REINVESTIGATING THE BEAUTY MATCH UP HYPOTHESIS AND SOCIAL
COMPARISON IN FOOD ADVERTISEMENTS
By
Juan Mundel

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Information and Media – Doctor of Philosophy
2018

	

ABSTRACT
REINVESTIGATING THE BEAUTY MATCH UP HYPOTHESIS AND SOCIAL
COMPARISON IN FOOD ADVERTISEMENTS
By
Juan Mundel
The use of attractive models as a means to grab consumers’ attention and influence their
product evaluations and purchase intentions is a common occurrence among advertisers.
Research shows that recurrent exposure to ads featuring models with idealized bodies can lead to
negative self-evaluations, development of eating disorders, and depression, among other negative
outcomes. Given their greater likelihood to express discontent with their bodies when compared
to their male counterparts, most of the literature on this issue has focused on women as the
population under study. However, a third of all individuals suffering from eating disorders in the
U.S. are males. This dissertation explores how males evaluate models featured in snack food
advertisements when their bodies conform (or not) with advertising industry norms, and the
effects of the pairing of different models with products perceived to be healthy (vs. unhealthy) on
participants’ evaluations of the self, the product, and the ad. Results showed a significant
interaction between exposure to idealized bodies in advertisements and upward social
comparison, indicating that respondents with higher social comparison scores had more negative
evaluations of the ads. Exposure to ads with idealized bodies predicts participants’ engagement
in upward social comparison. Further, our results show that upward social comparison was a
significant predictor of body dissatisfaction. Yet, overall participants had better evaluations of
the ads when presented with unhealthy foods and models with idealized bodies, which stresses
the need for guidelines for model casting in advertising.

	

Keywords: social comparison, gender, social media advertising, attractiveness, food advertising

Copyright by
JUAN MUNDEL
2018

ACKNOWLEDGEMENTS
I would like to thank, first of all, my mentors at West Virginia University, in particular
Dr. Diana Martinelli and Dr. Hogmin Ahn, for pointing me in the direction of a Ph.D. I am
equally grateful to Dr. Elizabeth Taylor Quilliam, who encouraged me to apply to the doctoral
program at Michigan State University during her visit to WVU, and who has been a member of
my guidance committee since my first day at Michigan State University.
I am grateful to the faculty at Michigan State University, for providing me with endless
opportunities to explore my abilities and guiding me through the doctoral program. I am deeply
honored to join the long list of Spartan alumni. Particularly, I dedicate special thanks to my
dissertation committee, from which I have learned so much in the research process. To Dr.
Saleem Alhabash, whose mentorship molded my research. Your guidance is invaluable. Dr.
Dawn Pysarchik, for the endless memories abroad, and for teaching me how to be a good
administrator. And to Dr. Esther Thorson, who without knowing me, trusted in my research
agenda.
Further, a special acknowledgment to my dissertation director, and graduate advisor, Dr.
Patricia Huddleston. Since my arrival at Michigan State University, Dr. Huddleston’s friendship
and guidance have been instrumental to my path in becoming a university professor. Thank you
for lending an ear, I am so fortunate to have had such a dedicated advisor.
A very special and heartfelt thank you to Kort, who is there every step of the way.
Wouldn’t have done it without your support. Lastly, special thanks to my mother, who always
told me “you can never fit enough knowledge in your brain” (el saber no ocupa lugar).

	

v

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... viii
LIST OF FIGURES ....................................................................................................................... ix
CHAPTER 1 ....................................................................................................................................1
INTRODUCTION ...........................................................................................................................1
CHAPTER 2 ....................................................................................................................................6
LITERATURE REVIEW ................................................................................................................6
Food Marketing....................................................................................................................6
Idealized Male Bodies in Advertising..................................................................................9
Social Media Advertising ..................................................................................................15
CHAPTER 3 ..................................................................................................................................19
CONCEPTUAL FRAMEWORK AND HYPOTHESIS DEVELOPMENT ...............................19
Social Comparison .............................................................................................................19
Matchup Effect of Spokesperson and Product Congruency ..............................................26
CHAPTER 4 ..................................................................................................................................33
METHODS ....................................................................................................................................33
Study 1 – Snack Pretest......................................................................................................33
Design ................................................................................................................... 33
Participants.............................................................................................................34
Measures and Procedures .......................................................................................34
Results ....................................................................................................................35
Study 2 – Stimulus Development ......................................................................................36
Measures and Procedures .......................................................................................36
Results ....................................................................................................................37
Study 3 – Model Pretest .................................................................................................... 38
Design ....................................................................................................................38
Participants.............................................................................................................39
Measures and Procedures .......................................................................................40
Results ....................................................................................................................40
Study 4 – Main Study ........................................................................................................44
Design ....................................................................................................................44
Participants.............................................................................................................45
Stimulus Materials .................................................................................................46
Procedures ..............................................................................................................47
Measures ................................................................................................................48
CHAPTER 5 ..................................................................................................................................51
RESULTS ......................................................................................................................................51
vi

Manipulation Checks .........................................................................................................51
Treatment Groups ..............................................................................................................53
Hypotheses Tests ...............................................................................................................54
CHAPTER 6 ..................................................................................................................................67
DISCUSSION AND CONCLUSIONS .........................................................................................67
Conclusions ........................................................................................................................72
Limitations .........................................................................................................................73
Implications........................................................................................................................75
APPENDICES ...............................................................................................................................77
Appendix A – Instrument ....................................................................................................78
Appendix B – Model ...........................................................................................................85
Appendix C – Correlations ..................................................................................................86
REFERENCES ..............................................................................................................................87

vii

LIST OF TABLES

Table 1: Product Pretest .................................................................................................................35
Table 2: Brand Pretest Results .......................................................................................................37
Table 3:	Descriptive information on models selected pretested for main study ............................39
Table 4: Muscle % Estimated Average by Subset ........................................................................41
Table 5: Fat % Estimated Average by Subset ................................................................................42
Table 6: Model Manipulation Check Results ................................................................................51
Table 7: Sample Description ..........................................................................................................54
Table 8: Mediation results of upward social comparison on the effects of exposure to idealized
bodies on self-esteem .................................................................................................................... 56
Table 9: Mediation results of upward social comparison on the effects of exposure to idealized
bodies on body dissatisfaction .......................................................................................................57
Table 10: Average Self-esteem and Body Dissatisfaction per condition and perceived model
attractiveness ..................................................................................................................................58
Table 11: Perceived product healthfulness by condition ...............................................................58
Table 12: Average Aad according to body type and snack healthfulness.......................................60
Table 13: Average AP according to body type and snack healthfulness........................................62
Table 14: Average VBI according to body type and snack healthfulness .....................................63
Table 15: Average PI according to body type and snack healthfulness ........................................65

viii

LIST OF FIGURES

Figure 1: Model Pretest Results .....................................................................................................43
Figure 2: Sample Ads Per Condition ............................................................................................47
Figure 3: AAd by condition ............................................................................................................60
Figure 4: Ad Repetition by Body Type by Snack Healthfulness Three Way Interaction
for AAd ...........................................................................................................................................61
Figure 5: Ad Repetition by Body Type by Snack Healthfulness Three Way
Interaction for VBI ........................................................................................................................ 63
Figure 6: Ad Repetition by Body Type by Snack Healthfulness Three Way
Interaction for PI ...............................................................................................................................

ix

CHAPTER 1
INTRODUCTION
Although increasing obesity and eating disorders rates in the United States are influenced
by a complex set of factors, such as more sedentary life-styles and individuals’ genetic
predisposition (Seiders & Petty, 2004), food marketing is often singled out as one of the main
culprits (Chandon & Wansink, 2011)1. Across product categories, food is still one of the most
heavily marketed products (Batada et al., 2008; Harris et al., 2010), and the food industry is
among the top ten advertisers in the U.S. media market (Chandon & Wansink, 2011). In
particular, scholars have found that most advertised products on television and the internet
almost entirely promote high-fat, -sugar, and –salt foods (Boyland, Harrold, Kirkham, et al.,
2008; Kelly, Halford, Boyland, et al., 2010). All of these contribute to higher obesity rates
(Boyland, Nolan, et al., 2016), which are linked to eating disorders. A number of studies explore
the effects of exposure to unhealthy food advertising. These studies have shown that, relative to
control conditions, participants’ food intake was greater after exposure to advertisements
(Halford, Gillespie, Brown, et al., 2004; Harris, Bargh, & Brownell, 2009; Harris, Speers,
Schwartz, & Brownell, 2012). In particular, these studies found that exposure to food
advertisements was positively related to amount of food consumed and caloric load intake.
The relationship between exposure to food ads and intake is of concern because often
times the message communicated in these ads is that unhealthy eating (e.g. frequent snacking on
calorie-dense and nutrient-poor food) is “normal, fun, and socially rewarding” (Chandon &
Wansink, 2011, p. 13). Cooke (2007) suggested that food ads are effective in shaping consumers’
food preferences given (1) the pervasiveness of ads, and (2) that they start shaping opinions in
1

	While we acknowledge that a number of articles first authored by B. Wansink had been retracted by peer reviewed journals,
there are no indications that Chandon’s work has been compromised. 	
1

early ages when these individuals have not yet fully developed food habits. Thus, advertisements
become a normative force that dictates what is acceptable in society.
Chandon and Wansink (2011) noted that even in the absence of health information in the
ad (i.e., caloric intake and nutrition information), people acquire expectations about the taste,
healthfulness, and social acceptability of a particular food item and its ingredients through
different peripheral cues presented in food advertising. These cues include branding, health
claims, and other stimuli, such as models or spokespersons. When developing ads, marketers
recognize that certain stimuli match-ups are more effective than others in communicating the
desired message (Fleck & Quester, 2007). In this regard, studies have shown that a match-up
between the characteristics of models used in advertisements and the characteristics of the
advertised product positively influence product, brand, and ad evaluation (Lynch & Schuler,
1994; Wright, 2016). Given the negative outcomes associated with exposure to unhealthy food
ads (e.g., obesity), and that using models in ads that match with the characteristics of the
products can result in positive evaluations, one should expect these ads to feature models with
preponderance of body fat. However, advertisers have long capitalized on attractive models
(Kolbe & Albanese, 1996; Hellmich, 2000; Pope et al., 2000), and in particular those who have
idealized bodies (i.e., flat stomachs and muscular build), because of the effectiveness of using
attractive rather than unattractive models in advertisements (Elliott & Elliott, 2007; Levine &
Murnen, 2009; Solomon, Ashmore, & Longo, 1982). In a comprehensive review, Joseph (1982)
showed that physically attractive models were liked more, were perceived more favorably, and
had a more positive outcome on product evaluation. In sum, this line of research claims that
“physically attractive models generated favorable attitudes toward advertised products because
the beauty of the model spills over directly onto product evaluations” (Kim & Sohn, 2016).

2

In a study focusing on professional athletes’ endorsement of food and beverages, Bragg
et al. (2013) found that out of 512 brands endorsed by different athletes, sporting goods
represented the largest category (28.3%), followed by food and beverages (23.8%). Among the
food products endorsed, 79% of them were energy-dense and nutrient-poor, and 93.4% of the
beverages had 100% of calories from added sugar. Dixon et al. (2014) found that this is a
common practice among fast food retailers, and that brands such as McDonalds often align
unhealthy products with images of “health, vitality and fame” (p. 212). These findings are a
paradox of the advertising industry: while the model (be it a celebrity, athlete, or other endorser)
has a body perceived to be healthy by social beauty standards, they are promoting the
consumption of foods and beverages that would prevent the general population to reach a similar
body mass composition. This is troublesome for a number of reasons. First, as Chandon and
Wansink (2011) point out, exposure to food ads featuring models with idealized bodies can lead
consumers to make judgments about the food’s healthfulness based on the model’s body such
that if an athlete endorses such product, the food item should be healthy. Second, in addition to
obesity, the literature indicates that the use of thin and muscular models in advertisements leads
to negative outcomes such as consumers’ body dissatisfaction, development of eating disorders,
and depression (Presnell et al., 2004). These results have often been explained through the lens
of social comparison theory (Festinger, 1954). This theory posits that individuals tend to
compare themselves to others, including models in the media, as a means to explore what is
acceptable within one’s environment.
The fit (or match) between models and products has been explored in abundance within
the marketing communications literature through the lens of the matchup hypothesis (e.g.,
Kamins, 1990; Lee, Kim, & Kim, 2014; Seiler, & Kucza, 2017). This framework posits that the

3

effectiveness of an endorser is contingent upon the perceived congruency between a person and
the product (Kamins, 1990). And while endorsers and products can match up on a number of
different attributes, physical attractiveness is the most commonly studied dimension (see Wright
2016 for a comprehensive list). Studies have also explored congruency between the model’s
expertise (Friedman & Friedman, 1979; Ohanian, 1990), sophistication (Batra & Homer, 2004),
credibility (Koerning & Page, 2002), trustworthiness (Kamins & Gupta, 1994), and masculinity
(Lynch & Schuler, 1994) with the advertised product. And while the matchup hypothesis has
received considerable scholarly attention, researchers seem to have ignored the congruence
between models with idealized bodies and perceived product healthfulness.
Extensive literature suggests several negative outcomes as a result of exposure to media
images of idealized thin and muscular models, including poor body image (Levine & Murren,
2009; Slater & Tiggemann, 2006), depression and suicide (Fildes et al., 2014), and body
dissatisfaction (Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999). The majority of the
literature focuses on the effects of representations of idealized bodies in the media on women’s
emotions and self-esteem (Slater & Tiggemann, 2006). However, consumer research focusing on
men’s processing of ads featuring models with idealized bodies remains limited (Elliott & Elliott,
2005). This phenomenon can be explained by women’s higher likeliness to discuss issues related
to, and seek help for, body image and eating disorders (Andersen, 2014; Elliott & Elliot, 2007),
and because the majority (70%) of individuals in the western countries that are dissatisfied with
their appearance are women (Diedrichs, Lee, & Kelly, 2011; Presnell et al., 2004). Yet, the
National Eating Disorders Association, the leading non-profit organization for those affected by
eating disorders, reports that in the US, about 10 million males will suffer from a clinically

4

significant eating disorder (ED) at some time in their life, including anorexia nervosa, bulimia
nervosa, and binge eating disorders.
While, in general, food advertising and promotion have been linked to individuals’ foodrelated decisions and preferences, snack food advertising should be of interest given that ads
promoting these products have been shown to increase snack consumption among children and
adults (Harris, Bargh, & Brownell, 2009), not only for the advertised brands, but for other
products in the market. Snack products account for 14% of the total advertising expenditures for
US food products (Story & French, 2004). Chandon and Wansink (2011) suggested that these
ads might promote short-term enjoyment goals in general, which is negatively related to longerterm healthy living goals. In addition to the pervasiveness of snack ads, Seiders and Petty (2004)
point out that greater availability (through vending machines) of (unhealthy) food (i.e., snacks)
has been found to increase food consumption, to the degree of having negative health outcomes
(Tordoff, 2002). Similarly, Cutler et al. (2003) claimed that individuals have increased the
frequency of their snacking habits compared to previous generations. Given that people acquire
expectations about the product through different peripheral cues used in food advertising, such as
spokespersons, and that using attractive models to advertise all sorts of products is a ubiquitous
practice (Bragg et al., 2013), this dissertation explores how males evaluate models featured in
snack food advertisements when their bodies conform (or not) with advertising industry norms,
and the effects of the pairing of different models with products perceived to be healthy (vs.
unhealthy) on participants’ evaluations of the self, the product, and the ad.

5

CHAPTER 2
LITERATURE REVIEW
Food Marketing
Several studies (e.g., Cairns, & Macdonald, 2016; Wartella, Lauricella, , Cingel, &
Connell, 2016) have explored the effects of food and beverage advertising on food consumption
and health outcomes, but the results of these studies offer mixed results. Food marketing can
influence consumer expectations through a number of strategies. Food marketers influence
consumers’ evaluations of the product’s quality and quantity through advertising and
promotions, branding, the convenience and salience of the purchase, the size and shape of
serving containers, and the atmospherics of the purchase and consumption environments
(Chandon & Wansink, 2011). There seems to be a consensus among scholars and policy makers
in that strategies employed by marketers to promote their products are, at the very least,
questionable (Boyland, Nolan, Kelly, et al., 2015). For example, a number of researchers have
noted that the food industry deflects attention from foods’ high caloric contents by emphasizing
physical activity in their ads (Folta et al., 2006).
As a result, food marketers have been charged with being responsible for a number of
illnesses (Quilliam, 2008), ranging from dental health problems to eating disorders (Livingstone
& Helsper, 2006). In the next paragraphs, the food marketing literature related to the objectives
of this dissertation are reviewed.
According to Boyland et al. (2015), an individual’s environment is a key driver of that
person’s diet. As such, obesogenic food environments (i.e. settings that promote unhealthy
dietary habits) are, in part, responsible for the increasing rates of obesity and eating disorders
(Swinburn, et al., 2011). This is particularly important given that almost all individuals in

6

Westernized societies are exposed to large amounts of unhealthy food advertising (Boyland et
al., 2015). But in addition to food advertising activities, increased food consumption is also
driven by convenience-related trends, such as heightened consumer purchasing of meals outside
the home and increased purchase of snacks through vending machines and kiosks (Cutler,
Glaeser, & Shapiro, 2003). Seiders and Petty (2004) highlight the pervasiveness of food
marketing and the intensive distribution and promotion of fast foods and snack foods (Harnack &
French, 2003) in all types of retail outlets, such as drug stores, convenience stores, mall food
courts, office vending machines, and airports, among others. This greater availability of food has
been found to increase consumption to the extent that it can become detrimental to one’s health
(Tardoff, 2002).
Studies have suggested that food marketing is responsible for an increased frequency of
snacking practices among U.S. respondents (Cutler et al., 2003). Snack foods are in a food
category that has sparked controversy among public health experts, nutritionists and public
interest organizations, given that they are frequently marketed with misleading nutritional
information (Seiders & Petty, 2004). For example, packaging and other marketing
communication materials present nutritional information based on one serving, when often the
package contains more than one. While these practices are troublesome, the fact that there are
few regulations about the inclusion of nutritional information for foods served in restaurants and
sold in vending machines (FDA, 2016), and none for ads is even more worrisome. Scholars
suggest that many consumers are unaware that they are consuming such a high proportion of the
recommended fat, sugar, and/or calories suggested for consumption per day (Berman & Wootan,
2003), and that advertising campaigns tend to favor displaying other information and appeals that
can lead to heightened ad effectiveness, such as featuring celebrity endorsers.

