USF Libraries
USF Digital Collections

Attitudinal antecedents of the first- and third-person effects of alcohol advertising on college students

MISSING IMAGE

Material Information

Title:
Attitudinal antecedents of the first- and third-person effects of alcohol advertising on college students
Physical Description:
Book
Language:
English
Creator:
Begin, Georgia
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Attitude toward the ad
Attitude toward the product
Dissertations, Academic -- Mass Communications -- Masters -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Data collected from a survey questionnaire via personal interviews among 488 college students was used to examine relationships among attitude toward alcohol advertising, attitude toward alcohol beverages, perceived influence of alcohol advertising on oneself, perceived influence of alcohol advertising on others, and attitude toward restrictions of alcohol advertising. Results supported the hypothesized direct effect of advertising attitudes and product attitudes on attitude toward restrictions. Also supported was the indirect effect of advertising and product attitudes on attitudes toward restricting alcohol advertising via such mediators as perceived influence of the ads on self and others using the theories of first- and third- person effects. Implications for future research, public policies, and marketing practices - including responsibility marketing - are discussed.
Thesis:
Thesis (M.A.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Georgia Begin.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 71 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001795811
oclc - 154709180
usfldc doi - E14-SFE0001586
usfldc handle - e14.1586
System ID:
SFS0025904:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Attitudinal Antecedents of the Firstand Third-Person Effect of Alcohol Advertising On College Students by Georgia Begin A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts School of Mass Communications College of Arts & Sciences University of South Florida Major Professor: Scott S. Liu, Ph.D. Kenneth Killebrew, Ph. D. Larry Leslie, Ph.D. Date of Approval: December 7, 2005 Keywords: attitude toward the ad, attitude toward the product Copyright 2006, Georgia Begin

PAGE 2

Acknowledgements I would like to thank Dr. Scott Liu for his continuous advice and support throughout this process and for the infinite help with the manuscript style and data analysis for this paper. I would also like to thank Drs. Ken Killebrew and Larry Leslie for their impartial input and valuable comments and corrections made to this paper. And many thanks to the student volunteer interviewers who conducted the survey and completed the questionnaires with each sample participant. I also would like to thank my family and friends for their encouragement and understanding during this undertaking and my husband for his patience and support with out which I would have had a difficult time finishing this paper.

PAGE 3

i Table of Contents List of Tables iii List of Figures iv Abstract v Chapter One Introduction 1 Overview 1 Statement of the Problem 2 Significance of the Study 6 Chapter Two Review of Related Literature 8 The Third-Person Effect 8 The First-Person Effect 10 The Behavioral Aspect of the Third-Person Effect 12 Research Questions and Hypotheses 15 Attitude 16 AttitudinalOverview 16 Attitudinal Antecedents to Perceived Influence 16 Attitude Toward the Ad 17 Attitude Toward the Product 20 Research Questions and Hypotheses 23 Chapter Three Research Design 25 The Structural Equation Model 25 Hypothesized Paths 26 Research Methodology 28 Selection of Sample 28 Survey Instrument 28 Data Gathering 29 Measures 29 Chapter Four Results 32 Structural ModelResults 32 Descriptive Results 32 Model Fitting and Test Hypothesis 34 Measurement Model Evaluation 35 Structural Model Results Analysis 35 In-Depth Key Path Analysis 39

PAGE 4

ii Chapter Five Discussion and Recommendations 44 Discussion 44 Implications 45 Recommendations 45 Limitations 48 References 49 Bibliography 55 Appendices 58 Appendix A Expanded Path Diagram 59 Appendix B Survey Questionnaire 60 Appendix C Survey Questions and Variables 62 Appendix D List of Research Questions and Hypotheses 64 Appendix E Frequency Distributions 66 Appendix F Descriptive Results 67 Appendix G Correlations 68 Appendix H Paired Sample Statistics 69

PAGE 5

iii List of Tables Table 1 Descriptive Statistics 33 Table 2 Measurement Model Results 34 Table 3 Structural Model Results 36 Table 4 Estimates of Indirect, Direct, and Total Effects 38

PAGE 6

iv List of Figures Figure 1 Structural Equation Model 25 Figure 2 Structural Model Results 32 Figure 3 Portion A of Path Diagram 39 Figure 4 Portion B of Path Diagram 40 Figure 5 Portion C of Path Diagram 41 Figure 6 Portion D of Path Diagram 42

PAGE 7

v Attitudinal Antecedents of the Firstand Third-Person Effect of Alcohol Advertising On College Students Georgia Begin Abstract Data collected from a survey questionnaire via personal interviews among 488 college students was used to examine relationships among attitude toward alcohol advertising, attitude toward alcohol beverages, perceived influence of alcohol advertising on oneself, perceived influence of alcohol advertising on others, and attitude toward restrictions of alcohol advertising. Results supported the hypothesized direct effect of advertising attitudes and product attitudes on attitude toward restrictions. Also supported was the indirect effect of advertising and product attitudes on attitudes toward restricting alcohol advertising via such mediators as perceived influence of the ads on self and others using the theories of firstand thirdperson effects. Implications for future research, public policies, and marketing practices-including responsibility marketing are discussed.

PAGE 8

1 Chapter One Introduction Overview Alcohol abuse is the leading drug problem among Americas youth today (OHara, 2003). Eighty percent of college students drink alcohol beverages, while 40% engage in heavy episodic drinking, 30% meet criteria for alcohol abuse and 6% for alcohol dependence (Collins, Ellickson, Hambarsoomians, & McCaffrey, 2005). Each year, an estimated half-million college students aged 18-24 suffer unintentional injuries while under the influence of alcohol, and at least 1,400 college student deaths a year are linked to alcohol (OHara, 2003). Furthermore, college students engage in heavy episodic drinking at higher rates than their same-aged peers who dont attend college. High-risk drinking also results in assaults, and other health and academic problems, and is a major factor in institutional property damage (Hingson, Hareen, Zakocs, Kopstein, & Wechsler, 2002). The consequences of excessive and underage drinking affect virtually all college campuses, college communities, and college students, whether they choose to drink or not. Knowing that this excessive alcohol consumption is a significant national problem, universities have undertaken a number of campaigns to reduce problem drinking, and despite stepped-up efforts on the part of college and universities over the past decade to change their drinking culture, a study shows that binge drinking is just as common on campuses today as it was in the early 1990s (Rimal & Real, 2005).

PAGE 9

2 Research shows that college students tend to harbor exaggerated perceptions about the prevalence of drinking in their midst; and the greater the prevalence perceived by students, the more likely they are to construe their own consumption as being normative (Perkins & Berkowitz, 1986). Statement of the Problem One factor leading to this perception of normalcy is the media exposure to alcohol marketing that is a constant in the lives of college students. People in this demographic, some of whom are not yet the legal drinking age, are exposed to more alcohol advertising in the forms of print, outdoor, radio, and television, and are the primary recipients of the newly-created viral and buzz-marketing, a grass-roots marketing campaign inherently targeting students on college campuses. Students who were highly exposed to media alcohol advertisements consumed more alcohol in all three categories (beer, wine, liquor) than those with low exposure (Collins et al., 2005). In fact, Collins et al. found that advertisements for distilled spirits had the most influence over college students to drink alcohol, even more so than peer pressure or colleague behavior. A 2002 study found a direct correlation of college-student bingedrinking to alcohol advertising, and has linked the level of binge-drinking on college campuses to high levels of alcohol advertising. The results indicate that the wet alcohol environment around college campuses including lower sale prices, more promotions, and alcohol advertising at both onand off-premise establishments was correlated with higher binge-drinking rates on campus. The more advertising and

PAGE 10

3 promotions there were on and around college campuses, the higher the binge-drinking rates. Advertisements had powerful effects on both drinking and non-drinking college students increasing perceptions of benefits, decreasing perceptions of risk, and making drinking more attractive (Blood & Snyder, 1991). A recent study suggests that the regulation of marketing practices such as advertising and promotion may be an important factor in reducing binge-drinking on college campuses (Kuo, Wechsler, Greenberg, & Lee, 2003). College administrators have a role in the attempt to control college-student binge-drinking rates by not allowing or accepting advertising and promotions dollars to be spent on campus. A 2003 survey of 700 college administrators revealed that only 50% of those colleges prohibited alcohol ads in campus newspapers for off-campus bars and clubs, however, most schools prohibited alcohol advertising at sporting events (Mitchell & Olson, 1981). Many groups are concerned with controlling youth exposure to alcohol advertising, and the alcohol industry claims not to recruit new consumers with their advertising, but only to solidify brand loyalty. The industrys own guidelines for marketing to youth have not lived up to the expectations of the public and groups concerned with protecting underage drinkers. The Distilled Spirits Council of the U.S. (DISCUS) advertising guidelines for protecting underage youth from exposure to messages includes placement that limits media vehicles to 50% or greater adult audience. In 1999, the FTC urged distilled spirits manufacturers and brewers to adhere to a standard of advertising placement that includes underage audience members to be 25% or less. The protection group Mothers Against Drunk Driving (MADD) wants the

PAGE 11

4 standard to be set at 10% youth audience, and the American Medical Association (AMA) desires a complete ban on alcohol ads (camy.org, 2003). According to the Center on Alcohol Marketing and Youth (CAMY), young people under the legal drinking age were the primary target market for alcohol ads in magazines in 2001, despite the industrys claim to the contrary (camy.org, 2003). Underage youth saw 95% more beer advertising than did adults aged 35 and up; and 25 alcohol brands placed all of their ads in youth-oriented magazines. As the population of the United States ages and alcohol consumption declines, the alcoholic beverage industry has a greater economic stake in recruiting young, heavy drinkers. Despite recent increases in federal support for drug education in the public school system and increasing public recognition of the seriousness of alcohol problems among youth, alcohol advertising continues to present drinking as normal, glamorous, and consequence-free, according to CAMY (camy.org, 2003). As a result, widespread concern exists among policymakers and the public about the potential effects of alcohol advertising on alcohol consumption and problems, especially among adolescents and youth. A recent national survey indicates that 67% of adults in the United States support banning liquor advertisements on television, and 61% favor banning beer and wine advertisements in this medium (Wagenaar, Murray, & Geban, 2000). Similarly, public health advocates routinely call for the strict regulation or even elimination of alcohol advertising; and initiatives at the community level frequently focus on reducing local alcohol advertising (Agostinelli & Grube 2002). Saffer and Dave (2002) found that alcohol advertising bans decrease alcohol consumption and that in implementing one [media] type of ban could reduce overall consumption by 5-8%. Comprehensive bans

PAGE 12

5 on all forms of advertising and promotions can eliminate options for substitution and be potentially more effective in reducing consumption. However, advertising cannot be reduced with limited bans which are likely to result in substitution to other available media (Saffer, 2002). The AMA, MADD, the American Academy of Pediatrics (AAP), and other groups in 1996 banded together and persuaded NBC to stop airing national television ads for liquor products only three months after the decision to air them. The fact that interest groups could sway media outlets to restrict a form of speech is especially interesting in mass communications research because of fears that any form of banning or censorship could result in exaggerated outcomes and set precedence for future justification of censorship. According to Saffer and Dave (2002), the focus on advertising bans is important because bans are a likely choice of public policy for the control of alcohol. Despite the efforts to counteract the pervasiveness of alcohol advertising, certain aspects of alcohol advertising itself appear to be particularly attractive to young people. These include the use of sports imagery, and the frequent use of image advertising in general as opposed to advertising that makes claims about the quality or production of the product. Not surprisingly, it is found that college students generally hold a positive attitude toward alcohol advertising (Wyllie, Fang, & Caswell, 1998b). The fact that more than 80% of college students drink alcoholic beverages suggest that their attitude toward alcohol products is all but negative (OHara, 2003). However, students are able to change their drinking attitudes relatively quickly if enlightened in a unique manner to the outcomes of excessive drinking, which was evidenced in a recent study by Jewell, Hupp, and Luttrell (2003). The investigation proved that when college students

PAGE 13

6 experience negative outcomes of alcohol consumption themselves, they can radically change their positive views of drinking and of alcohol products. In this study, students were asked to wear fatal-vision-goggles, a type of vision wear used by police and teachers to demonstrate to youth what its like when you are intoxicated. The students were then asked to perform sobriety and mock-driving tests in a controlled setting. Previous to the experiment, students attitudes on drinking and driving were collected and it was revealed that those who already drank alcohol had favorable attitudes toward drinking and driving. After wearing the goggles and performing the tasks, most subjects experienced a significant decline in their attitude toward drinking and driving and thus simply participating in the experiment had enough of an impact to change attitudes. More importantly, the students reduced their own consumption and they also become advocates for others to abstain from alcohol use, especially when expecting to drive. College students could have similar reactions to alcohol advertising when made aware of the effects of the ads on their own behavior and the behavior of other college students. It is possible that this primary target market for alcohol advertising would become advocates of restricting ads aimed at their peers which would place great pressure on the alcohol industry to change their promotional practices. Significance of the Study The purpose of this study is to answer the following questions: To what extent do college students find themselves susceptible to the influence of alcohol advertising? To what extent do college students attitude toward alcohol advertising and products in general affect their perceived influence of alcohol advertising? And to what extent does

PAGE 14

7 the perceived influence of alcohol advertising affect their views about restrictions of alcohol advertising? In what follows, we will first review the relevant theoretical constructs and reasoning. A distinction between first-person effect and third-person effect is made to facilitate understanding of the perceived influence of alcohol advertising. A structural equation model, which incorporates all hypothesized relationships among the key constructs, will then be presented, followed by a survey study designed to empirically test the model.

