ERROR LOADING HTML FROM SOURCE (http://ncf.sobek.ufl.edu//design/skins/UFDC/html/header_item.html)

ADVERTISING ANDROGYNY

Permanent Link: http://ncf.sobek.ufl.edu/NCFE004860/00001

Material Information

Title: ADVERTISING ANDROGYNY HOW MODELS' GENDER PRESENTATION IMPACTS YOUNG ADULTS' EVALUATIONS OF ADVERTISEMENTS AND THE EFFECTS OF AN INCLUSIVE INSTRUCTIONAL FRAME
Physical Description: Book
Language: English
Creator: Scherr, Arielle
Publisher: New College of Florida
Place of Publication: Sarasota, Fla.
Creation Date: 2013
Publication Date: 2013

Subjects

Subjects / Keywords: Gender
Advertising
Advertisement
Presentation
Instructional Frame
Body Image
Genre: bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This study investigated how the gender presentation of models in advertisements and an instructional frame emphasizing gender fluidity affected adolescents' and young adults' evaluations of these ads. Participants were exposed to five advertising images picturing a feminine female, a masculine female, a feminine male, a masculine male, and a pair of jeans (control). Participants then completed a survey measuring their evaluations of model attractiveness, ad effectiveness, ad relatability, self-attractiveness, sex role orientation, and media-inspired social comparison behavior. Participants exposed to the inclusive instructional frame rated the four model images as more effective and relatable than participants in the control group who did not receive the instructional frame. Participants exposed to the inclusive instructional frame also rated images of gender incongruent models as more relatable, while participants in the control condition (no frame) exhibited the opposite effect. Female participants rated images of gender incongruent models as more effective and relatable, while male participants exhibited the opposite effect. Gender incongruent participants rated all four model images as more relatable than gender congruent participants. Female participants had higher levels of media-inspired social comparison behavior than male participants, and there was a negative correlation between self-attractiveness ratings and media-inspired social comparison behavior. The findings of the current study demonstrate the feasibility of changing perceptions with the simple use of an inclusive framework. The application of this method as a potential tool for adolescent media literacy educations is discussed as a means of mediating the negative effects of stereotyped images in advertisements
Statement of Responsibility: by Arielle Scherr
Thesis: Thesis (B.A.) -- New College of Florida, 2013
Electronic Access: RESTRICTED TO NCF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE
Bibliography: Includes bibliographical references.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Local: Faculty Sponsor: Barton, Michelle

Record Information

Source Institution: New College of Florida
Holding Location: New College of Florida
Rights Management: Applicable rights reserved.
Classification: local - S.T. 2013 S3
System ID: NCFE004860:00001

Permanent Link: http://ncf.sobek.ufl.edu/NCFE004860/00001

Material Information

Title: ADVERTISING ANDROGYNY HOW MODELS' GENDER PRESENTATION IMPACTS YOUNG ADULTS' EVALUATIONS OF ADVERTISEMENTS AND THE EFFECTS OF AN INCLUSIVE INSTRUCTIONAL FRAME
Physical Description: Book
Language: English
Creator: Scherr, Arielle
Publisher: New College of Florida
Place of Publication: Sarasota, Fla.
Creation Date: 2013
Publication Date: 2013

Subjects

Subjects / Keywords: Gender
Advertising
Advertisement
Presentation
Instructional Frame
Body Image
Genre: bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This study investigated how the gender presentation of models in advertisements and an instructional frame emphasizing gender fluidity affected adolescents' and young adults' evaluations of these ads. Participants were exposed to five advertising images picturing a feminine female, a masculine female, a feminine male, a masculine male, and a pair of jeans (control). Participants then completed a survey measuring their evaluations of model attractiveness, ad effectiveness, ad relatability, self-attractiveness, sex role orientation, and media-inspired social comparison behavior. Participants exposed to the inclusive instructional frame rated the four model images as more effective and relatable than participants in the control group who did not receive the instructional frame. Participants exposed to the inclusive instructional frame also rated images of gender incongruent models as more relatable, while participants in the control condition (no frame) exhibited the opposite effect. Female participants rated images of gender incongruent models as more effective and relatable, while male participants exhibited the opposite effect. Gender incongruent participants rated all four model images as more relatable than gender congruent participants. Female participants had higher levels of media-inspired social comparison behavior than male participants, and there was a negative correlation between self-attractiveness ratings and media-inspired social comparison behavior. The findings of the current study demonstrate the feasibility of changing perceptions with the simple use of an inclusive framework. The application of this method as a potential tool for adolescent media literacy educations is discussed as a means of mediating the negative effects of stereotyped images in advertisements
Statement of Responsibility: by Arielle Scherr
Thesis: Thesis (B.A.) -- New College of Florida, 2013
Electronic Access: RESTRICTED TO NCF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE
Bibliography: Includes bibliographical references.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Local: Faculty Sponsor: Barton, Michelle

Record Information

Source Institution: New College of Florida
Holding Location: New College of Florida
Rights Management: Applicable rights reserved.
Classification: local - S.T. 2013 S3
System ID: NCFE004860:00001


This item is only available as the following downloads:


Full Text

PAGE 1

ADVERTISING ANDROGYNY: HOW MODELS' GENDER PRESENTATION IMPACTS YOUNG ADULTS' EVALUATIONS OF ADVERTISEMENTS AND THE EFFECTS OF AN INCLUSIVE INSTRUCTIONAL FRAME BY ARIELLE SCHERR A Thesis Submitted to the Division of Psychology New College of Flori da in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Michelle Barton Sarasota, Florida April, 2013

PAGE 2

ii Acknowledgements I would like to thank my partner, Tyler Whitson, for helping me to come up with my thesis topic and supporting me through the process every day afterwards. Thank you for the coffee, the computer repairs, the shoulder to cry on, the motivating speeches, the trips to the library, and so much more. I love you. Thanks to my mom for suppo rting and loving me for the past 22 years. I wouldn't be the person I am today without your guidance. Thank you to my grandmother, Beatrice, for being the sweetest woman I know and loving me unconditionally. Thank you to my incredible friends: Zo e Rayor Jessa Baker Moss, Emily Adams, Lacy Mroz, and Jordan Campbell. Zoe, thank you for believing in me and encouraging me when I thou ght I would never graduate. T hank you for the beautiful Bacc invitations and for helping to edit the photos used in my exper iment You are my best friend and I love you with all my heart. Jessa, thank you for offering to read my thesis, make edits, and accompany me to the library every single one of these final days; thank you for watching Buffy and Daria with me when I thoug ht this thesis would never be finished. Emily, thanks for always making me feel like a role model and accomplished feminist, especially when I was down on myself. Lacy and Jordan, thank you for dealing with t he thesising version of myself, I know she can be difficult sometimes. I would like to thank Andrew Swain and Jordan Campbell for being my beautiful models. I know it took a few tries, thanks for sticking around and helping me end up with a great final product. Thanks to my sponsor, Michelle Barton who put up with me through two years of trying thesis revisions. I would also like to thank Duff Cooper, who helped me re run my statistics over and over again until I got them right. Thank you to my other two committee members, Heidi Harley and Steven Graham, for spending their time and energy on me, I know you're busy. I would like to dedicate this project to my dad, who was taken from me too soon. I wish you were here tod ay to see the woman I've become and the accomplishments I've made. I love you Dad.

PAGE 3

iii Table of Contents ACKNOWLEDGEMENTS ii TABLE OF CONTENTS iii LIST OF TABLES v LIST OF FIGURES vi ABSTRACT vii LITERATURE REVIEW 1 Overview 1 How Adolescents Process Advertising 4 How Gende r is Portrayed in Advertising 11 Gender Stereotypes in Adolescence 11 Gender Stereotypes in Advertising 13 Sexuality in Advertising 19 How Gender Portrayals in Advertising Effects Adolescents' Evaluations of Gender 21 Pre occupation with Physical Attractiveness and the Body 25 Summary 30 THE CURRENT STUDY 31 Hypotheses 34 METHOD 35 RESULTS 41 DISCUSSION 50 REFERENCES 60

PAGE 4

iv TABLES 64 FIGURES 65 APPEN DICES 81

PAGE 5

v LIST OF TABLES TABLE 1: Condition Assignment by Bem SRI Classification

PAGE 6

vi LIST OF FIGURES FIGURE 1: Bem SRI Category Distribution FIGURE 2: Participant Gender Congruence Distribution FIGURE 3: Self Reported Gender Identity Distribution FIGUR E 4: Model Attractiveness by Participant Condition FIGURE 5: Advertisement Effectiveness by Participant Condition FIGURE 6: Advertisement Relatability by Participant Condition FIGURE 7: Model Attractiveness by Participant Sex FIGURE 8: Advertisement Effect iveness by Participant Sex FIGURE 9: Advertisement Relatability by Participant Sex FIGURE 10: Model Attractiveness by Participant Gender Congruence FIGURE 11: Advertisement Effectiveness by Participant Gender Congruence FIGURE 12: Advertisement Relatabilit y by Participant Gender Congruence FIGURE 13: Self Attractiveness Ratings by Participant Condition FIGURE 14: Self Attractiveness Ratings by Participant Sex FIGURE 15: Self Attractiveness Ratings by Participant Gender Congruence FIGURE 16: Interaction Betw een Self Attractiveness Ratings and Media Inspired Social Comparison Behavior

PAGE 7

vii ADVERTISING ANDROGYNY: HOW MODELS' GENDER PRESENTATION IMPACTS YOUNG ADULTS' EVALUATIONS OF ADVERTISEMENTS AND THE EFFECTS OF AN INCLUSIVE INSTRUCTIONAL FRAME Arielle Scherr New College of Florida, 2013 ABSTRACT This study investigated how the gender presentation of models in advertisements and an instructional frame emphasizing gender fluidity affected adolescents' and young adults' evaluations of these ads. Participants were e xposed to five advertising images picturing a feminine female, a masculine female, a feminine male, a masculine male, and a pair of jeans (control). Participants then completed a survey measuring their evaluations of model attractiveness, ad effectiveness ad relatability, self attractiveness, sex role orientation, and media inspired social comparison behavior. Participants exposed to the inclusive instructional frame rated the four model images as more effective and relatable than participants in the con trol group who did not receive the instructional frame. Participants exposed to the inclusive instructional frame also rated images of gender incongruent models as more relatable, while participants in the control condition (no frame) exhibited the opposi te effect. Female participants rated images of gender incongruent models as more effective and relatable, while male participants exhibited the opposite effect. Gender incongruent participants rated all four model images as more relatable than gender con gruent participants. Female participants had higher levels of media inspired social comparison behavior than male participants, and there was a

PAGE 8

viii negative correlation between self attractiveness ratings and media inspired social comparison behavior. The fi ndings of the current study demonstrate the feasibility of changing perceptions with the simple use of an inclusive framework. The application of this method as a potential tool for adolescent media literacy educations is discussed as a means of mediating the negative effects of stereotyped images in advertisements _____________________________ Dr. Michelle Barton Division of Social Sciences

PAGE 9

1 Advertising Androgyny: How Models' Gender Presentation Impacts Young Adults' Evaluations of Advertisements and the Effects of an Inclusive Instructional Frame Overview [T]wo trends the growth in advertising channels reaching children and the privatization of children's media use have resulted in a dramatic increase in advertising directly intended for the eyes and ea rs of children. It is estimated that advertisers spend more than $12 billion per year to reach the youth market and that children view more than 40,000 commercials each year. These figures represent dramatic increases over those from the 1970s. (America n Psychological Association [APA], 2004) In order to address this massive growth of advertising directed at America's youth, the APA assembled the Task Force on Advertising and Children under recommendation from the APA Board of Directors, the Board for the Advancement of Psychology in the Public Interest, the Committee on Children, Youth, and Families, and the Committee on Women in Psychology. They discovered that advertising for non nutritious foods is strongly linked to childhood obesity, and advertis ements for tobacco and alcohol contribute to youth smoking and drinking (APA, 2004). Consequently, advertising conveys strong messages to children and adolescents about which behaviors are acceptable and encouraged whether advertisers intend to communicat e these soci al norms or not. The societal messages conveyed may communicate normative behavior concerning consumption, dress, sexuality, interpersonal relationships, or gender norms.

