTRACTIVE UNLIKABLE:____:____:____:__ __:____:____:____:LIKABLE INSINCERE:____:____:____:__ __:____:____:____:SINCERE TRUSTWORTHY:____:__ __:____:____:____:____:_ ___: UNTRUSTWORTHY STRONG:____:____:____:__ __:____:____:____:WEAK UNREASONABLE: ____:____:____:_ ___:____:____:__ __: REASONABLE 4. Please indicate your agreement or disagreement with the following statements. (1)Voting for candidate Pete Sessions is a smart choice. totally disagree 1 2 3 4 5 6 7 totally agree (2)I think Pete Sessions woul d make a great politician. totally disagree 1 2 3 4 5 6 7 totally agree (3)If it were time to do so, I woul d vote for Candidate Pete Sessions. totally disagree 1 2 3 4 5 6 7 totally agree (4)I think voting for candidate Pete Sessions is a good idea. totally disagree 1 2 3 4 5 6 7 totally agree (5)I would vote for candidate Pete Sessions if the election were today. totally disagree 1 2 3 4 5 6 7 totally agree (6)Voting for Candidate Martin Frost is a smart choice. totally disagree 1 2 3 4 5 6 7 totally agree (7)I think Martin Frost w ould make a great politician. totally disagree 1 2 3 4 5 6 7 totally agree (8)If it were time to do so, I would vote for Candidate Martin Frost. totally disagree 1 2 3 4 5 6 7 totally agree

PAGE 80

80 (9)I think voting for candidate Martin Frost is a good idea. totally disagree 1 2 3 4 5 6 7 totally agree (10)I would vote for candidate Martin Frost if the election were today. totally disagree 1 2 3 4 5 6 7 totally agree (11)The sponsor of the political advertis ement I just watched was Pete Sessions. totally disagree 1 2 3 4 5 6 7 totally agree (12)The target of the political advertis ement I just watched was Martin Frost. totally disagree 1 2 3 4 5 6 7 totally agree (13)The message that I heard from Vonage was compelling. totally disagree 1 2 3 4 5 6 7 totally agree (14)I think the service offered by Vonage is expensive. totally disagree 1 2 3 4 5 6 7 totally agree (15)I would not buy Lays Potato chips because it has a lot of calories. totally disagree 1 2 3 4 5 6 7 totally agree (16)The message that I heard from Lays Potato chips was compelling. totally disagree 1 2 3 4 5 6 7 totally agree (17)I think Lays Potato chips is expensive. totally disagree 1 2 3 4 5 6 7 totally agree 5. Please tell us how well you think each of the words listed below describes the ad for PETE SESSIONS/MART IN FROST . Here we are interest ed in your thoughts about the ad, not the candidates. THIS AD IS BELIEVABLE not at all well 1 2 3 4 5 extremely well THIS AD IS INFORMATIVE not at all well 1 2 3 4 5 extremely well

PAGE 81

81 THIS AD IS INTERESTING not at all well 1 2 3 4 5 extremely well THIS AD IS IRRITATING not at all well 1 2 3 4 5 extremely well LIKED THE AD not at all well 1 2 3 4 5 extremely well ENJOYED THE AD not at all well 1 2 3 4 5 extremely well THIS AD IS WORT H REMEMBERING not at all well 1 2 3 4 5 extremely well 6. Please tell us how well yo u think each of the words liste d below describes the ad for VONAGE . Here we are interested in your thoughts about the ad, not the product. THIS AD IS BELIEVABLE not at all well 1 2 3 4 5 extremely well THIS AD IS INFORMATIVE not at all well 1 2 3 4 5 extremely well THIS AD IS INTERESTING not at all well 1 2 3 4 5 extremely well THIS AD IS IRRITATING not at all well 1 2 3 4 5 extremely well LIKED THE AD not at all well 1 2 3 4 5 extremely well ENJOYED THE AD not at all well 1 2 3 4 5 extremely well THIS AD IS WORT H REMEMBERING not at all well 1 2 3 4 5 extremely well 7. How likely would you be to vote for Pete Sessions? not at all likely 1 2 3 4 5 6 7 very likely 8. If you were voting in this el ection, how likely would you be to vote for Pete Sessions? not at all likely 1 2 3 4 5 6 7 very likely