PAGE 15

8 Chapter Two Review of Related Literature The Third-Person Effect Recent research in public opinion and mass communications points to the need to differentiate between the perceived influence of media on oneself (first-person effect), and the perceived influence on others (third-person effect). The same distinction is made in the present study between college students perceived influence of alcohol advertising on themselves and on other college students. The third-person effect perceptual hypothesis, first proposed by W.P. Davison (1983), predicts that individuals will perceive media messages to have a greater impact on other people than on themselves. The hypothesis has generated numerous studies in an effort to explain this phenomenon. Some researchers have argued that the thirdperson effect, at its heart, reflects a self-serving bias (Gunther & Mundy,1993; Gunther & Thorson,1992). In their meta-analysis, Paul, Salwen and Dupagne (2000) discussed varying sociological and psychological theories that have been used to explain the thirdperson effect and its consequences, including ego involvement, the elaboration likelihood model, the social categorization theory, attribution theory, and biased optimism. The comparison between self and other constitutes a form of unrealistic and biased optimism that is motivated by the need for ego enhancement (Dupagne & Salwen, 1999). The same motivation also may lead people to think that others are more

PAGE 16

9 likely to be harmed by the media; if by comparison, it enhances their view of themselves (McLeod, Detenber, & Eveland, 2001). The more negative a message is perceived, the wider the gap between its perceived influence on self and others (Eveland & McLeod, 1999). The significance of the third-person perception is in its tendency to lead individuals to advocate action to protect others from the perceived harmful influence of the media (Gunther, 1991). This feeling of others being more influenced by media tends to create paternalistic attitudes in people who fall victim to the third person effect; which leads to the desire by some to protect those people they feel are affected. Perloff (1996) notes that the third-person effect is likely to manifest itself when media messages advocate behavior that will not be beneficial for the self, or gives rise to the perception that it is not smart to be influenced by the message. The end result is that people surmise others to fall victim to medias influence while they do not. Wyllie (1997) also suggests that people are likely to consider themselves smarter and more resistant to a message when they feel the topic is one that has little benefit, or even potentially harmful consequences, for its audience. Similarly, Eveland and McLeod (2001) argue that the magnitude of the third-person effect perception is influenced by the social desirability of the message, the lower the social desirability of the message, the stronger the third-person effect. Aside from the feeling of protecting those more vulnerable to negative media influences, the third-person effect is also linked to a subjects belief that negative consequences wont happen to them (Paul, Salwen, & Dupagne, 2000). This process identifies peoples ability to distinguish between societal-others and personal-self-level effects from media, and it also identifies that media messages influence peoples

PAGE 17

10 perceptions of risk or harm (Tyler & Cook, 1984). The spectrum of what constitutes social desirability fluctuates with individuals and within the given community, but is a predictor and an independent variable of the third-person effect (Banning 2001; Eveland, Nathanson, Detenber, & McLeod, 1999). The First-Person Effect In contrast to the third-person effect, the first-person effect has been found to occur when the potential benefit from a message is high. That is, when media messages are positive and advocate beneficial outcomes, people tend to consider themselves just as influenced as others; and in some cases, they may anticipate even more effect on themselves. Gunther and Mundy (1993) point out that as interest in the message increases, so does the perceived influence on ourselves. Eveland and McLeod (1999) argue that ego enhancement is also responsible for the observed first-person effect where people view themselves as more persuaded by desirable media content. It is important when discussing third-person effect to distinguish between messages that are intended to inform, such as news, and those that are intended to persuade such as advertising, to recognize accepted social roles of each domain (Gunther & Thorson, 1992). Specifically, exposure to news is both sought-after and socially desirable, while advertising is usually an event to avoid. A national survey by DDB Needham Worldwide in 1989 revealed that sixty-six percent of respondents report that advertising insults their intelligence. Furthermore, people will perceive the two domains accordingly and will show domain-specific processing patterns. Since being

PAGE 18

11 persuaded by advertising is often viewed as detrimental, the general sphere of advertising would thus be likely to exhibit third person effect. However, Gunther and Thorson (1992) also noted that some advertising might induce a first-person effect. They examined estimates of influence on self and others in relation to ads that contained an emotional appeal and found that although self and others were equally affected by the positive-emotion message, subjects tended to recognize and admit more of an impact on themselves in these positive situations. Given young audiences often characterize alcohol advertising as fun, youthful, exciting, and thus emotional, one would expect such advertising to induce the first-person effect as well (David, Liu, & Meyer, 2004; Wyllie et al., 1998b). People who are experiencing the positive emotions while viewing messages should then be more likely to agree that the persuasion is not bad or undesirable, as well as admit that they themselves are more likely to be persuaded than others; in other words, a first-person effect. There are times when the desirability of content under study may depend on respondents predispositions regarding the issues (Salwen &Dupagne, 1999). A recent study discovered that the third-person effect occurred during negatively-stigmatized product ads six times more frequently than in ads for neutral products. A degree of social stigma might be said to be attached to any product depending on ones perspective; however, certain products have attained a more general negative societal stigma, as illustrated by major public service campaigns generating awareness of the negative aspects of these products such as cigarettes and alcohol (Banning, 2001). The designation of the term negative does not refer to the message or tone of the

PAGE 19

12 advertisement; it would be expected that ads are all positive because the selling nature of the ad; therefore the term negative refers only to a products standing from a societal perspective (Banning, 2001). In the present study, alcohol will be difficult to define as having either a positive or negative societal stigma. Excessive alcohol consumption is widely known to have damaging outcomes (negative); however, these products are held in high esteem by most college students (positive) as evidenced by the number of students who drink and their drinking statistics. The Behavioral Aspect of the Third-Person Effect The behavioral aspect of the third-person effect states that people will act on their perception of the media influencing other people by advocating restrictions on communication. From its inception, concern over the third-person effect stemmed from the possibility that strategic social action might be taken based on the over-estimation of media effects on others (Davidson, 1983). Several studies have linked the third-person effect perception to support for censorship (Gunther, 1991; Eveland & McLeod, 1999; Rojas, Shah, & Faber, 1996; Salwen, 1998). Rojas et al. (1999) argued that third-person effect perception involves two distinct effects individuals judgments of others susceptibility to media communications, and their beliefs about the severity of effect outcomes (Salwen &Dupagne, 1999). Researchers see the behavioral aspect of the third-person effect as the more socially relevant phenomenon because of the possible support for censorship (McLeod et al., 2001). According to McLeod et al. (2001), the third-person perception has been shown to be a meaningful predictor of peoples willingness to impose limits on certain types of communication.

PAGE 20

13 Studies have found in most instances, empirical support for a link between the third-person effect and censorship. Willingness to support censorship was attributable to the perception that others were not wholesome enough to resist immoral influences (Salwen, & Dupagne, 1999). In many studies, subjects supported limiting access to what was perceived as negative media content that they believed would affect others. In some instances, researchers concluded that this type of support for limiting access to media was due in part from a paternalistic attitudes and the need to feel like one is protecting others from harmful media effects (McLeod et al., 2001). The findings in each of these studies reaffirm robust support for the third-person perceptual hypothesis; however, Salwen and Dupagne (1999) found that in regards to the behavioral hypothesis, effect perceptions were issue-dependant (Salwen & Dupagne, 1999). In most cases, researchers selected issues or content presumed to be undesirable to believe (Salwen & Dupagne, 1999). Studies have linked greater thirdperson effect with support for censorship of rap music (Eveland andMcLeod, 1999), pornography, television violence (Rojas et al., 1996), and political campaigns (Salwen, 1998). According to Eveland and McLeod (1999), an issue that influences whether someone is willing to impose limits on expression may be related to the question, whos ox is being gored? Individuals may ask whether this content or message is one for which they would like to have access. Examples would be pornography or rap music. The people who buy and use pornography or rap music, it would seem, would be unwilling to allow it to be banned or censored. For the present study, college students will be asked about their attitudes toward censorship (or restrictions) of alcohol advertising in the variable attitude toward advertising restrictions, (ATTR).

PAGE 21

14 It is important to note here concerning the advertising of alcohol: censoring or restricting the advertisement does not have the same outcome as censoring rap music or pornography. Eliminating ads for alcohol does not mean that you will no longer be able to consume it as it would mean if rap music and pornography were censored. This factor may be an important element in determining attitudes toward censorship. People having attitudes in favor of censorship is important because they may support restrictions on media content and encourage policy makers to change laws favoring their position (Eveland &McLeod, 1999). It should also be noted that the firstand the third-person effect are often related. Specifically, the perceived influence of alcohol advertising on self may constitute the basis for the assessment of the influence on others. The reasoning is consistent with the hypothesis of looking-glass perception which finds individuals to project their own thoughts and feelings onto others: what I think must be what others think (Fields & Schuman, 1976). The looking glass perception is assumed to operate quite apart from the actual distribution of opinion. Relative to the current study, the hypotheses suggest that if college students experience the first-person effect of alcohol advertising, they would project the perceived influence onto others in the form of the third-person effect. Considering all the evidence in the firstand third-person effects, the following research questions and hypotheses are formed:

PAGE 22

15 Research Questions andHypotheses RQ1: What is the relationship between perceived influence of alcohol advertising on self (SELF) and on others (OTHERS)? H1: Perceived influence of alcohol advertising on self will be positively related to perceived influence on others. (SELF OTHERS; OTHERS SELF) RQ2: What is the relationship between perceived influence of alcohol advertising (on SELF and OTHERS) and attitude toward greater restrictions on alcohol advertising (ATTR)? H2-a: There will be a positive relationship between SELF and ATTR. (SELF ATTR) H2-b: There will be a positive relationship between OTHERS and ATTR. (OTHERS ATTR) RQ3: What is the relative influence of perceived effect of alcohol advertising on self (SELF) and others (OTHERS) on attitude toward restricting alcohol advertising (ATTR) H3: SELF ATTR OTHERS ATTR

PAGE 23

16 Attitude Attitudinal Overview Attitude can be described as an individuals internal evaluation of an object, such as a branded product (Mitchell & Olson, 1981). Attitudes are often considered relatively stable and enduring predispositions to behave, so they should be useful predictors of consumers activities with a product (Mitchell & Olson, 1981). According to previous studies, an ad affects consumers beliefs first; then influenced salient beliefs mediate the marketing variables effect on attitude, and attitude in turn, mediates subsequent effects on behavioral intention (Mitchell & Olson, 1981). Ajzen and Fishbein (1980) conceptualize attitude as having three forms of expression: cognitive (beliefs), affective (feelings) and conative (behavior). The most prevalent means of measuring [adolescents] attitudes appears to be an affective measure either in terms of preferences or liking (Phelps & Hoy, 1996). The attitudes to be examined in this study are attitude toward the ad (ATTA) which is measured by liking of alcohol ads in general, attitude toward the product (ATTP), measured by opinions and behaviors with alcohol, and attitude toward the restriction of alcohol advertising (ATTR). Attitudinal Antecedents to Perceived Influence Public attitudes toward advertising in general have been of interest to researchers for years. Advertising researchers have been interested in the impact of overall attitudes toward advertising on consumer behavior variables. Studies have

PAGE 24

17 suggested, for example, that consumers' attitudes toward individual advertisements are influenced by their attitudes toward advertising in general. People with more favorable feelings about advertising found specific advertisements more acceptable, informative, and enjoyable (Bartos & Dunn, 1974; Bauer & Greyser, 1968; Lutz, 1985). Consumers' overall positive attitude toward all advertising is also related positively to involvement with specific advertisements (James & Kover 1992). From a public policy perspective, concerns have been voiced that criticisms of advertising (i.e., it presents false and misleading information, it promotes undesirable values, it persuades people to buy things they do not need, etc.) may undermine its effectiveness or even lead to pleas for greater regulation (Calfee & Ringold, 1988; Pollay & Mittal, 1993). Attitude Toward the Ad According to Shen (1998), ads that are well-liked are more likely to be attended to and remembered than ads that are not, thus attributing to higher elaboration conditions and greater attitudinal affects guiding long-term behavior. This implies attitudes held about an ad can predict subsequent behavior related to the advertised product. A well-liked ad creates a well-liked product (Phelps & Hoy, 1996). In Petty & Cappicios (1983) study of advertising effects, findings revealed that subjects attitudes toward an advertised product were influenced more by their attitude toward the ad, or the ads likeability, than their thoughts about the actual product (Yates, 2001). Likeability, and is the single biggest predictor of advertising effectiveness which necessitates changes in attitudes and behavior; although studies reported by Blair & Rosenberg (1994) did not support this conclusion (Wyllie et al., 1998a).