PAGE 10

2 Advertisers rely on reproducing or manipulating what is socially accept able in order to sell products to their target market. Because adolescents are particularly preoccupied with social acceptance and the proliferation of social conventions (Lamb & Urberg, 1978), they may be especially susceptible to the impacts of advertis ing on behavior. For adolescents, the visual information conveyed in advertisements is the most important component (Covell, 1992; Covell, Dion, & Dion, 1994; Edens & McCormick, 2000). Adolescents revealed a greater preference for image based advertisi ng than quality or verbal based ads (Covell, 1992; Covell, et al., 1994). After viewing visually oriented advertisements, adolescents remembered more facts, created more inferences, demonstrated a greater emotional response (Edens & McCormick, 2000), and reported higher levels of persuasion (Covell, et al., 1994) than they did with quality oriented ads. Additionally, the peripheral visual information in advertisements became the central message for adolescents because they could not distinguish between i magery used to catch viewers' attention and imagery used to convey product information (Edens & McCormick, 2000). Female adolescents had an especially high preference for visually oriented advertisements (Covell, 1992; Covell, et al., 1994). Preoccupatio n with the visual imagery in advertisements makes the models and actors used in these ads especially important to adolescents. Researchers have discovered time and time again that the models and actors in advertisements are portrayed in gender stereotyp ical ways (Furnham & Paltzer, 2010; Ganahl, Prinsen, & Netzley, 2003; Kahlenberg & Hein, 2010; Mager & Helgeson, 2011; Nam, Lee, & Hwang; see also Das, 2011; Doring & Poschl, 2006; Eisend, 2010; Kim & Lowry, 2005; Wong & Chan, 2006), regardless of the medi um of the ad, the country it is

PAGE 11

3 aired or printed in, the year in which it was released, the target consumer demographic, or the product it is trying to sell. Females in these ads are subordinated through the use of sexualization (Mager & Helgeson, 2011; S tankiewicz & Rosselli, 2008), underrepresentation in comparison to population demographics (Ganahl, et al., 2003), restriction to domestic settings (Furnham & Paltzer, 2010; Kahlenberg & Hein, 2010), dependency upon males (Furnham & Paltzer, 2010; Mager & Helgeson, 2011), emphasis on beauty and social acceptance (Furnham & Paltzer, 2010; Kahlenberg & Hein, 2010), occupation of less space (Nam, et al., 2011), youth (Furnham & Paltzer, 2010; Ganahl, et al., 2003), and vulnerable posing (Mager & Helgeson, 2011 ; Nam, et al., 2011). Males, on the other hand, are afforded more variety in terms of representation (Furnham & Paltzer, 2010; Kahlenberg & Hein, 2010; Nam, et al., 2011) and higher levels of credibility and authority (Furnham & Paltzer, 2010; Ganahl, et al., 2003) in advertisements. Adolescent preoccupation with advertising imagery that is replete with gender stereotypes has encouraged researchers to build on these findings and determine how these representations affect adolescents. Lafky, Duffy, Stei nmaus, and Berkowitz (1996) found that adolescents exposed to stereotypical representations of women in advertisements were more likely to make gender stereotypical assumptions about a female who was not engaged in any task. They also found that the oppos ite was true when adolescents were exposed to counter stereotypical representations of women in advertisements. Martin and Kennedy (1993) discovered that, when adolescent girls viewed images of highly attractive models, their guidelines for attractiveness became inflated and they rated a moderately attractive girl as less attractive than girls exposed to

PAGE 12

4 images of moderately attractive models did. Loken and Peck (2005) exposed adolescent girls to images of thin and larger sized models and found that the g irls whose exposure was prefaced by instructions emphasizing a range of female bodies more reflective of the average woman rated both the larger models and themselves as more attractive. In each case, exposure to realistic representations of gender and th e body elicited a greater acceptance of models who deviate from gendered norms. However, it is unclear whether this model applies to visual gender markers on the body, like hairstyles and jewelry, because the topic has not yet been researched. Thus, the goal of this study was to examine the effects of visual gender identifiers in advertisements on adolescents' and young adults' evaluations of model attractiveness, self attractiveness, advertisement effectiveness, and advertisement relatability. In order to explore how gender presentation in advertising impacts adolescents' evaluations, it is important to understand adolescents' relationship with both advertising and gender stereotypes. Thus, the following literature review will provide an overview of fin dings on advertising, gender, adolescence, and the relationships between these variables. The review begins by describing how adolescents process advertisements, focusing on their preoccupation with visual imagery. Then, the literature review moves on to describe the gender stereotypes that are emphasized in this ad based visual imagery, which portray women from a particularly negative and limited perspective. This section is followed by a discussion of the harmful effects this proliferation of gender st ereotypes has on adolescents, including a preoccupation with an unattainable beauty ideal and an unrealistic perception of gender. How Adolescents Process Advertising

PAGE 13

5 Research has shown that adolescents do not differentiate between information disseminate d from advertising, entertainment, and news media (Lariscy, Reber, & Paek, 2010). As a result, advertising is a tool of informational power with influence equivalent to that of educational, entertainment, and current events programming. Adolescents also responded more strongly to visual information in advertisements in that they remember more specific facts, create more inferences, elicit a greater emotional response (Edens, & McCormick, 2000), report higher levels of persuasion, and reveal greater prefer ence for visually based advertising than quality or verbal based ads (Covell, Dion, & Dion, 1994). These trends are even more robust in female adolescents, who recalled more facts, made more inferences, detected more explicit claims, and had a more inten se emotional response to these advertisements than males did (Edens & McCormick, 2000). Covell (1992) notes that image or visually oriented advertisements depict a lifestyle; they are designed to emphasize the image that can be gained and/or projected by use of the productImage oriented advertisements, unlike quality oriented advertisements, provide attainment information to the individual who desires the image the advertisement promises. The more highly valued the image, the more persuasive the adver tisement is expected to be to those wishing such information. (p. 49) In order to observe adolescent reactions to advertising images, Edens and McCormick (2000) tested a group of 195 middle and high school students (with an average age of 14.7 years) on how they processed advertisements, specifically focusing on the ad's characteristics, processing objective, and the effects of the participants' gender on these variables. The researchers selected six print advertisements from teen

PAGE 14

6 magazines and coded whe ther they utilized more visual or verbal content. Half of the participants were assigned to the low elaboration condition in which they were instructed to remember each advertisement's wording and image along with as many details as possible. The other h alf were assigned to the high elaboration condition in which they were instructed to form an opinion about each ad as it was presented, use their imagination by closing their eyes and visualizing the product, and decide how they felt about both the product and the advertisement itself. The six ads were projected on a screen for 20 seconds at a time, after which participants completed questionnaires measuring memory of specific facts, number of inferences made, level of emotional response, detection of expl icit claims in the advertisement, and whether claims were perceived as misleading or irrelevant. Factual memory was measured using a list of four factual questions about each advertisement, some of which were incorrect, which participants had to answer wi th yes, no, or don't remember. Inference formation was measured using five statements of claims implied in each advertisement which participants judged as true, false, or don't know. An example of one of these inferential statements is "People should wea r cologne mainly when going out with the opposite sex" (p.454). Four of the five statements for each ad were inferences that could be made based on the verbal and visual information in the advertisement, but were not necessarily true. The fifth statement was the explicit assertion of the advertisement. Emotional response was measured by asking participants to rate the strength of three emotional descriptors evoked by each ad using a 7 point Likert scale from weak to strong. Edens and McCormick (2000) co ncluded that the peripheral visual information in print advertisements becomes the central message for adolescents. Although few

PAGE 15

7 participants detected the explicit claims of the advertisements used in the study, m ore than half of those who did no t detect these claims still reported that the advertisement "made sense to them" (Edens & McCormick, 2000, p. 458). Advertisements high in visual content stimulated the memory of more facts, and elicited more inferences and more emotional responses than those high in verbal content. There was also a gender effect, such that females recalled more facts, made more inferences, detected more explicit claims, and elicited a higher emotional response than males, and this gender gap widened in evaluations of visually heav y advertisements. The researchers concluded that visual images have a direct bearing on cognitive and affective responses more so than text used in advertising and adolescents cannot distinguish between visual imagery used to catch viewers' attention and imagery used to convey product information. The influence that visual imagery in advertisements has on adolescents is especially important because this imagery conveys norms of social desirability and adult behavior. Covell (1992) conducted a study in o rder to determine the appeal of image oriented advertisements to adolescents and preadolescents. The researcher showed 158 8 to 16 year olds print ads from current magazines selling clothing, personal care products, tobacco, and alcohol. Half of the adv ertisements in each of these categories were image oriented and half were quality oriented. Participants were shown 12 clothing and personal care advertisements one at a time and were asked to rate them based on how much they liked the ads and how much mo ney they would be willing to spend on the product, using a 5 point Likert scale. The tobacco and alcohol advertisements were presented in pairs of one quality oriented ad and one image oriented ad for the same product, and participants were asked to pick which advertisement was the most

PAGE 16

8 persuasive, which was their personal favorite, and which they presume their peers would like best. Covell also had participants complete the Imaginary Audience Scale, which measures preoccupation with social image. Across all product categories, females selected more image oriented advertisements and rated these ads higher than males. The 14 to 16 year old group demonstrated the highest preference for image oriented advertisements, while the youngest group 8 to 10 year o lds demonstrated the highest preference for quality oriented ads. The 11 to 13 year old group, on the other hand, was split on the basis of sex (with females preferring image based ads and males preferring those that were quality oriented) and on concern with social image (with those more preoccupied with social image preferring image based advertisements). Covell (1992) noted that, on average, females experience a growth spurt at age 11 while males experience their growth spurt at age 13. This developm ental shift puts females ahead of the curve, which may be why 11 to 13 year old girls' advertising preferences resemble those of 14 to 16 year olds' and males in the same age group more closely resemble 8 to 10 year olds. The author stated that with t he earlier physiological development in girls are found issues of developing identity: lower self esteem, more self consciousness, and most particularly concerns about popularity with the opposite sex (Elkind & Bowen, 1979; Enright, Lapsley, & Shukla, 1 979; Simmons & Rosenberg, 1975; Simmons, Brown, Bush, & Blyth, 1978). Thus it might be argued that the 11 through 13 year old girls share with the older age group an interest in information about lifestyle and peer popularity such as that offered in t he image advertisements (Covell, 1992, p. 57).

PAGE 17

9 Covell teamed up with two more researchers in 1994 to test 75 pairs of participants, each containing a 12 to 16 year old adolescent and one of their parents on how an image or quality oriented advertisement further affected preference and persuasion. The researchers presented each participant with 24 print advertisements 12 selling alcohol and 12 selling tobacco one at a time in random order. Half of the advertisements in each product category used an imag e oriented method, while the other half focused on product quality. Each participant rated the advertisements' appeal and persuasiveness using a 5 point Likert scale. The adolescents and adults in the study both found image oriented advertisements to be more persuasive than quality oriented ads. This finding is important, as Covell et al. (1994) noted, because visual imagery in advertisements often denotes socially desirable behavior or appearance, which is a matter of great importance to female adolesc ents. Once again, there was a gender effect such that adolescent girls preferred visually oriented ads over quality oriented advertisements, with adolescent boys showing no preference. They concluded that advertisements that send messages about what is s ocially desirable influence the attitudes of adolescent girls. Adolescents' and adults' shared preference for visually oriented advertisements is notable because, as Conger (1987) noted, "[a]dolescence isa time when many of the values, interests, pattern s of behavior, and ways of coping that will characterize one's adult lifestyle are formed, for good or ill" (p. 447). Adolescents' consistent preoccupation with visual imagery in advertisements makes models in these advertisements especially important. Th e images of men and women in advertisements are particularly salient to the adolescents viewing these ads,

PAGE 18

10 who often compare themselves to the models featured in them (Morrison & Shaffer, 2003; Tsai & Chang, 2007 ). Tsai and Chang (2007) investigated these claims in their experiment measuring the effect of the physical attractiveness of advertising models on male and female adolescents. They prefaced their study with a focus group of 30 female and 30 male undergraduate students who were shown photos of 10 models of their own gender and instructed to rate each model on their physical attractiveness. The most highly rated male and female models were categorized as highly attractive and the fifth highest rated male and female models were categorized as normal ly attractive. These four images were used in the main experiment to create identical advertisements for a fictional brand. The researchers recruited 240 18 to 19 year old undergraduate students who were each assigned to view an advertising image featur ing either a highly or normally attractive model of their own gender. Participants then completed a brief questionnaire measuring product attitude and purchase intention using five 5 point Likert scale items. Adolescents of both genders in Tsai and Chan g's (2007) experiment attributed higher product attitude and purchase intention to the advertisement with a normally attractive model of their own gender than the ad with a highly attractive model. Tsai and Chang concluded that this preference was due to adolescent self comparison to models in advertising. They argued that viewing highly attractive models in advertisements encourages a preoccupation with physical attractiveness and inflates the acceptable level of attractiveness to a degree that is unreal istic. Tsai and Chang further argued that this unattainable beauty ideal can lead to "negative affect stemm[ing] from the deflated self image of potential customers when they compared themselves to beautiful models"

PAGE 19

11 (p.835). Adolescent preoccupation with visual imagery in advertising is troubling because of the societal messages embedded within these images. Lafky, Duffy, Steinmaus, & Berkowitz (1996) explained: In the barrage of commercial messages that individuals in North America are exposed to daily, advertisers often present certain social ideals rather than life as it is. Advertising messages are created through presentations of idealized stereotypical portrayals of individuals. Through the use of simplistic images that ignore the complexities of modern lives, stereotypes thus become an essential short hand through which advertisers can readily communicate a product category and indicate for whom the product is intended. The convenience of using visual short hand is one reason advertisers use imag es that fit gender stereotypes. Stereotypes are the foundation of what have been called the "ideologies" of advertising. The "ideal of domesticity" and the "beauty ideal," for example, provide normative guidelines for portrayals and activities of women a s well as men (p. 380). This use of stereotypes by advertisers as short hand for target consumer demographic information results in an advertising industry saturated with gender stereotypical representations, which are especially limiting in regards to wom en (Furnham & Paltzer, 2010; Ganahl, Prinsen, & Netzley, 2003; Kahlenberg & Hein, 2010; Mager & Helgeson, 2011; Nam, Lee, & Hwang, 2011; Stankiewicz & Rosselli, 2008 ). How Gender is Portrayed in Advertising Gender Stereotypes in Adolescence

PAGE 20

12 In Browne's ( 1998) study on gender stereotypes in children's television, the author defined gender stereotypes as "general beliefs about sex linked traits (collections of psychological characteristics and behaviors characterizing men and women) and gender roles (activi ties differentially appropriate for men or women)" (p. 83). During adolescence, there is an especially strong push to adhere to gender stereotypes in order to be accepted by peer groups (Lamb & Urberg, 1978). The first component of adolescent thought dri ving this pressure to conform to gender norms is adolescent egocentrism (Elkind, 1967). Elkind established that adolescents fail to distinguish between their thoughts and the thoughts of others because they are preoccupied with the self. The author noted that adolescents create an imaginary audience that they perform for because they believe that others are just as critical of them as they are of themselves. The physical and psychological effects of puberty also play a large role in young adults' preoccu pation with adherence to rigid gender stereotypes. The emergence of secondary sex characteristics during this time period makes adolescents more concerned with their bodies than ever before. Puberty stimulates an awareness of the appearance of both onese lf and others, along with awareness that others are scrutinizing them as well. Lamb and Urberg (1978) noted that realization of this physical scrutiny causes adolescents to attempt to present themselves in a way that optimizes others' evaluations of them. They go on to say, "awareness of one's body and of the importance of physical (and perhaps behavioral) appearance in evaluations by others provokes a comparison of one's body with an ideal" (p. 192). Since ideals rely on perfection, adolescents often fa il to measure up to idealized notions of beauty or behavior and end up feeling inadequate. As a result, "fear of peer disapproval and a desire to be valued highly by same and opposite sex