7

In addition to the strategic distribution of unhealthy foods described above, food
marketing should be of concern to scholars and marketers alike given that the food industry is
among the top ten advertisers in the U.S. media market (Chandon & Wansink, 2011). In recent
years, in addition to the advertising efforts through traditional media (television, print, radio,
billboards), food marketing has also become a phenomenon in new media, such as social media
platforms. Chandon and Wansink (2011) point out that food marketers are diverting their budgets
from print and television advertising to social media. A number of market research companies
seem to echo these claims and have published rankings of the most active food brands on social
media. Sherwood (2015) notes that Starbucks, McDonalds, Taco Bell, Ben & Jerry’s, and Oreo
are among the top brands by followers on Instagram, and that other brands, such as KFC, are
often among the top brands by hashtag use. In all cases, the food displayed is high in caloriecontent, and often loaded with sugars. In a compilation of the top food brands in social media,
AdWeek also listed a number of brands that commercialize products that could be detrimental to
one’s health if consumed in excess, including Skittles (with 23,421,414 followers), Oreo
(43,000,000), Pringles (24,758,000), and M&M’s (10,000,000) 2, and others, such as Hershey’s,
Frito Lay, and Doritos. Therefore, this dissertation will use snack foods ads as the stimuli under
study.
Not only does this overload of food marketing through different channels contribute to
the development and growth of an obesogenic environment, but the cumulative exposure to food
ads also promotes more positive beliefs and attitudes towards those foods most heavily
advertised (Sixmith & Furnham, 2009). Thus, social media as the medium through which food
messages are disseminated is worth examining (1) because advertisers are increasing the
proportion of their budgets allotted to social media (Okazaki & Taylor 2013; Olmstead & Lu,
2

Updated number of followers for each brand as of Jan 26, 2018

8

2015; Saxena and Khanna 2013), (2) technology-savvy consumers tend to spend over 10 hours
per week on the Internet via their computers and 1.5 to 3 hours a week via mobile devices
researching products or services (Barnic, 2014), and (3) because social media are often used as a
means for social comparison (Nesi & Prinstein, 2015).
Idealized Male Bodies in Advertisements
Extensive literature has explored the representations of the female body in advertising
(Belkaoui & Belkaoui, 1976; Busby & Leichty, 1993; Englis, Solomon, & Ashmore, 1994;
Hawkings, Richards, Granley & Stein, 2010; Sohn & Youn, 2012). This line of research has
offered detailed information about women’s images in advertisements and recognized changes in
their portrayals. In general, scholars agree in that female models have become progressively
thinner and sexualized (Kim & Sohn, 2016; Yu, 2014). Another body of literature has examined
the depictions of male models in advertisements (Elliott & Elliott, 2005; Grau, & Zotos, 2016;
Kolbe & Albanese, 1996; Skelly & Lundstrom, 1981; Wolheter & Lammers, 1980), and Elliott
and Elliott (2005) point out that that since the 1980s, men’s bodies shown in ads offer a similarly
idealized body to that presented by women models, where men are shown as sex objects. These
findings should be of concern given that exposure to ads featuring models with idealized bodies
raises comparison standards for attractiveness and lower individual’s satisfaction with their own
attractiveness (Elliott & Elliot, 2005). Yet, most studies on body image dissatisfaction as a result
of exposure to mediated images have been centered on females because this group has
traditionally been more likely to express discomfort with their own bodies and have exhibited
higher rates of fasting, taking dietary supplements, vomiting or taking laxatives to lose weight or
to keep from gaining than men (Eaton et al., 2012). According to Brown et al. (2015), there is a
dearth of research that examines the effects of exposure to idealized images of models in ads and

9

the media among males. Thus, this dissertation will explore the extent to which men can be
negatively affected by idealized representations of male bodies, particularly for snack food ads
endorsed by male models with different body sizes and compositions. In the next paragraphs, the
representations of males in advertising are discussed.
According to Kolbe and Albanese (1996), the defining attributes of the American man
have remained virtually unchanged since the writings of Alexis de Tocqueville of his travels in
nineteenth-century America. These writings described the American man as individualistic,
insensible, cold, and implacable (Barker-Benfield, 1976). In this respect, in the author’s book
about male portrayals in Viagra ads, Baglia (2005) states that social communication (language,
symbols, and significant social performances and actions) creates norms, regularities, and
expectations about each actor in society. He goes on to note that masculinity is socially
constructed, and that the representations of males in the media (and advertisements) help institute
what the normal man should look and act like. Baglia (2005) bases the assessment on Goffman
(1963), who provided a depiction of the American man based on print ads in an influential piece
on gender advertising,. He wrote:
“[In] an important sense there is only one complete unblushing male in America: a
young, married, urban, northern, heterosexual Protestant father of college education, fully
employed, of good complexion, weight and height [emphasis added], and a recent record in
sports… Any male who fails to qualify in any of these ways is likely to view himself –during
moments at least- as unworthy, incomplete, and inferior…” (Goffman, 1963, p. 128).
As early as in 1963, Goffman made a similar observation to that of a number of more
contemporary researchers about the depiction of males in advertisements in today’s media: there
is a prevalence of ads depicting models with “washboard abdominal muscles, massive chests,

10

and inflated shoulders” (Pope et al., 2000, p. 34). This combination of muscularity and leanness
is difficult to accomplish with a regular exercising routine. In a content analysis of sole male
images in men’s magazines, Kolbe and Albanese (1996), found that across a number of
magazines (Esquire, GQ, and Business Week, among others), most models featured in ads had
bodies that were not “ordinary”, but those of strong and hard male icons, which the authors
considered to be an objectification of the male body. In particular, they found that there was an
overwhelming representation of bodies that were “strong, tough, and equipped for strenuous and
exacting physical activity” (p. 11) (i.e., mesomorph models), few models had endomorphic
bodies (i.e., soft and rounded), and even fewer had ectomorph bodies (i.e., thin and frail).
Similarly, Patterson and England (2000), also discuss a ubiquitous depiction of male
bodies in magazine ads, and showed that readers are often presented with a depiction of models
with idealized bodies (i.e., mesomorphic bodies). They further point out that endomorph models
(which is the body composition of the average American male; Hamblin, 2013) were rarely seen,
irrespective of product category, and that in those cases where endomorphic models were shown,
they tended to be the object of humor. In the case of ectomorphs, Patterson and England (2000)
suggested that models with this particular physique were common only in clothing advertising,
where products look better on slightly thinner men. Hellmich (2000) agrees with previous results
and claimed that “men are being bombarded with images of muscular, half-naked men on the
covers of men’s magazines” (p. 06D), which leads to men developing an “Adonis complex” and
eating disorders after being overexposed to idealized bodies in advertising (Elliott & Elliott,
2005). Diedrichs et al. (2011), summarize this review of literature and claim that the current
western idea of male beauty, based on the stereotypes reinforced through the media, consists of
muscular, tall and well-groomed men.

11

As Pope et al. (2000) indicated in their studies, ads feature idealized bodies to promote a
wide variety of products ranging from underwear to cars. And while previous literature
suggested that using highly attractive models in advertisements results in more positive ad,
product, and brand evaluations (see Belch, Belch, & Villarreal, 1987; Joseph, 1982 for reviews),
and heightened media coverage and brand impressions (Kalb, 2013), most recent research
literature contests the results (Bower, 2001; Bower & Landreth 2001). Studies have shown that
models with idealized bodies are only more effective than models that resemble the average
American male more closely in certain product categories (e.g. attractiveness-relevant products).
Of concern to this dissertation is the prevalence of male models with idealized bodies in
advertisements for unhealthy products, in particular, unhealthy food snacks. The snack food
segment is interesting given that the foods advertised promote energy-dense, nutrient poor foods,
which when paired with a model with an idealized body, send mixed messages about fitness,
health, and diet to consumers (Bragg et al., 2013; Dixon, Schully, Wakefield, et al., 2011b;
Nestle, 2006; World Health Organization, 2012). Further, these models are often portrayed as the
typical and ideal customer of the advertised brands, and look attractive and powerful while doing
so (Portwood-Stacer, 2014).
Recently, and paralleling campaigns that celebrated the shape of real women in ads, there
have been a number of initiatives to include real men in ads. In 2015, American Eagle released
an integrated campaign called #AerieReal. This ad caught the media’s attention for promoting
and celebrating male bodies of all types, using un-retouched photos in advertisements, with a
slogan emphasizing diversity, and claiming that: “the real you is sexy” (Mandell, 2016).
However, the company later admitted that they created the ad as an April Fool’s joke. Dove
released a campaign exploring how men view their own beauty. Unilever launched a commercial

12

for their male deodorant line (Axe) featuring models depicting a spectrum of body images. The
campaign, called “Find your magic”, aims at ending sexist stereotypes and dispelling traditional
portrayals of masculinity, while celebrating individuality (Weed, 2016). Similarly, Norwegian
textile brand Forsman & Bodenfors showcased male body diversity in an underwear campaign
by casting models with chest hair, beer-bellies, freckles, cellulite and tattoos (Jardine, 2015).
Yet, portraying realistic-looking bodies is not the norm when it comes to advertisements
for unhealthy foods (Bragg et al., 2013). For example, among the celebrities with idealized
bodies that have endorsed unhealthy foods, Oreo’s roster of endorsers included Eli and Peyton
Manning and Shaquille O’Neal; Peyton Manning for Papa John’s; LeBron James, Dwight
Howard, and Kobe Bryan for McDonald’s; and Miguel Cabrera, Anderson Silva, and David
Beckham for Burger King. But this phenomenon is not exclusive to athletes, a number of food
ads have also made use of models with idealized bodies to promote energy-dense, nutrient poor
foods. In 2013, Kraft’s released an integrated advertising campaign for Zesty Italian salad
dressing that caught the media’s attention for featuring a near-naked male model. According to
Business Insider, the product advertised was barely noticeable, and the periodical questioned the
effectiveness of the ad strategy to sell the product (Kalb, 2013). M&M’s released a 30 second ad
of a male model highly sexualized (see: https://youtu.be/6QMb16bxJvY). Bradley Cooper was
cast in a sexualized ad for Häagen-Dazs (see https://youtu.be/s3zY4Trz7po), an American ice
cream brand, and even brands for carbonated sodas have developed advertisements showing
half-naked men with idealized bodies.
Recognizing that little is known about the ideal male body size, Brierley and colleagues
(2016) identified the proportion of fat and muscle mass necessary for individuals to have a
healthy and attractive body, both for males and females. Brierley et al. (2016) claimed that while

13

previous studies had attempted to provide a characterization of participants’ perceptions of
healthy body mass indexes (i.e., weight scaled for height, BMI), these studies failed at
addressing the role of body fat (St. Onge, 2010) and muscle, which have distinct relationships
with health and are conflated in BMI, in determining perceived health and attractiveness.
Furthermore, Brierley et al. (2016) claimed that internalizing body ideals, as promoted by agents
such as the media, have an influence in individuals’ perceptions of attractive body compositions
for both genders, but even more so for women. Brierley et al. (2016) presented their participants
with a digital image of a male and asked participants to graphically manipulate the apparent fat
and muscle mass of the bodies presented to them to represent “healthy” and “attractive” male
bodies. Their results showed that, in general, male bodies were preferred when the subject
presented featured lower fat mass (M = 12.16 kg, SD = 4.69) than muscle mass (M = 63.27 kg,
SD = 4.06). Thus, given that the mean weight of male figures presented to the subjects was 79.29
kg, the men’s fat mass chosen for attractive and healthy male bodies fell within the range of
healthy body fat (8 to 21%; Frankfield et al., 2001; Gallagher et al., 2000). The guidelines
regarding body mass compositions for males offered by Brierley et al. (2016)’s study are deemed
an adequate reference for selecting the models employed in this dissertation given that it
addressed a previous limitation of the literature concerning BMI.
Rather than following body type classifications that rely solely on the individuals’ body
mass index (i.e., endomorph, ectomorph, and mesomorph), for this study we rely on more
comprehensive definitions. Given that BMI alone is not a reliable indicator of an individual’s
body composition (Frankfield et al., 2001), as it fails to take into account factors such as muscle
and fat mass, this dissertation follows the guidelines offered by Brierley et al. (2016)’s study. We
define models with idealized bodies as those whose muscularity accounts for at least 79% of

14

their body composition (with fat mass ranging between 8 and 21%). Further, we mimic St.
Onge’s (2010); this study will refer to as “average” models those male individuals with body fat
ranging between 22 and 40%, and muscle mass ranging between 60 and 75%. Overweight
individuals will be defined as those with fat mass above 40%. We recognize that these
classifications are not exhaustive, and that more research is needed to make Brierley et al.
(2016)’s study results generalizable. Yet, these indicators address previous limitations offered by
the BMI measures.
In sum, investigating the use of idealized (vs. average) bodies in food advertisements is
deemed meaningful given that research shows that (1) these snack ads make unhealthy foods
more attractive by associating them with attractive spokespersons/models, (Bragg et al., 2013;
Dixon, Schully, Wakefield, et al., 2011b; Nestle, 2006; World Health Organization, 2012), and
(2) exposure to mediated content featuring idealized bodies leads to poor body image (Levine &
Murren, 2009; Slater & Tiggemann, 2006), depression and suicide (Fildes et al., 2014), and body
dissatisfaction (Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999). The following section
reviews the role of digital advertising in the context of this dissertation.
Social Media Advertising
Social media has become a ubiquitous phenomenon of both individual’s lives and society
as a whole (Khang, Ki, & Ye, 2012). While social media sites (SNS), such as Facebook and
Twitter were originally used by many to peruse content for entertainment purposes, these
platforms are now an integral part of various activities such as education (e.g., Greenhow &
Lewin, 2016), employee communication (Brustein, 2010), and emergency communication
(Wukich, & Steinberg, 2016), among others. Since Facebook’s launch in 2007, there has been a
continuous growth in the use of social media for sharing user-generated content, such as

15

photography, videos, and podcasts (Khang, Ki, & Ye, 2012). According to a recent report by
Statista (2017), there are nearly four billion active social media users worldwide (considering
only the most popular social media outlets). Furthermore, as of 2016, daily social media usage of
global Internet users amounted to 118 minutes per day, up from 96 minutes in 2012 (Statista,
2017b). Scholars have noted that given the popularity of these platforms, advertisers have turned
to social media (SM) to create brand communities, reach their target populations, and augment
brand awareness and exposure (Kaplan & Haenlein, 2016).
In marketing communications, specifically, advertisers and marketers have turned to
social media recognizing the prospects of these platforms for delivering their campaign strategies
given SM’s potential for earned advertising (Khang, Ki, & Ye, 2012). To this regard, Alhabash
et al. (2017) claimed that social media amplifies the importance of the human factor, given that it
allows users to “respond, react, and generate online content that fits (or modifies) an existing
advertising strategy” (p. 285). Given that social media allows users not only to see the content,
but also share, engage with, and create content to be shared through these platforms (Scott,
2015), Alhabash et al. (2017) proposed a unique definition to differentiate social media
advertising from traditional approaches. They defined it as “any piece of online content designed
with a persuasive intent and/or distributed via social media platform that enables internet users to
access, share, engage with, add to, and co-create” (p. 286).
Allowing users to interact with content in various ways is important given that
advertising content that is passed along or shared among friends has shown more significant
effects on ad recall, brand awareness, and purchase intentions compared to traditional forms of
paid advertising (Bilton, 2010). Thus, it is not surprising that advertisers are progressively

16

allotting larger budgets to SM (Okazaki & Taylor 2013; Olmstead & Lu, 2015; Saxena &
Khanna 2013), with an estimated market worth $35.98 billion in 2017 (eMarketer, 2015).
Therefore, it seems safe to ascertain then that social media is an effective medium for
advertisers to deliver their messages to their various publics, and that social media might be more
effective than advertising via traditional outlets (Sass, 2015). There are several reasons to
support this claim: (1) the cost of ad placing on social media is much cheaper (vs. traditional
media outlets; Bhanot, 2012), (2) SM gives advertisers the tools to better quantify and reach
specific target audiences, and (3) the analytics tools associated with SM have resulted in machine
learning algorithms that provide better insights about different online and offline forms of ad
engagement (Alhabash, Mundel, & Hussain, 2017). In addition to these favorable outcomes,
social media advertising is deemed to be a meaningful area of study for this dissertation given
that social media is often used as a means for social comparison (Nesi & Prinstein, 2015).
In a study investigating the outcomes of social comparison to social media referents
among adolescent students, Nesi and Prinstein (2015) asked respondents to report on levels of
depressive symptoms (at baseline and after a year), technology use frequency, and social
comparison and feedback-seeking measures. Consistent with their hypotheses, the researchers
found that technology-based social comparison and feedback-seeking behaviors were associated
with depressive symptoms. Furthermore, they found that other factors, such as gender and level
of popularity, moderated the effects of social media use on social comparison, so that females
low on popularity suffered from these outcomes more than their counterparts. This dissertation
ascribes to the effects of social media exposure and use on social media comparison, and posits
that repeated exposure to social media content featuring attractive and fit individuals might result

17

in lower reported self-esteem and perceived self-attractiveness. Thus, social media advertising
serves as the stimuli under study.

18

CHAPTER 3
CONCEPTUAL FRAMEWORK AND HYPOTHESIS DEVELOPMENT

Social Comparison
Abundant research has explored the unintended consequences of idealized advertising
imagery on girls and women’s self-perceptions (Gulas & McKeage, 2000; Kim & Sohn, 2016;
Martin & Gentry, 1997; Richins, 1991). Groesz et al. (2002), for example, reviewed 25 studies
exploring the relationships between media exposure and negative effects on body image and
found that exposure to idealized bodies in advertising had an adverse effect on body satisfaction
when compared to control images. Social comparison theory, which posits that there is an innate
human need to compare oneself with others to gain accurate self-evaluations (Festinger, 1951), is
a theoretical framework often used by researchers to explain the consequences related to body
dissatisfaction as the result of media exposure (Gulas & McKeage, 2000). In the following
paragraphs, the literature on social comparison theory is reviewed.
Social comparison theory (SC) is based on Festinger’s earlier work, specifically, the
informal communication theory (ICT; 1950). ICT posited that when individuals are in groups,
they desire to conform with public opinion either because group consensus can provide a person
with confidence in one’s own opinion or because agreement is needed to achieve the group’s
goals (Suls & Wheeler, 2000). In 1951, Festinger introduced a new theory that built on ICT
called “A theory of social comparison processes”. The new theory (SC) focused on how
individuals use groups to evaluate their abilities and opinions against those of the other members
of the groups. That is, “observing those with similar abilities allows us to know what our own
possibilities for action in the environment are” (Suls & Wheeler, 2000, p. 5).