PAGE 25

18 A category of advertising with known high impact is one that creates positive emotions in the viewer; for instance, ads that arouse any type of moderate feelings have been shown to be associated with more positive brand attitudes and a greater intention to purchase than ads that do not create such emotions (Gunther & Thorson, 1992). As people experience positive emotions during a message, they are likely to retrieve positive material from memory which in turn influences decision-making (Gunther & Thorson, 1992). A 1992 study by Gunther and Thorson examined estimates on self and others in relation to product ads that contained an emotional appeal. The authors found that even though self and other should be equally influenced by a positive-emotion message, people tended to recognize and admit to the impact of such a message more so on themselves. According to the study: As emotional impact ratings increased, there was an apparent linear trend in estimated persuasive effects. The ad with the lowest emotion rating was seen to have the most negative impact. As emotion ratings increased, perceived impact moved steadily towards the positive, so that the ad rated highest on emotional content had the strongest perceived positive impact on the self (Gunther & Thorson, 1992). The bearing of experiencing positive emotions while viewing a message on the third-person effect should be to increase someones willingness to detect influences on themselves, thus possibly enabling a first-person effect. According to Gunther and Thorson (1992), perceived impact has a lot to do with how socially-desirable the respondents view the impact. Given that all ads treat their subject matter in a positive light, any expected attitude toward the ad (ATTA) should be a positive one unless subjects already have a negative attitude toward the product (ATTP). A positive ATTA would normally result

PAGE 26

19 in a first-person effect from the media exposure, (not a third-person effect), for morallyvalued messages, emotionally appealing ads and perhaps even for well-liked messages (Mcleod et al., 2001; Phelps & Hoy, 1996). We do know based on previous studies (David et al. 2004; Wyllie, 1997; Wyllie et al., 1998a) that alcohol ads are well-liked and sometimes college students favorite ads on television; also, we know alcohol use is prevalent on college campuses. Additionally known is that alcohol advertising and products have an overall negative societal stigma, so it may be reasonable to predict outcomes in our study to include both a first-person effect from alcohol advertising due to ad-liking, as well as third-person effect due to the negativity perceived from alcohol advertising. Several recent studies have found that adolescents who are exposed to greater amounts of alcohol advertising are more likely to use or intend to use such products (Garfield, Chung, & Rathouz, 2003). Researchers found a positive relationship between the preference for [alcohol] advertising and the intention to consume alcohol by teens (Kelly & Edwards, 1998). A recent study proved that positive responses to televised beer advertisements, measured by likability (ATTA), contributed to the quantities of alcohol being consumed by 18-29 year-olds (Wyllie et al., 1998b). The ads evaluated provided venerated outcomes for the test subjects such as peer group acceptance, appealing role models, and having fun. The respondents reacted more positively to the ads with these valued outcomes than those that were less relevant to them (Wyllie, 1997). The importance of ad liking, in keeping with Banduras Social Learning Theory, is that people are more likely to model behavior they believe has the outcomes they value. Attitude toward the ad research has found that audiences develop a generally

PAGE 27

20 good feeling about a brand when they like the commercials (Eveland & McLeod, 1999). The liking of an ad therefore creates scenarios in the consumers mind that they try to duplicate in their own lives. It is for this reason why teen and college student exposure to the enjoyable alcohol ads are under scrutiny. The subject matter in this study is uniquely different from previous research because we are considering a well-used, well-accepted product by the subjects of the current study (college students) which is also recognized to be a harmful substance when used irresponsibly and excessively, thus creating the potential outcomes of the study to be divergent (both firstand third-person effects). It is documented in previous studies (Wyllie, 1997; David et al., 2004) that adolescents and college students enjoy ads for alcohol products and desire to be in situations similar to those shown in ads. The advertisements are fun, youthful, and exciting. If you were to consider any other product, one which does not have a negative social stigma, you would be likely to hypothesize there to be a first-person effect from the ads based on research by David et al. (2004). However, considering the product examined in the study is alcohol, we expect there to also be a significant third-person effect, even though the subjects may have a positive attitude toward the ad (and may also experience first-person effect), results should be consistent with those found in third-person effect research. Attitude Toward the Product Past studies examining the phenomenon of attitude toward the brand have focused on specifically advertised brands, not product categories as a whole. The current study is somewhat exploratory in nature because subjects will not be asked

PAGE 28

21 about specific alcohol brands. Rather, we group together all alcohol brand-advertising into one contiguous group for the purposes of obtained an aggregate attitude toward alcohol products advertised, including any beverage that contains alcohol (beer, wine, hard liquor) and call this group attitude toward the product (ATTP). This is done not only for simplicity, but also to identify general attitudes harbored for any and all items in this group which can be said to have the same intended use, and similar conditions of purchase and consumption on college-campuses. Therefore, in this study, when discussing past findings of attitude toward the brand, we will identify this term as attitude toward the product (ATTP) for our own purposes. Many previous studies have tied together the concepts of ATTA and attitude toward the brand [ATTP] and have described ATTA to be a substantiated predictor and mediator of ATTP, a unidirectional relationship, for both familiar and unfamiliar brands (Phelps & Hoy, 1996). ATTA is thought to affect ATTP reflecting consumer perception of an affective and evaluative overlap between a brand and its message (Curlo & Chamblee, 1998). What this means is that if viewers dont have an opinion about a product before seeing an ad, the ATTA will affect ATTP (a well-liked ad will mean a well-liked product and vice versa) (Phelps & Hoy, 1996). Mitchell and Olson (1998) argued that previous findings in the relationship between ATTA-ATTP were flawed because researchers failed to account for prior brand attitude, they found evidence that although the ATTA-ATTP relationship was a well-documented one, it did not exist for familiar brands. However, Phelps and Hoy (1996) found ATTA to be a significant predictor of ATTP, for both unfamiliar and familiar brands, even when controlling for prior brand attitudes. Thorson (1991) concluded that despite prior brand

PAGE 29

22 attitude, ad liking predicted consequent brand liking, albeit, the relationship between ATTA-ATTP was stronger for unfamiliar brands. If ones attitude about an ad is positive, this will result in positive beliefs about the product in the ad. So strong is this relationship, that each prior study found at least some causal relationship from ATTA to ATTP. However, ATTA doesnt always predict ATTP, sometimes the paths of influence seem to work both ways, or in reverse. According to Mitchell and Olsen (1981), attitude toward the ad should be treated as a construct that is conceptually distinct from brand attribute beliefs and brand attitude, because the two constructs are inherently separate from one another. Wyllie (1997) found a reverse path of influence between ATTA and ATTP, where attitude toward the product predicted subsequent attitude toward the ad. It is possible that once someone becomes a drinker, they become more interested in and find alcohol advertising more appealing (Wyllie et al., 1998a). Wyllie (1997) found adolescents responses to alcohol advertising were a consequence of the initial liking of alcohol products. Researchers found that students who were already frequent drinkers (established ATTP) rated alcohol advertising as more effective, identified better with their portrayals, and regarded those portrayals as more desirable compared with students who were less frequent drinkers (Agostinelli & Grube, 2002). A higher drinking frequency also foretold less approval of anti-alcoholabuse public-service announcements. A college students positive ATTA of alcohol ads has been linked to greater influence of the ads on the self (a first-person effect) as proven in past studies (David et al., 2004).

PAGE 30

23 Will ATTP have similar outcomes? College students are expected to know that alcohol products are dangerous especially to their peers, yet most (more than 40%) engage in binge-drinking and thus probably have positive opinions of the product. What will be the implications of these conflicting cognitions? What is the relationship between general attitude toward alcohol products (ATTP) and the perceived influence of alcohol advertising on self and others? What can be expected to be found in the current study? What is the relationship between general attitude toward alcohol advertising (ATTA) and the perceived influence of alcohol advertising on self and other? This study seeks to determine what types of attitudes college students have toward drinking, alcohol advertising and alcohol products in general to clarify established opinions that may precede any first or third person effects from alcohol ads. The current study, exploratory in nature, seeks to discover the direct influence of ATTA and ATTP on college students desire to restrict alcohol ads, as well as the indirect influence on ATTR mediated through self and other variables. The following research questions and hypotheses were created to specifically define the relationships and correlations sought from the study: Research Questions and Hypotheses RQ4. What is the relationship between general attitude toward alcohol advertising and the perceived influence of alcohol advertising on self and others? H4-a: Perceived influence of alcohol advertising on self (SELF) will be positively related to attitude toward alcohol advertising (ATTA). (ATTA SELF) H4-b: Perceived influence of alcohol advertising on others (OTHERS) will be positively related to attitude toward alcohol advertising (ATTA). (ATTA OTHERS)

PAGE 31

24 RQ5. What is the relationship between general attitude toward (perception of) alcohol products and the perceived influence of alcohol advertising on self and others? H5-a: Perceived influence of alcohol advertising on self (SELF) will be positively related to attitude toward alcohol products (ATTP). (ATTP SELF) H5-b: Perceived influence of alcohol advertising on others (OTHERS) will be positively related to attitude toward alcohol products (ATTP). (ATTP OTHERS) RQ6. What is the direct relationship between ATTA and ATTP on ATTR? H6-a: There will be a negative relationship between ATTA and ATTR. H6-b: There will be a negative relationship between ATTP and ATTR. RQ7: What is the mediating role of SELF between the variables ATTA/ATTTP with ATTR? H7-a: The indirect relationship from ATTA to ATTR mediated through SELF, will be positive in both legs of the path (ATTA SELF ATTR). H7-b: The indirect relationship from ATTP to ATTR mediated through SELF, will be positive in both legs of the path (ATTP SELF ATTR). RQ8: What is the mediating role of OTHER between the variables ATTA/ATTTP with ATTR? H8-a: The indirect relationship from ATTA to ATTR mediated through OTHER, will be positive in both legs of the path (ATTA OTHER ATTR). H8-b: The indirect relationship from ATTP to ATTR mediated through OTHER, will be positive in both legs of the path (ATTP OTHER ATTR).

PAGE 32

25 Chapter Three Research Design X 16 X 15 X 13 X 12 X 14 X 11 X 10 X 9 X 8 X 7 X 6 X 5 X 4 X 3 X 2 X 1 ATTA ATTP OTHERS ATTR SELF The Structural Equation Model Figure 1 summarizes the hypothesized theoretical relationships among the variables in a path diagram. Each proposed relationship is clearly sketched with arrows indicating the hypothesized directional connection. The boxes around each circled variable represent the question numbers on the questionnaire that identify the value Figure 1. Structural Equation Model

PAGE 33

26 belonging to the variable. Questions were selected using a pre-tested questionnaire and represent valid measurements of each variable. Hypothesized Paths As the figure shows, attitude toward alcohol advertising (ATTA) and attitude toward alcohol products (ATTP) are two distinct but related attitudinal dimensions. Both of them have direct influence on attitude toward restrictions on alcohol advertising (ATTR). ATTA and ATTP also have indirect influence on ATTR via the perceived influence of alcohol advertising on oneself (SELF) and on other people (OTHERS). A direct transfer of effect from SELF to OTHERS is also recognized in the model. The model thus hypothesizes that the perceived effect of alcohol advertising on oneself (first-person effect) and the perceived effect of alcohol advertising on others (third-person effect) act as mediators of the relationship between attitude toward alcohol advertising and attitude toward alcohol advertising restrictions. They also act as mediators of the relationship between attitude toward alcohol products and attitude toward alcohol advertising restrictions. Consistent with the behavioral hypothesis of the third-person effect, the model predicts the significant influence of OTHERS on ATTR. For comparative purposes, the model also includes the causal link from SELF to ATTR. Since the third-person effect is more often used to justify support for media content restrictions than first-person effect, we would expect the OTHERS SELF

PAGE 34

27 In agreement with the looking glass perception hypothesis, a direct path leading from SELF to OTHERS is hypothesized in the model.Finally, the model hypothesizes a correlation between ATTA and ATTP. The correlation between these two variables implies the possibilities of indirect influence of ATTA on other variables via ATTP, and vice versa.