PAGE 82

82 9 . How likely would you be to vote for Martin Frost? not at all likely 1 2 3 4 5 6 7 very likely 10. If you were voting in this election, how lik ely would you be to vote for Martin Frost? not at all likely 1 2 3 4 5 6 7 very likely 11. From watching the ad, how much did you l earn about the personal qualities of candidate PETE SESSIONS ? A great deal 7:___: ___:___:___: ___:___:___:1 Very little 12. From watching the a d, how much di d you learn about the pers onal qualities of candidate MARTIN FROST ? A great deal 7:___: ___:___:___: ___:___:___:1 Very little 13. How much did you learn about the issue po sitions of candidate PETE SESSIONS from the ad you just saw? A great deal 7:___: ___:___:___: ___:___:___:1 Very little 14. How much did you learn about the issue positions of candidate MARTIN FROST from the ad you just saw? A great deal 7:___: ___:___:___: ___:___:___:1 Very little 15. Please give us your f eelings toward Candidates PETE SESSIONS and MARTIN FROST on this feeling thermometer. Ratings between 50 degrees and 100 de grees mean that you feel favorable and warm toward them. Ratings between 0 degrees and 50 mean that you don't feel favorable toward them and that you don't care too much for them. If you don 't feel particularly warm or cold, or you are unfamiliar with these candidates, you would rate them at the 50 degree mark. 0--------------------50 --------------------100 (1) Candidate PETE SESSIONS : _______ degrees Using the same scale, please rate the other candidate: (2) Candidate MARTIN FROST : ________degrees 16. Please list the specific points you remember about the ad with candidate PETE SESSIONS : 17. Please list the specific points you remember about the ad for LAYS POTATO CHIPS :

PAGE 83

83 18. Please list the specific points you remember about candidate MARTIN FROST : 19. Please list the specific points you remember about the ad for VONAGE : 20. What do you think was the purpose of this study? 21. My age is _______ years. 22. I am MALE FEMALE 23. Please indicate your ethnicity: African American Asian Caucasian/White Latin American Middle Eastern Native American Pacific Islander Other_______________________ 24. Please indicate what y ear you are in college: FRESHMAN SOPHOMORE JUNIOR SENIOR 25. Which of the following best represents your political beliefs? (circle one) (1) Democrat (2) Republican (3) Independent or ot her __________________ 26. Are you registered to vote? (1)Yes (0) No

PAGE 84

84 LIST OF REFERENCES Anand, Punam, and Brian Sternthal (1990), “Ease of Message Processing as a Moderator of Repetition Effects in Advertising,” Journal of Marketing Research , 27 (August), 345-353. Ansolabehere, Stephen D., and Shanto Iyengar (1995), Going Negative: How Political Advertisements Shrink and Polarize the Electorate , New York: Free Press. ____________ (1996), “The Craft of Political Ad vertising: A Progress Report,” in Political persuasion and attitude change , Diana C. Mutz, Paul M. Sniderman, and Richard A. Brody, eds., Michigan, Ann Arbor: Univer sity of Michigan Press, 101-122. ____________, and Adam Simon (1999), “Replicati ng Experiments Using Aggregate and Survey Data: The Case of Nega tive Advertising and Turnout,” The American Political Science Review , 93 (December), 901-909. Becker, Lee B., and John. C. Doolittle (1975), “How Repetition Affects Evaluations of and Information Seeking about Candidates,” Journalism Quarterly , 52 (Winter), 611-617. Belch, George E. (1982), “The Effects of Television Commercial Repetition on Cognitive Response and Message Acceptance,” Journal of Consumer Research , 9 (June), 56-65. Berlyne, Donald. E. (1970), “Novelt y, Complexity, and Hedonic Value,” Perception and Psychophysics , 8 (November), 279-286. Brickman, P., J. Redfield, A. A. Harrison, and R. Crandall (1972), “Drive and Predisposition as Factors in the Attitudinal Effects of Mere Exposure,” Journal of Experimental Social Psychology , 8, 31-44. Bruner II, Gordon C. and Paul J. Hensel (1996), Marketing Scales Handbook: A Compilation of Multi-Item Measures , Chicago: American Marketing Association. Cacciopo, John T., and Richard E. Petty (1979), “E ffects of Message Repetition and Position on Cognitive Response, Recall and Persuasion,” Journal of Personality and Social Psychology , 37 (January), 97-109. ____________ (1980), “Persuasiveness of Communicati ons is Affected by Exposure Frequency and Message Quality: A Theoretical and Em pirical Analysis of Persisting Attitude Change,” in Current Issues and Research in Advertising , J. H. Leigh and C. R. Martin Jr., eds., Michigan, Ann Arbor: Univers ity of Michigan Press, 97-122. Cohen, Jeremy and Robert G. Davis (1991), “Third -Person Effect and the Differential Impact in Negative Political Advertising,” Journalism Quarterly , 68 (Winter), 680-688. Cox, Dena S. and Anthony D. Cox (1988), “What Does Familiarity Breed? Complexity as a Moderator of Repetition Effects in Advertisement Evaluation,” Journal of Consumer Research , 15 (June), 111-116.

