Defying Beauty Norms: its Effect on Advertising Effectiveness and Women’s Mental Health The Effect of Advertising Models’ Body Size on Advertising Effectiveness and Women’s Body-Focused Anxiety and Mood

(1)

Defying Beauty Norms: Its Effect on Advertising

Effectiveness and Women’s Mental Health

The Effect of Advertising Models’ Body Size on Advertising Effectiveness and Women’s Body-Focused Anxiety and Mood

Zelia De Wulf (12389161)

Master’s Thesis

Master’s programme Communication Science Graduate School of Communication

Supervisor: Barbara Schouten June 28, 2019

(2)

Abstract

The thin idealized female body that is communicated in advertising sets unattainable standards for women and has shown to have detrimental effects on women’s physical and mental health. However, its effect on advertising effectiveness remains inconclusive. 157 women with mean age 23.82 (SD =.91) were exposed to a body scrub advertisement

featuring a thin or average sized model to investigate the effect of advertising models’ body sizes on women’s body-focused anxiety and mood, as well as advertisement attitude, brand attitude, and purchase intention. In addition, the moderating role of women’s subjective body size evaluation on the aforementioned relationships was investigated. The study thus consists out of a 2 x 2 quasi-experimental factorial design with the model’s body size (thin vs. average sized) and the participant’s subjective body size evaluation (negative vs. positive) as two factors. As expected, exposure to an average sized model resulted in a more positive

advertisement and brand attitude, as well as greater purchase intention, compared to exposure to a thin model. Counter to expectations, there was no significant effect of model’s body size on body-focused anxiety and mood. There was a significant interaction effect between model’s body size and subjective body size evaluation on mood, indicating the most negative mood was experienced by women with a negative body size evaluation who were exposed to a thin model. The findings thus point at the benefits of using average sized models in

advertising for promoting advertising effectiveness and avoiding a negative mood for women with a negative subjective body size evaluation.

(3)

Introduction

In today’s Western society, there is a strong cultural idea of female beauty which has increasingly become thinner, and significantly thinner than the average body size of the female population. While the average Dutch woman weighs 70kg, has a height of 1m68, and a waist size of 76cm, your typical fashion or editorial model weighs about 53kg and is 1m75 to 1m83 tall, with a waist measurement of around 58 cm. The desirability of a level of thinness that is impossible for many women to attain can result in feelings of body-focused anxiety, body dissatisfaction, low self-esteem, and eating disorders (Grabe, Ward, & Hyde, 2008). While these beauty standards are reinforced by many social influences, such as peers and parents (Keery, van den Berg, & Thompson, 2004), the mass media are described as the single most powerful transmitter of sociocultural ideals, due to their pervasiveness and reach (Tiggemann & McGill, 2004). The mass media plays a crucial role in constructing cultural and social ideals of slender beauty, and this includes advertising.

Beautiful and thin models serve as eye catching appeals under the motto that “beauty sells”. Research indicates that the positive features of such models can spill over to the product, and thus result in more favorable subsequent product evaluations (Baker & Churchill, 1977). However, because of the harm that these thin models can do to women’s body image and satisfaction, there is extensive societal criticism surrounding the use of extremely thin models in advertising (Hafner & Trampe, 2009). Companies have thus increasingly made use of more natural and curvy models in advertising, as studies show that such models elicit less body-focused anxiety than the use of thin models (Clayton, Ridgway, & Hendrickse, 2017; Dittmar & Howard, 2004; Halliwell & Dittmar, 2004; Yu, 2014). There is no consensus, however, on the effect of body size of model on advertising effectiveness. Some studies indicate no significant difference in advertising effectiveness between the use

(5)

while others have shown greater advertising effectiveness when using curvier models (Tsai & Chang, 2007; Yu, 2014).

A number of moderators have been tested in the relationship between type of model used in advertising and subsequent advertising effectiveness and body-focused anxiety to explain the different results described above. While thin-ideal internalization (Anschutz, Engels, Becker, & Van Strien, 2009; Diedrichs & Lee, 2011; Dittmar & Howard, 2004; Halliwell & Dittmar, 2004; Yu, 2014) and social comparison (Dittmar & Howard, 2004; Tiggemann & McGill, 2004; Yu, 2014) have been tested extensively, other possible factors that could influence the effect of models’ body size on body-focused anxiety and advertising effectiveness have not been researched yet. Some studies have looked at the effect of BMI of women, an objective measure of weight relative to height, as a moderator, but results are contradictory and inconclusive (Borau & Bonnefon, 2017; 1997Halliwell & Dittmar, 2004; Henderson-King & Henderson-King,). The lack of conclusive results could be due to the measure being objective, instead of a subjective measure of how one feels about their body size, since the size of your body does not determine how you feel towards it. Yet it is this feeling towards one’s own body that is more likely to impact the effect of models’ body sizes on body-focused anxiety. Investigating the moderating effect of women’s subjective body size evaluation might thus be fruitful, as it could indicate which women are more vulnerable to the effect of models’ body size on their self-image.

While previous studies about models’ body sizes have focused on advertising effectiveness and body-focused anxiety, there is a lack of research into how these images could affect women’s other emotions or overall mood. Previous studies investigating the effect of images featuring thin models on mood have not compared it to images featuring average sized models (Pinhas, 1999; Tiggemann & McGill, 2004) or have produced contradicting results (Anschutz et al., 2009). Since no conclusive results of the effect of

(6)

models’ body sizes on mood have been found, it is relevant to study it further in order to understand whether advertisement models’ body size could also have an impact on general emotions instead of solely body-focused anxiety. This way advertising strategists can be guided towards creating advertisements that don’t have aversive effects on women.