PAGE 35

28 Research Methodology Selection of Sample The sample consisted of 488 undergraduate and graduate students (201 males and 367 females) enrolled in a large southern university.Their mean age was 21.93 (SD = 3.13), with 31 percent age 20 or younger. Although 13% of the respondents claimed that they never drank, the majority (87.1%) of the sample said that they drank alcoholic beverages at least occasionally, and about 24% admitted that they drank often. Nearly 57% indicated that they had one to four drinks during the most recent social drinking occasion. The majority of underage (less than 21-years-old) respondents reported that they not only drink alcohol, but drink often. Survey Instrument The survey questionnaire was pre-tested on an undergraduate research methods class in the summer 2005 semester. The students were also asked to pre-test the draft questionnaire on other fellow students. Revisions were made by the primary researchers and the finalized questionnaire was developed using feedback from interviewers. The questionnaire used consisted of 25 Likert-scaled responses to questions about attitudes and thoughts concerning alcohol advertising, alcohol consumption, college student drinking patterns, alcohol advertising restrictions, and to whom the sample agreed that alcohol advertising would be most appropriate. Also included in the

PAGE 36

29 questionnaire were scaled answers to questions about the respondents current drinking behaviors, student status, age and gender. Data Gathering Trained interviewers took two weeks to personally interview the respondents during August, 2005.Each interview took an average of 15 minutes to complete. These interviewers represented the expected demographic of the actual sample and were so chosen to lessen the chance of exaggerated or false responses given for some of the sensitive questions [such as established drinking behavior] that underage drinkers may not want to share with an older adult interviewer. Training consisted of learning exact dialogue to be stated during the interview, as well as an approach to overcoming objections, and an overall plan for completing the questionnaire. All respondents were told that their participation was strictly voluntary, and that their response to the survey would remain confidential. No names or personal identifying information was gathered, therefore, answers were also completely anonymous. Measures The following list includes the key measures contained in the survey. All questions were developed as a result of performing two separate pretests on students and tailoring the final questions to minimize confusion and the subjects need to ask questions of the interviewer. These pre-tests were conducted weeks before the actual questionnaire was implemented.

PAGE 37

30 Attitude Toward Alcohol Advertising (ATTA). Three Likert-scaled (5: Strongly disagree, 1: strongly disagree) items were used to measure the construct: I like advertising for alcoholic beverages, I dont have a problem with advertising for alcoholic beverages, and I identify with the characters and situations portrayed in those ads. The Cronbachs alpha for ATTA was .84, indicating satisfactory inter-item reliability. Attitude Toward Alcohol Products (ATTP). This measure was based on two questions asking: I dont like liquor, beer or wine products, and I have a favorable opinion of most alcoholic beverages, on 5point Likert scales. The Cronbachs alpha was .67 for the measure. Perceived Influence of Alcohol Advertising on Oneself (SELF). Using a 4-point scale (4: A great deal of influence, 1: No influence at all), respondents were asked to indicate the extent to which alcohol advertising had an influence on their purchase of alcohol products, consumption of alcohol products, and attitude toward alcohol products. The Cronbachs alpha was .87. Perceived Influence of Alcohol Advertising on Others (OTHERS). Respondents were asked to indicate the extent to which alcohol advertising had an influence on other students purchase of alcohol products, consumption of alcohol products, and attitude toward alcohol products. Cronbachs alpha for this measure was .88

PAGE 38

31 Attitude Toward Restriction on Alcohol Advertising (ATTR). Respondents attitude toward alcohol advertising restrictions was measured by five Likert-scaled items: Alcohol companies shouldnt advertise to college students, People under the age of 21 shouldnt be exposed to alcohol advertising, There ought to be more restrictions on alcohol advertising, I support the idea of banning alcohol ads from some media forms, I support the idea of banning alcohol ads from all forms of media. The Cronbachs alpha for the five-item measure was .87.

PAGE 39

32 Chapter Four Results Structural Model Results Figure 2 is a pictorialdisplay of the descriptive results in the structural model diagram. In this model, every represented path was proven to be valid and significant, with the exception of the path between ATTA and OTHERS. In the following pages and sections, each relationship and each section of the diagram will be examined and the findings will be discussed. Descriptive Results Table 1 on the following page presents the means and standard deviations of all independent and dependent variables examined in this study. .241* .100* SELF ATTA ATTR OTHER ATTP -.126 .328* -.419** -.385** -.284* .784** .485** .123* **p<.01, *p<.05 Figure 2. Structural Model Results

PAGE 40

33 Table 1. Descriptive Statistics Cronbachs Variables Mean SD Alpha Attitude Toward Alcohol Advertising (ATTA) a 3.32 .95 .84 Like alcohol advertising 3.18 1.14 Have no problem with alcohol advertising 3.72 1.11 Identify with characters and situations in alcohol advertising 3.07 1.01 Attitude Toward Alcohol Products (ATTP) a 3.55 1.07 .67 Like liquor, beer or wine 3.24 1.19 Favorable opinion of alcoholic beverages 3.86 1.27 Perceived Influence of Alcohol Advertising on Self (SELF) b 2.02 .83 .87 Influence on purchase of alcohol products 2.04 .92 Influence on consumption of alcohol products 1.93 .90 Influence on attitude toward alcohol products 2.10 .95 Perceived Influence of Alcohol Advertising on Others (OTHER) b 2.94 .66 .88 Influence on purchase of alcohol products 2.93 .69 Influence on consumption of alcohol products 2.91 .76 Influence on attitude toward alcohol products 2.98 .76 Attitude Toward Restrictions of Alcohol Advertising (ATTR) a 2.78 .91 .87 More restrictions on alcohol advertising 2.70 1.06 Shouldnt advertise to college students 3.13 1.15 Shouldnt advertise to people under 21 3.00 1.14 Banning alcohol advertising from some media 2.93 1.22 Banning alcohol advertising from all media 2.14 1.06 a Items measured by 5-point Likert Scale (5: Strongly agree, 1: Strongly disagree)b Items measured by 4-point Scale (4: A great deal of influence, 1: No influence at all)Consistent with prior research, participants in the present study exhibited, on the average, a favorable attitude toward alcohol advertising in general (Mean ATTA = 3.32, SD = .95) as well as a favorable attitude toward alcohol products (Mean ATTP = 3.55, SD = 1.07). T-test results indicate both means are significantly different from the neutral point of 3 on the scale (t ATTA = 7.50, df = 487, p<.001; t ATTB = 11.45, df = 487, p<.001). On the other hand, respondents attitude toward restrictions on alcohol advertising was less than favorable (Mean ATTR = 2.78, SD = .91, t ATTR = 18.90, df = 487, p<.001). Also shown in Table 1 are measures of respondents perceived influence of alcohol advertising on themselves and on other college students. A comparison between the two shows clear support to the third-person hypothesis: Respondents

PAGE 41

34 tended to perceive greater influence of alcohol advertising on other college students (Mean OTHER = 2.94, SD = .66) than on themselves (Mean SELF = 2.02, SD = .83) (t = -22.14, df = 422, p<.001). Additional paired t-tests show significant self-other differences for product purchase (t = -19.34, df = 425, p<.001), product consumption (t = -21.62, df = 427, p<.001), and attitude toward products (t = -17.70, df = 428, p<.001). Model Fitting and Tests of Hypotheses The hypothesized relationships depicted in Figure 1 were examined through structural equation modeling (SEM) analysis with AMOS, which estimates parameters reflecting the proposed causal relations among constructs and the relationships between constructs and their indicators. Table 2. Measurement Model Results Standardized Standard Latent Constructs & Indicators Factor Loading Error Attitude Toward Alcohol Advertising (ATTA) Like alcohol advertising (X1 ) .857 ** ---Have no problem with alcohol advertising (X2 ) .833 ** .084 Identify with characters and situations in alcohol advertising (X3 ) .704 ** .094 Attitude Toward Alcohol Products (ATTP) Like liquor, beer or wine (X4 ) .693 ** ---Favorable opinion of alcoholic beverages (X5 ) .717 ** .078 Perceived Influence of Alcohol Advertising on Self (SELF) Influence on purchase of alcohol products (X6 ) .866 ** ---Influence on consumption of alcohol products (X7 ) .862 ** .044 Influence on attitude toward alcohol products (X8 ) .773 ** .047 Perceived Influence of Alcohol Advertising on OTHERS (OTHER) Influence on purchase of alcohol products (X9 ) .877 ** ---Influence on consumption of alcohol products (X10 ) .875 ** .051 Influence on attitude toward alcohol products (X11 ) .788 ** .056 Attitude Toward Restrictions on Alcohol Advertising (ATTR) More restrictions on alcohol advertising (X12 ) .790 ** ---Shouldnt advertise to college students (X13 ) .775 ** .085 Shouldnt advertise to people under 21 (X14 ) .583 .081 Banning alcohol advertising from some media (X15 ) .716 ** .094 Banning alcohol advertising from all media (X16 ) .793 ** .075 ** p<.01, p<.05

PAGE 42

35 Measurement Model Evaluation Standardized factor loadings and their standard errors for construct indicators are presented in Table 2. The indicator loadings for all constructs are generally high and statistically significant at the .05 level. Also, the standard errors are generally small, demonstrating acceptable validity of the measurement model. Structural Model Results Analysis An initial question is whether the structural equation analysis estimates for the model provide adequate fit to the data. Although the Chi-square test indicates lack of model fit (X 2 = 208.09, df = 81, p = .000), it should be noted that the Chi-square test is sensitive to large sample sizes, like the one employed in the present study.Our assessment of the models fit thus relied on other goodness-of-fit indices. Bryne (2001) suggests that models with GFI, AGFI, and CFI values greater than .90, and a RMSEA less than or equal to .10 be judged as providing a reasonable fit to the data. Similarly, Hu and Bentler (1999) recommend RMSEA values below .06 and TLI value of .95 or higher. In this study, all these goodness-of-fit measures (GFI = .952; AGFI = .918; CFI = .969, TLI = .954, RMSEA =.057) indicate that the model provides acceptable fit to the data.

PAGE 43

36 Table 3. Structural Model Results Standardized Standard Critical Path Path Coefficient Error Ratio ATTA SELF .328 ** .113 3.080 ATTA OTHERS .126 .086 1.122 ATTA ATTR .419 ** .104 4.144 ATT P SELF .241 .104 2.117 ATTP OTHERS .284 .082 2.288 ATTP ATTR .385 ** .101 3.377 SELF ATTR .123 .058 2.066 OTHERS ATTR .100 .065 2.063 SELF OTHERS .485 ** .047 7.510 ATTA ATTP .784 ** .058 8.876 X 2 = 208.09, df = 81, p = .000; GFI = .952; AGFI = .918; NFI = .951; CFI = .969; RMSEA = .057 ** p<.01, p<.05 Table 3 shows the estimates of structural model parameters. With the only exception of the link from ATTA to OTHERS, all other causal paths hypothesized between the latent constructs are statistically significant at the .05 level. These include the direct causal link between ATTA and ATTR (path = -.419, p<.01), and the indirect path from ATTA to ATTR via SELF (ATTA SELF =.328, p<.01; SELF ATTR=.123, p<.05). Thus, hypotheses 6a and 7a were supported; and hypothesis 8a was nullified, indicating a stronger first-person effect from ATTA. Similar to ATTA, ATTP also had a direct negative influence on ATTR (-.385, p<.01). Unlike that of ATTA, the influence of ATTP on ATTR was mediated by both SELF (ATTP SELF=.241, p<.05; SELF ATTR=.123, p<.05) and OTHERS (ATTP OTHERS= -.284, p<.05; OTHERS ATTR=.100, p<.05). This supports hypotheses 6-b, 7-b and 8-b, and implies that third-person effect is greater when ATTP is isolated.