PAGE 85

85 Crawford, Darlisa (2004), “Television Primar y Information Source for Most 2004 Voters,” (accessed on November 9th, 2006), [available at http://usinfo.state.gov/usinfo/ ]. D’Souza, Giles, and Ram C. Rao (1995), “Can Repeating an Advertisement more Frequently than the Competition Affect Brand Preference in a Mature Market?,” Journal of Marketing , 59 (April), 32-42. Fiske, Susan T. (1980), “Attention and Weight in Person Perception: The Impact of Negative and Extreme Behavior,” Journal of Personality and Social Psychology , 38 (6), 889-906. Garramone, Gina M. (1984), “Voter Re sponses to Negative Political Ads,” Journalism Quarterly , 61 (Summer), 250-259. ____________ (1985), “Effects of Negative Political A dvertising: The Roles of Sponsor and Rebuttal,” Journal of Broadcasti ng and Electronic Media , 34 (Spring), 147-159. ____________, and Sandra J. Smith (1984), “Reactions to Political Advertising: Clarifying Sponsor Effects,” Journalism Quarterly , 61 (Winter), 771-775. ____________, Charles K. Atkin, Bruce E. Pinkleton, a nd Richard T. Cole (1990), “Effects of Negative Political Advertising on the Political Process,” Journal of Broadcasting and Electronic Media , 34 (Summer), 299-311. Haddock, Geoffrey, and Mark P. Zanna (1997), “The Impact of Negative Advertising on Evaluations of Political Candidates: The 1993 Canadian Federal Election,” Basic and Applied Social Psychology , 19 (2), 204-223. Hamilton, David L. and Leroy J. Huffman (1971), “Generality of Impression-Formation Process for Evaluative and Nonevaluative Judgments,” Journal of Personality and Social Psychology , 20 (2), 200-207. ____________, and Mark P. Zanna (1972), “Differe ntial Weighting of Favorable and Unfavorable Attributes in Impression Formation,” Journal of Experimental Research in Personality , 6 (2-3), 204-212. Hill, Ronald P. (1989), “An Exploration of Vo ter Responses to Political Advertisements,” Journal of Advertising , 18 (4), 14-22. Janiszewski, Chris, and Tom Meyvis (2001), “E ffects of Brand Logo Complexity, Repetition, and Spacing on Processing Fluency and Judgment,” Journal of Consumer Research , 28 (June), 18-32. Johnson-Cartee, Karen S. and Gary Copeland (1987, May), “ Setting the Parameters of Good Taste: Negative Political Ad vertising and the 1986 elections ,” Paper presented at the International Communications Associat ions Convention, Montreal, Canada. ____________ (1989), “Southern Voter’s Reaction to Negative Political Ads in 1986 Election,” Journalism Quarterly , 66 (Winter), 888-893.

PAGE 86

86 ____________ (1991), Negative Political Advertising: Coming of Age , Hillsdale, NJ: Lawrence Erlbaum Associates. Kahn, Kim F., and John G. Geer (1994), “Creating Impressions: An Experimental Investigation of Political Advertising on Television,” Political Behavior , 16 (March), 93-116. ____________, and Patrick J. Kenney (1999), “Do Ne gative Campaigns Mobilize or Suppress Turnout? Clarifying the Relationship be tween Negativity and Participation,” The American Political Science Review , 93 (December), 877-889. Kaid, Lynda L. (1997), “Effects of the Televisi on Spots on Images of Dole and Clinton,” American Behavioral Scientist , 40 (August), 1085-1094. ____________ (2004), “Measuring Candidate Images with Semantic Differentials,” in Presidential Candidate Images , Kenneth L. Hacker, ed., Westport, CT: Praeger, 231-236. ____________, and John Boydston (1987), “An Experime ntal Study of the Effectiveness of Negative Political Advertisements,” Communication Quarterly , 35 (Spring), 193-201. ____________, Mike Chanslor, and Mark Hovind ( 1992), “The Influence of Program and Commercial Type on Political Advertising Effectiveness,” Journal of Broadcasting and Electronic Media , 36 (Summer), 303-320. ____________, and Mike Chanslor (1995), “Changing Candidates Images: The Effects of Television Advertising,” in Candidate Images in Presid ential Election Campaigns , Kenneth L. Hacker, ed., Westport, CT: Praeger, 83-97. Klinger, Mark R., and Anthony G. Greenwald (1 994), “Preferences Need No Inferences?: The Cognitive Basis of Unconscious Mere Exposure Effects,” in The Heart’s Eye: Emotional Influences in Perception and Attention , Paula M. Niedenthal and Shinobu Kitayama, eds., San Diego: Academic Press, 67-85. Kuhnhenn, Jim (2006), “Millions Spent on Negative Political Ads,” Retrieved November 9th, 2006 from The Associated Press. Lau, Richard R (1985), “Two Explanations for Negativity Effects in Political Behavior,” American Journal of Political Science , 29 (February), 119-138. ____________ (1982), “Negativity in Political Perception,” Political Behavior , 4 (4), 353-377. ____________, and Gerald M. Pomper (2002), “Effectiven ess of Negative Campaigning in U.S. Senate Elections,” American Journal of Political Science , 46 (January), 47-66. ____________, Lee Sigelman, Caroline Heldman, and Paul Babbitt (1999), “The Effects of Negative Political Advertisements: A Meta-Analytic Assessment,” The American Political Science Review , 93 (December), 851-875.