PAGE 21

13 peers thus combine to motivate a new commitment to conventional, s tereotyped sex roles" (p. 192). Gender Stereotypes in Advertising Researchers have set out to determine the prevalence of gendered themes in advertising for years, taking into account location, time, medium, source, and product category (Furnham & Paltzer 2010; Ganahl, et al., 2003; Kahlenberg & Hein, 2010; Mager & Helgeson, 2011; Nam, et al., 2011; see also Das, 2011; Doring & Poschl, 2006; Eisend, 2010; Kim & Lowry, 2005; Wong & Chan, 2006). They have all come to the same conclusion: advertisements ove rwhelmingly promote gender stereotypes. Mager and Helgeson (2011) analyzed 3,212 print advertisements from seven high circulation magazines in the United States published without interruption between 1950 and 2000 for gendered portrayals. These magazine s included two women's publications ( Cosmopolitan and Ladies' Home Journal ), two men's publications ( Esquire and Popular Mechanics ), two general editorial magazines ( Reader's Digest and National Geographic ), and one news magazine ( Time ). They coded the ad vertisements using five main coding sections: general counting categories, use of hands, function rating, ritualization of subordination, and licensed withdrawal. The general counting categories measured the number of males and females in each advertisem ent, the sex of models in suggestive' poses, and the sex of people represented only by body parts. The latter four categories are derived from Goffman's (1979) widely recognized rubric for measuring subtle portrayals of social position, sexuality, and se xism in advertisements. Use of hands is divided into two subcategories: ritualistic touch which is characterized by graceful placement of the hands, barely touching an item, or caressing items or people,

PAGE 22

14 and utilitarian touch characterized by grasping, ma nipulating, or holding an object or person. Function rating refers to a hierarchy of social position, namely an executive versus subordinate role, measured as performing an instructing role, receiving instruction or help, or being pictured in a traditiona lly male or female domain and looking either competent or unrealistic. Ritualization of subordination is another measure of a hierarchy of power, but with a stronger focus on the body rather than role. This variable was measured by analyzing the number o f people shown in images of deference, exposing their throat, in images of superiority, in physically low places, with obvious knee bending, in canting postures, in clownish postures, providing support or guidance, and receiving support or guidance. Licens ed withdrawal refers to engagement in actions that remove a person psychologically from their surroundings, which was measured by the number of people losing control', with fingers to their mouth, anxiously biting or sucking their finger or lips, in a fin ger to finger position, mentally drifting or looking into space, looking at a situation from behind something, and snuggling up. This withdrawal from a situation puts the withdrawn person in a position of dependence upon others around them for protection and prevents them from participating in what is happening around them. After analyzing the data, Mager and Helgeson (2011) found that nearly 90% of the models in suggestive, or sexualized, poses were female. Additionally, most of the models represented only by body parts were female. The authors assert that advertising images use the female body as a visual element and as decoration far more often than the male body is used for these same purposes. They also found that, over time, male models were les s likely to be represented using body parts but were more likely to be portrayed

PAGE 23

15 suggestively. This was consistent with their assertion that advertisers have relied on an increased use of sexuality to sell products over time. While males more often perfor med executive roles than females and females were instructed more often than males, these gender differences decreased with time creating a consistent trend towards gender balance in the area of role portrayals. The researchers also found that female mode ls were more often portrayed in positions of licensed withdrawal and utilizing ritualized touch than males. These results, taken as a whole, suggest that women are shown as dependent upon and subordinate to men and have been used increasingly as sexual ob jects in advertisements. Although male and female role positions have become more balanced overtly in function ranking, the more subtle indicators of cultural position, sexuality, and sexism such as ritualized subordination or licensed withdrawal continue to portray women as inferior to men. However, many researchers have ventured out of the realm of print media to analyze the gendered messages inherent in television advertisements. Ganahl, Prinsen, & Netzley (2003) analyzed 1,337 commercials across 63 h ours of prime time programming on ABC, CBS, and NBC airing in February of 1998. They coded for the sex of characters portrayed, whether these characters assumed a primary or secondary role, the characters' age, and the product being sold. The researchers found that females were underrepresented as primary characters, and even when secondary characters were included in the analyses, females were still underrepresented compared to the population. In the children and adolescent age group of 6 to 20 year ol ds, males represented over half of the primary character roles while the opposite was true in secondary roles. The authors noted that this disparity mirrors traditional gender stereotypes, which depict

PAGE 24

16 males as exuding power, status, and credibility over female subordinates. In the young adult age group of 21 to 50 year olds, men and women were equally portrayed as primary and secondary characters, even in stereotypically feminine product categories. In the adult age group of 51 years old and above, the majority of characters were male even though females outnumber males in this age group's population. Ganahl et al. stated that this disparity sends the message that men gain power and stature with age, while women lose it. When analyzing commercials by product type, the authors found that although food product commercials are a traditionally feminine category and marketing studies indicate that women make the majority of food purchase decisions males made up the majority of both primary and secondary cha racters in these ads. The authors noted that their findings "indicate that television commercials may be more irrelevant to female consumers today than they were 20 years ago because their characters are so out of date" (p. 550). Yet, this representation of females as simplistic and conventional is not limited to advertising aimed at adult consumers. Kahlenberg and Hein (2010) came to similar conclusions in a study analyzing 455 toy commercials aired on the children's network Nickelodeon in October of 20 04. They noted that the network has been praised for eschewing stereotypes in favor of strong female leads and an emphasis on multiculturalism, but the toy commercials aired on the channel send a more stereotypical message. The researchers analyzed the c ommercials for the type of toy being sold, the gender orientation of the toy (based on who was shown playing with it), the number of identifiable males and females under the age of 18, the age of the characters in the advertisement, the dominant type of in teraction (cooperative, competitive, parallel, or

PAGE 25

17 independent), the setting of the ad, and the color scheme of both the commercial and the toy. They found that most of the commercials showed only one gender and marketed the majority of the toys as gender specific. These fell in line with normative gender roles, with ads only showing girls mostly selling dolls and animal toys and ads only showing boys mostly selling transportation/construction toys, action figures, and sports toys. Toys that emphasized ma ke believe play were heavily gender segregated, with toys marketed towards boys encouraging them to compete for and maintain status through body movements and toys marketed towards girls providing lessons in domesticity, childcare, and grooming. Moreover, the make believe toys with a female target utilized sedentary positions, cooperation, and social closeness. Kahlenberg and Hein (2010) also found that ads only featuring girls mostly displayed cooperative play and never displayed competitive play, while ads featuring boys exhibited a wider range of interactions. Additionally, girls were more likely to be portrayed indoors, and in the few instances where they were shown outdoors the setting merely served as a backdrop for traditionally female activities t hat centered on family, friendship, and romance. Boys were shown playing in a diverse range of locations, which the authors state sends the message that males have more opportunities than females. The researchers also coded for pretend settings, which in cluded fantasy, imagined, or dreamt locations, and these settings did not include any girls. Commercials that featured boys only or mixed gender groups displayed a wide range of colors, but advertisements with a pastel scheme and pastel colored toys were only marketed towards an all female audience. Kahlenberg and Hein concluded that female children in toy advertisements are portrayed in more limited settings, interactions, and color palates than

PAGE 26

18 males. However, advertisements' endorsement of gender stere otypes is not limited to the United States. Researchers have discovered that stereotypical representations of men and women in advertising are pervasive internationally. Furnham and Paltzer (2010) compared data in a meta analysis of 30 studies on sex rol e stereotypes in television advertising published between 2000 and 2008. Their data spanned 11 countries in North America, Europe, Oceania, Asia, and Africa, and they measured gender themes in the ads' mode of presentation, end comments, credibility, role age, argument, reward type, product type, and background. They found that, across all cultures, there was a higher tendency for men to be placed in positions of authority through the use of voice overs, autonomous role designation (eg. interviewers, nar rators, or professionals), middle aged actors, and factual arguments. Females, on the other hand, were far more likely to be subordinated through the use of portrayal as a product user rather than an authority, dependent/familial role designation, younger actresses, and emphasis on a social/self enhancement reward. Additionally, women were more likely to be used in advertisements for body, home, and food products while men were more likely to be used in advertisements for car, sport, and alcohol products. Female actresses were more often seen in residential or home settings with children, while male actors were given more variability by being portrayed in leisure, occupational, and outdoor locations. Nam, Lee, and Hwang (2011) conducted a similar study measuring gendered portrayals in Korean fashion magazines targeting female adolescents. They discovered that females in these magazines were portrayed as smaller, utilized more feminine touch, caressed other people or objects more often, were ritually su bordinated more often, were

PAGE 27

19 more likely to be shown with childlike, cute, or pouting expressions, tilted their heads and bodies more often, bashfully bent their knees more often, exhibited more licensed withdrawal, were more likely to pose with an open mou th, withdrew their gaze more, were more likely to be withdrawn from the scene, were portrayed as less independent and self assured, were less likely to have a straight and confident gaze, and occupied less space in their postures than men. In short, women were portrayed as more submissive and withdrawn than men were. The male models in the advertisements were afforded more variability in their presentation and were less stereotypically portrayed than the female models. Although the aforementioned studie s are vastly different in their locations, mediums, target consumer audiences, and products, the body of research taken as a whole demonstrates that gender stereotypes are pervasive in advertising imagery. Females are especially limited in their represent ation, mostly being confined to domestic spaces and activities (Furnham & Paltzer, 2010; Kahlenberg & Hein, 2010) and subordinated to males in a number of ways (Furnham & Paltzer, 2010; Ganahl, et al., 2003; Mager & Helgeson, 2011; Nam, et al., 2011). Add itionally, women in advertisements are often sexualized to cater to the male gaze (Mager & Helgeson, 2011; Stankiewicz & Rosselli, 2008). Sexuality in Advertising One of the most fundamental ways that gender roles are portrayed in advertisements is throug h the sexualization of models. As previously discussed, Mager and Helgeson (2011) found that female models in ads were often displayed in sexually suggestive poses and were represented through the use of segmented body parts. In a

PAGE 28

20 study by Stankiewicz an d Rosselli (2008), the presentation of women in print advertisements was examined to determine the extent to which women were portrayed as sexual objects, victims, and aggressors and whether this varied by magazine type. The researchers found that half of all advertisements featuring women sexualized them, and this number increased to 63% in women's fashion magazines and 76% in men's magazines. Three quarters of advertisements that victimized female models sexualized them as well. This result coincides w ith Nam et al.'s (2011) finding that females in ads were often portrayed in childlike and vulnerable poses. Stankiewicz et al. asserted that this conflation of sexuality and victimization may create an association between the two that normalizes violence against women in the minds of consumers. Very few female models were portrayed as aggressors, but three quarters of these models were also sexualized. The researchers explained that the sexualization of women in advertising sends the message that women's bodies are constantly on display to be judged and a woman's value lies primarily in her appearance. Additionally, it implies that women have no sexual desire of their own, only the desire projected upon them by men. Consequently, the images in adverti sements adhere strongly to gendered stereotypes about how men and women act and look. American feminist writer, Vivian Gornick notes that: Advertisements depict for us not necessarily how we actually behave as men and women, but how we think men and wome n behave. This depiction serves the social purpose of convincing us that this is how men and women are or want to be, or should be, not only in relation to themselves but in relation to each other (Goffman, 1979, p. vii).

PAGE 29

21 This quote establishes a disconn ect between the stereotypes conveyed in advertisements and the true nature of the variability of men and women's behavior in society. This distorted representation of men and women can lead to unrealistically conventional expectations for members of eithe r sex (Lafky, et al., 1996). How Gend ered Portrayals in Advertising A ffects Adolescents' Evaluation of Gender As evidenced by the research previously discussed, advertising often serves to perpetuate existing gendered stereotypes, especially those concern ing women. Lafky, Duffy, Steinmaus, & Berkowitz (1996) exposed 125 high school students to advertisements picturing women stereotypically and non stereotypically, and then asked students to make inferences about a woman in a photo without any lifestyle id entifiers. They tested five classrooms, three of which viewed 10 advertisements with women in stereotypical roles and two of which viewed 10 advertisements picturing women in non traditional roles. In the stereotypical group, the ads displayed women comp leting tasks that place them in the private sphere of home and family, such as cleaning, preparing food, caring for children, and teaching. In the counter stereotypical group, the ads displayed women in the same roles as men acting with equal competence. These included women acting as doctors, engineers, businesswomen, and outdoor enthusiasts. Then, the researchers showed participants a neutral ad with a photo of a woman in a button down shirt and jeans not engaged in any obvious task. They were given a 12 item list of accomplishments or behaviors and asked to denote how likely they thought it was that the woman in the photograph engaged in these behaviors using a 5 point Likert scale. Six of them reflected traditional beliefs and six reflected non trad itional beliefs.