19

In particular, Festinger (1954) posited that humans have a drive for continual
improvement, and in order to decide on which changes need to be made, individuals compare
themselves to similar others. While Festinger did not pursue the development of this theory, a
number of researchers continued with this line of research (e.g., Gordon, 1966; Radloff, 1966;
Latané, Eckman, & Joy, 1966). Mettee and Smith (1977) posited that in addition to those who
are similar to us, individuals might also compare themselves to dissimilar others. This
proposition is of interest to this dissertation. This modification of SC theory is important given
that it operated under the assumption that the differences between some individuals and
dissimilar others might be greater than those identified when comparing the self to similar others.
In the preface to the Handbook of Social Comparison, Suls and Wheeler (2000) note that SC
theory furthered scholarly interest with Wills (1981) introduction of the notion of upward and
downward comparison. This broader view of SC, similar to Mettee and Smith’s (1977),
introduced the notion that in addition to self-evaluation, social comparison can also serve as a
means for self-enhancement and self-improvement goals (Gulas & McKeage, 2000; Wood,
1989).
In the study of (downward) social comparison, Wills (1981) posited that individuals can
increase their subjective well-being by making comparisons between the self and someone who
is worse off. Upward comparison refers to comparisons made with others considered above
oneself in regard to a particular dimension (i.e., an endomorph person comparing his body to that
of a mesomorph), which can lead to a decrease in well-being (Morrison, Kalin, & Morrison,
2004). Downward comparisons refer to those made against individuals judged as inferior on a
particular dimension (i.e., mesomorph comparing his body to an ectomorph), and is believed to
enhance an individual’s subjective well-being. Wills described a number of ways in which

20

downward comparisons can be made, including active derogation, causing harm to others, or
simply on a passive basis (i.e., taking advantage of chances to compare with people who are less
fortunate) (Suls & Wheeler, 2000).
The subsequent revisions and additions of SC theory led researchers to examine not only
individuals’ comparisons to other individuals in their circles, but also to other forces in the
environment, such as the media. While Wood (1989) saw the environment as an inactive
backdrop, more recent literature suggests that the social comparison process might be often
automatic (Lyubomirsky & Ross, 1997) and that the nature of the comparison might be forced by
the environment (Sirgy; 1998, Steil & Hay, 1997). According to Gulas and McKeage (2000), the
environment can force a variety of comparisons. For example, while in certain instances
exposure to an unfortunate other may encourage one to make downward comparisons, in other
separate instances the environment may also force upward comparisons with other individuals
who are present in a variety of situations (e.g., models in ads) who are more fortunate on one or
more dimensions. Thus, through the lens of social comparison, exposure to media images of
muscular males helps shape a standard of male attractiveness (Petty, 2011).
Scholars have noted that the images employed in advertising present a narrow view of
reality (Pollay, 1986; Richins, 1991, 1996). For models featured in ads, marketers often favor
attractive individuals (Bower, 2001) that have bodies impossible for most to attain (Petty, 2010;
Pope et al., 2000). Gulas and McKeage (2000) noted that consumers realize that the mass media
present the world in a “styled and idealized way” (p. 17). Yet, individuals still draw meaning
from these images (Hirschman & Thompson, 1997). Petty (2010) claims that the cultural
standard for the ideal body image, which is reinforced by the media, has created a multitude of
psychological problems (e.g. low self-esteem and eating disorders). Among the factors that have

21

an influence on body image formation, Petty (2010) considers social comparison as a central
contributor.
Tiggemann and McGill (2004) claim that: “the mass media, due to the pervasiveness and
reach, are probably the single most powerful transmitters of sociocultural ideals” (p. 24). They
list studies showing the relationships between exposure to mediated images and body
dissatisfaction and eating disorders. Similarly, Morrison et al. (2004), explored upward and
downward comparisons to images in the media and claimed that universalistic targets (i.e., the
mass media) exert more pressure to coincide with idealistic standards of attractiveness than other
targets (e.g., friends). Thus, making comparisons to mediated images with idealized bodies
should be of concern, particularly with advertisements, given that they often reinforce a standard
of beauty that is difficult to attain. Research shows that social comparison of physical appearance
often results in upward comparison (Morrison et al., 2004; Wheeler & Miyake, 1992), and that
this type of comparison results in negative self-evaluations (Suls, Martin, & Wheeler, 2002) and
lower perceived self-attractiveness (Morrison et al., 2004).
Recent research on upward and downward comparisons of the self to other individuals
has supported the notion that upward comparisons with mediated images can influence an
individual’s self-esteem, and perceived self-attractiveness, among other outcomes. For example,
Lanzieri and Hildebrant (2016) summarize a number of studies where males exposed to idealized
images of same-sex professional models report lower self-esteem and body satisfaction (vs.
controls). By contrast, research on downward comparisons shows mixed results (Suls, Martin, &
Wheeler, 2002). As Collins (2000) suggested, some people might make comparisons against
inferior others and cultivate more positive self-judgments due to aspirational and motivational
evaluations of the target. However, downward comparisons can also lead to negative judgments

22

and attitudes when an individual acknowledges that he could too look like the target to which the
person is being compared (Collins, 2000).
Morrison et al. (2004) explored the consequences of social comparison to sources in
different media outlets. The authors found that exposure to magazine and television content
featuring models with idealized bodies, and the degree to which men engaged in social
comparison, were predictors of participants’ perceived self-appearance, self-esteem, number of
diets to lose weight, use of pathogenic weight control practices, and use of steroids to increase
muscle mass. Similarly, in a comprehensive literature review on the effects of social comparison
on eating behaviors, Polivy and Pliner (2015) claim that abundant literature supports the
relationships between making comparisons between one’s weight and other individuals’ and
body dissatisfaction, and that, depending on the images, exposing participants to magazine ads
before eating can influence subsequent eating amounts (Mills, Polivy, Herman, & Tiggemann,
2002).
At the most basic level, exposure to models with idealized bodies in a persuasive
message has been found to result in lower self-esteem and body satisfaction. With this in mind,
we hypothesize:
H1: Participants will exhibit (H1a) greater body dissatisfaction, and (H1b) lower levels
of self-esteem when exposed to ads featuring idealized looking models than when
exposed to models with average or overweight bodies.
However, research shows that these negative effects experienced by participants tend to
be limited to a particular group of respondents, those who are heavier and have higher levels of
body dissatisfaction (Henderson-King & Henderson-King, 1997). In other words, although all
men are exposed to media content with idealized bodies, not all develop disorders related to their
23

bodies (Tiggemann & McGill, 2004). According to Hargreaves and Tiggemann (2009), men are
more likely to feel dissatisfied with their muscle size and tone and wish to be lean but highly
muscular. A number of experimental studies have researched the impact of exposure to idealized
muscular media images on men’s body image. Abrour and Martin Ginis (2006) showed that
participants’ trait muscle dissatisfaction moderated the negative effects associated with exposure
to idealized bodies. In particular, men with higher initial muscle dissatisfaction exhibited more
negative outcomes than those with lower scores. Furthermore, Hargreaves and Tiggemann
(2009), showed that the degree to which participants engage in social comparison mediated the
effects of the advertisements on men. Thus, we hypothesize that participants that engage in
upward comparisons will be more likely to report lower levels of self-esteem compared to other
participants.
H2: Upward social comparison will mediate the effect of idealized bodies in snack food
ads and (a) self-esteem, and (b) body dissatisfaction, such participants exposed to ads
featuring idealized bodies will report greater social comparison, which in turn will
negatively affect self-esteem and body satisfaction.

Recent studies support the relationships between exposure to idealized bodies on social
media and low self-esteem and perceived physical attractiveness among women (e.g. Neira,
Corey, & Barber, 2014; Perloff, 2014; Vogel, Rose, Roberts, & Eckles, 2014). Further, studies
show that technology-based social comparison and feedback-seeking behaviors on social media
were associated with depressive symptoms (Nesi & Prinstein, 2015). This line of research
supports the notion that exposure to idealized bodies on social media-based outlets have similar

24

social comparison and self-esteem outcomes as traditional media. However, these effects are not
yet clear in the context of a male audience.
In a study investigating social comparison on social media, Vogel and colleagues (2014)
suggested that social media sites, such as Facebook and Instagram, are suitable platforms for
engaging in social comparison. The authors posit that as social media sites are increasingly used
for a variety of social functions (e.g., sharing pictures with friends and following role models),
people use the information on these sites for social comparative functions, such as selfevaluation or self-enhancement. Furthermore, given that users have the ability to create a public
persona of their liking by selectively allowing content in their profiles, posting chosen pictures,
and describing themselves in a way that presents their ideal self-views (Rosenberg & Egbtert,
2011), the majority of the comparisons made on social media will be in upward direction.
Social comparison on social media is thus problematic given that (1) users on social
media tend to project positive rather than negative images of themselves, and (2) social media
users believe that other users are happier and more successful than themselves (Chou & Edge,
2012). Based on these premises, Vogel and colleagues (2014) conducted an experiment to
decipher whether or not (and in which direction) college students engaged in social comparison
through Facebook. Their results showed that frequency of Facebook use was (1) negatively
correlated with self-esteem (i.e., participants with more exposure to Facebook had lower selfevaluations); and (2) positively related to both types of social comparison (i.e., upward and
downward); and that participants who used Facebook the most engaged in social comparisons to
a greater extent. Interestingly, on average, Vogel et al.’s respondents reported engaging in more
upward (vs. downward) social comparisons on Facebook, and Vogel et al. (2014) concluded that

25

while upward comparisons on Facebook predicted lower self-esteem, downward comparisons
did not.
Of particular interest to this dissertation is Vogel et al. (2014)’s second study. The
authors explored whether manipulating user content on social media (to make the target of the
comparison seem to be healthier or unhealthier) had an impact on upward social comparison.
Their results showed that participants reported lower self-esteem and poorer self-evaluations
after exposure to social media content portraying “healthy” users. The results of Vogel et al.’s
(2014) study are in line with those of other researchers regarding upward comparison on social
media. Haferkamp and Kramer (2011) exposed participants to made-up social media profiles of
individuals deemed to be highly rated in attractiveness or unattractiveness. Their results showed
that following exposure to the upward comparison conditions, participants reported larger actualideal self-discrepancies than those in the downward comparison conditions. Similarly, studies
have found that different types of social media content, such as selfies, can act as sources for
social comparison and as a means to internalizing a standard of beauty. Chua and Chang (2016)
found that exposure to selfies on social media often resulted in participants’ lower self-esteem,
insecurity, and intentions to edit their self-presentation for peer approval.
Matchup Effect of Spokesperson and Product Congruency
Among the many visual components of an ad, the models featured in the ad are important
influencers of the consumer’s perception of the advertisement and the product (KnoblochWesterwick & Coates, 2006). The model featured in an advertisement can offer a number of cues
to consumers, including the intended users of a product (Knobloch-Westerwick & Coates, 2006),
the benefits resulting from product use (Lynch & Schuler, 1994), and trustworthiness of the
product (Kamins & Gupta, 1994). Thus, it is no surprise that considerable scholarly attention has

26

been paid to identifying which model’s characteristics are more effective in enhancing ad
effectiveness. In particular, abundant research has hypothesized that the effectiveness of an
advertisement featuring models is, in part, contingent upon some sort of congruence, or matchup,
between and endorser and a product.
This congruence has been explained through the lens of the matchup hypothesis (Kamins,
1990), which suggests that the effectiveness of the endorser is determined by its congruency with
certain product characteristics. Kamins’ (1990) work was informed by earlier studies that
provided support to the development and testing of the matchup hypothesis. Kanungo and Pang
(1973) investigated the fittingness between the model and kind of product promoted in ads by
manipulating the type of product (stereo, TV, car, and sofa) and featured model(s) (male, female,
and both genders present). In addition to showcasing the products with different models, to serve
as a baseline for comparing the effects of featuring a model in the ad, the authors included a
control where no models were included. Participants were asked to view the ads and respond to a
number of scales dealing with product attributes (e.g., attitudes toward the product). Their
results showed that ads with male models (vs. the other conditions) endorsing cars received more
favorable product evaluations from both male and female respondents. On the contrary, the
presence of female models produced more positive attitudes toward the product for sofa
advertisements among male respondents. For stereos, the ads created more favorable attitudes for
both male and female respondents when both male and female models were featured. On the
other hand, male subjects had higher product evaluations after being exposed to TV ads with no
models, while female respondents had more positive attitudes toward the product when the TV
sets were displayed with a male/female couple. Based on these results, Kanungo and Pang (1973)
were interested in understanding if the perceptions of the study’s stimuli were a function of the

27

product being perceived as either masculine or feminine by consumers. Thus, they asked a
second set of respondents to indicate whether they thought the four products were masculine,
feminine, equivalently masculine or feminine, or neither. Unsurprisingly, cars were viewed as
masculine, sofas as feminine, stereos being equally masculine and feminine, and TVs were not
viewed as masculine or feminine. In other words, the authors found that in general, a product
perceived to be masculine was a better fit with a male model, a feminine product with a female
model, and so on. Therefore, the authors concluded that if “there is a match or fittingness
between the product and a model in an advertisement, consumer evaluations of the product
advertised will be favorable because perceptual and attitudinal congruence will exist” (Lynch &
Schuler, 1994, p. 419). By contrast, attitudes toward the product will be lower if the stimuli are
perceived to be incongruent.
In order to test whether a matchup between endorser’s characteristics and perceived
product healthfulness affects participants’ perceptions of the product, we pose the following
hypothesis:
H3: Model body type in the ads will affect participants’ perceptions of product
healthfulness
In 1977, a study that explored the roles traditionally portrayed by women in advertising
(Peterson & Kerin, 1977) provided support to the notion of different levels of congruency
between the product and the model in the advertisement, as suggested by Kanungo and Pang
(1973). Their experiment involved both a feminine product (body oil) and a masculine product
(ratchet wrench set) with the same fictitious brand name (Vade). The ads that featured these
items had a professional model posing behind the product, with different levels of sensuality.
The data showed that the match between the seductive model and the body oil resulted in more
28

positive attitudes toward the advertisement (finding it more appealing) and the product when
compared with their other treatments. However, the nude model/wrench set condition was
perceived to be the least appealing. Building on this notion of a matchup between the product
characteristics and the model endorsing it, Kahle and Homer (1985) exposed 200 men and
women to ads for disposable razors with models that varied in their levels of perceived
attractiveness. Their results suggested that ads featuring an attractive model resulted in more
positive attitudes toward the product than those featuring an unattractive source.
Building on these studies, and particularly on those that focused the attention on the
match between a model’s attractiveness and the product characteristics, Solomon et al. (1992)
introduced the beauty matchup hypothesis to account for more subtle differentiations among
positive sources (i.e., models) that interact with product images to assess ad effectiveness than
the model introduced by Kamins (1990). Their study showed that particular products were
associated with specific types of good looks, such that Cosmopolitan magazine was better
advertised when using models that highlighted the sexual emphasis of the magazine, and
products such as Chanel perfume were better matched with models with a “classic beauty” look.
The beauty match up revision posited that a model’s particular type of beauty, and associated
image, is paired with a specific product to enhance the acceptance of the ad (vs. positing that
simple good looks are enough to influence product and ad evaluation). It also specifies that
consumers can differentiate among multiple types of good looks among different models, and
that in advertising, certain beauty ideals are more appropriately paired with certain products. A
plethora of research shows the effectiveness of using attractive rather than unattractive endorsers
and models in advertising and promotions (Baker & Churchill, 1977; Bloch & Richins, 1992;
Solomon, Ashmore, & Longo, 1992), lending support to the notion of the matchup hypothesis.

29

More recently, Till and Busler (2000) exposed respondents to ads with models that either
“matched” or “mismatched” the products endorsed. In particular, they investigated the relative
effectiveness of ads for an energy bar promoted by an athlete in the ad (considered to be a match)
or an actor (mismatch). Their results showed that respondents had more favorable evaluations
when there was a matchup than when the ad featured a mismatch. Similarly, Bower and Landreth
(2001) investigated the persuasive effects of manipulating the attractiveness level of the model
using beauty products. Their results showed that ad effectiveness depended on whether the
beauty product improved appearance (i.e., aesthetic purposes by enhancing beauty, such as
jewelry and colognes) or solved appearance problems (e.g., products that serve to fix or hide
beauty liabilities or flaws such as acne or dandruff), such that products that improved appearance
were better matched with more attractive endorsers. Choi and Rifon (2012) analyzed the
matchup among different celebrities and attractiveness (vs. technology related) products and
found similar results. They concluded that a matchup between the endorser and the product
enhances purchase intentions only through attitudes toward the advertisement and the brand, but
not directly toward the product. In 2016, Wright called for more recent studies that investigate
the matchup hypotheses in advertising and replicated the seminal work by Till and Busler
(2000). The study exposed participants to ads with either an attractive or an unattractive model
presented alongside an image of a fictional cologne brand called “Aerius 2.” Wright’s data
showed (with some differences to Till and Busler’s) that attitudes toward the brand were more
favorable when the model was attractive (vs. unattractive), but model attractiveness did not have
an effect on purchase intention.
However, there have been cases where the matchup effect was not observed. For
example, Kamins and Gupta (1994) explored whether the expertise of the endorser of a computer

30

system (with the ad claiming that the model was the company’s Chief Executive Officer) would
yield more positive evaluations because of the high degree of fit between the model and the
product. However, no significant differences were observed between the model’s influence in the
product evaluation when the advertised product was a shoe or the computer system. According to
Wright (2016), the lack of a matchup effect in certain studies reflects that perceived congruence
between products and models may be more dynamic and contextually dependent than previously
theorized. Wright (2016) cites Bower and Landreth’s (2001) study that found that ad
effectiveness (per the matchup hypothesis) depended on whether the advertised product
enhanced or solved appearance problems. In a similar line, Mundel & Huddleston (2017)
explored the effects of featuring models with different body sizes in ads for fast food brands
(McDonalds and Subway). It was expected that featuring endomorphic (vs. mesomorphic)
models in fast food ads would result in more positive attitudes toward the ad and toward the
product, as per the matchup hypothesis, and results showed that ad evaluations were more
positive for McDonalds when the model was endomorphic than mesomorphic. However, this
trend was reversed with Subway, where ad evaluations were more positive for ads with
mesomorphic models than for ads with endomorphic models (Mundel & Huddleston, 2017).
These results might indicate that when the product is perceived to be healthier, a model with a fit
body is better suited to advertise the product (and vice versa). Yet, there is a dearth of literature
examining the congruence between models with different body sizes and products with varied
levels of perceived healthfulness. Thus, it is considered meaningful to investigate how perceived
healthfulness of the advertised product affects participants’ evaluations of the model featured in
the ad, and in turn, of the product and the brand.

31

Given the need to further the understanding of the matchup effects between models and
different product types (Wright, 2016), and that there is an overrepresentation of models with
idealized bodies in advertising (Elliott & Elliott, 2007), this dissertation hypothesizes that social
media ads that match up the food’s healthfulness to the body composition of the model will
evoke more favorable responses, including ad, product, and brand attitude, as well as greater
viral behavioral intentions and purchase intentions.
H4: Participants’ (H4a) attitudes toward the ad and (H4b) the product, (H4c) viral
behavioral intentions, and (H4c) intentions to purchase the product will vary between
idealized, average, and overweight looking models at a greater magnitude when the ads
feature a healthy than an unhealthy snack.