In sum, this study will aim to close several gaps in this research field. It will examine the effect models’ body sizes on body-focused anxiety to substantiate previous scientific findings, and on advertising effectiveness (consisting out of advertisement attitude, brand attitude, and purchase intention) and mood to explain previous contradictory results. Additionally, it will close a gap in literature by investigating the moderating role of subjective body size evaluation in order to determine whether it can explain the lack of consistent previous results of the objective BMI measure. The study’s findings can provide insights into which type of models should be represented in advertisements in order to be economically effective and to prevent negative effects on women’s mental health. The study thus attempts to answer the research question: What is the effect of a model’s body size on women’s body-focused anxiety and mood, advertisement attitude, brand attitude, and purchase intention? And what is the moderating role of subjective body size evaluation?

Theoretical framework

The media’s beauty norms and its influence on women

Thin women and unrealistic body images are overrepresented in today’s mass media. Psychological theories, such as cultivation theory (Gerbner, Gross, & Morgan, 2002) and social learning theory (Bandura, 1977; Brown, 2002) state that repeated exposure to media content can lead one accepting to that content as reality. These thin ideals portrayed by the media thus lead to women accepting this ideal as normal, expected, and central to

(7)

the discrepancy between what the media communicates as the beauty norm and women’s actual body size can have detrimental effects on women. Meta-analyses have found that exposure to thin ideals in the media is indeed linked to women’s body image concerns and dissatisfaction with their bodies (Grabe et al., 2008; Groesz, Levine, & Murnen, 2002). Similarly, other studies have shown that women indicate lower body satisfaction (Clayton, Ridgway, & Hendrickse, 2017; Hargreaves & Tiggemann, 2004; Kim & Lennon, 2007) and greater body-focused anxiety (Dittmar & Howard, 2004; Halliwell, & Dittmar, 2004; Halliwell, Dittmar, & Howe, 2005) when exposed to idealized media images compared to non-idealized media images. Body image concerns and body dissatisfaction pose a significant public health concern, as they have been associated with psychological and physical health problems, such as disordered eating behaviors (Kim & Lennon, 2007; Polivy & Herman, 2002; Stice, 2002), low self-esteem (Furnham, Badmin, & Sneade, 2002; Kim & Lennon, 2007; Tiggeman, 2005), and depression (Brausch & Gutierrez, 2009; Stice, Hayward, Cameron, Killen, & Taylor, 2000). It is thus of importance to take a closer look at how models used in advertising specifically affect women’s mental health.

Advertising and Women’s Body-Focused Anxiety

Advertising is one way through which the thin ideal can be communicated, and subsequently can affect women’s body image. Body-focused anxiety is one such element of body image and implies “feelings of worry, and concern about one’s own body, or negative judgments about one’s own body image” (Borau & Bonnefon, 2017). Several studies have investigated body-focused anxiety as a consequence of being exposed to thin models in advertising. In these studies, researchers exposed women to advertisements featuring either thin models, averaged-sized models, or no models, in which they digitally stretched thin models to resemble an averaged sized model. The result of these studies consistently indicate

(8)

that women exposed to thin models in advertisements experience more body-focused anxiety than those in the averaged-sized model and control group (Dittmar & Howard, 2004;

Halliwell, & Dittmar, 2004; Halliwell, Dittmar, & Howe, 2005). Additionally, a study by Tiggemann and McGill (2004) showed that body-focused anxiety in women was higher after being exposed to thin models in advertising compared to body-focused anxiety before exposure. An explanation for the effect of models’ body sizes on body-focused anxiety is rooted in social comparison theory, first proposed by Leon Festinger (1954). Social

comparison theory suggests that people evaluate themselves through comparison with others. This includes comparing themselves with thin idealized models in the mass media, who they believe represent what is ideal and socially desirable (Sohn & Youn, 2013). There are two different directions of social comparison: upward comparison to others better off and downward comparison to others worse off. These comparisons result in different

consequences relating to self-perceptions and body-image. Whereas downward comparison tends to be self-enhancing, self-evaluation through upward comparisons with thin-idealized or unrealistic models in advertising can result in negative image and increased body-focused anxiety (Sohn & Youn, 2013), which has been repeatedly shown in the

aforementioned studies. Since there are clear negative effects experienced by women who are exposed to thin idealized images in advertising, this previous research suggests that using more realistic average-sized models may pose an effective public health intervention to lessen the negative psychological effects and promote body image (Diedrichs & Lee, 2011).

Hypothesis 1a: An advertisement featuring thin models will result in greater body-focused anxiety compared to an advertisement featuring averaged-sized models.