PAGE 44

37 Consistent with the behavioral hypothesis of the third-person effect, there was a weak but significant path leading from OTHERS to ATTR (path=.100, p<.05). That is, the greater the perceived influence of alcohol advertising on others, the more one would support restrictions on alcohol advertising. Somewhat surprisingly, such behavioral consequence was also observed between SELF and ATTR (path=.123, p<.05): The greater the perceived influence of alcohol advertising on oneself, the more one would support restrictions on alcohol advertising. This finding is not consistent with any previous third person effects research and represents an important discovery in firstperson effects. These figures support hypotheses 2-a and 2-b. To determine if the strength of the path from SELF to ATTR is significantly greater than the path from OTHERS to ATTR, we imposed equality constraints to the two paths. Results show that treating these two paths as equal did not significantly worsen the fit (the difference in X 2 between the model with and without the equality constraints is .08; with one degree of freedom this does not come remotely close to statistical significance). Thus, we may conclude the perceived influence on self and others exert equal amount of influence on ones attitude toward alcohol advertising restrictions.This nullifies hypothesis 3. As predicted, both paths leading from ATTA and ATTP to SELF were positive (.328 and .241, respectively), while the paths leading from ATTA and ATTP to OTHERS were negative (-.126 and -.284, respectively). In other words, the favorable ATTA and ATTP constituted the source of positive influence of alcohol advertising on oneself, but the source of negative influence on others, supporting theories of both a

PAGE 45

38 firstand third-person effect. These results support hypotheses 4-a and 5-a, but nullify hypotheses 4-b and 5-b. The significant and positive path (.485, p<.01) between SELF and OTHERS lends strong support to the looking glass perception hypothesis. That is, students who participated in this study used the perceived influence of alcohol advertising on themselves as the basis for their assessment of the influence of alcohol advertising on other students. The stronger the perceived influence on oneself, the stronger perceived influence on others. Hypothesis 1 is supported. Table 4. Estimates of Indirect, Direct, and Total Effects Total Total Total Indirect Effect Direct Effect Causal Effect ATTA ATTR -.249 -.419 -.668 ATTP ATTR -.280 -.385 -.665 ATTA SELF .189 .328 .517 ATTP SELF .257 .241 .498 ATTA OTHERS .251 -.126 .125 ATTP OTHERS .143 -.284 -.141 SELF ATTR .049 .123 .172 OTHERS ATTR ---.100 .100 SELF OTHER ---.485 .485 Considering the path ATTA ATTR, the indirect effects are calculated by multiplying the path coefficients for each path from ATTA to ATTR, thus: ATTA SELF ATTR is .328*.123 = .040 ATTA SELF OTHERS->ATTR is .328*.485*.100 = .016 ATTA OTHERS ATTR is -.126*.100 = -.013 ATTA ATTP SELF ATTR is .784*.241*.123 = .023 ATTA ATTP SELF OTHERS ATTR is .784*.241*.485*.100 = .009 ATTA ATTP OTHERS ATTR is .784*-.284*.100 = -.022 ATTA ATTP ATTR is .784*-.385 = -.302 Total indirect effect ATTA ATTR = .040+.016-.013+.023+.009-.022-.302 = -.249 Total direct effect ATTA ATTR = -.419 Total causal effect = (-.249)+(-.419) = -.668Table 4 presents the estimates of the indirect, direct and total effects. ATTA and ATTP showed the strongest total effects on ATTR (-.668 and -.665 respectively). ATTA and ATTP also had strong positive effects on SELF (.517 and .498, respectively), suggesting that as the liking for alcohol advertising and products increase, so will the perceived influence of alcohol advertising on oneself. In comparison, the influence of ATTA and ATTP on OTHERS appeared to be much weaker (.125 and

PAGE 46

39 .141 respectively). The total effect of SELF on ATTR (.172) was also greater than the effect of OTHERS on ATTR (.100). Together, these results seem to suggest that the perceived influence of alcohol advertising on oneself (i.e., the first-person effect) is more powerful than the perceived influence of alcohol advertising on others (i.e., the third-person effect). It should be noted, however, that the influence of SELF on ATTR consisted of the direct effect from SELF to ATTR (.123) and the indirect effect via OTHERS (.049). In-Depth Key Path Analysis Figure 3focuses on the paths between ATTA, SELF, and ATTR. Each path in this figure was validated and represents a significant relationship between variables. That is, ATTA not only has a direct and significant inverse relationship with ATTR, but also the path between ATTA and ATTR mediated through the SELF was a significant and valid path as well. In other words, as a persons attitude about an advertisement becomes more favorable, their support for limiting those ads becomes very unlikely as to be expected. However, when the same individual reflects upon the potential effects of alcohol advertising on themselves and their attitudes and behaviors, they do indeed agree that restricting such ads is a good idea, even though they may still have favorable .123* .328* -.419** SELF F ATTA ATTR Figure 3. Portion A of Path Diagram

PAGE 47

40 opinions of said ads. This particular relationship represents an interesting finding in that ones opinion on censorship can be manipulated. The respondents in this study, start out with attitudes against censorship in the simple case of treating ATTA as an isolated factor in determining ATTR. Afterfactoring in the effects of the ads on SELF. the sample begins to favor restrictions for advertising, and exhibit a first-person effect. Figure 4 adds the variable ATTP to the previous figure for an overall evaluation of the effects on SELF. We can see from this diagram that again, the direct inverse relationship between ATTP and ATTR is very strong. The outcome is expected and shows a favorable opinion about alcohol products makes it less likely the respondent will have attitudes supportive of censorship. Attitudes toward the product also showed significant signs of a positive relationship to perceived effects of alcohol advertising on SELF. In other words, there exists a causal link in the perception of advertising effects on SELF whenestablished product attitudes exist. Although the relationship discovered between SELF and ATTP was not as strong as the relationship between ATTA and SELF, there is still a causal link. SELF ATTA ATTR .123* .328* -.419** .241* ATTP -.385** Figure 4. Portion B of Path Diagram

PAGE 48

41 Figure 5 focuses on the paths between ATTP, OTHERS, and ATTR. Each path in this figure was validated and represents a significant relationship between variables. That is, ATTP not only has a direct and significant inverse relationship with ATTR, but also the path between ATTP and ATTR mediated through the OTHERS was a significant and valid path as well. In other words, as the respondents attitude of alcohol products become more favorable, their attitude toward restricting alcohol advertising became less favorable. This was an expected result and follows the logic that if you like products, you dont want their messages restricted. Also important is the favorable attitudes toward restricting such ads when ATTP is mediated through OTHERS proving a third-person effect. even though they enjoy the products, the respondents favored restricting ads that they perceived influenced OTHERS, shifting attitudes about restrictions toward the positive. .100* ATTR OTHER ATTP -.385** -.284* **p<.01, *p<.05 Figure 5. Portion C of Path Diagram

PAGE 49

42 Figure 6 takes a closer look at the respondents perception of effects on OTHERS. Several factors have bearing on what the sample believes effects OTHERS. One factor is the perceived effectson SELF. Consistent with the looking-glass theory, in this model, the SELF-OTHERS relationship is very real and significant, meaning that what we believe to be true with our selves, we also believe to be true with others. We see our perceptions and beliefs to be consistent with the opinions and beliefs of those around us. However, note from earlier results discussion that respondents believed OTHERS to be more effected by alcohol advertising than SELF; thus, the looking-glass perception findings strengthen the third-person effect in this study to an even greater degree by separating the effects on SELF from the effects on OTHERS when comparison opinions and beliefs seem to be equivalent. Figure 6 also highlights the very important finding that the respondents ATTA has no effect or relationship with perceived influence from the alcohol ads on OTHERS. In other words, the respondents opinion of the ads had no effect on the perceived effects of the ads on other people. This finding is extremely important in third-person effects research because previous ATTA .100* ATTR OTHER ATTP -.284* **p<.01, *p<.05 SELF -.126 .485** Figure 6. Portion D of Path Diagram

PAGE 50

43 findings have found favorable opinions of media content to lead to a first-person effect (happening here when ATTA has greater perceived effects on SELF than OTHERS).

PAGE 51

44 Chapter 5 Discussion and Recommendations Discussion The question of most importance to the researchers was: What will happen, a firstor third-person effect? It seems in this case that both events took place simultaneously. Overall, the third-person effects strength was greater than that of the first-person effect in most instances except when considering ATTA. Only the firstperson effect took place in relation to ones attitude toward the ad (and the sample had very positive attitudes toward alcohol ads). At the same time, there was also a significant third-person effect happening within the structural model which may be due to the nature of the product examined, alcohol. People recognized effects on self from the enjoyable ads and products, yet still wanted to protect those others that they felt were influenced a great deal by advertisingto drink or buy alcohol. Considering the influence perceived on self and others, some forms of restrictions were supported by the sample. These include restricting ads aimed at college students and protecting those under 21 from exposure to alcohol ads. Outright bans or complete censorship were not supported. Results of the present study showed that the perceived influence of alcohol advertising on self and others mediated the effects of attitude toward alcohol advertising and attitude toward alcohol products on college students views of restrictions of alcohol advertising. Unlike previous studies that often investigated the effects of alcohol

PAGE 52

45 advertising in a social vacuum, this study provides some initial evidence that alcohol advertising works, to a significant extent, by affecting how we think about others. Indeed, a more complex picture emerges when others are involved. Experimental research often assumes the direct influence of advertising on the individual; studies which took into account of the social context often find advertising to have both direct and indirect effects. Implications The present study also suggests that effective interventions to reduce alcohol abuse may require simultaneously addressing the two faces of alcohol advertising effects: Decrease the perceived positive, first-person effect of alcohol advertising, and increase its perceived negative, third-person effect. Communication designed to counter alcohol advertising should (1) strengthen and channel the negative consequences of alcohol abuse on significant others, while at the same time (2) reduce the positive feelings and emotions induced by alcohol advertising. In agreement with the emphasis on the third person, the general strategy of strengthening social norms against undesirable behavior has proven effective in relation to drinking-driving behavior. Recommendations Findings of the present study also provided the theoretical foundation for studying attitudinal antecedents to the firstand third-person effect in advertising. They suggest that the effect advertising achieves is not only due to any direct persuasive influence of the message itself, but also to the behavior of those persons who anticipate,

PAGE 53

46 or think they perceive, some reaction on the part of others, and behave differently as a result. People react to advertising depending on how they think other people understand the communication. In other words, peer and reference group pressure can be a powerful determinant in whether a person is likely to deny that a communication has had a persuasive impact on them. The findings presented here can be used by many organizations or government agencies, including consumer-protection groups, legislative bodies, and private and public industries that market alcohol beverages. 1. For Consumer Protection Groups: This study reveals important elements of communication that can be used to counter the effects of likeable alcohol advertisingan underage audience. These elements include: employing images or text to enlighten college students of the tactics used to sway their minds favorably toward the product, or using a call-to-action in messages asking the audience to participate in self-reflection after experiencing ads for alcohol products. The self-reflection should be to decide what effects, if any, had happened during the advertisement. Employing such elements would be most beneficial for counter-advertising, responsibility marketing or public service announcements. Each of these three categories of advertising spends far less money on advertising and marketing each year in the U.S., than do alcohol companies, but as this study proves, limited exposure (asking the students to self-reflect once) can still be effective, just one exposure to such an ad may be enough for most individuals, especially college students to turn to self-reflection and question the effects of said ads. As stated in the beginning of this study, if enough of the college student population forms a boycott of the products that target underage drinkers, these practices will

PAGE 54

47 change out of necessity. The college student market can be the most powerful voice for change in these practices as they represent the most important consumer market in this industry. More research should be done to discover to the lengths through which college students would go to enforce tighter marketing control policies. 2. For Governmental Groups or Legislative Bodies: This study provides evidence that some alcohol beverage companies are negligent in not only their advertising placement, but also in their methods employed to market their goods. Using image ads and always setting the scene for fun, friends and attracting the opposite sex, alcohol companies are lessthan responsible marketers of the benefits, uses or dangers of their products. These methods have been proven to be appealing to younger audiences which is why in previous studies the youth audience said that they couldnt wait to drink so that they too could have as much fun as the actors in the ads. Even in the midst of strict industry guidelines, several companies and brands have been found to consistently violate the marketing codes knowingthat once its in print, it is out there for all to see. There is certainly enough evidence to warrant further and severe restrictions on the marketing practices in this industry, and to enact harsher fines for those companies that continuously violate the marketing codes. 3. For Alcohol Beverage Companies:The companies could also use the findings in this report to continue their successful marketing practices to the college student population. Knowledge can also be had from this study on how to avoid potential restrictions on advertising practices by continuing to push in the direction of the adolescent consumers preferences, and touting the merits of free speech. As we discovered in this study, college students are very much against restricting this form of

PAGE 55

48 communication, until they realize what effects the ads have on themselves and their peers. Most students probably dont realize the effects of the ads on themselves and their peers unless asked to think about it. Probably the most important piece of information for alcohol beverage manufacturers to learn from this study is that there is a great need for accountable marketing practices that mix the idea of fun while keeping the underlying message of responsible drinking alive in all ads. Limitations Although the structural equation modeling (SEM) approach could be rather effectively used to examine the relationships among multiple variables simultaneously, as illustrated by the present study, it should be noted that ultimately the SEM analysis deals with correlation, not causation of variables (Everitt and Dunn 1991). The arrows in structural models do indeed reflect hypotheses about causation. However, many models may be consistent with a given dataset. SEM analysis merely illuminates the extent to which a particular model, derived from theory, is consistent with the pattern of correlations found in the data. The competing theories may be represented in separate path models with separate path analyses, or may be combined in a single path diagram, in which case the researcher is concerned with comparing the relative importance of different paths within the diagram. Future research should thus attempt to the test alternative models to better determine the validity of alternative theoretical explanations and predictions.