32

CHAPTER 4
METHODS

Four studies were conducted for this dissertation, three pretests to identify stimulus
products and test stimulus materials (e.g., brands and models), and the main study, an online
experiment.
Study 1 – Snack Pretest
Design
A major focus of this research was to identify those snacks that are perceived to be
beneficial or detrimental to one’s health. To identify and select the snack foods featured in the
advertisements, a pre-test was conducted among students from a Michigan State University. The
goal of this pre-test was to determine consumer familiarity with, past consumption of, and
attitudes toward different snacks, and to select stimulus snacks that consumers perceive to
promote either healthy or unhealthy food products. In order to pre-test meaningful snacks, in
other words, those that were appropriate for the healthy or unhealthy categories, two lists of
healthy and unhealthy snacks were identified to enhance the study’s external validity by using
real companies (brands). A list of the healthiest snacks in the U.S. was obtained from Lab42, a
market research company that published a white paper entitled “A closer look at healthy
snacking” (Lab42, 2013). This selection is considered appropriate given that the snacks listed
were chosen based on the foods’ caloric content, fat and sugar proportion, but also the sales
volumes of each snack product. The healthy products selected to be pre-tested were granola,
Greek yogurt, nuts, dried fruit, baby carrots, and hummus. To match these six healthy snacks, a

33

list of unhealthy snacks was identified. These snacks were chocolate cookies, chocolate spread,
chocolate bars, cheese puffs, cereal bars high in sugar and potato chips (Meyers, 2014).
Participants
Forty undergraduate male students were recruited from a student pool at Michigan State
University and offered extra credit for participation. Slightly more than half (51.4%) of this
convenience sample were juniors, followed by seniors (21.6%), sophomores (18.9%), and
freshmen (5.4%), and graduate students (2.7%). 59.5% of the sample was white, with most
respondents (64.8%) reporting affluent family incomes ($100,000 or more yearly). The mean age
was 22 years old, ranging between 19 and 31 years old.
Measures and Procedures
Students completed short self-administered questionnaires including questions measuring
familiarity with the snack foods described above (a seven-point semantic differential scale
anchored by 1= unfamiliar and 7= familiar), attitudes toward those products using a three item,
seven-point semantic differential scale (MacKenzie & Lutz, 1989), and perceptions of the
product as being healthy using a three item, seven-point semantic differential scale. Those snacks
with similar ratings on the attitude scales were selected for the main study. To assess attitudes for
the pretest snacks, the three attitude items were combined to form a scale for each product (α
ranged from .90 to .98). Similarly, the items related to healthfulness were combined (α ranged
from .74 to .97).

34

Results
Means and standard deviations for all these measures are presented in Table 1.
Table 1: Product Pretest
FAMILIARITY
M

SD

PAST
CONSUMPTION
M

SD

ATT
PRODUCT
M

PERCEIVED
HEALTHFULNESS

SD

M

SD

Granola
Greek
Yogurt

5.42

1.77

4.79

6.31

5.28

1.28

5.33

1.25

5.08

2.08

3.82

4.62

5.45

1.22

5.78

1.08

Nuts

5.79

1.66

5.24

6.11

5.39

1.51

5.38

1.39

Dried Fruits

2.24

1.68

1.82

1.81

4.66

1.45

4.95

1.51

Carrots

5.74

1.74

5.30

5.30

5.37

1.51

6.26

0.92

Hummus

2.73

2.04

2.59

3.41

4.39

1.81

5.09

1.58

Cookies
Chocolate
Spread

6.16

1.09

3.92

4.97

5.13

1.60

2.83

1.24

5.34

1.92

1.97

3.12

4.78

1.65

3.24

1.40

Chocolate Bars

6.27

1.07

3.84

4.68

5.11

1.60

2.70

0.95

Cheese Spread

6.26

1.35

5.00

5.73

4.96

1.71

2.87

1.10

Cereal Bars

4.81

2.01

2.30

2.95

4.61

1.20

3.48

0.98

Chips

6.37

0.91

5.82

6.67

5.02

1.48

2.70

1.04

Given these results, the snacks selected for the main study as being the healthiest were baby
carrots (M = 6.26, SD = 0.92), Greek yogurt (M = 5.78, SD = 1.08), and nuts (M = 5.38, SD =
1.39). The three unhealthy snacks selected for the main study were potato chips (M = 2.70, SD =
1.04), chocolate bars (M = 2.70, SD = 0.95), and cookies (M = 2.83, SD = 1.24). Paired samples
t-tests were run comparing participants’ attitudes toward the product to ensure there were no
significant differences between the pairs of healthy and unhealthy products: baby carrots/potato
chips (t (37) = 1.00, p > 0.05), Greek yogurt/ chocolate bars (t (36) = 1.23, p > 0.05), nuts/
cookies (t (37) = 1.02, p > 0.05).

35

Study 2 – Stimulus Development
To control for past brand exposure experience, this dissertation used foreign brands for
the products selected from the pre-test. The use of foreign brands (vs. brands available in the
respective market) was recommended given that it allows the researcher to have more control
over past brand exposure experience, therefore allowing an examination of the experimental
manipulation in isolation of previous attitudes toward the brand (Nelson, Keum, & Yaros, 2004;
Schneider & Cornwell, 2005).
Based on Ma’s (2013) procedures, we included 12 British brands and 12 Australian
brands for a pilot study. These included four potato chips brands (i.e., Walkers, Smith’s,
Nando’s, McCoy’s), four chocolate brands (i.e., Topic, Cadbury Dairy Milk, Haigh’s, Pod’s),
four cookies brands (i.e., Maryland, Drizzles, FOX’s, Arnott’s), two baby carrots brands (i.e.,
Organix, Um, Goodies, Only Organic), four yogurt brands (i.e., Liberté, Muller, Vaalia, Tamar
Valley), and four nut brands (i.e., Sensations, KP, Nobbys, Lucky). For the main study, the
brands with lowest scores in familiarity and a score closer to the mean in perceived healthfulness
were selected to control for variations in perceived brand healthfulness.
Measures and Procedures
40 undergraduate male students were recruited from a large Midwestern university and
offered extra credit for participating. Students completed a short online questionnaire including
questions measuring brand familiarity (a seven-point semantic differential scale anchored by
familiar/unfamiliar), a three item 7-point scale measuring attitudes towards the brand
(MacKenzie & Lutz, 1989), and a three item 7-point scale measuring and perceived brand
healthfulness (Bauer et al., 2013).

36

Table 2: Brand Pretest Results
Familiarity
M
SD
Arnotts
Maryland
Drizzles
FOX’s
Walkers
Smith’s
Nando’s
McCoys
Topic
Cadbury
Haighs
Pod’s
Liberté
Muller
Vaalia
Tamar Valley
Sensations
KP
Nobby’s
Lucky

2.50
2.41
2.76
2.65
3.32
2.68
2.56
2.65
2.50
3.47
3.32
2.82
3.44
3.97
3.29
3.35
3.06
2.79
2.88
3.24

1.81
1.73
1.92
1.86
2.18
1.82
1.74
1.67
1.81
2.08
2.23
1.88
1.97
2.08
2.14
2.23
1.95
1.89
1.89
2.12

AttB
M

SD

4.51
3.91
4.15
4.17
3.68
3.96
4.09
3.55
3.84
4.29
4.56
3.98
4.63
4.80
5.13
4.85
4.06
3.90
4.27
4.57

1.35
1.35
1.28
1.32
1.41
1.27
1.28
1.32
1.38
1.60
1.56
1.20
1.37
1.45
1.37
1.36
1.42
1.14
1.12
1.00

Perceived
Healthfulness
M
SD
3.86
3.87
3.82
3.75
3.31
3.66
3.70
3.44
3.68
3.75
4.20
3.61
4.90
5.23
5.25
5.22
3.59
3.90
4.55
4.89

1.24
1.34
1.39
1.29
1.36
1.20
1.30
1.33
1.36
1.40
1.28
1.22
1.37
1.36
1.39
1.17
1.37
1.17
1.12
1.06

Results
After successful reliability analyses (AttB, α = .92; Perceived Healthfulness, α = .91), the
items were averaged and combined into individual scales. Filtering the analyses by product type
(e.g., cookies), one-way analyses of variance were conducted among each product type to
uncover differences among the individual brands. The results are presented below per product:
-

Cookies: No significant differences were found in terms of familiarity (F(3,132) = .248, p
> .05), AttB (F(3,132) = 1.17, p > .05), or perceived healthfulness (F(3,132) = .06, p >
.05) among the cookie brands.
37

-

Potato Chips: No significant differences were found in terms of familiarity (F(3,132) =
.248, p > .05), AttB (F(3,132) = 1.17, p > .05), or perceived healthfulness (F(3,132) = .06,
p > .05) among the potato chip brands.

-

Chocolate bars: No significant differences were found in terms of familiarity (F(3,132) =
1.70, p > .05), AttB (F(3,132) = 1.68, p > .05), or perceived healthfulness (F(3,132) =
1.39, p > .05) among the chocolate bars.

-

Greek Yogurt: No significant differences were found in terms of familiarity (F(3,132) =
.74, p > .05), AttB (F(3,132) = .76, p > .05), or perceived healthfulness (F(3,132) = .53, p
> .05) among the Greek yogurt brands.

-

Nuts: No significant differences were found in terms of familiarity (F(3,132) = .34, p >
.05) or AttB (F(3,132) = 2.04, p > .05). However, a significant difference was found
between the brands in terms of perceived healthfulness (F(3,132) = 8.55, p = .000)
among nut brands. Thus, we selected KP, as it did not exhibit significant differences
compared to the other brands.

Given that the brands for the other three snacks did not exhibit significant differences, we
selected the brands based on convenience in terms of securing the snacks.
Study 3 – Model Pretest
Design
To ensure that the male models cast for the main study are perceived to have body sizes
that fit the three body types described (idealized, average, overweight), a pre-test was conducted
through Amazon’s Mechanical Turk. Before exposing participants to the models, an illustration
of the typical American male was presented to them as a reference (Spector, 2013). For each
model, participants were asked to estimate the percentage of body fat and muscle in each model.

38

Next, they were asked to indicate whether they believed the model had an overweight, average or
an idealized body. A total of 12 males with different body compositions were recruited for this
pretest. The table below describes their height (in inches), weight (in pounds), and waist size (in
inches), body fat and muscle percent.
Table 3: Descriptive information on models selected pretested for main study

Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7
Model 8
Model 9
Model 10
Model 11
Model 12

Height

Weight

Waist Size

Body Fat
%

Muscle %

6’
5’7”
5’11”
6’2”
6’
5”8’
5’11”
5’7”
6’1”
6’
5’11”
5’9”

170
230
160
175
225
187
170
197
182
180
215
167

30”
41”
31”
31”
36”
34”
30”
34”
30”
33”
36”
32”

9
25
13
11
16
18
9
16
8
15
17
15

91
75
87
89
84
82
91
84
92
85
83
85

Participants
60 male respondents were recruited through Amazon mTurk and offered a cash incentive
for participating. The majority of the sample was white (61.7%), with most respondents (51.7%)
reporting modest family incomes ($10,000 to $49,000 yearly). The mean age was 22 years old,
ranging between 19 and 31 years old. The mean age was 35 years old, ranging between 19 and
58 years old. Over a third of the sample reported having completed a bachelor’s degree (36%),
followed by some college classes but no degree (21.7%).

39

Measures and Procedures
Participants completed a short online questionnaire including questions measuring model
familiarity (a seven-point semantic differential scale anchored by familiar/unfamiliar), a three
item 7-point scale measuring attitudes towards the model (MacKenzie & Lutz, 1989), and two
questions related to the body composition of the model.
Results
After successful reliability analyses (AttM, α = .96), the items were averaged and
combined into individual scales. Comparisons were made through an ANOVA between the
models in the same body type category (i.e., overweight, average, and idealized) to ensure there
were no significant differences in terms of attitudes toward the model. Models 7, 10, and 12 were
discarded from future analyses because of their significantly different attitude scores toward the
model (F(659) = 29.41, p < .001).
The use of the slider bar to estimate percent of body fat and muscle did not prove
satisfactory, as participants’ estimates of muscle and fat percent were not reflective of the study’s
pre-set categories. The tables below reflect the means for groups in homogeneous subsets for
both muscle and fat percent as the result of one-way ANOVA analyses.

40

Table 4: Muscle % Estimated Average by Subset

Model 2
Model 5
Model 12
Model 8
Model 11
Model 6
Model 3
Model 9
Model 7
Model 1
Model 4
Sig.

Subset 1
Subset 2
Subset 3
40.71
42.81
45.21
47.01
47.01
47.56
47.56
52.80
52.80
52.80
53.56
53.56
53.56
59.31
59.31
59.63
59.63
61.75
65.01
.05
.06
.08

While the estimated muscle percent was lower than indicated in the guidelines offered by
Brierley et al. (2016)’s study, it is possible to find within the same subset the models that were
considered to belong to each body type category as reported below.
Similarly, while the data shows that participants were able to identify different subsets of
models according to their body fat, the estimated percent of body fat does not reflect the
guidelines described by Brierley et al. (2016). The subsets are presented below in Table 5.

41

Table 5: Fat % Estimated Average by Subset

Model 4
Model 9
Model 1
Model 3
Model 7
Model 6
Model 8
Model 11
Model 12
Model 5
Model 2
Sig.

Subset 1
15.58
16.71
17.71
21.16
22.63

0.139

Subset 2

21.16
22.63
26.46

0.542

Subset 3

26.46
31.91
32.16
32.68

0.294

Subset 4

Subset 5

31.91
32.16
32.68
37.06
0.586

49
1

Given that the slider use did not prove successful in accurately estimating body fat
percentage, chi-square tests were performed over the second body type measure. The tests
confirmed that there were no significant differences among models in the same body type
category, but scores for the models were significantly different across categories (X2(20) =
618.45, p = .000). Figure 1 below illustrates the selected models for the main study per category.

42

Figure 1: Model Pretest Results
Models with idealized bodies

Count = 48a

Count = 54a

Count = 45a

Count = 42b

Count = 49b

Models with average bodies

Count = 42b

43

Figure 1 (cont’d)
Models with overweight bodies

Count = 58c

Count = 32c

Count = 44c

Each subscript letter denotes a subset of model categories whose column proportions do not differ significantly from each other
at the .05 level.

Study 4 – Main Study
Design
To test the study’s hypotheses, this study used a 3 (model body type: idealized vs.
average vs. overweight) x 2 (Perceived snack healthfulness: healthy vs. unhealthy) x 3 (ad
repetition) mixed factorial design, with all factors except ad repetition manipulated between
subjects, as shown in the model found in Appendix B. Research suggests that the continued
pairing of the different stimuli shown in an ad (i.e., brand, model, product) through advertising
repetition increases the strength of the belief that the brand possesses the attributes advertised
(Fishbein, 1967; Olson & Mitchell, 1975). In addition, ad repetition could account for variations
in the effects of the ad. Thus, this study exposed participants to three similar advertisements.
Using a Qualtrics-administrated panel, participants were randomly assigned to view three

44

Facebook ads in one of the six experimental conditions. The model with hypothesized
relationships is presented in Appendix B – Model.

Participants
As in the previous studies, the population under study were U.S. male adults. Subjects
were recruited through a panel service operated by Qualtrics. This panel uses a double opt-in
process, where participants are initially prescreened by Qualtrics and invited to be in their
database. For each particular study, Qualtrics sends a second invitation where participants have
to agree to participate in the survey. As an incentive, panel members received compensation
through Qualtrics’ (final compensation to participant not disclosed by the panel administrator).
Four hundred and two U.S. male participants aged 20 to 85 participated in this study (M
= 51, SD = 15.89). Half of the sample was aged 50 or younger. Most participants had some
education, including some college but no degree (22.4%) or a bachelor’s degree (28.4%). Most
participants reported being employed (52.4%), and had a relatively low household income,
between $10,000 and $49,999 (44.3%), followed by $50,000 to $99,999 (32.8%). Ninety-four
percent of respondents were non-Hispanic, with the majority of the sample being white (86.1%).
Participants indicated frequent snacking (86%), with 53.5% of respondents indicating the
consumption of snacks four or more times a week, followed by two to three times a week
(32.8%). Furthermore, most participants indicated snacking once or twice a day (72.9%),
followed by three or four times a day (18.7%). Thus, the sample is deemed appropriate for the
study.
The panel service pre-screened the responses for missing data and delivered a data set
with less than 1% of partial responses. Thus, all 402 participants were included in the analysis.

45

Stimulus Materials
Based on the results of the pilot studies, stimulus materials (ads) were created for each of
the experimental conditions. The stimuli were Facebook advertisements for unhealthy (e.g.,
potato chips) and healthy (e.g., baby carrots) snacks. Each ad was shot outdoors, and featured the
model holding the product. The brand’s logo and a small picture of the product were
superimposed over the image. To allow participants’ identification of the stimuli as social media
ads, each ad mimicked the aspect of a Facebook post, including a frame with the options for
measuring participants’ evaluations of the post (i.e., comment, like, share). Each post had a
description that was adapted from other popular snack brands’ Facebook pages.
Based on the pre-test results, three ads were created for each experimental condition (i.e.,
healthy snack X model with idealized body, healthy snack X model with average body, healthy
snack X model with overweight body, unhealthy snack X model with idealized body, unhealthy
snack X model with average body, and unhealthy snack X model with overweight body). Each
ad featured a different model and product based on the appropriate categories. Sample ads are
shown below.

46

Figure 2: Sample Ads Per Condition

Healthy x Idealized

Healthy x Average

Healthy x Overweight

Unhealthy x Idealized

Unhealthy x Average

Unhealthy x Overweight

Procedures
The experiment was conducted using Qualtrics.com in August 2017. Upon providing
consent, participants answered screening questions (sex, weight, height, snacking behavior).
Participants were then randomly assigned by the online survey system to one of six experimental
conditions. In each condition, participants were exposed to three ads, each followed by questions
related to attitudes toward the ad, the brand, viral behavioral intentions, perceived product
healthiness, and perceived model fit with the particular advertisement and product. Following
exposure to the ads, participants answered questions regarding self-esteem, questions for the
body image assessment for obesity scale (BIA-O; Williamson et al., 2000), and muscularity
dissatisfaction. Participants reported demographic information (e.g., age, educational level,
47

household income). Description of the items follow. The questionnaire is shown in Appendix A
– Instruments. A correlation matrix is included in the Appendix C – Correlations.
Measures
Independent Variables
Endorser’s body type is defined as the general appearance of an individual human body
(Frederick, Sandhu, Morse, & Swami, 2016). Participants were randomly assigned to view three
Facebook ads featuring either a model with an idealized, average, or overweight body fat. We
based the distinction on the literature reviewed in the second chapter: models with idealized
bodies were high on muscularity (at least 79% of their body composition) and low on fat mass
(ranging between 8 and 21%). Average models had body fat ranging between 28 and 40%, and
muscle mass ranging between 60 and 75% (St. Onge, 2010). Overweight models had a fat body
mass above 40%. The pretest previously described was conducted on Amazon’s Mechanical
Turk to ensure the models selected for the study were perceived to have such body compositions.
Given that the hypotheses concerned differences among exposure to idealized bodies vs. nonidealized, a dichotomous variable was created where 0 = exposure to non-idealized bodies in
advertisements, 1 = exposure to idealized bodies in advertisements.
Product Healthfulness participants responded to three 7-point Likert-type items
anchored by “Unhealthy/Healthy”, “Harmful/Beneficial”, and “Unwholesome/Wholesome”
(Mundel & Huddleston, 2017). Reliability tests were successful, with Cronbach's α ranging
between .81 and .97.
Ad repetition is defined as the number of ads displayed to each participant. In each
condition, participants were exposed to three comparable ads, each featuring a unique
combination between product and model (or no model in the control condition).