(9)

Thin models’ effect on women’s mood

While there are some studies on thin models’ effect on women’s body-focused anxiety, other emotions women may experience, or their overall mood, has not been studied extensively and results remain inconclusive. Mood is defined as a conscious state of mind or predominant emotion (Merriam-Webster, 2019), and is thus temporary. The cognitive-mediational theory of emotion, proposed by Lazarus (1922) states that our emotions are determined by our appraisal of stimuli, which is immediate and can be conscious or

unconscious, and mediates the relationship between the stimulus and the emotional response. In relation to the current study, advertisements featuring models are the stimuli, which then lead to appraising this stimulus in a certain way. If one is presented with a thin-idealized model, one might appraise this model as being very different looking to oneself after engaging in social comparison, which then results in negative cognitions about one’s own body and subsequent negative emotions or mood. The limited research largely supports this notion, but contains some methodological issues rendering the results inconclusive. One study by Pinhas and colleagues (1999) found that viewing images of thin fashion models, compared to no models, had an immediate and negative effect on women’s mood, and Tiggemann and McGill (2004) found that women had a more negative mood after being exposed to thin models than before. Participants in both studies were not exposed to averaged-sized models, however, not allowing for a comparison between the effect of thin models and averaged size models on mood. Finally, one study by Anschutz and colleagues (2009) did compare thin models to average-sized, but they used television commercials (Dove versus Nivea) instead of print advertising and found the opposite effect. The findings indicated that women exposed to averaged-sized women in commercials resulted in a more negative mood compared to being exposed to thin women. The authors speculate that this unexpected finding could be due to women being more reminded of the thin beauty ideal by

(10)

the ‘Dove’ commercials, since these commercials explicitly refer to the beauty ideal, while the ‘Nivea’ commercials do not. Since there is no research comparing the effect of thin models to averaged-sized models on mood in print advertisements in a controlled manner, the current study will investigate this effect. Based on the research of thin models’ effect on mood as well as on body dissatisfaction and body image, we expect women exposed to thin models to experience a more negative mood than women exposed to average-sized models.

Hypothesis 1b: An advertisement featuring thin models will result in a more negative mood compared to an advertisement featuring averaged-sized models.

Advertising effectiveness of thin vs. average sized models

Advertising effectiveness is a broad term and has been defined in previous studies as consisting out of three components: attitude toward the advertisement, attitude toward the brand, and purchase intention of the product (Borau & Bonnefon, 2017; Diedrichs & Lee, 2011; Dittmar & Howard, 2004; Halliwell and Dittmar, 2004; Yu, 2014). The current study will thus measure the effects on all three variables.

Despite the evidence for the negative effects of thin advertising models on women’s body image and anxiety, the advertising and fashion industries continue to use such models. This is due to the concept that ‘beauty sells’ or ‘thinness sells’, since beauty has often

become synonymous to thinness in Western society. According to a spokesman from Premier Model Management, a London based model agency, "statistics have repeatedly shown that if you stick a beautiful skinny girl on the cover of a magazine you sell more copies." (Gillian, 2000, p. 7). Some studies support this ‘thinness sells’ concept, as previous studies have found that advertisements displaying attractive models with beautiful faces and thin bodies result in greater favorability of the advertisement and product, as well as greater purchase intention of

(11)

the product (Baker & Churchill, 1977; Kahle & Homer, 1985; Petroshius & Crocker, 1989), and result in greater trustworthiness of the model (Till & Busler, 2000). It is important to note however, that although these models were thin, they were compared to less attractive

individuals in advertisements, not to models with larger body sizes. Borau and Bonnefon (2017) did investigate the difference between body sizes of models in advertising on advertising effectiveness and found that larger models triggered repulsion in viewers with higher BMI, which hurt advertising performance. These findings of the advertising advantage of thin models have been inconsistently supported however, as a number of studies have found realistic models to be equally (Diedrichs & Lee, 2011; Dittmar & Howard, 2004; Halliwell & Dittmar, 2004; Yu, Damhorst, & Russell, 2011) or even more (Hafner &

Trampe, 2009; Sohn & Youn, 2013; Tsai and Chang, 2007; Yu, 2014) effective in advertising compared to thin models. These contradictory results could possibly be explained by the fact that in the studies where thin models were more effective, there was no comparison to average sized women at all or to average sized women with attractiveness being controlled.

Several studies that have focused on the effect of models’ body sizes on advertising effectiveness found that advertisements were equally effective, regardless of the models’ body size (Diedrichs & Lee, 2011; Dittmar & Howard, 2004; Halliwell & Dittmar, 2004; Yu, Damhorst, & Russell, 2011). These studies argued that this result was positive as it indicates that average-sized models can be used to avoid negative consequences on women while remaining just as effective. Additionally, some previous studies have pointed out the positive advertising effects of using average-sized models. Hafner and Trampe (2009) found that rounder advertising models resulted in more favorable responses when the product was evaluated reflectively, compared to thin models. For impulsive product evaluations the pattern was reversed. Research by Yu (2014) indicated that participants exposed to thin-idealized model images showed lower advertising effectiveness than those participants

(12)

exposed to non-idealized model images. Similarly, Sohn and Youn (2013) indicated that the use of an average-sized model generated more positive brand attitude and purchase intention over the use of either thin or plus-sized models.

There are two possible underlying reasons for why averaged-sized models can boost advertising effectiveness through advertisement attitudes. One explanation is based on social comparison theory. Upward comparison with a model that looks very different to oneself might result in negative feelings and a subsequent negative interpretation of the advertising message (Sohn & Youn, 2013; Yu, Damhorst, & Russell, 2011). A second, more plausible, explanation for the advertising effectiveness of averaged-sized models is that such

advertisements are liked due to its progressive message. Combatting old-fashioned beauty norms and promoting a body positive message by featuring realistic women of all shapes and sizes create positive attitudes toward the advertisement and increase campaign liking, which then spill over to the advertising effectiveness, which was shown by Hafner and Trampe (2009). Similarly, Anderson & Paas (2012) showed that thin models can hurt advertising effectiveness by evoking negative thoughts about the ethical character of the add. Based on the notions of social comparison theory, but primarily, the liking of advertisements which communicate a progressive message, I expect more positive attitudes toward the

advertisement and the brand, as well as greater purchase intention of the product, for advertisements featuring average-sized models, compared to advertisement featuring thin models.