PAGE 30

22 Lafky et al. (1996) found that students exposed to advertisements containing stereotypical images of women were more likely than those exposed to ads containing counter stereotypical images of women to agree that the woman: "always gets p ermission from her husband before getting involved in volunteer work," "prefers to let someone else volunteer to be the chairperson of a committee," and "has two children and is a housewife" (p. 384). Conversely, students exposed to non stereotypical ads were more likely to agree that she "is a criminal lawyer" (p. 385) than those in the other condition. There was also an effect of sex such that female students were more likely to agree that she "spends a part of every afternoon watching soap operas on te levision" (p. 384) than their male counterparts. Additionally, females disagreed that the woman "organized and administered a recycling campaign in her community of approximately 30,000 people" and "often helps her son with his trigonometry homework" (p. 385), while males were neutral on these items. The researchers also found an interaction between sex and condition in which males who were exposed to stereotypical representations of women were more likely to believe that she "performs most of the househol d chores such as cooking all the meals and cleaning for her family" (p. 384) than males in the counter stereotypical exposure condition. These findings indicate that adolescent exposure to stereotypical representations of women promote more stereotypical expectations for women, at least short term. Morrison and Shaffer (2003) developed a study in order to better determine how nontraditional presentation of men and women in advertising impact traditionally gender oriented and non traditionally gender orie ntated individuals. In their first experiment, 246 18 to 23 year old undergraduate students were randomly assigned to either observe

PAGE 31

23 traditional advertisements in which females endorsed stereotypically feminine products and males endorsed stereotypically masculine products, or nontraditional advertisements in which females endorsed stereotypically masculine products and males endorsed stereotypically feminine products along with control advertisements in which the products were gender neutral. Each parti cipant also completed the BEM Sex Role Inventory, a measurement of their sex role orientation, and was consequently categorized as either traditional (feminine females or masculine males) or non traditional (feminine males, masculine females, or androgynou s/undifferentiated individuals). Participants rated each advertisement's effectiveness through measures of an overall advertisement preference, product attitude scale, and purchase intentions. The experimenters found that the traditional participants pre ferred the traditional advertisements over the non traditional ones and found them to be more effective, while the non traditional participants reported a preference for the non traditional advertisements. However, the researchers felt unsatisfied with th eir results and chose to conduct a second experiment which included a self referencing variable to determine for certain whether the participants' preferences truly reflected an identification with the advertisement's spokesperson. In the second experim ent, half of the newly recruited pool of 242 participants were assigned to the self referencing condition and the other half were given no instructions relating to self referencing. Those in the self referencing condition were instructed to imagine themse lves in the situation portrayed in the advertisements and think back to a time in which they used or could have used each product. Instead of using assignment to traditional or non traditional condition, participants were merely

PAGE 32

24 assigned to an advertiseme nt utilizing a spokesperson of their own sex. The traditional advertisements were done away with and each participant was shown three non traditional advertisements and three control advertisements, which promoted a gender neutral product. Participants i n the self referencing condition rated non traditional advertisements as more effective than those in the condition without self referencing instruction. Additionally, traditionally oriented participants without self referencing instruction rated non trad itional advertisements as less favorable than those in the self referencing condition. Non traditionally oriented participants exhibited no self referencing effect. Although these results illuminate some interesting connections between advertising relata bility and gender role orientation, the elimination of traditionally oriented advertisements in the second experiment provide no means of comparison for self referencing within gender orientation groups. Unfortunately, the advertising industry does not pr ovide many non traditional representations of gender so researchers like Rouner and her colleagues (2003) chose to replicate the proliferation of stereotypical gender themes in their studies to determine their impact. In order to determine the thought pro cess of adolescents while observing sexist or sexual representations in advertising, Rouner, Slater, and Domenech Rodriguez (2003) showed six advertisements to a group of 385 high school students and asked the adolescents to respond to each ad. Prior to t he experiment, they conducted a content analysis of 72 television advertisements and chose six ads that were representative of the general level of sexism and sexual themes encountered in modern American advertising. Adolescents were incredibly perceptive of unnecessary nudity, unrealistic gender representations, and the use of females as background images or sexual objects.

PAGE 33

25 However, one quarter of the participants still produced at least one comment in their responses about females in a derogatory and se xist manner. Female participants provided more commentary about sexist and sexual content in beer advertising, in addition to more commentary in general, than male participants did. Rouner et al. (2003) concluded that advertisements communicate norms and expectations about sex for adolescents that are unrealistic. As a result of self socialization and gender schema processes, which assert that females need to emphasize their image, female adolescents were more attentive to messages about gender norms whi le males focused more on attractiveness and sexual imagery. Although this research indicates that adolescents reproduce some of the gendered messages communicated by advertisements, it also reveals that adolescents have begun critiquing these messages as well. Even though stereotypical representations of women in advertising promote more stereotypical expectations for females, the research has demonstrated mediating effects for a preference or assumption of conventional gender norms (Lafky et al., 1996; M orrison & Shaffer, 2003). In Lafky's study, exposure to counter stereotypical representations of women promoted counter stereotypical expectations for an unrelated female. In Morrison and Shaffer's study, a self referencing frame promoted higher preferen ce for non traditional representations of gender in advertisements. In the following section, Martin and Kennedy (1993) determine whether this effect persists in reference to physical attractiveness. Preoccupation with Physical Attractiveness and the Body Advertisements often place an emphasis on female beauty and promote unrealistic standards of physical attractiveness in order to further their own goal of selling products.

PAGE 34

26 The impact on young women and girls is quite detrimental, and Martin and Kennedy (1993) conducted an experiment in order to determine just how damaging it is. The groups of 4 th 8 th and 12 th grade girls who participated in the experiment were assigned to either the highly attractive condition where they viewed six advertisements fea turing highly attractive models, the moderately attractive condition where they viewed the same six advertisements with the model substituted for a moderately attractive one, or the control condition where they viewed the advertisements with non human stim uli. Participants completed a questionnaire that contained the self esteem and self perception of physical attractiveness subscales from The Self Perception Profile for Children and The Self Perception Profile for Adolescents, along with a measure of soci al comparison to advertising models created by the researchers. Self perceptions of physical attractiveness were negatively correlated with social comparison behavior and positively correlated with self esteem. The 8 th and 12 th grade girls had a higher tendency to compare themselves to models than the 4 th graders and, when asked to rate the attractiveness of a moderately attractive sorority girl, those in the high attractiveness conditions for the two older grades rated her as far less attractive than t hose in the moderately attractive conditions. They also found that self perceptions of physical attractiveness decreased with age. Martin and Kennedy (1993) concluded that the schema for physical attractiveness in older girls is more developed than in yo ung girls. Preadolescent girls' social comparison behavior increases with age, and this tendency is greater in girls with low self esteem and low self perceptions of physical attractiveness. The findings above about social comparison behavior are intere sting, but they

PAGE 35

27 only illuminate details about its interaction with other variables in females. Luther (2009) conducted a study measuring the relationship between public self consciousness, self esteem, appearance satisfaction, and ad inspired social compa rison behavior in Japanese adolescent females and males. The researcher distributed a survey to 334 15 to 17 year old adolescents at a public high school in Japan tha t measured for perceptions of the importance of physical attributes for a successful car eer or to be an ideal spouse, acceptance of various artificial means to enhance appearance, ad inspired social comparison behavior, public self consciousness, self esteem, and appearance satisfaction. Nentl, Wilson, and Faber's ad inspired social comparis on behavior scale, the Fenigstein Public Self Consciousness Scale, and Rosenberg's Self Esteem Inventory Scale were all adapted for the study; the remaining scales were developed by the experimenter. Luther found that half of males and 67% of females repo rted moderate to high levels of advertisement inspired social comparison behavior. Both male and female participants reported that good looks and a slender body were important for people of either gender to possess in order to have successful careers and be good spouses. The strongest predictor of increased ad inspired social comparison behavior in all participants was a high level of public self consciousness. Social comparison behavior was negatively associated with appearance satisfaction in females, but positively associated in males. Additionally, there was a positive correlation between social comparison behavior and acceptance of diet pills and cosmetic surgery for girls and acceptance of muscle inducing drugs, diet assisting drugs, and cosmetic s urgery for boys. The author believed that adolescents', especially female adolescents', conflation of physical attractiveness with success may be related to the media's emphasis on physical attributes. This preoccupation with physical

PAGE 36

28 appearance has come up again and again in the research (Furnham & Paltzer, 2010; Kahlenberg & Hein, 2010; Lamb & Urberg, 1978; Luther, 2009; Mager & Helgeson, 2011; Martin & Kennedy, 1993; Stankiewicz & Rosselli, 2008; Tsai & Chang, 2007; Xie, et al., 2006), but how much do adolescents criticize this information in advertisements? Hobbs, Broder, Pope, and Rowe (2006) conducted a study directly measuring the level of criticism adolescent and preadolescent girls produced in response to print and television weight loss advertis ing. Three quarters of participants recognized customer testimonials and before and after photos as persuasive techniques employed by the advertisers. The girls often questioned the authenticity of the before and after photos presented in the advertiseme nts. Over half of the participants discredited claims of rapid weight loss through the use of anecdotal evidence, and half discredited claims of permanent weight loss without diet and exercise by citing nutritional information. However, few participants recognized doctors' endorsements, claims of scientific proof, and claims that a product was 'safe' and 'natural' as persuasive techniques. Additionally, only slightly over one fourth of participants identified the target audience of the commercials and ev en fewer recognized that certain visual techniques were used to establish credibility. Although 90% of participants noted that they identified with characters portrayed in the advertisements, very few were cognizant of the use of emotion to sell products. None of the participants demonstrated a satisfactory recognition of omissions advertisers left out, such as health concerns, product risks, and unsubstantiated implications. The authors concluded that a large number of girls demonstrate some ability to respond critically to weight loss advertising, but more needs to be done in the way of media literacy.

PAGE 37

29 One way of enhancing media literacy and critical observation of advertising is by presenting participants with an instructional frame that alerts them t o certain themes prior to viewing advertisements. Loken and Peck (2005) wanted to determine how an instructional frame that activates and supports non traditional female stereotypes about body size would affect how adolescent females rated attractiveness of larger sized models, attractiveness of thinner sized models, self attractiveness, self esteem, and self referencing in regards to advertisements with either larger or thinner sized models. Eighty two 14 to 18 year old female high school students were shown four print advertisements, two of which featured a thin model and two of which featured a larger sized model. Participants in the traditional condition were told beforehand that they would be shown advertisements from popular women's magazines while those in the non traditional condition were told that they would be shown advertisements from a new young women's magazine that uses advertisements featuring heavier women than the traditional model in an effort to be more reflective of average sized wome n. Measures of model attractiveness, self attractiveness, self referencing, and self esteem were administered to participants. The self referencing variable wa s broken down into general self referencing ("I could relate myself to the ad"), positive self referencing ("This ad made me think positive thoughts about myself"), and negative self referencing ("This ad made me think negative thoughts about myself"). Adolescents in both conditions rated the thin models equally in attractiveness, however those in the non traditional condition rated the larger models as more attractive than those in the traditional condition did. Girls in the non traditional condition also reported higher ratings of self attractiveness and self esteem, and rated the larger models a s more relatable than those in the traditional

PAGE 38

30 condition. When the researchers controlled for the effects of positive and negative self referencing, the relationship between frame and self esteem was no longer significant, indicating that the discrepancy in self esteem levels was indeed related to a reaction to the stimuli rather than differing levels prior to assignment to condition. Participants generated more negative thoughts when exposed to larger models in the traditional condition than in the non t raditional one. The results of this study provide information about the positive effects that advertisers can have on adolescent girls' self perception. The authors establish that their findings "are consistent with the theory that a nontraditional instr uctional frame that breaks with normative expectations invites more elaborative thinking with regard to self (self referencing)" (p. 863). This study illuminated how adolescents process the implicit normative information passed along in advertisements and how an experimenter can manipulate the way in which this information is processed. Summary Researchers have established that the visual information in advertisements is important to adolescents (Covell, 1992; Covell et al., 1994; Edens & McCormick, 2000; Lariscy et al., 2010). They have also determined that, across a wide variety of advertisement categories, the visual information in ads conveys gender stereotypes in regards to sex role (Furnham & Paltzer, 2010; Ganahl et al., 2002; Kahlenberg & Hein, 20 10; Mager & Helgeson, 2011), sexualization of women (Mager & Helgeson, 2011; Stankiewicz & Rosselli, 2008), and body language (Kahlenberg & Hein, 2010; Mager & Helgeson, 2011; Nam et al., 2011). Exposure to these stereotypes leads to stereotypical assumpt ions and expectations, at least short term and in reference to females (Lafky et

PAGE 39

31 al., 1996). However, exposure to images of thinner and larger sized advertising models that were accompanied by an inclusive instructional frame led to higher model attractiv eness and self attractiveness ratings in adolescents than occurred without an inclusive instructional frame (Loken & Peck, 2005). This effect seems to be the result of the widely recognized phenomenon of self comparison to advertising models in adolescent s (Luther, 2009; Martin & Kennedy, 1993; Mor ris on & Shaffer, 2003; Tsai & Chang, 2007). However, there is still a gap in the research when it comes to measuring these issues as they relate to the visual gendered identifiers of hairstyles and fashion. The Current Study The current study utilized the lessons learned from Loken and Peck's (2005) experiment and applied them to gendered presentation in advertising, rather than body size. None of the studies included in the literature review above measure the impact of readily observable gender identifiers, such as hairstyles and fashion, on adolescent evaluations of advertising models. Instead, they focus on gender stereotypes as they are communicated through sex roles (Furnham & Paltzer, 2010; Ganahl, et al. 2002; Kahlenberg & Hein, 2010; Mager & Helgeson, 2011), sexualization of women (Mager & Helgeson, 2011; Stankiewicz & Rosselli, 2008), and body language (Kahlenberg & Hein, 2010; Mager & Helgeson, 2011; Nam, et al., 2011). As a result, this study attemp ted to fill that gap in the literature. Loken and Peck's (2005) finding that an inclusive instructional framework impacts participants' self attractiveness, self esteem, and model attractiveness ratings is taken into account and this study utilized the s ame frames as they relate to gender presentation. Additionally a pre and post test for self attractiveness was included in the