48

Muscularity Dissatisfaction: attitudinal preoccupations with muscle mass (McCreary et
al., 2004). The seven-item muscularity attitudes subscale of the Drive for Muscularity Scale
(DMS, McCreary, Sasse, Saucier, & Dorsch, 2004) was used to measure muscularity
dissatisfaction. Respondents indicated to what extent the items applied to them in a scale from 1
to 7 (always, never).
Control Variables
Unhealthy snack consumption: to control for possible differences accounted for the
frequency of consumption of these snacks, we included this variable in the model.
Mediators
Social Comparison: to assess participants’ upward or downward comparison we
followed the procedures in Vogel et al. (2014)’s study. Participants responded to the following
questions: “When comparing yourself to others on Facebook, to what extent do you focus on
people who are better off than you?”, “When comparing yourself to others on Facebook, to what
extent do you focus on people who are worse off than you?”, “When comparing yourself to
others on social media advertisements, to what extent do you focus on people who are better off
than you?”, “When comparing yourself to others on social media advertisements, to what extent
do you focus on people who are worse off than you?” (1 = not at all; 7 = a great deal). Two
variables were created, one for upward social comparison, and one for downward social
comparison. The scales were reliable, with a Cronbach's α of .93.
Dependent Variables
Attitudes towards the Ad (Aad) and Product (AP): participants’ predisposition to
respond in a (un)favorable manner to a particular ad, product, or other stimuli (Choi, Miracle, &
Biocca, 2001). Three seven-point semantic differential scales were used to measure attitude

49

toward the ad, the product, and the brand: negative/positive, bad/good, and
unfavorable/favorable (Choi, Miracle, & Biocca, 2001; Coulter & Punj, 2007; MacKenzie &
Lutz, 1989). The original scale, formed by the unweighted sum of four brand-related semantic
differential items, had a Cronbach's α of .92 (Coulter & Punj, 2007). Reliability tests for these
variables were successful, with Cronbach's α ranging between .97 and .98.
Viral behavioral intentions: this study used measures developed by Alhabash and
McAlister (2014). Virality is defined as whether participants think that: (1) the ad is worth
sharing; (2) they would recommend the ad to others; (3) they would like; (4) share; (5) comment
on the ad on Facebook, (6) would like the Facebook page of the advertised brand, and (7) would
post about the brand on their own Facebook pages. All items were measured on a 7-point Likerttype scale anchored by “Strongly Disagree” to “Strongly Agree.” Reliability analyses were
successful. Cronbach's α ranged between .97 and .98.
Rosenberg Self-Esteem Scale: participants’ general feelings about themselves
(Rosenberg, 1965). This dissertation used the Rosenberg self-esteem Scale (Rosenberg, 1965).
Participants indicated their agreement with 10 statements on 7-point Likert scales (1 = not at all
true, very true), such as “All in all, I am inclined to feel that I am a failure”. Initial reliability was
not successful. After performing a factor analysis, four items were dropped. The final variable,
consisting of six items, had a Cronbach's α of .81.
Purchase Intention: participants responded to the following question: what are the
prospects that you would purchase the product in the ad in the next 7 days? Three 7-point
semantic differential scales (e.g., unlikely/likely) were used to indicate their intent. Reliability
analyses were successful. Cronbach's α ranged between .95 and .96

50

CHAPTER 5
RESULTS

Manipulation Checks
Given that the use of slider bars to estimate percent of body fat and muscle failed to
provide robust results, this manipulation check was based on the model body type measure,
where participants were prompted to identify the model as having either an idealized, average, or
overweight body. Results of the chi-square analysis (χ2 (20) = 618.45, p < .001) are reflected
below.
Table 6: Model Manipulation Check Results
Perceived Model Body Type
Condition
Overweight

OVERWEIGHT
Count

71b

10c

25.3

79

29.7

39.60%

53.00%

7.50%

15a

91b

28a, b

25.3

79

29.7

11.20%

67.90%

20.90%

8a

75b

51c

25.3

79

29.7

6.00%

56.00%

38.10%

Count

76

237

89

Expected Count

76

237

89

% within Condition
Count
Expected Count
% within Condition
Idealized

Count
Expected Count
% within Condition

Total

IDEALIZED

53a

Expected Count
Average

AVERAGE

% within Condition
18.90%
59.00%
22.10%
Each subscript letter denotes a subset of Model Body Type categories whose column proportions do not differ
significantly from each other at the .05 level.

The results of the table above show that overall, regardless of body type, participants
tended to classify most of the models as having an average body. For the average condition, the
majority of participants (67%) indicated that the models had an average body, which confirms
51

the manipulation for the “average” experimental condition. In the overweight conditions, 39% of
participants believed the models had an overweight body, while the majority (53%) indicated the
models had an average body. Similarly, in the idealized models condition, 38% of the
respondents indicated the models had an idealized body, while the majority (56%) indicated the
models had an average body. These results were surprising given that a pre-test was conducted to
identify models that were suitable for the main study. Possible methodological reasons and the
effects of these results are discussed in the limitations section. However, it is important to note
that other researchers have stressed the difficulties that male samples have in judging body
dimensions associated with obesity (Elliott & Elliott, 2007). A survey by Cancer Research
carried out in 2012 examined 2000 adults and found that only seven percent of males correctly
judged their own weight, and that there seems to be a phenomenon where bigger sizes are
becoming the new normal, which leads to people not being able to recognize health problems
related to weight (Parry, 2014). The results herein lend support to men’s limited capacity to
identify different body compositions as being healthy or unhealthy, which calls for more research
to identify cues that would assist men in more accurately evaluating body composition.
To check for significant differences among perceived healthfulness of the products, the
scores for the three products shown by condition were averaged into a single item ranging from 1
(unhealthy) to 7 (healthy). An independent sample t-test was conducted comparing the scores of
those participants in healthy and unhealthy foods conditions. Results showed a significant
difference (t (390) = 10.29, p = .000) that confirmed the manipulation was successful. Products
in the healthy foods conditions were seen as healthier (M = 4.83, SD = 1.25) than in the
unhealthy foods conditions (M = 4.25, SD = 1.31). Further analyses were performed to determine
whether or not there were differences among each of the products shown to participants. An

52

ANOVA with repeated measures with a Greenhouse-Geisser correction was calculated, the mean
scores for product healthfulness were statistically significantly different (F(1.97, 9.61) = 12.13, p
< .001). In the healthy conditions, Organix carrots (M = 4.97, SD = 1.51) had higher scores for
perceived product healthfulness than KP peanuts (M = 4.41, SD = 1.59) and Liberté yogurt (M =
4.83, SD = 1.61). In the unhealthy conditions, FOX cookies (M = 4.40, SD = 1.46) received
higher scores than Cadbury chocolate (M = 4.34, SD = 1.47) and Walkers chips (M = 4.23, SD =
1.62).
Treatment Groups
To ensure that the six treatment groups were similar in terms of age, educational level,
and BMI, a series of Kruskal-Wallis H tests, which allow comparison of more than two
independent groups, were conducted.
A Kruskal-Wallis H test showed that there was no statistically significant difference in
age between the different treatments, χ2(5) = 2.54, p = 0.769. Similarly, there were no significant
differences in BMI (χ2(5) = 6.42, p = 0.267), or education (χ2(5) = 3.45, p = 0.630).

53

Table 7: Sample Description
Variables
Gender
Male
Age, mean years ± SD
BMI
Healthy
Average
Overweight
Education
High school graduate
Associate's degree
Some college no degree
Professional degree
Bachelor's degree
Graduate Degree
Ethnicity
Hispanic
White/Caucasian
Asian
Black/African American
Other
Income
Less than $10,000
$10,000 to $49,999
$50,000 to $99,999
$100,000 to $149,999
$150,000 or more

n
(N = 402)

%

402
50 ± 15.89

100%

130
153
119

32.3%
38.1%
29.6%

82
54
90
5
114
154

20.4%
13.4%
22.4%
1.2%
28.4%
38.4%

24
346
20
29
7
24
178
132
46
13
24

6%
86.1%
5%
7.2%
1.7%
6%
44.3%
32.8%
11.4%
3.2%
6%

Hypotheses Tests
To test H1, which concerned changes in participants reported self-esteem and body
dissatisfaction according to the body type of the model in the ad (i.e., average vs. idealized vs.
overweight), the dependent variables were submitted to a one-way ANOVA. Results showed no
significant differences between condition means for either self-esteem (H1a, F(2,399) = .586, p >
.05) or for body dissatisfaction (H1b, F(2,399) = 1.18, p > .05). Thus, H1 was not supported.

54

Given that the literature (Hargreaves & Tiggemann, 2009), shows that the degree to
which participants engage in social comparison mediates the effects of advertisements on men,
we conducted a mediation analysis on the effects of social comparison on self-esteem and body
dissatisfaction. To test H2, regarding whether social comparison mediated the effect of exposure
to idealized bodies on self-esteem (H2a) and muscularity dissatisfaction (H2b), PROCESS Macro
(Model 4; Hayes, 2013) was used. Rather than following a more traditional approach as in
Barron and Kennedy (1986) with a Sobel test, which is considered a very conservative test with
very low power (MacKinnon, Warsi, & Dwyer, 1995), we used the indirect effect as the measure
of mediation (Hayes & Rockwood, 2017). Hayes and Scharkow (2013) recommend using
bootstrapping, a non-parametric method based on resampling with replacement. With the release
of PROCESS Macros, which made macros such as SOBEL obsolete, a confidence interval for
the indirect effect is computed, and it is checked to determine if zero is in the interval. If not,
then it is assumed that the indirect effect is significant. It is important to note that bootstrapping
does not require the assumption that paths a and b in the mediation model are uncorrelated.
To test the hypotheses, a dichotomous independent variable (i.e., exposure to ads with
idealized bodies = 1, exposure to ads with non-idealized bodies = 0) was used. Given that model
attractiveness was significantly correlated with the dependent variables, perceived model
attractiveness was added to the model as a covariate.
For H2a, indirect effects were tested using a bootstrap estimation approach with 1000
samples (Shrout & Bolger, 2002). These results indicated the indirect coefficient was not
significant (b = -.02, SE =.01, 95% CI = -.07, .01). Therefore, these results do not support the
mediation hypothesis (Hayes & Rockwood, 2017). Results of the different paths are presented in
Table 8

55

Table 8: Mediation results of upward social comparison on the effects of exposure to idealized
bodies on self-esteem
Paths
Path a

Coefficient
(b)
-.361

t (df = 399)

p

-1.92

.06

.028

.83

.40

-.011

-.09

.92

-.001

-.01

.99

Exposure to idealized bodies à
upward social comparison

Path b
Upward social comparison à Selfesteem

Path c
Exposure to idealized bodies on
self-esteem controlling for upward
social comparison

Patch c’
Exposure to idealized bodies à
Self-esteem

For H2b, indirect effects were tested using a bootstrap estimation approach with 1000
samples (Shrout & Bolger, 2002). These results indicated the indirect coefficient was not
significant (b = -.039, SE =.02, 95% CI = -.1004, -.0042). Therefore, these results do not support
the mediation hypothesis. However, other approaches to mediation analyses, such as Barron and
Kenney’s, could focus on the partial mediation explained by exposure to idealized bodies and
upward social comparison, and through upward social comparison and the self-esteem. Results
are presented in Table 9.

56

Table 9: Mediation results of upward social comparison on the effects of exposure to idealized
bodies on body dissatisfaction
Paths
Path a

Coefficient (b)
-.361

t (df = 399)
-1.92

P
.05

.11

3.52

.001

.09

.76

.44

.13

1.10

.26

Exposure to idealized bodies à
upward social comparison

Path b
Upward social comparison à
Body dissatisfaction

Path c
Exposure to idealized bodies on
body dissatisfaction controlling
for upward social comparison

Patch c’
Exposure to idealized bodies à
Body dissatisfaction

Further analyses were conducted to evaluate whether perceived model attractiveness
influenced the results reported above. Simply put, we were curious to understand whether
models perceived to be more attractive could influence participants’ reported self-esteem and
body dissatisfaction. A 2 X 3 (attractiveness X model body type) factorial analysis of variance
tested the effects of perceived model attractiveness and condition on self-esteem. Results
indicated a non significant main effect for attractiveness, F(1,205) = 3.34, p > .05. The main
effect for the model body type was not significant either, F(2,205) = .28, p > .05. Similarly, the
interaction between the two factors was not significant, F(2,205) = .26, p > .05, indicating that
perceived model attractiveness was not a source of significant variation for participants’ selfesteem. Results are shown on Table 10.
Further, a 2 X 3 (attractiveness X model body type) factorial analysis of variance tested
the effects of perceived model attractiveness and condition on body dissatisfaction. Results
indicated a non significant main effect for attractiveness, F(1,205) = .63, p > .05. The main effect
for the model body type was not significant either, F(2,205) = .04, p > .05. Similarly, the
interaction between the two factors was not significant, F(2,205) = .63, p > .05, indicating that
57

perceived model attractiveness was not a source of significant variation for participants’ body
dissatisfaction.
Table 10: Average Self-esteem and Body Dissatisfaction per condition and perceived model
attractiveness
Perceived
Attractiveness
Not attractive

Attractive

Model Body
Type
Overweight Body
Average Body
Idealized Body
Overweight Body
Average Body
Idealized Body

Self-esteem
M
SD
5.27
1.26
5.17
1.14
5.50
1.13
5.60
1.28
5.68
1.19
5.67
1.42

Body Dissatisfaction
M
SD
3.90
1.21
3.86
0.90
3.74
1.44
3.94
1.21
3.96
1.18
3.99
1.10

H3 concerned whether or not the conditions would influence participants’ perceptions of the
products’ healthfulness. A one-way ANOVA was conducted comparing participants’ ratings of
product healthfulness by condition. Results showed a significant difference (F(5,396) = 2.69, p =
.02), indicating that in the unhealthy conditions, participants perceived the products to be less
healthy when the model was overweight than in the other conditions. On the other hand, in the
healthy products conditions, participants perceived the snacks to be healthier when the model
had an idealized body than in the other conditions. H3 was supported. Results are summarized in
Table 11.
Table 11: Perceived product healthfulness by condition
Idealized
Healthy
Products
Unhealthy
Products

Average

Overweight
M
SD

M

SD

M

SD

4.97

1.32

4.62

1.54

4.62

1.20

4.40

1.38

4.40

1.35

4.17

1.36

58

The last set of hypotheses dealt with the main effects of product healthfulness and models’ body
type and their interaction on attitudes toward the ads (H4a), attitudes toward the products (H4b),
viral behavioral intentions (H4c) and intentions to purchase the products (H4d). To analyze these
relationships, we used a 2 (product healthfulness) x 3 (models’ body type) x 3 (message
repetition) repeated measures ANOVA.
Attitudes toward the Ad. The analysis revealed that there was not a significant main effect of
message repetition on Aad, F(2, 792) = 1.60, p = .20. The main effect of model body type on Aad
was significant, F(2, 396) = 4.09, p < .05, η2p = .02. Participants expressed more favorable
attitudes towards the ad when exposed to models with idealized bodies (M = 4.73, SD = .12) than
in ads with average (M = 4.48, SD = .12) or overweight models (M = 4.24, SE = .12). Results are
illustrated on Figure 3. The main effect of snack healthfulness was not significant, F(1, 396) =
1.19, p > .05. The effect of model body type x snack healthfulness interaction on Aad was not
significant, F(1, 396) = 0.12, p > .05. Additionally, a quadratic trend three-way interaction
among model body type, snack healthfulness, and message repetition was not significant, F(2,
396) = .91, p > .05. H4a was not supported. Average means for the six experimental conditions
are presented on Table 12.

59

Figure 3: AAd by condition

Table 12: Average Aad according to body type and snack healthfulness
Body Type
Overweight body
Average body
Idealized body

Snack
Healthfulness
Unhealthy
Healthy
Unhealthy
Healthy
Unhealthy
Healthy

60

M
4.15
4.32
4.45
4.51
4.61
4.84

SE
0.17
0.17
0.17
0.17
0.17
0.17

Figure 4: Ad Repetition by Body Type by Snack Healthfulness Three Way Interaction for AAd

Attitudes toward the Product. The analysis revealed that there was a significant main effect of
message repetition on AP, F(2, 792) = 4.13, p = .02, η2p = .01. The main effect of model body
type, however, was not significant, F(2, 396) = 2.35, p > .05. The main effect of snack
healthfulness was not significant either, F(1, 396) = 1.23, p > .05. The effect of model body type
61

x snack healthfulness interaction on AP was not significant, F(2, 396) = 0.11, p > .05.
Additionally, a quadratic trend three-way interaction among model body type, snack
healthfulness, and message repetition was not significant, F(2, 396) = .18, p > .05. H4b was not
supported. Average means for the six experimental conditions are presented on Table 13.
Table 13: Average AP according to body type and snack healthfulness
Body Type
Overweight body
Average body
Idealized body

Snack
Healthfulness
Unhealthy
Healthy
Unhealthy
Healthy
Unhealthy
Healthy

M
4.31
4.53
4.59
4.66
4.70
4.88

SE
0.17
0.17
0.17
0.17
0.17
0.17

Viral Behavioral Intentions. The analysis revealed that there was not a significant main effect
of message repetition on VBI, F(1.98, 785.11) = 1.90, p > .05. The main effect of model body
type, was not significant, F(2, 396) = 0.42, p > .05. The main effect of snack healthfulness was
not significant either, F(1, 396) = 1.23, p > .05. The effect of model body type x snack
healthfulness interaction on VBI was not significant, F(2, 396) = 0.74, p > .05. However, a
quadratic trend three-way interaction among model body type, snack healthfulness, and message
repetition was significant, F(2, 396) = 4.54, p = .01, η2p = .02. Overall, participants exhibited
stronger viral behavioral intentions when exposed to ads for healthy products featuring a model
with an average body, than when exposed to unhealthy products with an average model. H4c was
partially supported. Average means for the six experimental conditions are presented on Table
14.

62

Table 14: Average VBI according to body type and snack healthfulness
Body Type
Overweight body
Average body
Idealized body

Snack
Healthfulness
Unhealthy
Healthy
Unhealthy
Healthy
Unhealthy
Healthy

M
1.88
1.92
1.81
2.23
1.95
2.09

SE
0.16
0.16
0.16
0.16
0.16
0.16

Figure 5: Ad Repetition by Body Type by Snack Healthfulness Three Way Interaction for VBI

63

Figure 5 (cont’d)

Purchase Intentions. The analysis revealed that there was a significant main effect of message
repetition on PI, F(2, 792) = 7.02, p = .001, η2p = .02. The main effect of model body type,
however, was not significant, F(2, 396) = 0.26, p > .05. The main effect of snack healthfulness
was not significant either, F(1, 396) = 2.44, p > .05. The effect of model body type x snack
healthfulness interaction on PI was not significant, F(2, 396) = 1.42, p > .05. Additionally, a
quadratic trend three-way interaction among model body type, snack healthfulness, and message
repetition was significant, F(2, 396) = 4.20, p < .05, η2p = .02. Overall, participants reported
stronger purchase intentions when exposed to ads for healthy snacks than for unhealthy snacks.
Further, those exposed to models with overweight bodies in the healthy snacks condition
expressed stronger PI than any of the other conditions. H4d was partially supported. Average
means for the six experimental conditions are presented on Table 15.