Furthermore, it is hypothesis that there will be a relationship between advertisement attitude, brand attitude, and purchase intention. It is hypothesized that the effects of the model’s body size on advertisement attitude, due to reasons described above, will affect, or spill over to, the brand attitude, a phenomenon known as the “spillover effect” (Balachander & Ghose, 2003). Additionally, according to the theory of planned behavior (Ajzen, 1991),

(13)

attitude toward behavior, subjective norms, and perceived behavioral control, together shape an individual's behavioral intentions. This means that the attitude toward the advertisement and brand mediate the relationship between type of model used and purchase intention of the product. Based on the extent to which the advertising message is progressive, the spillover effect, and the theory of planned behavior we expect that model’s body size will influence advertisement attitude, which will in turn influence brand attitude, which in turn will influence purchase intention.

Hypothesis 1c, d, e: An advertisement featuring thin models will result in c) less positive attitude toward the advertisement and d) the brand, as well as e) less purchase intention, compared to an advertisement featuring averaged-sized models.

Hypothesis 2: There will be a serial mediation between models’ body size, advertisement attitude, brand attitude and purchase intention, meaning model’s body size will affect attitude towards the advertisement, which will in turn affect attitude towards the brand, which will in turn affect purchase intention.

Subjective Body Size Evaluation

Body image is a component of, and central to, the self-concept. Self-concept is the concept that an individual has of himself as a physical, social, and spiritual moral being (Gecas, 1982), but also in relation to family and work (Fitts & Roid, 1964). According to Carl Rogers (1959) the concept is related to several other ‘self’ constructs, such as self-image, self-esteem, and the ideal-self. Self-image is how one views him- or herself, which is subjective and can vary significantly from reality (Rosenberg, 2015). Body-image is a component of self-image and is defined as a “multidimensional construct that embraces a

(14)

person’s conscious perception of their physical self, including the thoughts and feelings that result from that perception” (Irvine et al., 2019, p.2).

After performing a meta-analysis, Cash and Deagle (1997) identified two components to body image: i) a perceptual component, that involves how an individual judges the

dimensions of their physical appearance, and ii) an attitudinal component which encompasses the feelings an individual has towards that perception of body size and shape. For the current study, the attitudinal component is of interest, specifically one’s attitude toward one’s body size. Since body-image is often not an accurate representation of reality, body-image measures can differ from objective body size measures, such as one’s body mass index (Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999). Studies researching BMI as a moderator have found varied results. One study has found thin idealized media images to increase body satisfaction only in heavier women (Henderson-King & Henderson-King, 1997), another study found average-sized models to trigger repulsion in viewers with higher BMI, which hurt advertising performance (Borau & Bonnefon, 2017), and other research has failed to find a systematic impact of body mass index of the viewer on advertising

effectiveness and body-focused anxiety (Halliwell & Dittmar, 2004). Investigating the effect of women’s subjective evaluation of their body size, meaning how negative or positive they feel about their body size, instead of their actual body-size, might be more fruitful. Results about BMI as a moderator have been inconclusive, possibly because someone’s actual body size does not determine how good or bad one feels about their body, while it is this feeling that is more likely to moderate the effect of a model’ body size on body-focused anxiety and advertising. Investigating this subjective measure will thus allow to observe whether

advertising models impact body-focused anxiety, mood, and advertisement attitude differently depending on the viewer’s evaluation of her own body size, and can offer a solution for inconclusive results of previous studies that studied an objective measure. One

(15)

study by Posavac and colleagues (1998) found that the negative effects, such as weight concerns, after exposure to thin media images, compared to neutral images, were higher in women who had the stable trait of body-dissatisfaction. This result is due to women engaging in social comparison whereby female perceivers assess their appearance relative to society’s standards as depicted in the media (Posavac, Posavac, & Posavac, 1998). If there is a large discrepancy between one’s body and that of the model in the advertisement, negative consequences as the ones discussed earlier are more likely to result. Thus, for women who have a negative evaluation of their body-size, being confronted with an idealized model will result in more negative consequences, such as body-focused anxiety, than women who have a more positive evaluation of their body-size. Extrapolating this result, we may also assume that women who have a negative evaluation of their body-size will also have a more negative mood than women with a more positive evaluation of their body-size when presented with a thin model, although no research exists yet. Research on women’s own body evaluation as a moderator for the relationship between advertising models’ body size and advertising

effectiveness is also limited. One study by Yu, Damhorst, and Russel (2011), found however, that individuals’ body images indirectly influenced brand attitudes through their perceptions of similarity with model images in the social comparison process. The following hypotheses of subjective body-size evaluation can thus be made:

Hypothesis 3: The effects of thin models on a) increased body focused anxiety and b) a negative mood, and c) a negative attitude towards the advertisement, will be stronger for women with a more negative subjective body size evaluation. The effects of average-sized models on a) decreased body-focused anxiety and b) positive mood, as well as increasing c) positive attitude towards the advertisement, will be stronger for women with a more negative body size evaluation.

(16)

Figure 1. Study design model.

Methods

Design

The study consists out of a 2 x 2 quasi-experimental factorial design with the model’s body size (thin vs. average sized) and the participant’s subjective body size evaluation (negative vs. positive) as two factors.