PAGE 56

49 References Agostinelli, G., & Grube, J. W. (2002). Alcohol Counter-Advertising and the Media: A Review of Recent Research Alcohol Research and Health, 26 (1), 15-23. Ajzen, I., & Fishbein, M. (1980).Understanding Attitudes and Predicting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall. Andsager, J., Weintraub A.E., & Pinkleton, B. E., (2002). Gender as a Variable in Interpretation of Alcohol-Related Messages. Communication Research, 29 (3), 246-469. Banning, S.A., (2001). Do You See What I See?: Third Person Effects on Public Communication Through Self-Esteem, Social Stigma, and Product Use. Mass Communication & Society, 4 (2), 127-147. Bartos, R. & Dunn, T. (1974). Advertising and Consumers: New Perspectives, New York: American Association of Advertising Agencies. Bauer, R. A., & Greyser, S.A. (1968). Advertising in America: The Consumer View, Boston: Harvard University, Graduate School of Business Administration. Blair, M., &Rosenberg, K. E. (1994). Convergent Findings Increase Our Understanding of How Advertising Works. Journal of Advertising Research, 34 35-45. Blood, D. and L. B. Snyder (1991). Marketing Alcohol to Young People: Effects, Responses, Evaluations, and Prospects. Paper prepared for the WHO international technical meeting on Marketing and Promotion of Alcohol to Young People, Valencia, Spain, May 7-9, 2002. Byrne, B. M. (2001). Structural equation modeling with AMOS: Basic concepts, applications, and programming, Mahwah, NJ: Erlbaum. Center on Alcohol Marketing and Youth (2003). Alcohol Advertising On Television, 2001 to 2003: More of the Same. Georgetown University, Washington D.C. Center for Science in the Public Interest (CSPI) Survey (2001), Retrieved August 31, 2005 from http://www.cspinet.org/booze/liquorads_poll.html Calfee, J.E. & Ringold, D.J. (1988). Consumer Skepticism and Advertising Regulation: What do the Polls Show? Advances in Consumer Research,15. M.J. Houston ed., Provo, UT: Association for Consumer Research, 244-248.

PAGE 57

50 Chang, C. (2002), Self-Congruency as a Cure in Different Advertising-Processing Contexts. Communication Research, 29 (5), 503-536. Collins, R.L., Ellickson, P.L., Hambarsoomians, K., & McCaffrey, D.F. (2005). Does Alcohol Promote Adolescent Drinking? Results from a Longitudinal Assessment. Addiction, 100 (2), p.235. Curlo, E., & Chamblee, R. (1998). Ad Procession and Persuasion: The Role of Brand Identification. Psychology & Marketing, 15 (3), 279-299. David, P., Liu, K., & Myer, M. (2004). Methodological Artifact or Persistent Bias: Testing the Robustness of the Third-Person Effects for Alcohol Messages. Communication Research, 31 (2), 206-233. Davidson, W.P. (1983). The Third-Person Effect in Communication. Public Opinion Quarterly, 47 1-15. Eveland, W.P., & McLeod, D.M. (1999). The Effect of Social Desirability on Perceived Media Impact: Implications for Third-Person Perceptions. International Journal of Public Opinion Research, 11 (4), 315-333. Eveland, W.P., Nathanson, A.I., Detenber, B.H., & McLeod, D.M. (1999). Rethinking the Social Distance Corollary: Perceived Likelihood of Exposure and the ThirdPerson Perception. Communication Research, 26 (3), 275-302. Everitt, B. S., & Dunn, G. (1991). Applied multivariate data analysis, London: Edward Arnold. Fields, J. M. & H. Schuman (1976). Public Beliefs about Beliefs of the Public. Public Opinion Quarterly, 40 427-448. Garfield, C.F., Chung, P.J., & Rathouz, P.J. (2003). Alcohol Advertising in Magazines and Adolescent Readership. Journal American Medical Association, 289 (18), 2424-2429. Gunther, A. C. (1991). What We Think Others Think. Communication Research, 18 (3), 355-372. Gunther, A.C.,& Mundy, P. (1993). Biased Optimism and the Third-Person Effect. Journalism Quarterly, 70 58-67. Gunther, A.C.,& Thorson, E. (1992). Perceived Persuasive Effects of Product Commercials and Service Announcements. Communication Research, 19 (5), 574596.

PAGE 58

51 Hingson R. W., Heeren T., Zakocs R.C., Kopstein A., & Wechsler H. (2002). Magnitude of alcohol-related mortality and morbidity among U.S. college students ages 18. Journal of Studies on Alcohol 63 (2):136. Hu, L. & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6 (1), 1-55. James, W. L., & Kover, A. J. (1992). Observations: Do overall attitudes toward advertising affect involvement with specific advertisements? Journal of Advertising Research, 32 (5), 78-83. Jewel, J., Hupp, S., & Luttrell, G. (2004). The Effectiveness of Fatal Vision Goggles: Disentangling Experiential Versus Onlooker Effects. Journal of Alcohol and Drug Education, 48 63-84. Kelly, K.J., & Edwards, R.W. (1998,Spring,). Image Advertisements for Alcohol Products: Is There Appeal Associated With Adolescents Intention to Consume Alcohol. Adolescence,33(129), 47-59. Kelly, K.J., Detenber, B.H., & Eveland, Jr., W.P. (1999 August). Behind the ThirdPerson Effect: How People Generate Media Impact Assessments and Link Them to Support for Censorship. Paper presented at the annual meeting of the AEJMC, New Orleans, LA. Kuo, M., Wechsler, H., Greenberg, P., & Lee, H. (2003). The Marketing of Alcohol to College Students: The Role of Low Prices and Special Promotions. American Journal of Preventative Medicine, 25 (3), 204-211. Lutz, R.J.(1985). Affective and Cognitive Antecedents of Attitude towards the Ad: A conceptual framework. In L. Allowed, & A. Mitchell (Ed.), Psychological Processes and Advertising Effects: Theory, Research, and Applications (pp.45-63). Hillsdale, NJ: Erlbaum. McLeod, D. M., Detenber, B.H. & Eveland, Jr., W.P. (2001, December). Behind the Third-Person Effect: Differentiating Perceptual Processes for Self and Other. Journal of Communication, 678-694. Mitchell, A.A., & Olson, J. (1981). Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude. Journal of Marketing Research, 8 318-332. Nelson, J. P. (2003). Advertising Bans, Monopoly, and Alcohol Demand: Testing for Substituion Effects using State Panel Data. Review of Industrial Organization, 22 1-25.

PAGE 59

52 OHara, J.A. III (2003). Drops in the Bucket: Alcohol Industry Responsibility Advertising on Television in 2001. Center on Alcohol Marketing and Youth, Georgetown University, Retrieved June 2004 from: http://camy.org/research/drops0203/ Paul, B., Salwen, M.B., & Dupagne, M. (2000). The Third-Person Effect: A MetaAnalysis of the Perceptual Hypothesis. Mass Communication & Society, 3 (1), 5785. Perkins, H.W. &. Berkowitz, A.D. (1986). Perceiving the Community Norms of Alcohol Use Among Students: Some Research Implications for Campus Alcohol Education Programming. International Journal of Addiction, 21 (9-10), 961-977. Petty, C. E. &. Cacioppo, J.T. (1983). Central and Peripheral Routes to Persuasion: Application to Advertising. In L. Percy & A. Woodside (Ed.), Advertising and Consumer Psychology (3-23). Lexington, MA: Lexington Books. Perloff, R.M. (1989). Ego-Involvement and the Third-Person Effect of Televised News Coverage. Communication Research, 16 236-262. Perloff, R.M. (1996). Perceptions and Conceptions of Political Media Impact: the Third-Person effect and Beyond. Psychology of Political Communication 177-179. Phelps, J.E., & Grubbs Hoy, M. (1996). The Aad-Ab-PI Relationship in Children: The Impact of Brand Familiarity and Measurement Timing. Psychology & Marketing, 13 (1), 77-105. Pollay, R.W., & Mittal, B. (1993, July). Here's the Beef., Factors, Determinants, and Segments in Consumer Criticism of Advertising. Journal of Marketing, 57 99114. Rimal, R.N., & Real, K. (2005). How Behaviour are Influenced by Percieved Norms: A Test of the Theory of Normative Social Behavior. Communication Research, 33 (3), 389-414. Rojas, H., Shah, D.V., & Faber, R.J. (1996). For the Good of Others: Censorship and the Third-Person Effect. International Journal of Public Opinion Research, 8 163186. Saffer, H. (2002). Alcohol Advertising and Youth. Journal of Studies on Alcohol, 63 (14), 173-181. Saffer, H., & Dave, D. (2002). Alcohol Consumption and Alcohol Advertising Bans. Applied Economics, 1325 1325-1334.

PAGE 60

53 Salwen, M.B. (1998). Perceptions of Media Influence and Support for Censorship: The Third Person Effect in the 1996 Presidential Election. Communication Research, 25 259-285. Salwen, M.B., & Dupagne, M. (1999). The Third-Person Effect: Perceptions of the Medias Influence and Immoral Consequences. Communication Research, 26 (5), 523-549. Shah, D.V., Faber, R.J., & Youn, S. (1999). Susceptibility and Severity: Perceptual Dimensions Underlying the Third-Person Effect. Communication Research, 26 (2), 240-267. Shen, F. (1998). Assessing Advertising Effectiveness: A Comparison of Two RealTime Measures of Ad-Liking. Paper presented at the annual AEJMC conference, Baltimore, MD, August, 1998. Snyder, L.B., Fleming Milici, F., Mitchell, E.W., & Proctor, D.C.B. (2000). Media, Product Differences and Seasonality in Alcohol Advertising in 1997. Journal of Studies on Alcohol, 61 (6), 896-913. Thorson, E. (1991). Likability: Ten Years of Academic Research. ARF Copy Research Workshop, New York. Tyler, T.R. & Cook, F.L. (1984). The Mass Media and Judgements of Risk: Distinguishing Impact on Personal and Societal Levels. Journal of Personality and Social Psychology, 74 693-708. Wagenaar, A.C., Murray, D.M., & Geban, J.P. (2000). Communities Mobilizing for Change: Outcomes from a Randomized Community Trial. Journal of Studies on Alcohol, 61 (1):85-94. Wechsler, H., Kelley, K., Weitzman, E.R., San Giovanni, J.P., & Seibring, M. (2000). What Colleges Are Doing About Student Binge Drinking A Survey of College Administrators. Journal of American College Health, 48 (5), 219-226. Wyllie, A. H. (1997). Considering Interacting Factors in the Third-Person Effect: Argument Strength and Social Distance. Journal and Mass Communications Quarterly, 74 (3), 557-564. Wyllie, A.H., Fang, J., & Casswell, S. (1998a). Responses to Televised Alcohol Advertisements Associated with Drinking Behavior of 10-17-Year-Olds. Addiction 93 (3), 361-371.

PAGE 61

54 Wyllie, A.H., Fang, J., & Casswell, S. (1998b). Positive Responses to Televised Beer Advertisements Associated with Drinking and Problems Reported by 18 to 29-YearOlds. Addiction 93 (5), 749-760 Yates, B.L. (2001). Media Literacy and Attitude Change: Assessing the Effectiveness of Media Literacy Training on Childrens Responsibility Advertising. Paper presented at annual AEJMC conference,2001.

PAGE 62

55 Bibliography Agostinelli, G., & Grube, J.W. (2002). Alcohol counter-advertising and the media: a review of recent research. Alcohol Research and Health, 26 (I),15-21. Andasager, J.L., Austin, E.W. & Pinkleton, B.E. (2001, March). Questioning the value of realism: young adults processing of messages in alcohol-related public service announcements and advertising. Journal of Communication 121-128. Andsager, J., Weintraub-Austin, E., Pinkleton, B. (2002). Gender as a Variable in Interpretation of Alcohol-Related Messages. Communication Research, 29 (3), 609647. Author unknown (2003). Debate over broadcast liquor ads still fizzing. New Zealand Financial Times Information May 8, 2003. Banning, S.A. (2001). Do you see what I see?: third-person effects on public communication through self-esteem, social stigma, and product use. Mass Communication & Society, 4 (2), 127-147. Banning, S.A. (2001). Third-person effect suppressor variables in program evaluations. Web Journal of Mass Communications Research, 4 :3. Banning, S. (2002). The Credibility Connection: Discovery of a Connection Between Credibility and the Third-Person Effect with Newspaper Stimuli. Paper submitted to the Communication Theory and Methodology Division of the Association for Education in Journalism and Mass Communication, Baton Rouge. Chapin, J.R. (1999). Third-person perception and sexual risk taking among minority atrisk youth. Mass Communication & Society, 2(3/4), 163-173. Cho, C.H. (1999). How advertising works on the WWW: modified elaboration likelihood model. Unpublished paper, University of Texas at Austin. Dupagne, M., & Salwen, M.B. (1999). The third-person effect: perceptions of the medias influence and immoral consequences. Communication Research, 26 (5), 523-549. Eveland, Jr., W.P. (1997). Interactions and nonlinearity in mass communication: connecting theory and methodology. Journalism & Mass Communications Quarterly, 74 (2), 400-416.