PAGE 40

32 methodology because, although previous researchers have measured this variable, none have established a direction of causality (Loken & Peck, 2005; Luther, 2009; Martin & Kennedy, 1993). Research has established that the attractiveness of advertising models can influence a multitude of variables (Martin & Kennedy, 1993; Tsai & Chang, 2007), and participant ratings of model attra ctiveness can be influenced by instructional frame (Loken & Peck, 2005), so this measure was included in the current study. Morrisson and Shaffer (2003) found that participant ratings of advertising effectiveness were impacted by the sex role orientation of participants and advertising models, so this study included a measure of ad effectiveness to determine if the trend perseveres in reference to gender presentation. Moreover, research has shown that adolescents have particularly high ratings of ad insp ired social comparison behavior (Luther, 2009), and this tendency to compare oneself to advertising models has been cited as a possible influence on ratings of ad effectiveness (Luther, 2009; Morrison & Shaffer, 2003; Tsai & Chang, 2007) and a correlate to ratings of self attractiveness (Luther, 2009; Martin & Kennedy, 1993). As a result, the current study included an experimental measure of ad relatability for each advertising image and a global measure of media inspired social comparison behavior. Morri son and Shaffer found that participants with a traditional sex role orientation preferred advertisements depicting traditional sex roles, while participants with a non traditional sex role orientation preferred advertisements depicting non traditional sex roles. The current study included Bem's (1974) measure of sex role orientation to determine if this pattern persists in reference to gender presentation. Lastly, this study takes into account the focus on female adolescents in the literature (Hobbs, et a l., 2006; Loken & Peck, 2005; Martin & Kennedy, 1993) and

PAGE 41

33 includes both male and female participants in order to get a more holistic view of the impact of gender presentation of advertising models on adolescents. This study's aim was to determine whether a n instructional frame that promotes the fluidity of gender expression would influence participants' ratings of model attractiveness, ad relatability, perceived ad effectiveness, and self attractiveness for both gender congruent and gender incongruent prese ntation in advertising. Gender congruent and incongruent presentation was presented through the use of stereotypically masculine and feminine hairstyles and necklaces. These hairstyles and necklaces were selected in accordance with the results of a pilot study on 11 male and female undergraduate students conducted by the experimenter. Participants were shown photos of 11 hairstyles taken from a hairstyle image database and four necklaces taken by the experimenter and asked to rate them on a 5 point Liker t scale from very feminine' to very masculine'. The most feminine and most masculine hairstyle and necklace were selected for use in this study. The hairstyles selected were a bob with side swept bangs and flipped out ends 1 for the feminine condition, and a slicked back style 2 for the masculine condition. The necklace selected for the feminine condition contained a filigree heart charm 3 and the necklace selected for the masculine condition contained a military dog tag 4 Both charms were strung on iden tical chains. For the main experiment, participants were presented with either neutral instructions that explained the experiment plainly or inclusive instructions that emphasized the acceptance of gender fluidity. Then, they were presented with 1 M = 1.455, SD = 0.688 2 M = 4.545, SD = 0.522 3 M = 1.636, SD = 0.703 4 M = 3.818, SD = 0.874

PAGE 42

34 advertisi ng images featuring gender congruent and incongruent models of both sexes along with one control image that only featured the product, with five images in total. For each advertisement, participants rated the model's attractiveness, how relatable they fou nd the advertisement, and whether they would buy this product. After all stimuli were presented, participants completed a questionnaire measuring their media inspired social comparison behavior, along with a sex role inventory and demographic information. Participants also completed a measure of self attractiveness before the instructional frame was introduced and after images were viewed and evaluated. Hypotheses It was predicted that participants in the experimental condition (those given an inclusiv e instructional frame) would rate gender incongruent models as more attractive (H1) and the ads picturing these models as more effective (H2) and more relatable (H3) than congruent models, with participants in the control condition exhibiting the opposite effect. Martin and Kennedy's (1993) study found that non traditional participants with respect to gender preferred non traditional representations of models in advertising over traditional ones. The opposite was true for traditional participants. As a result, it was hypothesized that participants whose sex role orientation was incongruent with their sex would rate gender incongruent models as more attractive (H4) than congruent models and ads featuring gender incongruent models as more effective (H5) a nd relatable (H6) than those featuring congruent models, with congruent participants exhibiting the opposite effect. In line with Loken and Peck's (2005) findings that adolescent females exposed to

PAGE 43

35 an inclusive instructional frame with respect to body siz e rated themselves as more attractive than those exposed to a neutral frame, it was predicted that participants in the experimental condition would have a greater increase in ratings of self attractiveness than those in the control (H7). It was also predi cted that female participants would have lower self attractiveness ratings (H8) and higher social comparison behavior scores (H9) than male participants. Finally, it was predicted that higher social comparison behavior scores would coincide with lower rat ings of self attractiveness (H10). Method Participants Forty nine 18 to 20 year old undergraduate students at a small undergraduate liberal arts college in Southwest Florida were recruited for this study. One participant's data were thrown out because h er survey was incomplete. The final participant pool of 48 students included 11 males and 37 females. Materials Stimuli Images. For the purpose of this study, five stimuli images were created by the researcher. A male and female model were photographe d by the researcher in both masculine and feminine attire. In order to control for variables besides gender presentation, the researcher chose models who fit the normative standard for models in the American fashion industry: both were Caucasian, thin, ta ll, tan, able bodied, conventionally attractive, and had chin length brown hair. Additionally, the models were styled in an identical manner: both wore a white t shirt and snug fitting denim jeans without shoes. In all stimuli images, the models posed wi th their legs spread shoulder

PAGE 44

36 width apart, hands on their hips, facing the camera against the same outdoor background. Gender presentation was exhibited by using two identifiers: hairstyle and jewelry. For the masculine images, the models' hair was slick ed back and they wore a dog tag necklace. For the feminine images, the models' hair was swooped to the side and flipped out on the ends, and they wore a filigree heart necklace. For the control image, a photo of the jeans was interposed on top of the backg round imagery used in the other four images. Each photo was edited to deemphasize the background imagery, switch the photo from color to black and white, frame the models identically, and include a brand name for a fictional denim company invented by the experimente r called "Cool Blues." See Appendix A for the five images. Measures Experimental Measures. The experimental measures were developed by the researcher for use in this study in order to determine participants' reactions to the five advertising i mages. Each image, except for the control image, was followed by three items measured on a five point Likert scale. The first item, which was not included for the control image, asked "How attractive do you think this model is?" and was rated from 1 (ver y attractive) to 5 (very unattractive). The second item asked respondents "How much do you relate to this advertisement?" and was rated from 1 (not at all) to 5 (very much). The third item asked participants "How effective do you find this advertisement? and was rated from 1 (not effective at all) to 5 (very effective). After the items for each image, participants were asked to circle the numbers of the images that they thought were the best advertisements. This item was followed by an open response qu estion asking respondents to explain why they chose the images. See Appendix B for

PAGE 45

37 the experimental measures. Instructional Frames. Participants were given one of two sets of instructional frames: a neutral frame or an inclusive frame. The neutral frame acted as the control condition, and this frame made no mention of gender. It stated: In this study, I am interested in learning your opinions about advertising. In particular, I will be showing you some ads from popular teen and young adult magazines an d would like you to answer some questions about each one. Since the models that appear in these ads may also be important for understanding an ad's attractiveness, I will also be asking you about the models in the ads and your reactions to them. The inclu sive frame acted as the experimental condition, and this frame emphasized fluidity of gender presentation in the images. It stated: In this study, I am interested in learning your opinions about advertising. In particular, I will be showing you some ads from a NEW MAGAZINE for teens and young adults that emphasizes gender non conformity and uses models that reflect the range of gender identities present in the population. Since the models that appear in these ads may also be important for understanding a n ad's attractiveness, I will also be asking you about the models in the ads and your reactions to them. Both frames were adapted from the frames used by Loken and Peck (2005). For the purpose of this study, the experimental frame was modified from the or iginal version (which discussed body size) to mention gender non conformity and a range of gender identities. Additionally, the researcher changed the word we' to I' and modified the

PAGE 46

38 verbs thereafter to reflect this change in pronouns. Bem Sex Role Inve ntory. The Bem Sex Role Inventory (Bem, 1974) is a measure of participants' embodiment of traditionally sex typed attributes. It classifies participants as either feminine sex typed, masculine sex typed, androgynous, or undifferentiated. The measure was developed by Sandra Bem in 1974 to account for psychological androgyny (or, embodiment of both feminine and masculine qualities). Measures of sex typed behavior and characteristics prior to Bem's scale placed femininity and masculinity at opposite ends of one dimension, creating an inverse relationship between them. The Bem SRI measures feminine sex typed and masculine sex typed attributes on two separate dimensions, thus allowing for participants to score high or low on both subscales to account for a mu ltidimensional representation of sex typing. Bem developed an original version and a short version of the Bem SRI; for the purpose of this study, the original version was used as it has been deemed a more accurate measure of sex typing. The original scal e is comprised of 60 personality characteristics: 20 feminine sex typed characteristics (e.g. affectionate, sensitive to the needs of others), 20 masculine sex typed characteristics (e.g. ambitious, assertive), and 20 neutral characteristics that serve as filler items (e.g. truthful). Participants indicate how well the characteristics describe themselves on a 7 point Likert scale from 1 (never or almost never true) to 7 (always or almost always true). Participants were classified on the basis of a median split into one of the four sex typed categories. Based on normative data on Stanford students of both sexes, the median raw score for the femininity dimension is 4.90 and the median raw score for the masculinity dimension is 4.95 (this is the mean score on the 20 items for that dimension). If participants score higher than the

PAGE 47

39 median, they are classified as high in that dimension, and if they score lower than the median, they are classified as low in that dimension. A high score on both dimensions yield s classification as androgynous, a low score on both dimensions yields an undifferentiated classification, and a high score on one dimension and low on the other yields a feminine or masculine classification. Self Attractiveness Scale. The self attractive ness scale (SAS) was developed by Loken and Peck (2005) to measure participants' ratings of self attractiveness using a multiple item scale. The scale is comprised of five items: "I feel I have a number of good physical features," "I have a positive attit ude toward my body," "I think that I am pretty attractive," "On the whole I am satisfied with the way I look," and "At times, I think I am not attractive at all" (reverse scored). Participants rate these items on a 7 point Likert scale from 1 (strongly di sagree) to 7 (strongly agree). Ratings on the items are averaged for a measure of self attractiveness ranging from a minimum possible score of 1 to a maximum score of 7 ( =.86). See Appendix C for this measure. Media Inspired Social Comparison Behavior S cale. The media inspired social comparison behavior scale used in this study was originally developed by Nentl (1998). The scale consisted of seven items measured on a 5 point Likert scale from 1 (strongly disagree) to 5 (strongly agree). In Nentl's ver sion, 4 was labeled as "strongly agree" and 5 was labeled as "undecided" which was given a value of zero when scoring. In this study, the experimenter moved the undecided label to 3 and the strongly agree label to 5 in order to more closely reflect the la bels on other scales used in the survey. This labeling method was the same adaptation used by Luther (2009). Responses on the seven items were summed for a measure of media inspired social comparison behavior ranging

PAGE 48

40 from a minimum possible score of 7 to a maximum score of 35 ( =.86). See Appendix D for this measure. Demographic Questionnaire. The demographic questionnaire was created by the experimenter for use in this study and it is comprised of six items. These items asked participants for their se x, race, age, and self identified gender orientation. There were two additional questions which asked participants whether or not they were aware that the Bem SRI measures one's gender orientation, and whether or not they knew one or both of the models in the stimuli images. Since the models were students at the small liberal arts college that participants attended, this was included in order to determine the effects of personal acquaintance or friendship with the models on their ratings of attractiveness effectiveness, and relatability. See Appendix E for this measure. Procedure All participants were recruited through the student email forum. Once students responded to the email, the experimenter scheduled appointments with prospective participants at their convenience. Once 3 days worth of appointments were scheduled, the researcher put the names of the female and male participants into two separate list randomizers. The first half of the names on the list was assigned to the experimental condition and the second half was assigned to the control condition. Participants were assigned to one of 10 image sequences developed using a Latin square in order of their appointment. A separate PowerPoint presentation and questionnaire packet was developed for each of these 10 sequences to account for order effects of the images. The experiment was held in a small classroom on the liberal arts college's campus where participants were tested one at a time. Each participant was given an informed

PAGE 49

41 consent form, w hich they read, signed, and returned to the experimenter. The experimenter then provided participants with the questionnaire packet. This packet contained survey items in the following order: the Self Attractiveness Scale (SAS), the instructional frame, the experimental measures, another copy of the SAS to serve as a post test, the Bem Sex Role Inventory, the Media Inspired Social Comparison Behavior survey, and the demographics questionnaire. Participants were instructed to fill out the measure on the f irst page (the SAS pre test) and read the instructions on the second page (the instructional frame), after which they notified the experimenter for further directions. She explained to participants that they would be completing a paper questionnaire while viewing a PowerP oint presentation on the laptop placed on the table in front of them and read the following script aloud: You are about to view five advertising images. Each image will remain on the screen for 7 seconds. After each image disappears, ans wer the questions that correspond to the image. Then, click through to the next slide. You may begin. Participants clicked through the images at their own pace and completed the questionnaire while the experimenter sat unobtrusively on the other side of the room. After participants completed the survey, they were debriefed, thanked for their participation, and given a $5 gift card to Target. Results Sample Overview Participants were assigned to condition by sex and appointment time. After three days worth of appointments were scheduled, the experimenter divided participants' names by sex and entered the female and male lists separately into a list randomizer. The