(17)

Table 1. Study conditions

Subjective body size evaluation

Model’s body size Negative Positive

Thin N = 40 N = 40

Average sized N = 37 N = 40

Participants

Convenience sampling was utilized to recruit participants, as I posted the link to the survey on my social media platforms and requested women to complete the survey. This technique has great reach, is very easy, and saves time because participants are selected on the basis of availability and willingness to take part. However, this technique is prone to bias, as individuals who volunteer might have different characteristics from those who do not (Etikan, Musa, & Alkassim, 2016). Two hundred and forty-two women started the survey, but 85 had to be removed because they did not complete it. A final sample of 157 women were used to conduct the analyses (see table 1 for the number of respondents in each condition).

Procedure

The study took the form of an online experiment, with a cover story about advertising and consumer personality. Participation was voluntary and the survey first asked for informed consent (Appendix A), followed by an introduction to the research and a demographics section asking for age, gender, ethnicity, country of upbringing, and education level. Participants who indicated they were male were guided to the exit page. Participants were then presented with items to assess their subjective body size evaluation. Among these items were eight filler items about the participant’s personality to fit the cover story (e.g.: “I like

(18)

being at the center of attention”). Participants were then told that they would see an

advertisement for a body scrub brand and would have to evaluate it later. The next part of the questionnaire includes the experimental manipulation. Participants were randomly assigned to an advertisement featuring either a thin or an averaged sized model. After viewing the ad, participants were asked questions about their attitudes towards the advertisement, their attitudes towards the brand, and their intentions to purchase the product. Participants were then asked to indicate how anxious they are feeling “right now” about various aspects of their lives. Embedded within these questions was the shortened version of the PASTAS focusing on weight-related body parts (see below for more information; Reed, Thompson, Brannick & Sacco, 1991). The eight filler items include participants rating how anxious they felt about “my personality,” “my intelligence,” and “my social life”. Participants were then asked presented with five items to assess their mood. In order to check whether participants were aware of the true purpose of the study, they were asked to state what they thought the specific purpose of the study was. Finally, they were asked to what extent they believed the model was thin or curvy, as a manipulation check. The last page of the survey included a debriefing of the study’s true purpose as well as an email address that participants could contact if they would like more information about the results of the study.

Materials

Images and Advertisements. Each model condition was represented by an

advertisement that showed a fictional body scrub brand (Appendix B). It was a made-up brand in order to avoid any bias towards the product, which could have impacted the results. Adobe PhotoShop was used to cut an image of a body scrub and alter it so that the brand was unrecognizable, after which it was placed on the new advertisements. Advertising slogans were added but kept bland to ensure that the image was the main point of focus of the

(19)

advertisement. The only difference between the two conditions was the size of the model featured in the advertisement. Based on a study by Halliwell and Dittmar (2004), the average sized model condition featured a model with approximately to UK dress size 14, U.S. size 12-14, a 76 cm waist, which is based on a survey of women’s body sizes in the UK that showed that size 14 is the average size for British women (Aldridge, 2000). To create the advertisements for the thin model condition, PhotoShop was used to compress the size of the model’s body. Compressing a picture of one model, rather than using different models, allows all factors except body size to be controlled (such as attractiveness). The model was compressed to a UK size 8, U.S. size 2, with a waist size of 58 cm, following the study by Halliwell and Dittmar (2004), as these measurements resemble the measurements of the first 20 models on a leading model agency website (www.womenmanagement.com).

Pilot Study: Attractiveness and Body Size of the Models and Believability of the Advertisement. In order to check that the thin and average-size models were indeed perceived as having different body sizes, and that the manipulation did not affect the attractiveness of the models, fifteen women were recruited to rate either the thin or the average-sized model. Each model was rated twice on a 7-point Likert scale, based on the pilot study scales used in a study by Dittmar and Howard (2004). The first one measured attractiveness, ranging from very unattractive (1) to very attractive (7), and the second one assessed body size, ranging from very thin (1) to very curvy (7). An independent samples t-test was conducted with the model’s body size (thin vs. average) as independent variable, and body size rating as

dependent variable. A significant effect of model’s body size was found on body size rating, t(13) = -3.89, p = .002. Participants perceived the average sized model stimulus as

significantly more curvaceous (M = 5.14, SD = 1.21) than the thin model stimulus (M = 2.75, SD = 1.16). This indicates that the manipulation of model’s body size was successful. An insignificant effect of model’s body size was found on attractiveness rating, t(13) = 1.82, p =

(20)

.092. Participants thus did not perceive the thin model stimulus as significantly more

attractive (M = 6.13, SD = .64) than the average sized model stimulus (M = 5.39, SD = 1.11). Finally, participants were also presented with the statement “This advertisement appears realistic and professional” to which they were asked to respond on a 7-point Likert scale from strongly disagree (1) to strongly agree (7). The mean was 6.20 (SD = .56), indicating that the advertisement appeared highly realistic and professional. An independent samples t-test also indicated that there is no significant difference in realistic and professional appearance between the ad featuring the thin model (M = 6.38, SD = .52) and the ad featuring the average sized model (M = .58, SD = .22), t(13) = 1.33, p = .207.

Measures

Demographic data. Age, ethnicity, country of upbringing, and education level were self-reported by the participants.