PAGE 63

56 Eveland, W. & McLeod, D. (1999). The Effect of Social Desirability on Percieved Media Impact: Implications for Third-Person Perceptions. International Journal of Public Opinion Research, 11 (4), 315-333. Fursich, E. & Roushanzamir, E.P.L. (2001). Corporate expansion, textual expansion: commodification model of communication. Journal of Communication Inquiry, 25( 4), 375-395. Gaziano, C. (1983). The knowledge gap: an analytical review of media effects. Communication Research,10(4), 447-486. Gunther, A., & Thorson, E. (1992). Perceived Persuasive Effects of Product Commercials and Public Service Announcements: Third-Person Effects in New Domains Communication Research,19(5), 574-596. Gunther, A., (1991). What We Think Others Think: Cause and Consequence in the Third-Person Effect. Communication Research, 18 (3), 355-372. Huh, J., Delorme, D., & Reid, L. (2004). The Third-Person Effect and Its Influence on Behavorial Outcomes in a Product Advertising Context: The Case of Direct-toConsumer Prescription Drug Advertising. Communication Research, 31 (5), 568599. Hu, Y.-W. (1999). The Third-Person Effect: Social Cognition or Academic Creation? Paper submitted to the Communication Theory and Methodology Division of the Association for Education in Journalism and Mass Communication, New Orleans, August 4-7, 1999. Hu, Y.-W. (2000). Communication Behavior as a Critical Factor on the Third-Person Effect. Paper submitted to the Communication Theory and Methodology Division of the Association for Education in Journalism and Mass Communication, Phoenix, Az. August 9-12. Kelly, K.J., & Edwards, R.W. (1998, Spring). Image advertisements for alcohol products: is there appeal associated with adolescents intention to consume alcohol. Adolescence, 33 (129), 47-56. Leone, R. (?). Offense and harm as predictors in a third-person effect variation study. Research paper, Stonehill College. Mcleod, D.M., Detenber, B.H., & Eveland, Jr., W.P. (2001, December). Behind the thirdperson effect: differentiating perceptual processes for self and other. Journal of Communication 678-693.

PAGE 64

57 Neuwirth, K., Frederick, E., & Mayo, C. (2002). Person-Effects and HeuristicSystematic Processing. Communication Research, 29 (3), 320-359. Neuwirth, K., & Frederick, E. (2002). Extending the framework of the third-, first-, and second-person effects. Mass Communications & Society, 5 (2), 113-140. Paek, H.-J., Pan, Z., Sun, Y., Abisaid, J., Houden, D., (2005). The Third-Person Perception as Social Judgement: An Exploration of Social Distance and Uncertainty in Percieved Effects of Political Attack Ads. Communication Research, 32 (2), 143170. Paul, B., Salwen, M.B., & Dupagne, M. (2000). The third-person effect; a meta-analysis of the perceptual hypothesis. Mass Communication & Society, 3 (1), 57-85. Petty, R.E., Wegener, D.T., Fabrigar, L.R. (1997). Attitudes and attitude change. Annual Review of Psychology,48, 609-647. Salwen, M. & Dupagne, M. (1999). The Third-Person Effect: Perceptions of the Medias Influence and Immoral Consequences. Communication Research,26(5), 523-549. Shen, F. (1998). Assissing advertising effectiveness: a comparison of two real-time measures of ad liking. Paper presented at the annual AEJMC conference, Baltimore, MD, August 1998. Snyder, L.B., Milici, F.F., Mitchell, E.W., & Proctor, D.C.B. (2000). Media, product differences and seasonality in alcohol advertising. Journal of Studies on Alcohol, 61 (6), 896-712. Stephenson, M.T., Benoit, W.L., & Tschida, D.A. (2001). Testing the mediating role of cognitive responses in the elaboration likelihood model. Communication Studies, 52 (4), 324-337. White, H.A. (1997). Considering interacting factors in the third-person effect: argument strength and sd. Journalism & Mass Communications Quarterly, 74 (3), 557-564. Wood, W. (2000). Attitude change: persuasion and social influence. Annual Review of Psychology, 51 539-570. Worsnop, R.L. (1997). Alcohol advertising: should liquor be advertised on radio and TV? CQ Researcher, 7 (10), 219-220. Yates, B.L. (2001). Media literacy and attitude change: assessing the effectiveness of media literacy training on childrens response to advertisements. Meeting papers. Yoon, K., Kim, C.H., Kim, M.S. (1998). A cross-cultural comparison of the effects of source credibility on attitudes and behavioral intentions. Paper presented to the advertising division of the AEJMC.

PAGE 65

58 Appendices

PAGE 66

59 X 16 X 15 X 13 X 12 X 14 X 11 X 10 X 9 X 8 X 7 X 6 X 5 X 4 X 3 X 2 X 1 ATTA ATTP OTHERS ATTR SELF Appendix A Expanded Path Diagram Figure 1. Path Diagram of Hypothesized Relationships ATTA: Attitude toward Alcohol Advertising, ATTP: Attitude toward Alcoholic Beverage Products SELF: Perceived Influence of Alcohol Advertising on Self OTHERS: Perceived Influence of Alcohol Advertising on Others ATTR: Attitude toward Restrictions of Alcohol Advertising

PAGE 67

60 Appendix B Survey Questionnaire Were conducting a study of students perception of advertising. Would you mind helping us by answering a few questions? If REFUSED---- Thank you for your time. If YES---- Great! ----Go to the first question. Let me begin by asking how you feel about some statements regarding alcohol advertising, including ads forbeer, wine, wine coolers, liquor, and mixed drinks. Please tell us whether you Strongly Agree, Agree, Neither Agree nor Disagree, Disagree, or Strongly Disagree with each statement. Strongly Agree Agree Neither Agree nor Disagree Disagree Strongly Disagree D/K 1. I dont have a problem with alcohol advertising 5 4 3 2 1 9 2. I like alcohol advertising 5 4 3 2 1 9 3 I identify with the characters and situations portrayed in alcohol advertising 5 4 3 2 1 9 4. Alcohol advertising increases my desire to use the product 5 4 3 2 1 9 5. Alcohol advertising makes drinking fun 5 4 3 2 1 9 6. I have a favorable opinion of most alcoholic beverages 5 4 3 2 1 9 7. I don't like liquor, beer or wine products 5 4 3 2 1 9 8. I am concerned about alcohol advertising directed at college students 5 4 3 2 1 9 9. I think alcohol consumption is a serious issue for college students 5 4 3 2 1 9 10. Some college students dont know how much drinking is too much 5 4 3 2 1 9 11. Alcohol consumption can lead to dangerous outcomes 5 4 3 2 1 9 Wed like to know the extent to which alcohol advertising has an influence on you personally. Do you think alcohol advertising has a great deal of influence some influence very little influence or no influence at all : A great deal Some Very little No influence of influence influence influence at all D/K 12. On your purchase of alcohol products4 3 2 1 9 13. On your consumption of alcohol products4 3 2 1 9 14. On your attitude toward alcohol products4 3 2 1 9 Now wed like to ask you the extent to which alcohol advertising may influence other people. Lets look at USF students first. Do you think alcohol advertising has a great deal of influence some influence very little influence or no influence at all : A great deal Some Very little No influence of influence influence influence at all D/K 15. On USF students purchase of alcohol products4 3 2 1 9 16. On USF students consumption of alcohol products4 3 2 1 9 17. On USF students attitude toward alcohol products4 3 2 1 9

PAGE 68

61 Appendix B (Continued) Survey Questionnaire Now wed like you to think of students at other universities in Florida. Do you think alcohol advertising has a great deal of influence some influence very little influence or no influence at all : A great deal Some Very little No influence of influence influence influence at all D/K 18. On the purchase of alcohol products by students at other universities4 3 2 1 9 19. On the consumption of alcohol products by students at other universities4 3 2 1 9 20. On their attitude toward alcohol products4 3 2 1 9 Please tell us whether you: Strongly Agree, Agree, Neither Agree nor Disagree, Disagree, or Strongly Disagree with the following statements: Strongly Agree Agree Neither Agree nor Disagree Disagree Strongly Disagree D/K 21. Alcohol companies shouldnt advertise to college students 5 4 3 2 1 9 22. People under the age of 21 shouldnt be exposed to alcohol advertising 5 4 3 2 1 9 23. There ought to be more restrictions on alcohol advertising 5 4 3 2 1 9 24. I support the idea of banning alcohol ads from some media forms 5 4 3 2 1 9 25. I support the idea of banning alcohol ads from all forms of media 5 4 3 2 1 9 Finally, wed like to ask a few more questions about you. These are for research purposes only and all information will be kept strictly confidential.26. How often do you drink liquor, wine or beer? Do you often sometimes rarely or never drink? [4] often [3] sometimes [2] rarely [1] never (if never, go to Q 28) 27. Think about your last social drinking occasion, how many drinks did you consume on that occasion? [1] 0 [2] 1 to 2 [3] 3 to 4 [4] 5 to 6 [5] 7 to 8 [6] 9 to 10 [7] 11 or more 28. Are you presently a freshman sophomore junior senior or graduate student at USF? [1] Freshman [2] Sophomore [3] Junior [4] Senior [5] Graduate [6] Other (specify) ____________ 29. What is your age, please? __________ (record in years) 30. Gender of respondent: (1) Male (2) Female (circle answer) Okay, this completes our survey, Thank you very much for your patience and help. Have a great day.

PAGE 69

62 Appendix C Survey Questions and Variables 1. Attitude Toward Alcohol Advertising (ATTA)Q1: I don't have a problem with alcohol advertising (PRO B) Q2: I like alcohol advertising (LIKEA) Q3: I identify with the characters and situations portrayed in alcohol advertising (IDENT) Q4: Alcohol advertising increases my desire to use the product (DESIRE) Q5: Alcohol advertising makes drinking fun (FUN) 2. Attitude Toward Alcoholic Beverage Products (ATTP) Q6: I have a favorable opinion of most alcoholic beverages (FAVOR) Q7: I don't like liquor, beer or wine products (Reverse Coded) (LIKEP)3. Drinking Problem Perception (DPP)Q8: I am concerned about alcohol advertising directed at college students (CONCERN) Q9: I think alcohol consumption is a serious issue for college students (SERIOUS) Q10: Some college students don't know how much drinking is too much (TOOMUCH) Q11: Alcohol consumption can lead to dangerous outcomes (DANGER)4. Effect of Alcohol Advertising on Self (SELF) Q12: On your purchase of alcohol products (SBUY) Q13: On your consumption of alcohol products (SUSE) Q14: On your attitude toward alcohol products (SATT)5. Effect of Alcohol Advertising on Other USF Students (OUSF) Q15: On USF students' purchase of alcohol products (USFBUY) Q16: On USF students' consumption of alcohol products (USFUSE) Q17: On USF students' attitude toward alcohol products (USFATT)6. Effect of Alcohol Advertising on Other College Students (OCOLL) Q18: On the purchase of alcohol products by students at other universities (OBUY) Q19: On the consumption of alcohol products by students at other universities (OUSE) Q20: On their attitude toward alcohol products (OATT)

PAGE 70

63 Appendix C (Continued) Survey Questions and Variables 7. Restrictions (ATTR) or Ban on Alcohol Advertising (BAN) Q21: Alcohol companies shouldn't advertise to college students (NOADV) Q22: People under the age of 21 shouldn't be exposed to alcohol advertising (UNDER) Q23: There ought to be more restrictions on alcohol advertising (RESTRICT) Q24: I support the idea of banning alcohol ads from some media forms (BANSOME) Q25: I support the idea of banning alcohol ads from all forms of media (BANALL)8. Frequency of Alcohol Consumption (FREQ) Q26: How often do you drink liquor, wine or beer? Do you often, sometimes, rarelv, or never drink? (FREQ)9. No. of Drinks Last Occasion (DRINKS) Q27: Think about your last social drinking occasion, how many drinks did you consume on that occasion? (DRINKS)10. Student Status (STATUS) Q28: Are you presently a freshman, sophQ!!!Q@, iunior, senior, or qraduate student at USF? (STATUS)11. Age (AGE) Q29: What is your age, please? (AGE) 12. Gender (GENDER) Q30: Gender of respondent (GENDER)