64

Table 15: Average PI according to body type and snack healthfulness
Body Type
Overweight body
Average body
Idealized body

Snack
Healthfulness
Unhealthy
Healthy
Unhealthy
Healthy
Unhealthy
Healthy

M
3.61
4.29
3.78
3.82
3.86
3.95

SE
0.21
0.21
0.21
0.21
0.21
0.21

Figure 6: Ad Repetition by Body Type by Snack Healthfulness Three Way Interaction for PI

65

Figure 6 (cont’d)

66

CHAPTER 6
DISCUSSION AND CONCLUSIONS

While it is widely recognized that food marketing is one of the main culprits for
increasing obesity and eating disorders in the United States (Chandon & Wansink, 2011),
researchers are becoming increasingly aware of the risks associated with exposure to foodrelated content in the image driven social media environment. In addition to the normative body
images projected by peers, the mainstream media, and traditional ads, social media sites
reinforce messages supporting thin and muscular bodies with images of idealized bodies. SNS go
beyond the portrayal of a particular type of beauty, and users accompany images showcasing
idealized bodies with disturbing hashtags (e.g., #thinspogram #thighgap or #bonespo) and
inspirational messages (e.g., “skip dinner, be thinner”). This is problematic because while social
media amplifies behaviors associated with eating disorders (e.g., body dissatisfaction, low selfesteem), there is also a constant stream of ads promoting foods that are high in caloric content,
high in sugars and saturated fats, and low in nutrients (Seiders & Petty, 2004). This dissertation
further summarized research on food distribution showing that unhealthy foods make their way
into different spheres of a person’s life through vending machines and distribution among
cafeterias and convenience shops (Chandon & Wansink, 2011). The need for more research, both
for foods contributing to obesogenic outcomes and body representations in advertisements,
becomes apparent.
In this dissertation, participants’ responses to healthy and unhealthy snack ads were
explored using two main frameworks: social comparison (Festinger, 1951) and the matchup
hypothesis (Kamins, 1990). Social comparison operates under the assumption that humans have

67

a drive for continual improvement, and for an individual to know what changes need to be made,
individuals compare themselves with other agents in the environment. Previous research has
shown that social comparison of physical appearance often results in upward comparison, and
that this type of comparison results in negative self-evaluations. We examined whether upward
social comparison would mediate the relationship between participants’ exposure to idealized
bodies and their reported self-esteem and body dissatisfaction. The mediational analyses, as well
as the analyses of variance, showed no significant differences in self-esteem when participants
engaged in upward social comparison. These results highlight the need for more research to
reach more definitive conclusions about the effect of mediated images reflecting a particular
standard of beauty on male self-esteem. While a number of studies showed that participants
reported lower levels of self-esteem after being exposed to idealized bodies (e.g., Petty, 2010;
Tiggemann & McGill, 2004), this dissertation does not lend support for this hypothesis. Other
lines of research had suggested that the negative scores on self-esteem as result of exposure to
idealized bodies could be explained by the respondents’ BMI (Henderson-King & HendersonKing, 1997). However, we did not find significant differences in participants’ self-esteem
between groups of respondents with different BMI. The study outcomes might be a function of
the models cast for the stimuli. Models with more defined bodies might trigger different
evaluations among male participants. It is also possible that the male population might be less
susceptible to changes in self-esteem as the result of exposure to ads than female samples, which
are frequently the subject under study in social comparison research.
Further, we tested the relationship between exposure to idealized bodies and upward
social comparison on participants’ body dissatisfaction. The mediation analyses did not lend
support for full mediation. However, we found that participants who were exposed to ads with

68

idealized bodies reported higher upward social comparison scores than those in the other
conditions. Our results also show that upward social comparison was a significant predictor of
body dissatisfaction, which is in line with previous research on upward social comparison
(Morrison et al., 2004; Suls, Martin, & Wheeler, 2002; Wheeler & Miyake, 1992).
Further, we were concerned with the effects of perceived attractiveness on participants’
self evaluations. Previous research has shown that physically attractive models generate more
favorable attitudes because the beauty of the model has an effect on product evaluations (Kim &
Sohn, 2016). Morrison et al. (2004) suggested that universalistic targets (i.e., the mass media)
elicit greater pressure to conform to idealistic standards of attractiveness than particularistic
targets (i.e., friends). However, we did not find attractiveness to explain variance in our
hypotheses. Future studies could benefit from using a single model whose body is digitally
manipulated to avoid differences in model attractiveness. Overall, the results suggest that men
might be less likely to experience negative feelings towards themselves as the result of media
exposure than was reported by Morrison and colleagues (2004).
The match up hypothesis posits that consumers tend to have more favorable evaluations
of ads and products when the cues in the ad “match” or fit well together (Landreth, 2001). Based
on our own preliminary study (Mundel & Huddleston, 2017) and the literature (Till & Busler,
2000; Wright, 2016), we expected participants to perceive healthy snacks as being healthier
when the model had a fit body (i.e., a match between endorser and product type) than in the other
conditions. As hypothesized, results showed that participants perceived the snacks to be less
healthy when the model was overweight than in the other conditions. Further, participants
perceived snacks to be healthier when the ads featured a model with an idealized body than in
the other conditions.

69

Further, we expected participants to report more positive attitudes towards the ads and the
products when exposed to models whose weight matched the healthfulness of the stimuli. In line
with previous studies, that showed that attractive models often result in more favorable ad and
product evaluations (Kim & Sohn, 2016), results showed that participants expressed more
favorable attitudes towards the ad when exposed to models with idealized bodies than when
exposed to ads in the average and overweight models conditions. However, the results indicate
that perceived healthfulness did not significantly interact with the body conditions, and that
perceived healthfulness was not a significant predictor of participants’ product or ad evaluations.
Thus, the contribution to the literature of the match up portion of this dissertation is in the
discovery of further limiting conditions of the advantages of models with idealized bodies.
Conversely, using models that reflect the average American male did not hurt ad or product
evaluation for snack foods social media ads. Further we compared models with idealized bodies
with more ecologically valid counterpoints, such as models with average and overweight body
types, which can be useful for future studies investigating the role of average-looking models on
participants’ self evaluations.
Future studies should investigate the importance of healthy eating and lifestyles among
males to understand the relationship between product type (e.g., healthy vs. unhealthy) and selfesteem, body dissatisfaction, and ad and product evaluation. We noted boundaries for the match
up effect between the independent variables (i.e., product healthfulness and model body type)
and the outcome variables. This is not surprising given the number of cases in the literature
where the matchup effect was not observed. For example, Kamins and Gupta (1994) showed that
matching an endorser expertise to electronic products did not offer significantly better product
evaluations than when featuring other types of endorsers.

70

We also explored the effects of exposing participants to social media ads promoting
healthy (vs. unhealthy) snack foods featuring models with different body sizes. Participants’
behavioral intentions to engage with the stimuli- (Facebook ads-viral behavioral intentions-VBI)
were low across all conditions. However, we found a significant quadratic trend interaction
among model body type, snack healthfulness and message repetition, which showed that
participants reported stronger VBI for healthy snacks in ads featuring models with average
bodies than in the other conditions. The low reported VBI might be the result of the stimuli.
Given that the ads were modeled after the brands’ social media posts, the advertisements might
not have been unique or compelling enough to influence the participants’ intentions to engage
with them. Another possible explanation lies in men’s reticence to discuss body-related issues,
which could result in lower intentions to interact with ads that could trigger weight and health
conversations. Further, the lack of a familiar brand name could have resulted in the overall low
VBI.
In regards to implications for marketing communications practitioners, this dissertation
showed that it is important for marketers to understand that the models cast for advertisements
exhort an influence on consumers’ perceptions of the product healthfulness as a result of the
model’s body type. We showed that participants perceived healthy snacks to be healthier when
exposed to models with idealized bodies, and unhealthier when exposed to unhealthy snacks with
overweight models. Thus, it is important for marketers to be aware of the influence the model
selected for a particular ad can have a positive or negative effect on product evaluations.
The results show that overall, participants in the healthy conditions exhibited more
positive attitudes and purchase intentions for products when the model featured in the ad had an
idealized body, yet these ratings were not always significantly different from those exposed to

71

ads with models featuring non-idealized bodies. This speaks about more credible and authentic
ads from food brands, where consumers could still favor an unhealthy product when presented
with an ad with models with body compositions that relate to the consumption of this type of
food. In the last few years, brands often known for offering unhealthy products have been taking
corrective steps as the result of the pressure of activists, and more health-conscious consumers
are shifting their business to brands that offer them alternatives (such as McDonalds burger
customization and salad options). Should these pressures spill over to ad regulation, marketers
could opt for including more diverse models that are congruent with the perceived healthfulness
of their product portfolio. Further, marketers operating in the food industry could opt for selfregulating as a gesture of good will before governmental intervention, which could result in a
stronger relationship with consumers.
Conclusions
Food marketing research shows that in the absence of health information in
advertisements, such as caloric intake and nutrition information, consumers develop evaluations
of the advertised product and its ingredients through different peripheral cues present in the ad
(Chandon & Wansink, 2011). These cues include branding, health claims, and stimuli such as
colors and the people featured in the ad. This dissertation showed that overall, participants
perceived snack foods to be healthier when exposed to models with idealized bodies in the
healthy products conditions, and unhealthier when exposed to models with non-idealized bodies
in the unhealthy products conditions.
Further, advertisers often feature attractive models in their ads, particularly those with
idealized bodies (Elliott & Elliott, 2007; Levine & Murnen, 2009) because of the effectiveness of
using attractive rather than unattractive models in their promotional material. And while more

72

policy groups voice their concerns about the negative consequences associated with these
practices on consumers’ self-evaluations and subsequent eating disorders, a hard reality looms
over the results of this study. Under First Amendment rights, advertisers and marketers have the
freedom to best promote their products. Thus, it seems unlikely that a regulation could easily be
set in place to help prevent using idealized bodies to promote unhealthy foods.
However, just as alcohol marketers have opted for self-regulation in regards to the age of
the models featured in advertisements as a means to prevent underage drinking behaviors as the
result of identification between the consumer and the model, food brands can choose their own
strategies to promote their products. Given that scholars have posited that exposure to ads
featuring models with idealized bodies can lead to depression and body dissatisfaction among
males, the results of this study suggest that the use of models with non-idealized bodies in
unhealthy food ads might be as effective as the “traditional” approach for some products, such as
snack foods.
Limitations
There are a number of limitations worth noting. First, the results of the manipulation
check for model body type are somewhat surprising given that the perception of the treatments
had been successful during the model pretest. However, Sigall and Mills (1998) claim this type
of outcome is not unusual for experimental designs. As per these researchers’ approach, the issue
is not that the experimental apparatus failed, but rather that the manipulation itself did not
achieve the intended effects on the independent variable within subjects. Furthermore, Sigall and
Mills (1998) claim that not seeing the desired effect on a manipulation check is not a threat to the
validity of the results, as although non-significant, there was an effect of the independent
variables on the outcome variables. Following Sigall and Mills (1998), if (1) there is an effect on

73

the dependent variables, (2) the researcher controls for other variables in the model (covariates),
and (3) excluding the sampling cases who do not pass the manipulation check does not affect the
results, there is no issue with the experimental design. Even more, according to John et al.
(2012), proceeding with an analysis without those who failed the check could be considered
cherry-picking and is a questionable practice. Future studies would perhaps obtain more robust
results by having the survey panel drop the cases in which the stimuli did not achieve the
intended effects on the subjects. Certain respondent panels, such as Qualtrics, allow the
researcher to set parameters on whether or not to include such cases. However, this was not an
option for this dissertation given limited financial resources. The results also might be explained
by a lack of enough of a difference in the two model body types to trigger significant differences
among conditions.
Second, while the studies described here used real foreign brands as stimuli and
respondents obtained through a panel (vs. a student sample), the surveys described above were
not random samples of the adult male population of the United States. Participants were
randomly sampled from the virtual panels used for the various studies, but the panels themselves
are not random. Future research with a more representative sample would be desirable. Future
studies would benefit from using a wider range of products to assess whether or not the results in
this dissertation would hold true for other product categories. It is possible that participants were
not necessarily driven to evaluate the ads and focused on the snacks themselves because of the
high frequency of snacking practices among this sample.
Another limitation of the current research related to the use of different pools of
participants for each study is discussed in the methods section. While in some cases we used
student samples, in other we used recruiting services, which might offer different outcomes.

74

These decisions responded to economic restrictions related to using a recruiting panel for testing
each outcome. Other researchers would benefit from sampling within the same platform.
Further, when designing the ads, some models were shown looking at the camera
directly, while others were shown from the side. This could have affected participants’
perceptions of the model’s body type. The last notation has to do with casting models from
different races. Future studies should control for model race in their analyses.
Implications
This research has both practical and theoretical implications. In this study, we did not
find support for the mediational effects of social comparison on men’s self-evaluations. While
previous research (e.g., Elliott & Elliott, 2005) indicate that idealized bodies raise comparison
standards for male’s own bodies and lower individual’s satisfaction with their own
attractiveness, this dissertation seems to show a boundary for the effects of social comparison
when it comes to comparing oneself to models in snack food advertisements. Further, while most
studies on body image dissatisfaction as a result of exposure to mediated images have been
centered on females because of their likelihood of expressing discomfort with their own bodies
than men (Eaton et al., 2012), this dissertation shows partial results suggesting that men can be
negatively affected by idealized representations of male bodies, particularly for snack food ads
endorsed by male models with different body sizes and compositions. This research lays the
ground for further analysis of the male population.
While the results did not vary significantly, in some cases the direction of the results
leaned in the direction hypothesized. Overall, participants rated healthy products to be healthier
when exposed to idealized bodies than in the other conditions, and unhealthy products to be
unhealthier when exposed to overweight individuals. Finally, our results showed that

75

respondents who engage in upward social comparison expressed less favorable attitudes toward
the ad in the idealized bodies conditions. We stress the need for greater corporate responsibility
from advertising agencies and food marketers. These actors could benefit from following a
similar approach to the Children’s Food and Beverage Advertising Initiative (CFBAI), which
promotes self-regulation among food marketers for non-nutritious food among children.
Thus, as activists and associations such as the National Eating Disorder Association seek
to promote regulation of models in the media, and as marketers and advertisers look to adapt to
the expectations of their consumers, this dissertation offers insight on how to better approach the
over-representation of models with idealized bodies across food ads.

76

APPENDICES

77

Appendix A – Instrument

INTRODUCTION
Thank you for agreeing to take part in our study.
In this study, you we are going to ask you to answer questions about certain brands and products, your
satisfaction with your own body, perceptions about different social media ads, and demographic
information. When you’re ready, please hit the continue button (>>).
PART ONE: SCREENING QUESTIONS
We want to make sure that you are eligible for this study. Please answer the following question:
c
c
c
c

1. What is your gender?
Female
Male
Transgender
Otherwise not listed
2.
3.
4.
5.

To the best of your knowledge, what is your current weight in lbs.?
To the best of your knowledge, what is your current height in feet and inches?
If you were to think of an ideal weight for yourself, what would that weight be in lbs.?
If you were to think of an ideal height for yourself, what would that height be in feet and
inches?

PART TWO: SNACKING BEHAVIOR
In this section, you will be asked a number of questions related to your snacking habits.
Please think about what you ate during the past week. You are going to mark the column that shows, on
average, how many times you ate that particular snack. If you did not eat this food or drink this beverage
during the past week, please mark "never or less than 1 per month."

How often do you have snack foods?
c
c
c
c
c

Never
Monthly or less
2-4 times a month
2-3 times a week
4 or more times a week

How many snacks do you have on a typical day?
c
c
c
c
c

1 or 2
3 or 4
5 or 6
7 or 9
10 or more

How often did you eat these foods in the past week?
78

Never
or
less
than
1 per
week

1 per
week

2-4
5-6
per
per
week week

1
per
day

2-3
per
day

Low-fat or non-fat potato chips, tortilla chips and corn chips
c
c
c
c
c c
(such as Baked Lays, Reduced-fat Doritos, Fat-Free Pringles)
Regular potato chips, tortilla chips, corn chips and puffs (such
c
c
c
c
c c
as all flavors of Ruffles, Lay's, Pringles, Doritos, Fritos,
Cheetos)
Other salty snacks (like cheese nibs, Chex mix, gold fish
c
c
c
c
c c
crackers, Ritz Bitz)
Candy, including chocolate, candy bars, jelly bellies,
c
c
c
c
c c
gummies and Lifesavers (do not include cookies)
Low or nonfat frozen desserts such as low fat ice
c
c
c
c
c c
cream, frozen yogurt, popsicles, & sherbet
c
c
c
c
c c
Regular ice cream & milkshakes (include all flavors)
How often did you eat a serving of vegetables such as green
c
c
c
c
c c
salad, peas, green beans or corn? (do not count fried potatoes
or French fries)
How often did you eat a serving of fruit such as a banana,
c
c
c
c
c c
apple or grapes? (do not count juices)
(Neuhouser, Lilley, Lund, & Johnson, 2009) – Pearson correlations for items ranging from 0.56 to 0.87
PART THREE: AD EVALUATION
In this section, you will evaluate different Facebook ads.
[Each participant will be exposed to a total of FACEBOOK 3 ads; either for a healthy or unhealthy
product (snack foods – already pretested) with either an average or fit model] Ads will be produced for
the experiment and pre-test
Please evaluate the specific AD that you just saw using the scale from 1 to 7.
Negative
Bad
Unfavorable

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

Positive
Good
Favorable

Please evaluate the PRODUCT that you just saw featured in the ad using the scale from 1 to 7.
Negative
Bad
Unfavorable

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

Positive
Good
Favorable

Please evaluate the PRODUCT that you just saw featured in the ad using the scale from 1 to 7.
Unhealthy

c

c

c

c

79

c

c

c

Healthy

4+
per
dau

c
c
c
c
c
c
c
c

Harmful
Unwholesome

c
c

c
c

c
c

c
c

c
c

c
c

c
c

Beneficial
Wholesome

Please indicate your agreement or disagreement with the following statements regarding what you
could possibly do in relation to the ADVERTISEMENT you’ve just seen. Use the scale from
1=Strongly Disagree to 7=Strongly Agree to rate each statement.
Strongly
Strongly
Disagree
Agree
1
2
3
4
5
6
7
c c c c c c c
This ad is worth sharing with others
c c c c c c c
I will recommend this ad to others
c c c c c c c
I will “like” this ad on Facebook
c c c c c c c
I will “share” this ad on Facebook
c c c c c c c
I will “comment” on this ad on Facebook
c c c c c c c
I will “follow” the brand on Facebook
Please use the following bipolar adjectives to rate your intentions to purchase the (PRODUCT) that
you just saw using the scale from 1 to 7.
Unlikely
Improbable
Impossible

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

Likely
Probable
Possible

Please indicate your agreement or disagreement with the following statements in relation to the
ADVERTISEMENT you’ve just seen. Use the scale from 1=Strongly Disagree to 7=Strongly Agree
to rate each statement.
Strongly
Strongly
Disagree
Agree
1
2
3
4
5
6
7
c c c c c c c
In this ad, the man and product together have clear meaning to me
c c c c c c c
It is obvious to me why they used this man to advertise this product
c c c c c c c
It is clear to my why the advertiser put this man with this product
I understand what the advertiser was trying to say to me about this product
c c c c c c c
by using this particular man
c c c c c c c
I see a connection between the man in the ad and the product
Because of the man used with this particular product in the ad, I had to
c c c c c c c
really question why it was there (R)
c c c c c c c
The man’s body size is acceptable for an advertisement
c c c c c c c
The man’s body size is suitable for an advertisement
c c c c c c c
The man’s body size is appropriate for an advertisement
PART FOUR: BIA-O
Below, you can to see a number of body silhouettes presented in random order. We need you to select the
silhouette that most accurately depicts your body size as you perceive it to be. Please be honest,
accurate, and definitive. You must choose only one silhouette by clicking on the button on the left side of
the image that most accurately depicts your current body size.