Subjective body-size evaluation. This is a stable trait variable and was measured through 3 items: “I feel good about my body size”, “I am confident about my body size”, and “I like the way I look naked”. The items were evaluated on a 7 point Likert scale that ranged from definitely disagree (1) to definitely agree (7) (Appendix C). An exploratory factor analysis indicated that the scale was unidimensional (only one component with Eigenvalue above 1.00), explaining 88.14% of the variance in the three original items. The scale also proved to be reliable as indicated by a Cronbach’s Alpha of .93 (M = 4.19, SD = 1.61). Since this a quasi-experimental variable, a median split was performed to create a negative and a positive subjective body-size evaluation. Participants with an evaluation below the median of 4.66 were included in the negative subjective body-size evaluation group (N = 78), and women with an evaluation greater or equal then 4.66 were included in the positive subjective body-size evaluation group (N = 80).

(21)

Body-focused anxiety. The Physical Appearance State and Trait Anxiety Scale (PASTAS: Reed, Thompson, Brannick & Sacco, 1991) was used to measure Body-Focused Anxiety (Appendix D). The scale assesses anxiety associated with various body parts, and in the current study, the scale is administered as a state measure of anxiety. Since only the weight related items are of interest, seven weight-related items were selected from the original 16-item scale. Participants rated on a five-point scale how anxious they felt “right now”, ranging from not at all (1) to extremely (5), about “the extent to which I look overweight”, “my thighs”, “my buttocks”, “my hips”, “my stomach”, “my legs”, and “my waist”. This weight related subscale has been proven reliable in previous research with reliabilities of .90 or above (Anschutz et al., 2009; Halliwell & Dittmar, 2004; Tiggemann & McGill, 2004). An exploratory factor analysis indicated that the scale was unidimensional, explaining 64.46% of the variance in the three original items. The scale also proved to be reliable as indicated by a Cronbach’s Alpha of .91 (M = 4.45, SD = 1.22).

Mood. Following Heinberg and Thompson (1995) and Tiggemann and McGill (2004), participants were asked to indicate how they felt “right now” on the mood dimensions of anxiety, depression, happiness, anger, and confidence, on a 5-point Likert scale from not at all (1) to extremely (5) (Appendix E). Heinberg and Thompson (1995) proved the 5-item scale to be reliable, which was demonstrated in this study as well with a Cronbach’s Alpha of .70 (M = 3.69, SD = .58).

Attitude towards the advertisement. This outcome variable was assessed by the question “What is your opinion of the advertisement?” Participants responded on five seven-point semantic differentials: unfavorable-favorable, unappealing-appealing, unpleasant-pleasant, bad-good, and unlikable-likable (Appendix F). These items measuring attitude toward the advertisement have previously been proven reliable with a Cronbach’s Alpha of .97 (Spears & Singh, 2004). An exploratory factor analysis indicated that the scale for

(22)

attitude towards advertisement was unidimensional, explaining 81.53% of the variance in the three original items. The scale also proved to be reliable as indicated by a Cronbach’s Alpha of .94 (M = 4.58, SD = 1.31).

Attitude towards the brand was measured through the question “Describe your overall reaction and feeling about the brand described in the ad”. Participants again responded on five seven-point semantic differentials: unfavorable-favorable, unappealing-appealing, unpleasant-pleasant, bad-good, and unlikable-likable (Appendix F). The items measuring attitude towards the brand have been proven reliable in previous research (α =.97; Spears & Singh, 2004), and also proved to be reliable in this study with a Cronbach’s alpha of .94 (M = 4.45, SD = 1.22), as well as unidimensional, explaining 87.72% of the variance.

Purchase intent was measured with a single item: “How likely would you be to purchase it on your next shopping trip?”. Responses were recorded on a seven-point Likert scale ranging from very unlikely (1) to very likely (7) (Appendix F).

Awareness of the study’s true purpose. This was assessed by asking the subjects what they thought the true purpose of the study was and if anything stood out to them. The

questions were open-ended, and participants provided a written response.

Manipulation Check. In order to check whether the models in the two advertisements did indeed differ in perceived body size, women were asked to rate the body size of the model in the advertisement on a 6-point Likert scale ranging from very thin (1) to very curvaceous (7). This item is based on a study by Dittmar and Howard (2004).

Results

Descriptive Statistics

A total of 157 women completed the entire survey. Their mean age was 23.82 (SD =.91), 94.3% of the sample was white and 82.2% has been living in Europe the majority of

(23)

their lives, 13.9% in North America, and 3.8% in Africa. Of the participants 67.5%

completed a Bachelor’s degree or higher. The descriptive statistics of each outcome variable for each condition are displayed in Table 2.

Table 2. Descriptive statistics of all outcome variables for the four conditions

Outcome variables Thin model, negative body size evaluation N = 40 Thin model, positive body size evaluation N = 40 Average sized model, negative body size evaluation N = 37 Average sized model, positive body size evaluation N = 40 Total N= 157 Body-focused anxiety, M (SD) 3.24 (.74) 1.92 (.55) 3.09 (.82) 2.05 (.72) 2.56 (.92) Mood, M (SD) 3.41 (.53) 3.94 (.45) 3.71 (.62) 3.83 (.55) 3.69 (.58) Advertisement attitude, M (SD) 3.79 (.92) 4.22 (1.25) 5.21 (1.16) 5.17 (1.31) 4.58 (1.13) Brand attitude, M (SD) 3.97 (1.08) 4.24 (1.12) 4.78 (1.34) 4.87 (1.15) 4.45 (1.22) Purchase intention, M (SD) 3.05 (1.60) 3.47 (1.52) 4.43 (1.71) 4.48 (1.95) 3.85 (1.79)

Randomization- and Manipulation Check

In order to check if participants’ age and education level was comparable over the four conditions, a One Way ANOVA was conducted with the four conditions as independent variable and the control variables as dependent variable. Participants’ mean age did not significantly differ across the four conditions, F(3, 153) = .93, p = .427, and neither did

(24)

education level, F(1, 153) = .19, p = .906. A χ2 - test indicated that ethnicity of the participants did not differ significantly across the four conditions, X2_{(6, 157) = 10.68, p} =.099, and neither did country of upbringing, X2_{(6, 157) = 8.86, p = .182. The control} variables were thus evenly distributed among the four conditions.