PAGE 87

87 Leventhal, Howard, and David L. Singer (1964 ), “Cognitive Complexity, Impression Formation and Impression Change,” Journal of Personality , 32 (June), 210-226. Martinez, Michael D., and Tad Delegal (1990), “The Irrelevance of Negative Campaigns to Political Trust: Experimental and Survey Results,” Political Communication and Persuasion , 7, 25-40. McLeod, Jack M. and Daniel G. McDonald (1985), “Beyond Simple Exposure: Media Orientations and their Imp act on Political Processes,” Communication Research , 12 (1), 333. Meirick, Patrick (2002), “Cognitive Responses to Negative and Comparative Political Advertising,” Journal of Advertising , 31 (Spring), 49-62. Merritt, Sharyne (1984), “Negative Political Advertising: Some Empirical Findings,” Journal of Advertising , 13 (3), 27-38. Miller, Richard L. (1976), “Mere Exposure, Psychological Reactance and Attitude Change,” Public Opinion Quarterly , 40 (Summer), 229-233. Nordhielm, Christie L. (2002), “The Influence of Level of Pr ocessing on Advertising Repetition Effects,” Journal of Consumer Research , 29 (December), 371-382. Pinkleton, Bruce (1997), “The Effects of Ne gative Comparative Polit ical Advertising on Candidate Evaluations and Advertisi ng Evaluations: An Exploration,” Journal of Advertising , 26 (Spring), 19-29. Rethans, Arno J., John L. Swasy, and Lawren ce J. Marks (1986), “Effects of Television Commercial Repetition, Receive r Knowledge, and Commercial Length: A Test of the Two-Factor Model,” Journal of Marketing Research , 23 (February), 50-61. Roddy, Brian L., & Gina M. Garramone (1988). “A ppeals and Strategies of Negative Political Advertising,” Journal of Broadcastin g and Electronic Media , 32 (Fall), 415-427. Sawyer, Alan G. (1981), “Repetition, Cognitive Responses and Persuasion,” in Cognitive Responses in Persuasion , Richard E. Petty, Thomas M. Ostrom, and Timothy C. Brock, eds., Hillsdale, NJ: Erlbaum, 237-261. Seamon, J. G., Williams, P. C. Crowley, M. J., & Ki m, I. J. (1995). The mere exposure effect is based on implicit memory: Effects of stimulus type, encoding conditions, and number of exposures on recognition and affect judgments. Journal of Experi mental Psychology , 21, 711-721. Shapiro, Michael A., and Robert H. Rieger (199 2), “Comparing Positive and Negative Political Advertising on Radio,” Journalism Quarterly , 69 (Spring), 135-145. Stang, David J. (1973), “Six Theo ries of Exposure and Affect,” JSAS Catalogue of Selected Documents in Psychology , Manuscript # 148, 3 , 126.

PAGE 88

88 ____________ (1974), “Methodological Factors in Mere Exposure Research,” Psychological Bulletin , 81 (December), 1014-1025. Stewart, Charles J. (1975), “V oter Perception of Mud-Slinging in Political Communication,” Central States Speech Journal , 26 (Winter), 279-286. Surlin, Stuart H., and Thomas F. Gordon (1977), “How Values Affect Attitudes toward Direct Reference Political Advertising,” Journalism Quarterly , 54 (Spring), 89-98. TNS Media Intelligence (accessed on N ovember 9th, 2006), [available at http://www.tnsmicmag.com/ ]. Zajonc, Robert B. (1968), “Attit udinal Effects of Mere Exposure,” Journal of Personality and Social Psychology Monographs , 9 (2, Part 2), 1-27. Zinkhan, George M., William B. Locander, and James H. Leigh (1986), “Dimensional Relationships of Aided Recall and Recognition,” Journal of Advertising , 15 (1), 38-46.

PAGE 89

89 BIOGRAPHICAL SKETCH Juliana de Brum Fernandes was born and rais ed in Sao Leopoldo, Br azil. She graduated from Unisinos University, Brazil, in 2003, with a bachelor’s degree in journalism. She recently completed her Master of Arts in Mass Communication, specializing in political campaigning. For the futu re, she intends to pursue a docto ral degree in the field of journalism and communications.