80

Next, you will see the same body silhouettes, and they will also be presented to you in random order.
Please select the silhouette that most accurately depicts the body size that you would most prefer.
Again, be honest, accurate, and definitive. You must choose only one silhouette.

PART FIVE: SOCIAL COMPARISON, SELF-ESTEEM
Please evaluate each of the following statements using the scale from 1 = Not at all to 7 = A great
deal
Not at all

Social Comparison on SNS
When comparing yourself to others on Facebook, to
what extent do you focus on people who are better off
than you?
When comparing yourself to others on Facebook, to
what extent do you focus on people who are worse off
than you?

1

2

3

5

5

6

A
great
deal
7

c

c

c

c

c

c

c

c

c

c

c

c

c

c

Please evaluate each of the following statements using the scale from 1 = Not at all to 7 =

81

Extremely

Self-Esteem (Rosenberg, 1965)
On the whole, I am satisfied with myself
At times I think I am no good at all
I feel that others respect and admire me
I feel that I have a number of good qualities
I am able to do things as well as most other people
I feel I do not have much to be proud of
I certainly feel useless at times
I feel that I'm a person of worth, at least on an equal
plane with others
I wish I could have more respect for myself
All in all, I am inclined to feel that I am a failure
I take a positive attitude toward myself

Not at
all
1
2

Extremely
3

5

5

6

7

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c
c
c
c
c
c
c

c

c

c

c

c

c

c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

c
c
c

Items 2, 5, 6, 8, 9 are reverse scored. Score: “Strongly Disagree” 1 point, “Disagree” 2 points, “Agree” 3 points, and “Strongly
Agree” 4 points. Sum scores for all ten items. Scores should be on a continuous scale. Higher scores indicate higher self-esteem.

Please evaluate each of the following statements using the scale from 1 = Always to 7= Never
Always
1
2
MBDS: Male Body Dissatisfaction Scale (Hallsworth et al.,
2005)
I think my chest is just the right size
I think my thighs are just the right size
I like the shape of my buttocks
I think my stomach is just the right size
I think my chest is too small
I feel satisfied with the shape of my body
I think my biceps are just the right size
I think my thighs are too small
I think that my biceps are too small
PART SIX: DEMOGRAPHIC QUESTIONS
What is your marital status?
c Married
c Widowed
c Divorced
c Separated
c Never married

82

c
c
c
c
c
c
c
c
c

c
c
c
c
c
c
c
c
c

Never
3

4

6

5

7

c
c
c
c
c
c
c
c
c

c
c
c
c
c
c
c
c
c

c
c
c
c
c
c
c
c
c

c
c
c
c
c
c
c
c
c

c
c
c
c
c
c
c
c
c

Which of the following best describes the area in which you live?
c Urban
c Suburban
c Rural
c Other, please specify
Please indicate your educational background
c Less than high school
c High school graduate
c Some college no degree
c Associate’s degree, occupational
c Associate’s degree, academic
c Bachelor’s degree
c Master’s degree
c Professional degree
c Doctoral degree
Please indicate your occupation
c Management, professional and related
c Service
c Sales and office
c Farming, fishing, and forestry
c Construction, extraction, and maintenance
c Production, transportation, and material moving
c Government
c Retired
c Unemployed
c Student
Which statement describes best your current employment status?
c Working (paid employee)
c Working (self-employed)
c Not working (temporary layoff from a job)
c Not working (looking for work)
c Not working (retired)
c Not working (disabled)
c Not working (other)
c Prefer not to answer
Are you….
c Hispanic
c Non-Hispanic
Which of the following best describes your ethnic background?
c White/Caucasian
c Black, African American, or Negro
c American Indian
c Alaska Native
c Asian
c Native Hawaiian or Other Pacific Islander
c Other, please specify: _________________

83

What is your family annual income?
c Less than $10,000
c $10,000 to $49,000
c $50,000 to $99,999
c $100,000 to $149,999
c $150,000 or more

84

Appendix B – Model

85

Appendix C – Correlations

Ad
Attitudes

Ad
Attitu
des
1

Purchase
Intention
s
.708**

Perceived
Model
Attractiveness
.572**

Unhealthy
Snack
Cons.
.293**

SelfEstee
m
.224*
*

Body
Dissatisfa
ction
0.023

Househo
ld
Income
-0.012

0

0

0

0

0.642

0.818

40
2
.55
3*
*
0

402

402

402

402

402

402

.752**

.502**

.272**

.263*
*

-0.003

-0.052

0

0

0

0

0.946

0.299

40
2
1

402

402

402

402

402

402

.684**

.484**

.478**

0.062

.098*

0

0

0

.215*
*
0

0.215

0.049

402

402

402

402

402

402

Pearson
Correlation
Sig. (2-tailed)

1

.413**

.359**

0

0.029

0

0

.226*
*
0

0.996

0.567

N

402

402

402

402

402

402

1

.195**

.176*
*

0.047

0.083

0

0

0.352

0.098

402

402

402

402

Pearson
Correlation
Sig. (2-tailed)

1

.180*
*
0

0.063

.117*

0.207

0.019

N

402

402

402

402

Pearson
Correlation
Sig. (2-tailed)

1

-.191**

0.019

0

0.707

N

402

402

402

Pearson
Correlation
Sig. (2-tailed)

1

0.069

N

402

Pearson
Correlati
on
Sig. (2-tailed)
N

Product
Attitudes

402

Pearson
Correlation

Product
Attitude
s
.911**
0
402
1

Sig. (2-tailed)
N
VBI

402

Pearson
Correlation
Sig. (2-tailed)
N

Purchase
Intentions

Perceived
Model
Attractiveness

V
BI
.61
5*
*
0

40
2

Pearson
Correlation
Sig. (2-tailed)
N

Unhealthy
Snack Cons.

Self-Esteem

Body
Dissatisfaction

Household
Income

402

0.168
402

Pearson
Correlation
Sig. (2-tailed)

1

N

402

** Correlation is significant at the 0.01 level (2tailed).
* Correlation is significant at the 0.05 level (2tailed).

86

REFERENCES

87

REFERENCES
Alhabash, S., Mundel, J., & Hussain, S. A. (2017). Social media advertising: Unraveling the
mystery box. In S. Rodgers & E. Thorson (Eds.), Digital advertising. Routledge.
Alhabash, S., Nieves-Pizarro, Y., Mundel, J., Deng, T., Huang, G., Lynch, K., Nelson, M.,
Timmons, R., and Kanver, D. (unpublished manuscript). Cultural Adaptation in U.S. and
Mexican Beer Ads: The Moderating Effect of Automatic Bias Against Hispanics on EyeTracking Measures.
Andersen, A. (2014). A brief history of eating disorders in males. Current findings on males with
eating disorders, 4-10.
Baglia, J. (2005). The Viagra ad venture: Masculinity, marketing, and the performance of sexual
health. New York: Peter Lang.
Baker, M. J., & Churchill Jr, G. A. (1977). The impact of physically attractive models on
advertising evaluations. Journal of Marketing research, 538-555.
Barker-Benfield, G.J. (1976). The Horrors of the Half-Known Life: male Attitudes Toward
Women and Sexuality in Nineteenth-Century America. New York: Harper & Row.
Barnic, B. (2014, November 26). The affluents’ affinity for digital. Retrieved from
http://www.mediapost.com/publications/article/238819/the-affluents-affinity-fordigital.html?utm_source=newsletter&utm_medium=email&utm_content=headline&utm_
campaign=78232
Batada, A., Seitz, M. D., Wootan, M. G., & Story, M. (2008). Nine out of 10 food
advertisements shown during Saturday morning children’s television programming are
for foods high in fat, sodium, or added sugars, or low in nutrients. Journal of the
American Dietetic Association, 108(4), 673-678.
Bauer, H. H., Heinrich, D., & Schäfer, D. B. (2013). The effects of organic labels on global,
local, and private brands: More hype than substance?. Journal of Business Research,
66(8), 1035-1043.
Bhanot, S. (2012). Use of social media by companies to reach their customers. SIES Journal of
Management, 8(1), 47.
Belkaoui, A., & Belkaoui, J. M. (1976). A comparative analysis of the roles portrayed by women
in print advertisements: 1958, 1970, 1972. Journal of Marketing Research, 168-172.
Bennet, S. (2013). The Top 20 Food Brands in Social Media [INFOGRAPHIC]. Retrieved from
http://www.adweek.com/socialtimes/top-20-social-food-brands/478658

88

Berman, R., & Wootan, M. (2003). Should chain restaurants be required to publish nutrition
labels on their menus. CQ Researcher, 95-97.
Belch, G. E., Belch, M. A., & Villarreal, A. (1987). Effects of advertising communications:
Review of research. Research in marketing.
Bilton, N. (2010). 22 Percent of Internet Time is Social, Nielsen Says. New York Times.
Bloch, P. H., & Richins, M. L. (1992). You look “mahvelous”: The pursuit of beauty and the
marketing concept. Psychology & Marketing, 9(1), 3-15.
Boyland, E. J., Harrold, J. A., Kirkham, T. C., & Halford, J. C. G. (2008). Does television food
advertising affect children's food preferences? Appetite, 51(3), 760.
Bower, A. B. (2001). Highly attractive models in advertising and the women who loathe them:
The implications of negative affect for spokesperson effectiveness. Journal of
Advertising, 30(3), 51-63.
Bower, A. B., & Landreth, S. (2001). Is beauty best? Highly versus normally attractive models in
advertising. Journal of Advertising, 30(1), 1-12.
Boyland, E. J., Nolan, S., Kelly, B., Tudur-Smith, C., Jones, A., Halford, J. C., & Robinson, E.
(2016). Advertising as a cue to consume: a systematic review and meta-analysis of the
effects of acute exposure to unhealthy food and nonalcoholic beverage advertising on
intake in children and adults. The American journal of clinical nutrition, ajcn120022.
Bragg, M. A., Yanamadala, S., Roberto, C. A., Harris, J. L., & Brownell, K. D. (2013). Athlete
endorsements in food marketing. Pediatrics, 132(5), 805-810.
Brierley, M.-E., Brooks, K. R., Mond, J., Stevenson, R. J., & Stephen, I. D. (2016). The Body
and the Beautiful: Health, Attractiveness and Body Composition in Men’s and Women’s
Bodies. PLoS ONE, 11(6), e0156722. http://doi.org/10.1371/journal.pone.0156722
Brinol, P., & Petty, R. E. (2009). Source factors in persuasion: A self-validation approach.
European Review of Social Psychology, 20(1), 49-96.
Brustein, J. (2010). Keeping Closer Eye on the Employees’ Social Networking. New York Times.
Busby, L. J., & Leichty, G. (1993). Feminism and advertising in traditional and nontraditional
women's magazines 1950s-1980s. Journalism & Mass Communication Quarterly, 70(2),
247-264.
Brown, J. D., & L'Engle, K. L. (2009). X-rated sexual attitudes and behaviors associated with US
early adolescents' exposure to sexually explicit media. Communication Research, 36(1),
129-151.

89

Cairns, G., & Macdonald, L. (2016). Stakeholder insights on the planning and development of an
independent benchmark standard for responsible food marketing. Evaluation and
program planning, 56, 109-120.
Chandon, P., & Wansink, B. (2011). Is food marketing making us fat? A multi-disciplinary
review. Faculty & Research. SSRN. Retrieved from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1854370
Choi, S. M., & Rifon, N. J. (2012). It is a match: The impact of congruence between celebrity
image and consumer ideal self on endorsement effectiveness. Psychology & Marketing,
29(9), 639-650.
Chou, H. T. G., & Edge, N. (2012). “They are happier and having better lives than I am”: the
impact of using Facebook on perceptions of others' lives. Cyberpsychology, Behavior,
and Social Networking, 15(2), 117-121.
Collins, R. L. (2000). Among the better ones. In Handbook of social comparison (pp. 159-171).
Springer US.
Cooke, L. (2007). The importance of exposure for healthy eating in childhood: a review. Journal
of human nutrition and dietetics, 20(4), 294-301.
Cruwys, T., Platow, M. J., Rieger, E., & Byrne, D. G. (2013). The development and validation of
the Dieting Intentions Scale (DIS). Psychological assessment, 25(1), 264.
Cutler, D. M., Glaeser, E. L., & Shapiro, J. M. (2003). Why have Americans become more
obese?. The Journal of Economic Perspectives, 17(3), 93-118.
De Houwer, J. (2009). How do people evaluate objects? A brief review. Social and Personality
Psychology Compass, 3(1), 36-48.
De Houwer, J., Gawronski, B., & Barnes-Holmes, D. (2013). A functional-cognitive framework
for attitude research. European Review of Social Psychology, 24(1), 252-287.
De Houwer, J., Geldof, T., & De Bruycker, E. (2005). The implicit association test as a general
measure of similarity. Canadian Journal of Experimental Psychology/Revue canadienne
de psychologie expérimentale, 59(4), 228.
Diedrichs, P. C., Lee, C., & Kelly, M. (2011). Seeing the beauty in everyday people: A
qualitative study of young Australians’ opinions on body image, the mass media and
models. Body Image, 8(3), 259-266.
Dixon, H., Scully, M., Kelly, B., Chapman, K., & Wakefield, M. (2014). Can counter-advertising
reduce pre-adolescent children's susceptibility to front-of-package promotions on
unhealthy foods?: Experimental research. Social Science & Medicine, 116, 211-219.

90

Dixon H, Scully M, Wakefield M, Kelly B, Chapman K, Donovan R. (2011b). Parent’s
responses to nutrient claims and sports celebrity endorsements on energy-dense and
nutrient- poor foods: an experimental study. Public Health Nutr.;14(6):1071–1079
Eaton, D. K., Kann, L., Kinchen, S., Ross, J., Hawkins, J., Harris, W. A., ... & Lim, C. (2006).
Youth risk behavior surveillance—United States, 2005. Journal of school health, 76(7),
353-372.
Elliott, R., & Elliott, C. (2005). Idealized images of the male body in advertising: a readerresponse exploration. Journal of Marketing Communications, 11(1), 3-19.
eMarketer (2015). Social Network Ad Spending to Hit $23.68 Billion Worldwide in 2015.
Retrieved from http://www.emarketer.com/Article/Social-Network-Ad-Spending-Hit2368-Billion-Worldwide-2015/1012357
FDA. (2016). Menu and Vending Machines Labeling Requirements. Retrieved from
https://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformati
on/LabelingNutrition/ucm217762.htm
Festinger, L., Back, K. W., & Schachter, S. (1950). Social pressures in informal groups: A study
of human factors in housing (Vol. 3). Stanford University Press.
Festinger, L. (1954). A theory of social comparison processes. Human relations, 7(2), 117-140.
Fildes, J., Robbins, A., Cave, L., Perrens, B. and Wearring, A. 2014, Mission Australia’s 2014
Youth Survey Report, Mission Australia.
Fleck, N. D., & Quester, P. (2007). Birds of a feather flock together… definition, role and
measure of congruence: An application to sponsorship. Psychology & Marketing, 24(11),
975-1000.
Forehand, M. R., & Perkins, A. (2005). Implicit assimilation and explicit contrast: A set/reset
model of response to celebrity voice-overs. Journal of Consumer Research, 32(3), 435441.
Frankenfield, D. C., Rowe, W. A., Cooney, R. N., Smith, J. S., & Becker, D. (2001). Limits of
body mass index to detect obesity and predict body composition. Nutrition, 17(1), 26-30.
Frederick, D. A., Sandhu, G., Morse, P. J., & Swami, V. (2016). Correlates of appearance and
weight satisfaction in a US National Sample: Personality, attachment style, television
viewing, self-esteem, and life satisfaction. Body image, 17, 191-203.
Friedman, H. H., & Friedman, L. (1979). Endorser effectiveness by product type. Journal of
advertising research, 19(5), 63-71.

91

Folta, S. C., Goldberg, J. P., Economos, C., Bell, R., & Meltzer, R. (2006). Food advertising
targeted at school-age children: A content analysis. Journal of nutrition education and
behavior, 38(4), 244-248.
Friese, M., Smith, C. T., Koever, M., & Bluemke, M. (2016). Implicit Measures of Attitudes and
Political Voting Behavior. Social and Personality Psychology Compass, 10(4), 188-201.
Gallagher, D., Heymsfield, S. B., Heo, M., Jebb, S. A., Murgatroyd, P. R., & Sakamoto, Y.
(2000). Healthy percentage body fat ranges: an approach for developing guidelines based
on body mass index. The American journal of clinical nutrition, 72(3), 694-701.
Gibson, B. (2008). Can evaluative conditioning change attitudes toward mature brands? New
evidence from the Implicit Association Test. Journal of Consumer Research, 35(1), 178188.
Goethals, G. R., & Darley, J. M. (1977). Social comparison theory: An attributional approach.
Social comparison processes: Theoretical and empirical perspectives, 259-278.
Gordon, B. F. (1966). Influence and social comparison as motives for affiliation. Journal of
Experimental Social Psychology, 1, 55-65.
Grau, S. L., & Zotos, Y. C. (2016). Gender stereotypes in advertising: a review of current
research. International Journal of Advertising, 35(5), 761-770.
Greenhow, C., & Lewin, C. (2016). Social media and education: reconceptualizing the
boundaries of formal and informal learning. Learning, Media and Technology, 41(1), 630.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). Measuring individual differences in
implicit cognition: the implicit association test. Journal of personality and social
psychology, 74(6), 1464.
Greenwald, A. G., & Nosek, B. A. (2001). Health of the Implicit Association Test at age 3.
Zeitschrift für Experimentelle Psychologie, 48(2), 85-93.
Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect of experimental presentation
of thin media images on body satisfaction: A meta-analytic review. International Journal
of eating disorders, 31(1), 1-16.
Haferkamp, N., & Krämer, N. C. (2011). Social comparison 2.0: Examining the effects of online
profiles on social-networking sites. Cyberpsychology, Behavior, and Social
Networking, 14(5), 309-314.
Halford, J. C., Gillespie, J., Brown, V., Pontin, E. E., & Dovey, T. M. (2004). Effect of television
advertisements for foods on food consumption in children. Appetite, 42(2), 221-225.