PAGE 50

42 first half of the names on the list was assigned to the experimental condition and the s econd half was assigned to the control. Since participants were assigned to condition before they were categorized by the Bem SRI, analyses are included below outlining their distribution according to Bem classification and condition. On the Bem SRI, part icipants' classifications were predominantly undifferentiated with the rest evenly split between female sex typed and male sex typed and one participant being classified as androgyn ous (s ee Figure 1 ). In all subsequent analyse s using Bem SRI categories, the single androgynous participant was excluded. According to participants' Bem categorization and self identified sex, slightly over half were gender congruent and slightly less than half were gender incongruent (see Figure 2 ) On the self report measur e of gender identity, one quarter of participants identified as very feminine, one third identified as feminine, one fifth identified as androgynous, one fifth identified as masculine, and only three participants identified as very masculine (see Figure 3 ) For female participants, half were assigned to the control condition ( n = 18) and half were assigned to the experimental condition ( n = 19) For males, half were assigned to the control condition ( n = 5) and half were assigned to the experimental condit ion ( n = 6). Of participants who were categorized as female sex typed on the Bem SRI, three quarters were assigned to the control condition and one quarter were assigned to the experimental condition. Of those categorized as male sex typed on the Bem SRI half were assigned to the control condition and half were assigned to the experimental condition. Of those categorized as undifferentiated on the Bem SRI, one third were assigned to the control condition and two thirds were assigned to the experimental

PAGE 51

43 condition (see Table 1) A chi square analysis was conducted on the Bem distribution of participants by condition without the outlying androgynous participant and it was found that the data approached significance, 2 (2 47) = 5.2661, p = 0.0719. Model Attractiveness, Ad Effectiveness, Ad Relatability, Model Gender Congruence, and Instructional Frame In the first hypothesis, it was predicted that participants in the experimental condition would rate gender incongruent models as more attractive than con gruent models, with participants in the control condition exhibiting the opposite effect. A 2 (condition) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of model attractiveness See Figure 4 for means and standard deviations of model attractiveness ratings by participant condition. There was no significant main effect for participant condition, F (1, 46) = 0.009, p = 0.927, no significant main effect for model gender congruence, F (1, 46 ) = 0.161, p = 0.690, and no significant interaction between the two variables, F (1, 46 ) = 0.733, p = 0.396. The first hypothesis was not supported. The second hypothesis stated that participants in the experimental condition would rate gender incongruent mo dels as more effective than congruent models, with participants in the control condition exhibiting the opposite effect. A 2 (condition) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of ad e ffectiveness See Figure 5 for mean ratings of ad effectiveness by participant condition. There was a significant main effect of participant condition, F (1, 46) = 5.642, p = 0.022, such that there were significantly higher ratings of ad effectiveness fro m participants in the experimental condition ( M = 2.680, SD =

PAGE 52

44 0.171) than participants in the control condition ( M = 2.120, SD = 0.179). There was no significant main effect for model gender congruence, F (1, 46 ) = 0.010, p = 0.921, and no significant int eraction between the two variables, F (1, 46) = 0.969, p = 0.330. The second hypothesis was not supported, although participants in the experimental condition gave all four images picturing models higher ratings of ad effectiveness than participants in th e control group. The third hypothesis stated that participants in the experimental condition would rate gender incongruent models as more relatable than congruent models, with participants in the control condition exhibiting the opposite effect. A 2 (cond ition) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of ad relatability See Figure 6 for mean ratings of ad relatability by participant condition. There was a significant main effect of par ticipant condition, F (1, 46) = 6.920, p = 0.012, such that there were significantly higher ratings of ad relatability from participants in the experimental condition ( M = 2.470, SD = 0.145) than participants in the control condition ( M = 1.957, SD = 0.151 ). There was no significant main effect for model gender congruence, F (1, 46 ) = 0.760, p = 0 .388. There was a significant interaction between the two variables, F (1, 46) = 6.903, p = 0.012 such that participants in the control group gave lower ratings of ad relatability to incongruent models than any of the other three groups The third hypothesis was supported, participants in the experimental condition gave all four images picturing models higher ratings of ad relatability than participants in the c ontrol condition, the lowest levels of ad relatability were those of incongruent models rated by participants in the control condition, and the highest were of the incongruent models in the experimental condition.

PAGE 53

45 Model Attractiveness, Ad Effectiveness, Ad Relatability, Model Gender Congruence, and Participant Sex Although no hypotheses were developed for the influence of model gender congruence and participant sex on ratings of model attractiveness, ad effectiveness, and ad relatabili ty, the researcher ex plored these relationship s A 2 (participant sex) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of model attractiveness See Figure 7 for mean ratings of model attractiveness by participant sex. There was no significant main effect for participant sex, F (1, 46) = 0.012, p = 0.913, no significant main effect for model gender congruence, F (1, 46 ) = 0.266, p = 0.608, and no significant interaction between the two variables, F (1, 46 ) = 0.078 p = 0.781. A 2 (participant sex) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of ad effectiveness See Figure 8 for mean ratings of ad effectiveness by participant sex. There was no si gnificant main effect for participant sex, F (1, 46) = 0.000, p = 0.992, and no significant main effect for model gender congruence, F (1, 46 ) = 3.985, p = 0.052. There was a significant interaction between the two variables, F (1, 46) = 15.493, p < 0.001 such that males rated the congruent images as more effective while females rated the incongruent images as more effective. A 2 (participant sex) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on rating s of ad relatability See Figure 9 fo r mean ratings of ad relatability by participant sex. There was no significant main effect for participant sex, F (1, 46) = 1.405, p = 0.242, and no significant main

PAGE 54

46 effect for model gender congruence, F (1, 46 ) = 0.8 96, p = 0.349. There was a significant interaction between the two variables, F (1, 46) = 11.095, p = 0.002 such that males rated the congruent images as more relatable while females rated the incongruent images as more relatable Model Attractiveness, Ad Effectiveness, Ad Relatability Model Gender Congruence, and Participant Gender Congruence The fourth hypothesis stated that participants whose sex role orientation (Bem SRI) was incongruent with their sex would rate gender incongruent models as more at tractive than gender congruent ones, with gender congruent participants exhibiting the opposite effect. A 2 (participant congruence) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of model at tractiveness See Figure 10 for mean ratings of model attractiveness by participant gender congruence. There was no significant main effect for participant gender congruence, F (1, 46) = 0.035, p = 0.852, no significant main effect for model gender congr uence, F (1, 46 ) = 0.023, p = 0.880, and no significant interaction between the two variables, F (1, 46) = 3.024, p = 0.089. The fourth hypothesis was not supported. The fifth hypothesis stated that participants whose sex role orientation was incongruent with their sex would rate ads featuring gender incongruent models as more effective than those featuring gender congruent ones, with gender congruent participants exhibiting the opposite effect. A 2 (participant congruence) x 2 (model gender congruence) m ixed ANOVA with repeated measures on the second factor was conducted on ratings of ad effectiveness See Figure 11 for mean ratings of ad effectiveness by participant gender congruence. There was no significant main effect for participant

PAGE 55

47 congruence (alt hough it approached significance) F (1, 46) = 3.715, p = 0.060, no significant main effect for model gender congruence, F (1, 46 ) = 0.005, p = 0.944, and no significant interaction between the two variables, F (1, 46) = 0.162, p = 0.689. The fifth hypoth esis was not supported. The sixth hypothesis stated that participants whose sex role orientation was incongruent with their sex would rate ads featuring gender incongruent models as more relatable than those featuring gender congruent ones, with gender con gruent participants exhibiting the opposite effect. A 2 (participant congruence) x 2 (model gender congruence) mixed ANOVA with repeated measures on the second factor was conducted on ratings of ad relatability See Figure 12 for mean ratings of ad relat ability by participant gender congruence. There was a main effect of participant congruence, F (1, 46) = 06.064, p = 0.018, such that there were significantly lower ratings of ad effectiveness from congruent participants ( M = 1.938, SD = 0.164) than incon gruent participants ( M = 2.429, SD = 0.138) There was no significant main effect for model gender congruence, F (1, 46 ) = 0.435, p = 0.513, and no significant interaction between the two variables, F (1, 46) = 2.605, p = 0.113. The sixth hypothesis was not supported although gender incongruent participants gave higher ratings of ad relatability to all four images picturing models than those of gender congruent participants Self Attractiveness Scores In the seventh hypothesis, it was predicted that the self attractiveness scores of participants in the experimental condition would increase from pre test to post test more so than the scores of participants in the control condition. A 2 (participant condition) x 2 (self attractiveness test) mixed ANOVA wit h repeated measures on the second factor was

PAGE 56

48 conducted. See Figure 13 for mean ratings of self attractiveness by participant condition. There was no significant main effect for condition, F (1, 46) = 2.896, p = 0.096, no significant main effect for self attractiveness test, F (1, 46 ) = 0.037, p = 0.849, and no significant interaction between the two variables, F (1, 46) = 0.133, p = 0.717. The seventh hypothesis was not supported. Additional analyses were conducted in order to measure how participant s ex and participant gender congruence related to the difference between self attractiveness pre and post test scores. First, a 2 (participant sex) x 2 (self attractiveness test) mixed ANOVA with repeated measures on the second factor was conducted. See F igure 1 4 for m ean ratings of self attractiveness by participant sex. There was no significant main effect for participant sex, F (1, 46) = 0.185, p = 0.669, no significant main effect for self attractiveness test, F (1, 46 ) = 0.145, p = 0.705, and no sign ificant interaction between the two variables, F (1, 46) = 0.145, p = 0.705. Next, a 2 (participant gender congruence) x 2 (self attractiveness test) mixed ANOVA with repeated measures on the second factor was conducted. See Figure 15 for mean ratings o f self attractiveness by participant gender congruence. There was no significant main effect for participant congruence, F (1, 46) = 1.441, p = 0.236, no significant main effect for self attractiveness test, F (1, 46 ) = 0.006, p = 0.939, and no significan t interaction between the two variables, F (1, 46) = 0.590, p = 0.446. The eighth hypothesis stated that female participants would have lower pre test ratings of self attractiveness than male participants. An independent sample t test measuring self attr activeness pre test scores by participant sex was conducted in order to test this hypothesis, and it was found to be non significant, t (46) = 0.51, p = 0.6105.

PAGE 57

49 Female participants ( M = 4.4324, SD = 1.0451) did not score significantly differently from mal es ( M = 4.2182, SD = 1.6958) on the self attractiveness pre test. The eighth hypothesis was not supported. Media Inspired Social Comparison Behavior In the ninth hypothesis, it was predicted that female participants would have higher social comparison behavior scores than male participants. An independent sample t test measuring social comparison score by participant sex was conducted to test this hypothesis, and it was found to be significant, t (46) = 2.46, p = 0.0176. The ninth hypothesis was suppo rted: female participants had significantly higher levels of social comparison behavior ( M = 23.0270, SD = 6.3530) than male participants ( M = 17.5455, SD = 6.9190) An independent sample t test measuring social comparison score by participant gender cong ruence was also conducted, but it was found to be non significant, t (46) = 1.23, p = 0.2236. Congruent participants ( M = 23.2000, SD = 7.4382) and incongruent participants ( M = 20.7500, SD = 6.2812) had media inspired social comparison scores that were n ot significantly different. The final hypothesis stated that pre test ratings of self attractiveness would be negatively correlated with levels of social comparison behavior. A Pearson r test measuring self attractiveness pre test scores and social compa rison behavior scores was conducted to test this hypothesis, and it was found to be significant, r (46) = 0.29977, p = 0.0385. To verify that these findings were not influenced by this study's small sample size, a Spearman r test on the same data was con ducted and also found to be significant, r s (46) = 0.31975, p = 0.0267. See Figure 16 for the scatterplot of social comparison behavior and self attractiveness. The tenth hypothesis was supported, self attractiveness

PAGE 58

50 pre test scores and social compariso n scores were negatively correlated. Discussion The goal of this study was to determine the effects of model gender performance in advertisements and an inclusive instructional frame on adolescents' and young adults' ratings of model attractiveness, ad effectiveness, ad relatability, and self attractiveness. The 10 hypotheses established by the researcher in the current study will be outlined below, along with the results pertaining to each one and a discussion of the implications of these results. Inst ructional Frame In the researcher's first hypothesis, it was predicted that participants in the experimental condition would rate gender incongruent models as more attractive than congruent models, with participants in the control condition exhibiting the opposite effect. This hypothesis was rejected; there was no significant difference in model attractiveness scores between the two conditions. This is surprising because Loken and Peck (2005) found that an inclusive instructional frame caused adolescent females to rate larger sized models as more attractive than they did without the inclusive frame. The current study replaced body size with gender presentation in order to break with normative expectations of advertising models' appearance. It's possible that the effects of instructional frame on ratings of model attractiveness do not carry over to issues of gender presentation, which would explain these results. The second and third hypotheses predicted the same relationship between instructional frame a nd model gender congruence in reference to advertisement effectiveness and advertisement relatability instead of model attractive ness. The

PAGE 59

51 hypothesis predicting an interaction between condition and model gender congruence impacting ad effectiveness was no t supported but participants exposed to an inclusive instructional frame produced higher ratings of effectiveness for all four model containing images than did participants in the control condition. The hypothesis pertaining to ad relatability was fully supported in that participants exposed to an inclusive instructional frame rated gender incongruent models as more relatable than congruent models with participants in the control condition exhibiting the opposite effect. There was also a main effect of c ondition such that participants given the inclusive frame rated all four model containing images as more relatable than participants in the control group did. Although Loken and Peck's (2005) findings were not mirrored in this study in reference to model attractiveness, the same main effects of instructional frame were exhibited for effectiveness and relatability and the same interaction of normative appearance and instructional frame were exhibited for relatability. The lack of an attractiveness effect w ith regards to the current study as compared to Loken and Peck's experiment is surprising. One major methodological difference between their study and the current study is the inclusion of male participants and male models, while Loken and Peck only used females, although the researcher is unsure how this sex effect would impact the results. The lack of significance with regards to model attractiveness across variables in this study will be examined further later in the discussion. Morrison and Shaffer ( 2003) used a self referencing instructional frame in their study to measure how effective participants found non traditional gender representations in advertisements. They found that participants in the self referencing condition rated all the advertiseme nts as more effective than those in the control condition, and this effect