In order to check that the thin and average-size models were indeed perceived as having different body sizes, an independent samples t-test was conducted with the model stimulus as independent variable, and body size rating as dependent variable. A significant effect of model’s body size was found on body size rating, t(155) = -15.46, p = .000.

Participants perceived the average-sized model stimulus as significantly more curvaceous (M = 5.19, SD = .92) than the thin model stimulus (M = 2.79, SD = 1.03). This indicates that the manipulation of model’s body size was successful.

Hypotheses Testing

To test the different hypotheses an ANCOVA was computed with the model’s body size and subjective body size evaluation as independent variables, and body-focused anxiety, mood, advertisement attitude, brand attitude, and purchase intention as dependent variables

Hypotheses 1a, b, c, d, e

The ANCOVA indicated that there is no significant effect of model’s body size on body-focused anxiety, F(1, 152) = .003, p = .954, η2_{p < .001. Body-focused anxiety was slightly} higher for women exposed to a thin model (M = 2.58, SD = .92) compared to women exposed to an average-sized model (M = 2.55, SD = .93), but the difference was not significant, thus rejecting hypothesis 1a.

When the ANCOVA was repeated with mood as the dependent variable, again no significant effect of model’s body size was found, F(1, 1152) = .062, p = .804, , η2_{p = .001.}

(25)

Women exposed to a thin model displayed a slightly more negative mood, (M = 3.67, SD = .56) compared those exposed to a curvy model (M = 3.71, SD = .62), but this difference was not significant, and hypothesis 1b was thus rejected.

When repeating the ANCOVA with advertisement attitude as the dependent variable, there was a significant effect of model’s body size, F(1, 152) = 39.88, p < .001, η2_{p = .208,} thus confirming hypothesis 1c. The participants’ attitude toward the advertisement was significantly more positive when it featured an average-sized model (M = 5.19, SD = 1.23) compared to a thin model (M = 4.00, SD = 1.11).

Hypothesis 1d was also confirmed as there was a significant effect of model’s body size on brand attitude, F(1, 152) = 15.48, p < .001, η2_{p = .089. Women had a significantly} more positive attitude towards the brand featuring the average sized model (M = 4.82, SD = 1.24), compared to the advertisement featuring a thin model (M = 4.10, SD = 1.10).

The ANCOVA analysis also indicated a significant effect of model’s body size on purchase intention of the product, F(1, 152) = 19.71, p < .001, η2_{p = .112, confirming}

hypothesis 1e. Women exposed to the advertisement featuring an averaged sized model were significantly more likely to purchase the product on their next shopping trip (M = 4.45, SD = 1.83), compared to women who were exposed to an advertisement featuring a thin model (M = 3.26, SD = 1.57).

(26)

Figure 2 The effect of Model's body size on Advertisement attitude, Brand attitude, and Purchase

intention, and overall Advertising effectiveness

Hypothesis 2

In order to check the hypothesis of serial mediation from model’s body size to advertisement attitude, to brand attitude, to purchase intention, PROCESS by Andrew Hayes was used. Using model number 6, purchase intention was added as the dependent variable, model’s body size as the independent variable, advertisement attitude as mediator 1, and brand attitude as mediator 2. The results of the serial mediation analysis can be found in Table 2 and Figure 3, and indicate that advertisement attitude significantly mediates the path from model’s body size to brand attitude and from model’s body size to purchase intention, but brand attitude did not significantly mediate the path from advertisement attitude to purchase intention. Hypothesis 2 was thus partially confirmed.

0.00 1.00 2.00 3.00 4.00 5.00 6.00

Thin model Average sized model

Me

an Advertising attitude

Brand attitude Purchase intention

(27)

Table 2. Serial mediation analysis: Direct and indirect effects of Model’s body size on Advertising

attitude, Brand attitude, and Purchase intention.

Direct effect (Unstandardized coefficient (SD) [LLCI, ULCI]) Advertisement attitude Brand attitude Purchase intention Model’s body size 1.18 (.19)

[2.24, 3.40]* -.14 (.14) [-.42, .15] .58 (.28) [.02, 1.14]* Advertisement attitude .73 (.06) [.62, .83]* .38 (.16) [.07, .69]* Brand attitude .22 (.16) [-.09, .53] Indirect effect

Purchase intention via Advertisement attitude

Purchase intention via Brand attitude

Purchase intention via Advertisement attitude and Brand attitude

Model’s body size .45 (.19) [.08, .85]*

-.03 (.05) [-.17, .03]

.19 (.14) [-.08, .48]

Abbreviations: LLCI = lower limit confidence interval, LUCI = upper limit confidence interval

Figure 3. Serial mediation model of Model’s body size, Advertisement attitude, Brand attitude, and

Purchase intention

(28)

Hypothesis 3a, b, c

Results indicated that there was no significant interaction effect of model’s body size and subjective body size evaluation on body-focused anxiety, F(1, 152) = 1.71, p = .192, η2_p = .009, thus disconfirming hypothesis 3a.