92

Hamblin, J. (2013). This Is the Average Man's Body. The Atlantic. Retrieved from
https://www.theatlantic.com/health/archive/2013/10/this-is-the-average-mansbody/280194/
Harnack, L., & French, S. (2003). Fattening up on fast food. Journal of the Academy of Nutrition
and Dietetics, 103(10), 1296.
Harris, J. L., Bargh, J. A., & Brownell, K. D. (2009). Priming effects of television food
advertising on eating behavior. Health psychology, 28(4), 404.
Harris, J., Schwartz, M., & Brownell, K. Fast Food FACTS: Evaluating Fast Food Nutrition and
Marketing to Youth, 2010. [WWW document]. URL http://www. rwjf.
org/files/research/20101108fffactsreport. pdf
Harris, J. L., Speers, S. E., Schwartz, M. B., & Brownell, K. D. (2012). US food company
branded advergames on the internet: children's exposure and effects on snack
consumption. Journal of Children and Media, 6(1), 51-68.
Hawkins, N., Richards, P. S., Granley, H. M., & Stein, D. M. (2004). The impact of exposure to
the thin-ideal media image on women. Eating disorders, 12(1), 35-50.
Hayes, A. F., & Rockwood, N. J. (2017). Regression-based statistical mediation and moderation
analysis in clinical research: Observations, recommendations, and
implementation. Behaviour research and therapy, 98, 39-57.
Hayes, A. F., & Scharkow, M. (2013). The relative trustworthiness of inferential tests of the
indirect effect in statistical mediation analysis: Does method really
matter?. Psychological science, 24(10), 1918-1927.
Hellmich, N. (2000) Body fixation may be muscling out health, USA Today, 19 September, p.
06D.
Hirschman, E. C., & Thompson, C. J. (1997). Why media matter: toward a richer understanding
of consumers' relationships with advertising and mass media. Journal of Advertising,
26(1), 43-60.
Hodgkins, C. E., Raats, M. M., Fife-Schaw, C., Peacock, M., Gröppel-Klein, A., Koenigstorfer,
J., ... & Gibbs, M. (2015). Guiding healthier food choice: systematic comparison of four
front-of-pack labelling systems and their effect on judgements of product healthiness.
British Journal of Nutrition, 113(10), 1652-1663.
Holmes, G. R., Spears, N., & Blankson, C. (2013). An Investigation of Match-Up Effects:
Influential Sources of Fit and the Generative Role of Imagination. Journal of Current
Issues & Research in Advertising, 34(1), 151-165.

93

Howlett, E. A., Burton, S., Bates, K., & Huggins, K. (2009). Coming to a restaurant near you?
Potential consumer responses to nutrition information disclosure on menus. Journal of
Consumer Research, 36(3), 494-503.
Huajian, C. (2003). A Review On Implicit Association Test. Advances in Psychological Science,
3, 016.
Jardine, A. (2015). ‘Real’ Guys Pose in underwear in Dove-Like Fashion Campaign. Creativity.
Retrieved from http://creativity-online.com/work/dressmann-underwear-for-perfectmen/43310
John, L. K., Loewenstein, G., & Prelec, D. (2012). Measuring the prevalence of questionable
research practices with incentives for truth telling. Psychological science, 23(5), 524-532.
Joseph, W. B. (1982). The credibility of physically attractive communicators: A review. Journal
of advertising, 11(3), 15-24.
Kahle, L. R., & Homer, P. M. (1985). Physical attractiveness of the celebrity endorser: A social
adaptation perspective. Journal of consumer research, 11(4), 954-961.
Kalb, I. (2013). Why Kraft's Sexy Ad For Italian Dressing Totally Backfired. Business Insider.
Retrieved from http://www.businessinsider.com/earth-to-kraft-your-salad-dressing-is-notsexy-2013-6
Kamins, M. A. (1990). An investigation into the “match-up” hypothesis in celebrity advertising:
When beauty may be only skin deep. Journal of advertising, 19(1), 4-13.
Kamins, M. A., & Gupta, K. (1994). Congruence between spokesperson and product type: A
matchup hypothesis perspective. Psychology & Marketing, 11(6), 569-586.
Kanungo, R. N., & Pang, S. (1973). Effects of human models on perceived product quality.
Journal of Applied Psychology, 57(2), 172.
Kaplan, A. M., & Haenlein, M. (2016). Higher education and the digital revolution: About
MOOCs, SPOCs, social media, and the Cookie Monster. Business Horizons, 59(4), 441450.
Khang, H., Ki, E. J., & Ye, L. (2012). Social media research in advertising, communication,
marketing, and public relations, 1997–2010. Journalism & Mass Communication
Quarterly, 89(2), 279-298.
Kelly, B., Halford, J. C., Boyland, E. J., Chapman, K., Bautista-Castaño, I., Berg, C., ... &
Grammatikaki, E. (2010). Television food advertising to children: a global perspective.
American Journal of Public Health, 100(9), 1730-1736.

94

Kim, H., & Sohn, S. H. (2016). Impact of Advertising Model's Body Size on Perceived Product
Quality: Category-Based Induction Approach. Journal of Promotion Management, 22(1),
151-174.
Knobloch-Westerwick, S., & Coates, B. (2006). Minority models in advertisements in magazines
popular with minorities. Journalism & Mass Communication Quarterly, 83(3), 596-614.
Koernig, S. K., & Page, A. L. (2002). What if your dentist looked like Tom Cruise? Applying the
match-up hypothesis to a service encounter. Psychology & Marketing, 19(1), 91-110.
Kolbe, R. H., & Albanese, P. J. (1996). Man to man: A content analysis of sole-male images in
male-audience magazines. Journal of Advertising, 25(4), 1-20.
Lab42. (2013). A closer look at healthy snacking. Retrieved from
http://lab42.com/infographics/a-closer-look-at-healthy-snacks
Latané, B., Eckman, J., & Joy, V. (1966). Shared stress and interpersonal attraction. Journal of
Experimental Social Psychology, 1, 80-94.
Lanzieri, N., & Hildebrandt, T. (2016). Using Objectification Theory to Examine the Effects of
Media on Gay Male Body Image. Clinical Social Work Journal, 44(1), 105-113.
Lee, S., Kim, W., & Kim, E. J. (2014). The extended match-up hypothesis model: the role of
self-referencing in athlete endorsement effects. International Journal of Sport
Management and Marketing, 15(5-6), 301-321.
Levine, M. P., & Murnen, S. K. (2009). " Everybody Knows That Mass Media Are/Are Not
[Pick One] A Cause of Eating Disorders": A Critical Review Of Evidence For A Causal
Link Between Media, Negative Body Image, And Disordered Eating In Females. Journal
of Social and Clinical Psychology, 28(1), 9.
Livingstone, S., & Helsper, E. J. (2006). Does advertising literacy mediate the effects of
advertising on children? A critical examination of two linked research literatures in
relation to obesity and food choice. Journal of communication, 56(3), 560-584.
Lynch, J., & Schuler, D. (1994). The matchup effect of spokesperson and product congruency: A
schema theory interpretation. Psychology & Marketing, 11(5), 417-445.
Lyubomirsky, S., & Ross, L. (1997). Hedonic consequences of social comparison: a contrast of
happy and unhappy people. Journal of personality and social psychology, 73(6), 1141.
Ma, M. (2013). Effect of online review modality on online shoppers' attitude and behavior.
Michigan State University.
Maison, D., Greenwald, A. G., & Bruin, R. H. (2004). Predictive validity of the Implicit
Association Test in studies of brands, consumer attitudes, and behavior. Journal of
consumer psychology, 14(4), 405-415.
95

Mandell, S. (2016). American Eagle Wants You to Know that ‘The Real You is Sexy’. Toleroad.
Retrieved from http://www.towleroad.com/2016/03/aerie-man/
Martin, M. C., & Gentry, J. W. (1997). Stuck in the model trap: The effects of beautiful models
in ads on female pre-adolescents and adolescents. Journal of Advertising, 26(2), 19-33.
Mettee, D. R., & Smith, G. (1977). Social comparison and interpersonal attraction: The case for
dissimilarity. Social comparison processes: Theoretical and empirical perspectives, 69,
101.
Meyers, D. (2014). America's Unhealthiest Packaged Snack Foods. The Daily Mail. Retrieved
from http://www.thedailymeal.com/eat/americas-unhealthiest-packaged-snack-foods
Mills, J. S., Polivy, J., Herman, C. P., & Tiggemann, M. (2002). Effects of exposure to thin
media images: Evidence of self-enhancement among restrained eaters. Personality and
Social Psychology Bulletin, 28(12), 1687-1699.
Morrison, T. G., Kalin, R., & Morrison, M. A. (2004). Body-image evaluation and body-image
investment among adolescents: A test of sociocultural and social comparison theories.
Adolescence, 39(155), 571.
Mundel, J., & Huddleston, P. (2017). Seeing beauty in the “average Joe”: effects of “real” male
models on social media food advertisements. Unpublished Manuscript.
Neuhouser, M. L., Lilley, S., Lund, A., & Johnson, D. B. (2009). Development and validation of
a beverage and snack questionnaire for use in evaluation of school nutrition policies.
Journal of the American Dietetic Association, 109(9), 1587-1592.
Neira, B., Corey, J., & Barber, B. L. (2014). Social networking site use: Linked to adolescents'
social self-concept, self-esteem, and depressed mood. Australian Journal of Psychology,
66(1), 56-64.
Nelson, M. R., Keum, H., & Yaros, R. A. (2004). Advertainment or adcreep game players’
attitudes toward advertising and product placements in computer games. Journal of
Interactive Advertising, 5(1), 3-21.
Nesi, J., & Prinstein, M. J. (2015). Using social media for social comparison and feedbackseeking: gender and popularity moderate associations with depressive symptoms. Journal
of abnormal child psychology, 43(8), 1427-1438.
Nestle M. Food marketing and childhood obesity—a matter of policy. N Engl J Med.
2006;354(24):2527–2529

96

Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers'
perceived expertise, trustworthiness, and attractiveness. Journal of advertising, 19(3), 3952.
Okazaki, S., Taylor, C. (2013) Social media and international advertising: Theoretical challenges
and future directions. International Marketing Review, 30 (1), 56-71.
Olmstead, K., & Lu, K. (2015). Digital News Revenue: Fact Sheet. Pew Research Center.
Retrieved from http://www.journalism.org/2015/04/29/digital-news-revenue-fact-sheet/
Panel, NHLBI Obesity Education Initative Expert. (1998). Clinical guidelines on the
identification, evaluation, and treatment of overweight and obesity in adults. Bethesda,
MD: National Heart, Lung, and Blood Institute.
Park, J., Felix, K., & Lee, G. (2007). Implicit attitudes toward Arab-Muslims and the moderating
effects of social information. Basic and Applied Social Psychology, 29(1), 35-45.
Patterson, M. & England, G. (2000). Body work: depicting the male body in magazines, in
Proceedings of the Academy of Marketing Annual Conference, University of Derby, CD
ROM.
Payne, B. K., Cheng, C. M., Govorun, O., & Stewart, B. D. (2005). An inkblot for attitudes:
affect misattribution as implicit measurement. Journal of personality and social
psychology, 89(3), 277.
Payne, B. K., Govorun, O., & Arbuckle, N. L. (2008). Automatic attitudes and alcohol: Does
implicit liking predict drinking?. Cognition & Emotion, 22(2), 238-271.
Perloff, R. M. (2014). Social media effects on young women’s body image concerns: Theoretical
perspectives and an agenda for research. Sex Roles, 71(11-12), 363-377.
Perry, L. (2014). Nearly all obese people are in denial about their size: 90% don't think they have
a weight problem because bigger sizes are becoming the new 'normal'. Daily Mail.
Retrieved from http://www.dailymail.co.uk/health/article-2833165/90-obese-people-dont-think-weight-problem.html#ixzz4xPH3zUv3
Peterson, R. A., & Kerin, R. A. (1977). The female role in advertisements: Some experimental
evidence. The Journal of Marketing, 59-63.
Petty, K. N. (2010). Implicit Theories, Dissatisfaction, and Risk Behavior: A New Theoretical
Approach to Body Image (Doctoral dissertation, University of Houston).
Pollay, R. W. (1986). Quality of life in the padded sell: Common criticisms of advertising's
cultural character and international public policies. Current Issues and Research in
Advertising, 9(1-2), 173-250.

97

Pope, H., Phillips, K. A., & Olivardia, R. (2000). The Adonis complex: The secret crisis of male
body obsession. Simon and Schuster.
Portwood-Stacer, L. (2014). Targeting Women – Fast Food Advertising. Advertising and Society
2014. Retrieved from https://advsoc2014.wordpress.com/2014/04/27/targeting-womenfast-food-advertising/
Presnell, K., Bearman, S. K., & Stice, E. (2004). Risk factors for body dissatisfaction in
adolescent boys and girls: A prospective study. International Journal of Eating
Disorders, 36(4), 389-401.
Quilliam, E.T. (2008). Happy meals, happy parents: Food marketing strategies and corporate
social responsibility. ProQuest.
Radloff, R. (1966). Social comparison and ability evaluation. Journal of Experimental Social
Psychology, 1, 6-26.
Richins, M. L. (1991). Social comparison and the idealized images of advertising. Journal of
consumer research, 18(1), 71-83.
Richins, M. (1996). Materialism, Desire and Discontent. Marketing and consumer research in
the public interest, 109, 132.
Rosenberg, M. (1965). Society and the adolescent self-image (Vol. 11, p. 326). Princeton, NJ:
Princeton university press.
Rosenberg, J., & Egbert, N. (2011). Online impres- sion management: Personality traits and
concerns for secondary goals as predictors of self-presenta- tion tactics on Facebook.
Journal of Computer- Mediated Communication, 17, 1–18. doi:10.1111/ j.10836101.2011.01560.x
Rothermund, K., & Wentura, D. (2004). Underlying processes in the implicit association test:
dissociating salience from associations. Journal of Experimental Psychology: General,
133(2), 139.
Saxena, A., & Khanna, U. (2013). Advertising on social network sites: A structural equation
modelling approach. Vision: The Journal of Business Perspective, 17(1), 17-25.
Schneider, L. P., Systems, B., & Cornwell, T. B. (2005). Cashing in on crashes via brand
placement in computer games: The effects of experience and flow on memory.
International Journal of Advertising, 24(3), 321-343.
Scott, D. M. (2015). The new rules of marketing and PR: How to use social media, online video,
mobile applications, blogs, news releases, and viral marketing to reach buyers directly.
Hoboken, NJ: John Wiley & Sons.

98

Seiders, K., & Petty, R. D. (2004). Obesity and the role of food marketing: A policy analysis of
issues and remedies. Journal of public policy and marketing, 23(2), 153-169.
Seiler, R., & Kucza, G. (2017). Source Credibility Model, Source Attractiveness Model And
Match-Up-Hypothesis–An Integrated Model. Economy & Business Journal, 11(1), 1-15.
Sherwood, C. (2015). 2015’s Top Food Brands on Instagram and What They’re Doing Right.
Room 214. Retrieved from http://room214.com/social/2015s-top-food-brands-oninstagram-and-what-theyre-doing-right
Sigall, H., & Mills, J. (1998). Measures of independent variables and mediators are useful in
social psychology experiments: But are they necessary?. Personality and Social
Psychology Review, 2(3), 218-226.
Sirgy, M. J. (1998). Materialism and quality of life. Social indicators research, 43(3), 227-260.
Skelly, G. U., & Lundstrom, W. J. (1981). Male sex roles in magazine advertising, 1959–1979.
Journal of Communication, 31(4), 52-57.
Slater, A., & Tiggemann, M. (2006). The contribution of physical activity and media use during
childhood and adolescence to adult women's body image. Journal of health psychology,
11(4), 553-565.
Sohn, S. H., & Youn, S. (2013). Does she have to be thin? testing the effects of models' body
sizes on advertising effectiveness. Atlantic Journal of Communication, 21(3), 164-183.
Solomon, M. R., Ashmore, R. D., & Longo, L. C. (1992). The beauty match-up hypothesis:
Congruence between types of beauty and product images in advertising. Journal of
advertising, 21(4), 23-34.
St-Onge, M. P. (2010). Are Normal-Weight Americans Over-Fat?. Obesity, 18(11), 2067-2068.
Statista. (2017). Most famous social network sites worldwide as of January 2017, ranked by
number of active users (in millions). Retrieved from
https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-ofusers/
Statista. (2017b). Daily time spent on social networking by internet users worldwide from 2012
to 2016 (in minutes). Retrieved from https://www.statista.com/statistics/433871/dailysocial-media-usage-worldwide/
Steil, J. M., & Hay, J. L. (1997). Social comparison in the workplace: A study of 60 dual-career
couples. Personality and Social Psychology Bulletin, 23, 427-438.

99

Stice, E., & Shaw, H. (2003). Prospective relations of body image, eating, and affective
disturbances to smoking onset in adolescent girls: how Virginia slims. Journal of
consulting and clinical psychology, 71(1), 129.
Story, M., & French, S. (2004). Food advertising and marketing directed at children and
adolescents in the US. International Journal of Behavioral Nutrition and Physical
Activity, 1(1), 3.
Suls, J., Martin, R., & Wheeler, L. (2002). Social comparison: Why, with whom, and with what
effect?. Current directions in psychological science, 11(5), 159-163.
Suls, J., & Wheeler, L. (2000). A selective history of classic and neo-social comparison theory.
In Handbook of social comparison (pp. 3-19). Springer US.
Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-Dunn, S. (1999). Exacting beauty:
Theory, assessment, and treatment of body image disturbance. American Psychological
Association.
Till, B. D., & Busler, M. (2000). The match-up hypothesis: Physical attractiveness, expertise,
and the role of fit on brand attitude, purchase intent and brand beliefs. Journal of
advertising, 29(3), 1-13.
Tordoff, M. G. (2002). Obesity by choice: the powerful influence of nutrient availability on
nutrient intake. American Journal of Physiology-Regulatory, Integrative and
Comparative Physiology, 282(5), R1536-R1539.
Vogel, E. A., Rose, J. P., Roberts, L. R., & Eckles, K. (2014). Social comparison, social media,
and self-esteem. Psychology of popular media culture, 3(4), 206.
Wartella, E. A., Lauricella, A. R., Cingel, D., & Connell, S. (2016). Children and Adolescents:
Television, Computers, and Media Viewing. In Encyclopedia of mental health. Academic
Press, Elsevier.
Weed, K. (2016, June 23). Keith Weed talks to Cannes Lions TV. Retrieved from
https://www.youtube.com/watch?v=T_pa80IDhyk&feature=youtu.be
Wells, L. E. (1976). Self-esteem: Its conceptualization and measurement (Doctoral dissertation,
ProQuest Information & Learning).
Wiers, R. W., & Stacy, A. W. (2006). Implicit cognition and addiction. Current Directions in
Psychological Science, 15(6), 292-296.
Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological
bulletin, 90(2), 245.

100

Wood, J. V. (1989). Theory and research concerning social comparisons of personal attributes.
Psychological bulletin, 106(2), 231.
World Health Organization. A framework for implementing the set of recommendations on the
marketing of foods and non-alcoholic beverages to children. 2012. Available at:
www.who.int/dietphysicalacti- vity/framework_marketing_food_to_child- ren/en/.
Wolheter, M., & Lammers, H. B. (1980). An analysis of male roles in print advertisements over
a 20-year span: 1958-1978. NA-Advances in Consumer Research Volume 07.
Wukich, C., & Steinberg, A. (2016). 11 Social Media for Emergency Management. Social Media
for Government: Theory and Practice, 140.
Wright, S. A. (2016). Reinvestigating the Endorser by Product Matchup Hypothesis in
Advertising. Journal of Advertising, 45(1), 26-32.
Yu, U. J. (2014). Deconstructing college students’ perceptions of thin-idealized versus
nonidealized media images on body dissatisfaction and advertising effectiveness.
Clothing and Textiles Research Journal, 0887302X14525850.

101