PAGE 60

52 was even stronger in reference to non traditional advertisements. The current study found the same impact on effectiveness ratings with an inclusive instructional frame and expande d those results to include relatability ratings. Loken and Peck's (2005) assertion that "a nontraditional instructional frame that breaks with normative expectations invite[s] more elaborative thinking with regard to self (self referencing)" (p. 863), alo ng with Morrison and Shaffer's results, can be applied to the current study to explain the effects of the inclusive instructional frame on ratings of effectiveness and relatability. If the non traditional instructional frame invited participants to self r eference, then the tendency for participants in the inclusive frame condition to relate more to incongruent models is quite logical. Additionally, Morrison and Shaffer's assertion that self referencing results in higher perceptions of advertising effectiv eness was confirmed by the main effect of instructional frame on effectiveness. Participant Sex No predictions were made in the current study regarding the relationship between participant sex and model attractiveness, advertisement effectiveness, and adv ertisement relatability. However, participant sex made an impact on some of these ratings nonetheless. Female participants gave higher ratings of both ad effectiveness and ad relatability to images of gender incongruent models than male participants did. Rouner, Slater, and Domenech Rodriguez (2003) discovered that adolescent female participants in their study provided more commentary about sexist and sexual content in beer advertisements than adolescent male participants did. The authors asserted that pervasive norms emphasizing female beauty cause females to be more attentive to messages about gender norms than males are. This heightened awareness of gender representations in

PAGE 61

53 advertisements from females might explain why women gave higher effectivenes s and relatability ratings to ads picturing gender incongruent models than males did. Participant Gender Congruence The fourth hypothesis in the current study stated that participants whose sex role orientation (Bem SRI) was incongruent with their sex wo uld rate gender incongruent models as more attractive than congruent ones, with gender congruent participants exhibiting the opposite effect. This hypothesis was rejected; there were no significant effects of participant gender congruence on model attract iveness ratings. The fifth and sixth hypotheses stated that participants whole sex role orientation was incongruent with their sex would rate ad images picturing gender incongruent models as more effective and relatable than ones picturing congruent model s, with gender congruent participants exhibiting the opposite effect. The hypothese s regarding effectiveness and relatability ratings were rejected. Although the interaction between participant gender congruence and ad relatability was not significant, t here was a main effect of congruence such that gender incongruent participants rated all four advertisements picturing models as more relatable than gender congruent participants did. Morrison and Shaffer's (2003) study may shed some light on these resul ts, as they also looked at participant gender congruence dictated by Bem classification and participant sex and its effect on ratings of advertisements. In their experiment, they found that traditional (ie. gender congruent) participants preferred traditi onal gender representations in advertisements and found them more effective, while non traditional (ie. gender incongruent) participants preferred non traditional representations in ads. Morrison and Shaffer's second experiment in the same study demonstra ted that

PAGE 62

54 participants given a self referencing instructional frame gave higher ratings of advertisement effectiveness to non traditional ads than those in the control condition did. When they examined this relationship more closely, they found that tradi tional participants in the self referencing condition rated non traditional ads as more favorable than those in the control condition, but non traditional participants had no self referencing effect. Mor ris on and Shaffer noted that non traditional partici pants exhibit no self referencing effect, "possibly because these individuals view a wider range of activities and behaviors as appropriate for each sex and are perhaps more inclined than traditional participants to spontaneously self reference when viewin g advertisements more commonly associated with the other sex" (p.272 3). As a result, gender incongruent participants may be more likely to spontaneously relate to advertisements regardless of the gender presentation of the people featured in them than ge nder congruent participants are, and this effect was observed in the current study. Self Attractiveness The seventh hypothesis in the current study predicted that participants exposed to an inclusive instructional frame would display a greater increase in self attractiveness ratings from pre test to post test than participants in the control group. This hypothesis was rejected; although participants in the experimental condition did exhibit a modest increase from pre test ( M = 4.104) to post test ( M = 4.1 60), the results were far from gaining significance. In Loken and Peck's (2005) study, adolescent girls exposed to an inclusive instructional frame were more likely to rate themselves as more attractive than those in the control group. However, their exp eriment did not include pre and post test measures so the effect they observed may have been the result of condition assignment.

PAGE 63

55 Another explanation might be that administering the same self attractiveness scale twice in the current study over a short pe riod of time caused participants to repeat the answers they gave on the pre test in the post test survey. Future replications of this study might benefit from administering the pre test in advance of the main experiment to lessen the memory effect. An al ternative option may be to utilize alternative forms of a self a ttractiveness measure using different but equivalent items. The current study also measured the relationship between self attractiveness and both participant sex and gender congruence, but the se analyses did not gain significance either. The eighth hypothesis asserted that self attractiveness pre test scores would be higher among female participants than male participants, but the difference between males' ( M = 4.218) and females' ( M = 4.432) scores did not gain significance so the hypothesis was rejected. Media Inspired Social Comparison Behavior The ninth hypothesis of the current study asserted that females would have higher levels of social comparison behavior than males. This hypothesi s was supported; females exhibited media inspired social comparison behavior levels 30% higher than those of males. These findings are consistent with those of Luther's (2009) study in which half of males and 67% of females reported moderate to high level s of advertisement inspired social comparison behavior. In the current study, participants scored in the moderate range with the mean for males located on the lower end of moderate and the female mean on the higher end. Covell et al. (1992) noted that, with the earlier physiological development in girls are found issues of developing identity: lower self esteem, more self consciousness, and most particularly concerns about popularity with the opposite sex

PAGE 64

56 (Elkind & Bowen, 1979; Enright, Lapsley, & Shukla 1979; Simmons & Rosenberg, 1975; Simmons, Brown, Bush, & Blyth, 1978)" (p. 57). The potential impact of this heightened awareness of social pressures may explain the high levels of media inspired social comparison behavior that females exhibited in this study. The relationship between participant gender congruence and media inspired social comparison behavior was also measured, although there was no hypothesized direction of the effect. Gender congruent participants ( M = 23.200) displayed higher levels of social comparison than incongruent participants ( M = 20.750), but this relationship never reached significance. The tenth and final hypothesis of the current study asserted that there would be a negative correlation between self attractiveness pre te st scores and media inspired social comparison behavior levels. This hypothesis was supported; the relationship was significant and consistent with the previous research. Martin and Kennedy's (1993) study determined that self perceptions of physical attr activeness were negatively correlated with social comparison behavior and self esteem in adolescent and pre adolescent girls. Tsai and Chang (2007) found that adolescents attributed higher purchase intention and product attitude to advertisements featurin g normally attractive models than those featuring highly attractive models. The authors assert that this relationship was due to self comparison to models in advertisements: viewing highly attractive models inflates acceptable levels of attractiveness and this unattainable beauty ideal results in a deflated self image in adolescents. Additionally, Luther (2009) found that social comparison behavior was negatively associated with appearance satisfaction in females but positively associated in males. The re sults observed in this study confirmed

PAGE 65

57 the existing data. Even Luther's effect of sex is consistent when considering the female majority of the sample in the current study. Limitations and Directions for Future Research The biggest limit to the generaliz ability of the results of this study was the sample size. The small sample size and especially the small sample of male participants call into question the validity of its results when generalizing to a larger population. Additionally, the fact that all participants were recruited from a small liberal arts college with a reputation for feminist activism may have skewed the sample's perception of gender presentation and fluidity from the start. The college has two prominent feminist activism groups, a gen der queer club, and a transgender support group, as well as a successful gender studies department which cross references classes in a multitude of disciplines. The existence of these institutions within the college educating students on the socially cons tructed nature of gender and promoting acceptance of gender incongruent (ie. transgender or gender queer) individuals may have impacted their opinions of non conforming gender expression. Future replications of the current study will be more effective if they recruit participants from a broader population, collect a larger sample, and ensure that the balance between male and female participants nears equivalence. Another limitation of the current study was the use of student models. Although both of the models used in the stimuli images had done some modeling work in the past, they were students at the same college where participants were recruited. Participants' friendship or acquaintance with the models may have impacted their perception of the images, as a sizable minority of participants ( n = 17) answered yes' to the question asking whether they knew or recognized the models. However, the researcher attempted

PAGE 66

58 to minimize this possibility of acquaintance by using fourth year students residing off cam pus as the models and first and second year students who are generally required to reside on campus as the participants. Other factors concerning the creation of the stimuli images may be at work here, as well. The existing research establishes that view ing highly attractive models inflates adolescents' guidelines for attractiveness (Martin & Kennedy, 1993; Tsai & Chang, 2007). Participants' exposure to mainstream media containing images of highly attractive individuals may have inflated their perception s of attractiveness, which would explain why none of the analyses including model attractiveness gained significance. The appearance of professional models ubiquitous in mainstream media is enhanced through the use of expert styling, professional photogra phy, and extensive photo editing. Since the researcher created all the images used in this study, the quality of the images was not equivalent to those being created by the fashion industry. The use of real advertisements in the current study would have been ideal, but representations of gender incongruent models in the advertising industry are rare and difficult to obtain. The benefit of images created specifically for the current study allows for complete control of the content, with the sole manipulat ions being those developed by the experimenter. The use of researcher created images establishes a much tighter experimental design and, as a result, future researchers would benefit from the use of professional photographers, models, and a graphic design er to create the advertisements for their experiment. Nevertheless, the results of this study contribute to the existing research on the impact of gender representations in the media on adolescents and young adults. Females' heightened awareness of sexis m in advertising (Rouner, Slater, & Domenech Rodriguez,

PAGE 67

59 2003) and tendency to exhibit higher levels of social comparison behavior (Covell, Dion, & Dion, 1992; Luther, 2009) were both confirmed in the current study. Gender incongruent participants' higher levels of relatability to models in advertisements (Morrison & Shaffer, 2003) were also consistent with previous research. The impact of non traditional instructional frames on perceptions of advertisements (Loken & Peck, 2005; Morrison & Shaffer, 2003) w as established once again in the current study. Also, the negative correlation between self attractiveness and social comparison behavior (Luther, 2009; Martin & Kennedy, 1993; Tsai & Chang, 2007) was confirmed. When taken together, these results indicat e a need to reframe discussions of gender representations in the media among adolescents to include acceptance of a wider range of expression in order to mediate some of the negative effects these stereotypical portrayals have on young consumers.

PAGE 68

60 Referen ces American Psychological Association. (2004). Report of the APA Task Force on Advertising and Children Washington, D. C.: Wilcox, B., Cantor, J., Dowrick, P., Kunkel, D., Linn, S., & Palmer, E. Bem, S. L. (1974). The measurement of psychologica l androgyny. Journal of Consulting and Clinical Psychology 42 (2), 155 62. Browne, B. A. (1998). Gender stereotypes in advertising on children's television in the 1990s: A cross national analysis. Journal of Advertising 27 (1), 83 96. Conger, J. J. (1987). Behavioral medicine and health psychology in a changing world. Child Abuse & Neglect 11 443 53. Covell, K. (1992). The appeal of image advertisements: Age, gender, and product differences. Journal of Early A dolescence 12 273 87. Cove ll, K., Dion, K L., & Dion, K. K. (1994). Gender differences in evaluations of tobacco and alcohol advertisements. Canadian Journal of Behavioral Science 26 (3), 404 20. Das, M. (2011). Gender role portrayals in Indian television ads. Sex Roles 64 208 22. Doring, N., & Poschl, S. (2006). Images of men and women in mobile phone advertisements: A content analysis of advertisements for mobile phone communication systems in selected popular magazines. Sex Roles 55 173 85. Edens, K. M., & Mc Cormick, C. B. (2000). How do adolescents process advertisements? The influence of ad characteristics, processing objective, and gender. Contemporary Educational Psychology 25 450 63. Eisend, M. (2010). A meta analysis of gender roles in adverti sing. Journal of the

PAGE 69

61 Academy of Marketing Science 38 418 40. Elkind, D. (1967). Egocentrism in adolescence. Child Development 38 (4), 1025 34. Furnham, A., & Paltzer, S. (2010). The portrayal of men and women in television advertisements: An up dated review of 30 studies published since 2000. Scandinavian Journal of Psychology 51 216 36. Ganahl, D. J., Prinsen, T. J., & Netzley, S. B. (2003). A content analysis of prime time commercials: A contextual framework of gender representation. Sex Roles 49 (9/10), 545 51. Goffman, E. (1979). Gender advertisements (1 st Harper colophone ed.). New York, NY: Harper & Row. Hobbs, R., Broder, S., Pope, H., & Rowe, J. (2006). How adolescent girls interpret weight loss advertising. Health Ed ucation Research 21 (5), 719 30. Kahlenberg, S. G., & Hein, M. M. (2010). Progression on Nickelodeon? Gender role stereotypes in toy commercials. Sex Roles 62 830 47. Kim, K., & Lowry, D. T. (2005). Television commercials as a lagging social indic ator: Gender role stereotypes in Korean television advertising. Sex Roles 53 (11/12), 901 10. Lafky, S., Duffy, M., Steinmaus, M., & Berkowitz, D. (1996). Looking through gendered lenses: Female stereotyping in advertisements and gender role exp ectations. Journalism & Mass Communication Quarterly 73 (2), 379 88. Lamb, M. E., & Urberg, K. A. (1978). The development of gender role and identity. In M. E. Lamb (Ed.), Social and Personality Development (178 99). New York, NY: Holt, Rinehart, and Winston.