PAGE 71

64 Appendix D List of Research Questions and Hypotheses RQ1: What is the relationship between perceived influence of alcohol advertising on self (SELF) and on others (OTHERS)? H1: Perceived influence of alcohol advertising on self will be positively related to perceived influence on others. (SELF OTHERS; OTHERS SELF) RQ2: What is the relationship between perceived influence of alcohol advertising (on SELF and OTHERS) and attitude toward greater restrictions on alcohol advertising (ATTR)? H2-a: There will be a positive relationship between SELF and ATTR. (SELF ATTR) H2-b: There will be a positive relationship between OTHERS and ATTR. (OTHERS ATTR) RQ3: What is the relative influence of perceived effect of alcohol advertising on self (SELF) and others (OTHERS) on attitude toward restricting alcohol advertising (ATTR) H3: SELF OTHERS OTHERS ATTR RQ4. What is the relationship between general attitude toward alcohol advertising and the perceived influence of alcohol advertising on self and others? H4-a: Perceived influence of alcohol advertising on self (SELF) will be positively related to attitude toward alcohol advertising (ATTA). (ATTA SELF) H4-b: Perceived influence of alcohol advertising on others (OTHERS) will be positively related to attitude toward alcohol advertising (ATTA). (ATTA OTHERS) RQ5. What is the relationship between general attitude toward (perception of) alcohol products and the perceived influence of alcohol advertising on self and others? H5-a: Perceived influence of alcohol advertising on self (SELF) will be positively related to attitude toward alcohol products (ATTP). (ATTP SELF)

PAGE 72

65 Appendix D (Continued) List of Research Questions and Hypotheses H5-b: Perceived influence of alcohol advertising on others (OTHERS) will be positively related to attitude toward alcohol products (ATTP). (ATTP OTHERS) RQ6. What is the relationship (direct and indirect) between ATTA and ATTP on ATTR? H6-a: There will be a negative relationship between ATTA and ATTR. H6-b : There will be a negative relationship between ATTP and ATTR.RQ7: What is the mediating role of SELF between the variables ATTA/ATTTP with ATTR? H7-a: The indirect relationship from ATTA to ATTR mediated through SELF, will be positive in both legs of the path (ATTA SELF ATTR). H7-b: The indirect relationship from ATTP to ATTR mediated through SELF, will be positive in both legs of the path (ATTP SELF ATTR). RQ8: What is the mediating role of OTHER between the variables ATTA/ATTTP with ATTR? H8-a: The indirect relationship from ATTA to ATTR mediated through OTHER, will be positive in both legs of the path (ATTA OTHER ATTR). H8-b: The indirect relationship from ATTP to ATTR mediated through OTHER, will be positive in both legs of the path (ATTP OTHER ATTR).

PAGE 73

66 Appendix E Frequency Distributions Frequency Table: FREQ Drinking Frequency Frequency Percent Valid Percent Cumulative Percent Never 59 12.9 12.9 12.9 Rarely 107 23.4 23.4 36.2 Sometimes 183 40 40 76.2 Valid Often 109 23.8 23.8 100 Total458 100 100 Frequency Table: Drinks per Occasion Frequency Percent Valid Percent Cumulative Percent 0 67 14.6 14.7 14.7 1 to 2 140 30.6 30.7 45.4 3 to 4 121 26.4 26.5 71.9 5 to 6 54 11.8 11.8 83.8 7 to 8 32 7 7 90.8 9 to 10 12 2.6 2.6 93.4 Valid 11 or more 30 6.6 6.6 100 Total456 99.6 100 Missing 99 2 0.4 Total 458 100 Frequency Table: Student STATUS Frequency Percent Valid Percent Cumulative Percent Freshman 31 6.8 6.8 6.8 Sophomore 76 16.6 16.6 23.4 Junior 128 27.9 27.9 51.3 Senior 178 38.9 38.9 90.2 Graduate 41 9.0 9.0 99.1 Other 4 0.9 0.9 100.0 Valid Total458 100.0 100.0 Frequency Table: Gender Frequency Percent Valid Percent Cumulative Percent Male188 41.0 41.3 41.3 Female 267 58.3 58.7 100.0 Valid Total455 99.3 100.0 Missing .99 3 0.7 Total458 100.0

PAGE 74

67 Appendix F Descriptive Results Descriptive: AGE Descriptive Statistics N Range Minimum Maximum Mean AGE457 26 17 43 21.93 Descriptive: ATTA N Range Min Max Mean SD PROB 456 4 1 5 3.7149 1.1261 LIKEA454 4 1 5 3.152 1.1494 IDENT 453 4 1 5 2.5828 1.1426 DESIRE 453 4 1 5 2.4371 1.1205 FUN 452 4 1 5 2.5442 1.1858 Descriptive: ATTP N Range Min Max Mean SD FAVOR 455 4 1 5 3.2484 1.1974 LIKEP 454 4 1 5 2.1498 1.2951 Descriptive: DPP N Range Min Max Mean SD CONCERN 449 4 1 5 2.882 1.1695 SERIOUS 455 4 1 5 3.5956 1.1062 TOOMUCH456 4 1 5 4.1645 0.8854 DANGER 458 4 1 5 4.321 0.847 Descriptive: SELF N Range Min Max Mean SD SBUY 453 3 1 4 2.0221 0.9286 SUSE 453 3 1 4 1.8985 0.899 SATT 454 3 1 4 2.0969 0.9603

PAGE 75

68 Appendix G Correlations Crltn: ATTA, ATTP, DPP, SELF, OUSF, OCOLL, BAN Mean Std. Deviation N ATTA3.1079 0.9335 448 ATTP3.551 1.0825 451 DPP 3.7388 0.7896 446 SELF 2.003 0.8341 451 OUSF 2.9415 0.6825 427 OCOLL 2.9681 0.6851 407 BAN 2.7761 0.9265 435 Correlations ATTA ATTP DPP SELF OUSF OCOLL BAN ATTAPearson Correlation 1 Sig. (2-tailed) ATTPPearson Correlation 0.625 1 Sig. (2-tailed) 0 DPP Pearson Correlation -0.46 -0.405 1 Sig. (2-tailed) 0 0 SELF Pearson Correlation 0.559 0.375 -0.225 1 Sig. (2-tailed) 0 0 0 0 OUSF Pearson Correlation 0.002 -0.081 0.252 0.366 1 Sig. (2-tailed) 0.962 0.097 0 0 OCOLL Pearson Correlation -0.065 -0.142 0.271 0.241 0.787 1 Sig. (2-tailed) 0.194 0.004 0 0 0 BAN Pearson Correlation -0.562 -0.508 0.56 -0.229 0.187 0.223 1 Sig. (2-tailed) 0 0 0 0 0 0

PAGE 76

69 Appendix H Paired Sample Statistics T-Tests: SELF vs. OTHER USF STUDENTS Pair Samples Statistics Mean N Std. Deviation Pair 1 SBUY 2.028 426 0.9301 USFBUY 2.941 426 0.7121 Pair 2 SUSE 1.909 428 0.8968 USFUSE 2.911 428 0.7631 Pair 3 SATT 2.110 429 0.9544 USFATT2.960 429 0.7739 Pair 4 SELF 2.010 423 0.8343 OUSF 2.939 423 0.6848 Paired Samples Statistics Mean N Std. Deviation Pair 1 SBUY 2.039 407 0.936 OBUY 2.985 407 0.719 Pair 2 SUSE 1.919 408 0.904 OUSE 2.939 408 0.770 Pair 3 SATT 2.128 406 0.963 OATT 2.983 406 0.749 Pair 4 SELF 2.027 401 0.836 OCOLL 2.971 401 0.686 Paired Samples Test t df Sig. (2-tailed) Pair 1 SBUY-OBUY -18.232 406 0 Pair 2 SUSE-OUSE -19.722 407 0 Pair 3 SATT-OATT -16.082 405 0 Pair 4 SELF-OCOLL -20.014 400 0 Paired Samples Test t df Sig.(2-tailed) Pair 1 SBUY-USFBUY -19.336 425 0.0 Pair 2 SUSE-USFUSE-21.615 427 0.0 Pair 3 SATT-USFATT-17.696 428 0.0 Pair 4 SELF-OUSF -22.138 422 0.0

PAGE 77

70 Appendix H (Continued) Paired Sample Statistics Paired Differences 95% Confidence Interval of the Difference Mean SD Std. Error Mean Lower Upper t Pair 1 SBUY-USFBUY -0.901 0.973 0.046 -0.991 -0.811 -19.736 Pair 2 SUSE-USFUSE-0.980 0.965 0.045 -1.069 -0.891 -21.666 Pair 3 SATT-USFATT-0.853 0.997 0.047 -0.945 -0.762 -18.300 Pair 4 SELF-OUSF -0.092 0.863 0.041 -0.998 -0.839 -22.580 Pair 5 SBUY-OBUY -0.094 1.033 0.050 -1.033 -0.838 -18.863 Pair 6 SUSE-OUSE -1.000 1.038 0.050 -1.098 -0.902 -20.085 Pair 7 SATT-OATT-0.857 1.064 0.051 -0.957 -0.756 -16.756 Pair 8 SELF-OCOLL -0.934 0.932 0.045 -1.023 -0.846 -20.708 Pair 9 USFBUY-OBUY -0.049 0.523 0.025 -0.098 0.001 -1.931 Pair 10 USFUSE-OUSE-0.028 0.534 0.026 -0.078 0.023 -1.078 Pair 11 USFATT-OATT-0.014 0.574 0.028 -0.068 0.040 -0.503 Pair 12 OUSF-OCOLL-0.031 0.437 0.021 -0.073 0.010 -1.478 Paired Sample Tests df Sig. (2-tailed) Pair 1 SBUY-USFBUY 453 0.000 Pair 2 SUSE-USFUSE454 0.000 Pair 3 SATT-USFATT456 0.000 Pair 4 SELF-OUSF 449 0.000 Pair 5 SBUY-OBUY 433 0.000 Pair 6 SUSE-OUSE 434 0.000 Pair 7 SATT-OATT432 0.000 Pair 8 SELF-OCOLL 426 0.000 Pair 9 USFBUY-OBUY 431 0.054 Pair 10 USFUSE-OUSE433 0.282 Pair 11 USFATT-OATT431 0.615 Pair 12 OUSF-OCOLL425 0.140

PAGE 78

71 Appendix H (Continued) Paired Sample Statistics Paired Samples Statistics Mean N SD Std. Error Mean Pair 1 SBUY 2.044 454 0.9220 0.0433 USFBUY 2.945 454 0.6963 0.0327 Pair 2 SUSE 1.932 455 0.8916 0.0418 USFUSE 2.912 455 0.7516 0.0352 Pair 3 SATT 2.114 457 0.9506 0.0445 USFATT2.967 457 0.7587 0.0355 Pair 4 SELF 2.024 450 0.8265 0.0390 OUSF 2.943 450 0.6700 0.0316 Pair 5 SBUY 2.058 434 0.9263 0.0445 OBUY 2.993 434 0.7013 0.0337 Pair 6 SUSE 1.943 435 0.8975 0.0430 OUSE 2.943 435 0.7613 0.0365 Pair 7 SATT 2.127 433 0.9550 0.0459 OATT 2.984 433 0.7405 0.0356 Pair 8 SELT 2.041 427 0.8255 0.0400 OEOLL 2.976 427 0.6711 0.0325 Pair 9 USFBUY 2.949 432 0.6895 0.0332 OBUY 2.998 432 0.6930 0.0334 Pair 10 USFUSE 2.915 434 0.7504 0.0360 OUSE 2.942 434 0.7622 0.0366 Pair 11 USFATT2.968 432 0.7609 0.0366 OATT 2.982 432 0.7397 0.0356 Pair 12 OUST 2.950 426 0.6683 0.0324 OEOLL 2.981 426 0.6633 0.0321


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001795811
003 fts
005 20070712133723.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 070712s2005 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001586
040
FHM
c FHM
035
(OCoLC)154709180
049
FHMM
090
P90 (ONLINE)
1 100
Begin, Georgia.
0 245
Attitudinal antecedents of the first- and third-person effects of alcohol advertising on college students
h [electronic resource] /
by Georgia Begin.
260
[Tampa, Fla] :
b University of South Florida,
2005.
3 520
ABSTRACT: Data collected from a survey questionnaire via personal interviews among 488 college students was used to examine relationships among attitude toward alcohol advertising, attitude toward alcohol beverages, perceived influence of alcohol advertising on oneself, perceived influence of alcohol advertising on others, and attitude toward restrictions of alcohol advertising. Results supported the hypothesized direct effect of advertising attitudes and product attitudes on attitude toward restrictions. Also supported was the indirect effect of advertising and product attitudes on attitudes toward restricting alcohol advertising via such mediators as perceived influence of the ads on self and others using the theories of first- and third- person effects. Implications for future research, public policies, and marketing practices including responsibility marketing are discussed.
502
Thesis (M.A.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 71 pages.
590
Adviser: Scott S. Liu, Ph.D.
653
Attitude toward the ad.
Attitude toward the product.
690
Dissertations, Academic
z USF
x Mass Communications
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1586