PAGE 70

62 Lariscy, R. W., Paek, H. J. (2010). Examination of media channels and types as health information sources for adolescents: Comparisons for Black/White, male/female, urban/rural. Journal of Broadcasting & Electronic Media 54 (1), 102 20. Loken, B., & Peck, J. (2005). The effects of instructional frame on female adolescents' evaluations of larger sized female models in print advertising. Journal of Applied Social Psychology 35 (4), 850 68. Luther, C. A. (2009). Importance placed on physical attractiveness and advertisement inspired social comparison behavior among Japanese female and male teenagers. Journal of Communication 59 279 95. Mager, J., & Helgeson, J. G. (2011). Fifty years of advertising images: Some changing perspectives on role portrayals along with enduring consistencies. Sex Roles 64 238 52. Martin, M. C., & Kennedy, P. F. (1993). Advertising and social comparison: C onsequences for female preadolescents and adolescents. Psychology & Marketing 1 0 (6), 513 30. Morrison, M. M., & Shaffer, D. R. (2003). Gender role congruence and self referencing as determinants of advertising effectiveness. Sex Roles 49 (5/6), 265 75. Nam, K., Lee, G., & Hwang, J. S. (2011). Gender stereotypes depicted by Wes tern and Korean advertising models in Korean adolescent girls' magazines. Sex Roles 64 223 37. Nentl, N. J. (1998). Media cultivation: The impact of the media on beliefs and attitudes about beauty (Unpublished doctoral dissertation). Universi ty of Minnesota, Minneapolis, Minnesota.

PAGE 71

63 Rouner, D., Slater, M. D., & Domenech Rodriguez, M. (2003). Adolescent evaluation of gender role and sexual imagery in television advertisements. Journal of Broadcasting & Electronic Media 47 (3), 435 54. S tankiewicz, J. M., & Rosselli, F. (2008). Women as sex objects and victims in print advertisements. Sex Roles 58 579 89. Tsai, C. C., & Chang, C. H. (2007). The effect of physical attractiveness of models on advertising effectiveness for male an d female adolescents. Adolescence 42 (168), 827 36. Xie, B., Chou, C. P., Spruijt Metz, D., Reynolds, K., Clark, F., Palmer, P. H., Gallaher, P., Sun, P., Guo, Q., & Johnson, C. A. (2006). Weight perception and weight related sociocultural and beh avioral factors in Chinese adolescents. Preventive Medicine 42 229 34. Wong, K., & Chan, K. (2006). A gender portrayal of children's television commercials in mainland China. International Advertising and Communication 4 319 41.

PAGE 72

64 Table 1 Con dition Assignment by Bem SRI Classification Frequencies (%) Bem Category Experimental Control Female Sex Typed 4 10 (28.57%) (71.43%) Male Sex Typed 8 6 (57.14%) (42.86%) Undifferentiated 13 6 (68.42%) (31.58%) Androgynous 0 1 (0%) (100%)

PAGE 73

65 Figure 1 Bem SRI Category Distribution. !"# !$# !$# !# !"#$%"&$'()"$*+,"+-(./$ %&'())*+*&,(-,*'# .*/-0*#1*23456*'# 7-0*#1*23456*'# 8&'+9:5&9;<#

PAGE 74

66 Figure 2 Participant Gender Congruence Distribution. =># =?# 01.-23241+-$5"+6".$7(+8.9"+3"$ @9&:+;*&,# A&B9&:+;*&,#

PAGE 75

67 Figure 3 Self Reported Gender Identity Distribution. !=# !C# "# C# D# %"):;'"4(.-"6$5"+6".$*6"+-2-/$ E*+5#.*/(&(&*# .*/(&(&*# 8&'+9:5&9;<# 7-
PAGE 76

68 Figure 4 Model Attractiveness Ratings by Participant Condition. On a scale where 1 is "very attractive" and 5 is "very unattractive". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# H26*+(/*&,-0# @9&,+90# <(6")$=--.13-2,"+">>$ 01.-23241+-$7(+62-2(+$ <(6")$=--.13-2,"+">>$?/$01.-23241+-$ 7(+62-2(+$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 77

69 Figure 5 Advertisement Effectiveness Ratings by Participant Condition. On a scale where 1 is "not effective at all" and 5 is "very effective". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# H26*+(/*&,-0# @9&,+90# =6,".-2>"#"+-$@::"3-2,"+">>$ 01.-23241+-$7(+62-2(+$ =6,".-2>"#"+-$@::"3-2,"+">>$?/$ 01.-23241+-$7(+62-2(+$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 78

70 Figure 6 Advertisement Relatability Ratings by Participant Condition. On a scale where 1 is "not at all" relatable and 5 is "very much" relatable. ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# H26*+(/*&,-0# @9&,+90# =6,".-2>"#"+-$'")1-1?2)2-/$ 01.-23241+-$7(+62-2(+$ =6,".-2>"#"+-$'")1-1?2)2-/$?/$ 01.-23241+-$7(+62-2(+$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 79

71 Figure 7 Model Attractiveness Ratings by Participant Sex. On a scale where 1 is "very attractive" and 5 is "very unattractive". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# .*/-0*# 7-0*# <(6")$=--.13-2,"+">>$ 01.-23241+-$%"&$ <(6")$=--.13-2,"+">>$?/$01.-23241+-$%"&$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 80

72 Figure 8 Advertisement Effectiveness Ratings by Participant Sex. On a s cale where 1 is "not effective at all" and 5 is "very effective". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# .*/-0*# 7-0*# =6,".-2>"#"+-$@::"3-2,"+">>$ 01.-23241+-$%"&$ =6,".-2>"#"+-$@::"3-2,"+">>$?/$ 01.-23241+-$%"&$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 81

73 Figure 9 Advertisement Relatability Ratings by Participant Sex. On a scale where 1 is "not at all" relatable and 5 is "very much" relatable. ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# .*/-0*# 7-0*# =6,".-2>"#"+-$'")1-1?2)2-/$ 01.-23241+-$%"&$ =6,".-2>"#"+-$'")1-1?2)2-/$?/$ 01.-23241+-$%"&$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 82

74 Figure 10 Model Attractiveness Ratings by Participant Gender Congruence. On a scale where 1 is "very attractive" and 5 is "very unattractive". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# @9&:+;*&,# A&B9&:+;*&,# <(6")$=--.13-2,"+">>$ 01.-23241+-$5"+6".$7(+8.9"+3"$ <(6")$=--.13-2,"+">>$?/$01.-23241+-$ 5"+6".$7(+8.9"+3"$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 83

75 Figure 11 Advertisement Effectiveness Ratings by Participant Gender Congruence. On a scale where 1 is "not effective at all" and 5 is "very effective". ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# @9&:+;*&,# A&B9&:+;*&,# =6,".-2>"#"+-$@::"3-2,"+">>$ 01.-23241+-$5"+6".$7(+8.9"+3"$ =6,".-2>"#"+-$@::"3-2,"+">>$?/$ 01.-23241+-$5"+6".$7(+8.9"+3"$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 84

76 Figure 12 Advertisement Relatability Ratings by Participant Gender Congruence. On a scale where 1 is "not at all" relatable and 5 is "very much" relatable. ?# ?FG# !# !FG# =# =FG# D# DFG# $# $FG# G# @9&:+;*&,# A&B9&:+;*&,# =6,".-2>"#"+-$'")1-1?2)2-/$ 01.-23241+-$5"+6".$7(+8.9"+3"$ =6,".-2>"#"+-$'")1-1?2)2-/$?/$ 01.-23241+-$5"+6".$7(+8.9"+3"$ I*&'*+#@9&:+;*&,# I*&'*+#A&B9&:+;*&,#

PAGE 85

77 Figure 13 Self Attractiveness Ratings by Participant Condition. On a scale where 1 represents the minimum level of self attractiveness and 5 represents the maximum level. ?# !# =# D# $# G# J# C# H26*+(/*&,-0# @9&,+90# %"):;=--.13-2,"+">>$ 01.-23241+-$7(+62-2(+$ %"):;=--.13-2,"+">>$'1-2+8>$?/$ 01.-23241+-$7(+62-2(+$ K+*34*<,# K9<,34*<,#

PAGE 86

78 Figure 14 Self Attractiveness Ratings by Participant Sex. On a scale where 1 represents the minimum level of self attractiveness and 5 represents the maximum level. ?# !# =# D# $# G# J# C# .*/-0*# 7-0*# %"):;=--.13-2,"+">>$ 01.-23241+-$%"&$ %"):;=--.13-2,"+">>$?/$01.-23241+-$%"&$ K+*34*<,# K9<,34*<,#

PAGE 87

79 Figure 15 Self Attractiveness Ratings by Participant Gender Con gruence. On a scale where 1 represents the minimum level of self attractiveness and 5 represents the maximum level. ?# !# =# D# $# G# J# C# @9&:+;*&,# A&B9&:+;*&,# %"):;=--.13-2,"+">>$ 01.-23241+-$5"+6".$7(+8.9"+3"$ %"):;=--.13-2,"+">>$?/$01.-23241+-$ 5"+6".$7(+8.9"+3"$ K+*34*<,# K9<,34*<,#

PAGE 88

80 Figure 16 Interaction between Self Attractiveness Ratings and Media Inspired Social Comparison Behavior. On a scale where 1 represents t he minimum level of self attractiveness and 5 represents the maximum level, and 7 represents the minimum level of media inspired social comparison behavior and 35 represents the maximum level. ?# !# =# D# $# G# J# C# C# "# !!# !D# !G# !C# !"# =!# =D# =G# =C# ="# D!# DD# DG# %"):$=--.13-2,"+">>$'1-2+8>$ %(321)$7(#41.2>(+$!"A1,2(.$ %"):;=--.13-2,"+">>$1+6$%(321)$ 7(#41.2>(+$!"A1,2(.$

PAGE 89

81 Appendi x A

PAGE 90

82

PAGE 91

83 Appendix B Rate each image tha t appears on the screen using the following scales. When you are done filling out the three scales for each image, look up at the experimenter. Image 1 How attractive d o you think this model is? very attractive very unattractive 1 2 3 4 5 How much do you relate to this advertisement? not at all very much 1 2 3 4 5 How effective do you find this advertisement? not effective at all very effective 1 2 3 4 5 Image 2 How attracti ve d o you think this model is? very attractive very unattractive 1 2 3 4 5 How much do you relate to this advertisement? not at all very much 1 2 3 4 5 How effective do you find this advertisement? not effective at all very effective 1 2 3 4 5 Image 3 How attractive d o you think this model is? very attractive very unattractive 1 2 3 4 5 How much do you relate to this advertisement? not at all very much 1 2 3 4 5 How effective do you find this advertisement? not effective at all very effective 1 2 3 4 5

PAGE 92

84 Image 4 How much do you relate to this advertisement? not at all very much 1 2 3 4 5 How effective do you find this advertisement? not effective at all very effective 1 2 3 4 5 Image 5 How attractive d o you think this model is? very attractive very unattractive 1 2 3 4 5 How much do you relate to this advertisement? not at all very much 1 2 3 4 5 How effective do you find this advertisement? not effective at all very effective 1 2 3 4 5 Please circle the numbers of the t wo images that you think are the best advertisements. (in no particular order) 1 2 3 4 5 Please explain why you chose these images below: (Note: On this version of the survey, the control image was presented fourth.)

PAGE 93

85 Appendix C Answer the following questions to the best of your ability: 1. I feel I have a number of good physical features. strongly disagree 1 2 3 4 5 6 7 strongly agree 2. I have a positive attitude toward my body. strongly disagree 1 2 3 4 5 6 7 strongly agree 3. I think that I am pretty attractive. strongly disagree 1 2 3 4 5 6 7 strongly agree 4. On the whole, I am satisfied with the way I look. strongly disagree 1 2 3 4 5 6 7 strongly agree 5. At times, I think I am not attractive at all. strongly disagree 1 2 3 4 5 6 7 strongly agree

PAGE 94

86 Appendix D Please answer the following questions to the best of your ability by circling the answer of your choice: 1. I frequently compare myself to models and actors in TV and magazines. strongly dis agree undecided strongly agree 1 2 3 4 5 2. I am likely to compare myself to an actor or model if (he, she) is surrounded by beautiful (women, men) or appears to be very popular. strongly disagree undecided strongly agree 1 2 3 4 5 3. I use (men, women) in ads in magazines and TV as a yardstick to measure how good I look. strongly disagree undecided strongly agree 1 2 3 4 5 4. I am likely to compare myself to an actor or model if (he, she) has coloring similar to mine (i.e., eyes, hair, skin color). strongly disagree undecided strongly agree 1 2 3 4 5 5. I am likely to compare myself to a model or actor if (he, she) is around my age. strongly disagree undecided strongly agree 1 2 3 4 5 6. I am likely to compare myself to a model or actor if my girl(boy)friend thinks (he, she) is appealing. strongly disagree undecided strongly agree 1 2 3 4 5 7. I like to try and copy a "look" I see in TV or magazines, especially if a lot of my friends think it's cool.

PAGE 95

87 strongly disagree undecided strongly agree 1 2 3 4 5

PAGE 96

88 Appendix E Please answer the following questions to the best of your ability: 1. What is your sex? (Circle one) Male Female 2. What is your race/ethnicity? (Circle all options that you identify with) Anglo Saxon / White African American / Black Latino(a) / Hispanic Asian / Pacific Islande r Middle Eastern Native American Other ___________ 3. How old are you? _____ 4. Where on the following scale do you think you fall? (Circle one) very feminine androgynous very masculine 1 2 3 4 5 5. Did you know prior t o this test that the Bem SRI (the self attributes questionnaire in which you rated a list of adjectives) tests for gender role orientation? (Circle one) Yes No 6. Did you recognize or do you know one or both of the models in the photos previously s hown? (Circle one) Yes No


ERROR LOADING HTML FROM SOURCE (http://ncf.sobek.ufl.edu//design/skins/UFDC/html/footer_item.html)