There was a significant interaction effect of model’s body size and subjective body size evaluation on mood, F(1, 152) = 4.52, p = .035, η2_{p = .019, illustrated in Figure 4. The} direction of the interaction effect was not entirely as hypothesized however. It was predicted that women’s mood would be more positive when exposed to an averaged sized model compared to a thin model, but that this difference in means would be greater for women with a negative subjective body size evaluation than for women with a positive subjective body size evaluation. The results showed that women with a negative subjective body size evaluation indeed had a more positive mood when exposed to an average sized model (M = 3.59, SD = .67) compared to a thin model (M = 3.41, SD = .53), but that women with a positive body size evaluation had a more positive mood when exposed to a thin model (M = 3.94, SD = .45) compared to when exposed to an average sized model (M = 3.83, SD = .55). Women with a positive subjective body size evaluation displayed significantly higher means in mood (M = 3.88, SD = .50) compared to women with a negative subjective body size evaluation (M = 3.49, SD = .60), regardless of whether they were exposed to a thin or averaged sized model.

(29)

Figure 4. The interaction effect of Model's body size and Subjective body size evaluation on Mood

Hypothesis 3c was disconfirmed as there is no significant interaction effect of

model’s body size and subjective body size evaluation on advertisement attitude (F(1, 152) = 1.46, p = .228), η2_{p = .010.}

Discussion

The aim of the present study was to examine the impact of advertising models’ body size on women’s body-focused anxiety and mood, as well attitude towards the advertisement, the brand, and purchase intention. Furthermore, the moderating role of subjective body size evaluation was investigated, as well as the serial mediation between model’s body size, advertising attitude, brand attitude, and purchase intention. Results indicated that the use of a thin model resulted in more negative attitude towards the advertisement and brand, as well as lower purchase intention, compared to the use of an average sized model. There was no significant difference however, between the two types of body sizes of the models on

body-3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4

Thin model Averaged sized model

Av er ag e M oo d Negative subjective body size evaluation Positive subjective body size evaluation

(30)

focused anxiety and mood of the participants. Subjective body size evaluation also did not have a significant moderating effect on the relationship between models’ body size and the outcome variables, except for on mood. Finally, there was no full serial mediation between the variables, but there was a significant mediation effect of advertisement attitude between model’s body size and brand attitude, and between model’s body size and purchase intention. In sum, models’ body type affect advertisement attitude, brand attitude, and purchase

intention, but not body-focused anxiety and mood.

Discussion of the results

Effect of model’s body size on advertisement attitude, brand attitude, and purchase intention, and the moderating role of subjective body size evaluation

In line with hypothesis 1, the use of an average sized model in the advertisement resulted in c) a more positive attitude toward the advertisement, d) the brand, and e) greater purchase intention, compared to the use of a thin model. Scientific research on the advertising effectiveness of advertisements featuring thin vs. average sized models have come to varied conclusions. Some studies have found thin models to be more effective (Baker & Churchill, 1977; Borau & Bonnefon, 2017;) than regular looking women, whereas the majority of studies have found average sized models to be equally (Diedrichs & Lee, 2011; Dittmar & Howard, 2004; Halliwell & Dittmar, 2004; Yu, Damhorst, & Russell, 2011), or more, effective than thin models (Hafner & Trampe, 2009; Sohn & Youn, 2013; Tsai and Chang, 2007; Yu, 2014). The results of the present study are in line with the latter, as using average sized models in advertisements proved to be beneficial to the effectiveness of the

advertisement. The inconsistency with studies that showed more effectiveness of thin models can be explained by the fact that those studies did not control the attractiveness between the two conditions. Women who were exposed to the advertisement featuring an average sized

(31)

model liked the advertisement and the brand more and were more willing to purchase the product on their next shopping trip, compared to women exposed to a thin model. Two possible theoretical explanations were provided in the theoretical framework to explain this positive effect of average sized models on advertisement and brand attitude and purchase intention: social comparison theory and the progressive message of the advertisement. Since no significant effect of model’s body size was found on body-focused anxiety in women, we can speculate that it is unlikely that negative feelings towards one’s own body, due to social comparison processes, provide an explanation for why the advertisements featuring thin models were not as effective. It is more likely that the progressive message of the

advertisement featuring the average sized model is what led to the more positive attitude toward the advertisement compared to the one featuring a thin model. Women nowadays tend to support and like advertisements that step away from traditional beauty standards and are all inclusive, embracing the wide spectrum of beauty of all women (Kemp, 2018; Mabry-Flynn & Champlin, 2018). Campaigns that are defying the norm and redefining what it means to be beautiful are popular and appreciated tremendously this day and age, thus providing a possible explanation for the positive attitude toward the advertisements featuring average sized models, and the less positive attitude towards the advertisements featuring thin idealized models.

No moderation effect of subjective body size evaluation was found on the relationship between model’s body size and advertisement attitude. In line with the aforementioned explanation, if the progressive message of the advertisement featuring average sized models is the underlying mechanism that causes positive attitudes towards the advertisement, then the subjective body size evaluation one is less likely to impact this relationship, than if social comparison tendency was the underlying mechanism. How one evaluates their own body can

(32)

be separate from having a positive attitude toward progressive messages about body positivity.

Effect of model’s body size on women’s body-focused anxiety and mood, and the moderating role of subjective body size evaluation

The result of the present study indicated that the models’ body size did not impact the amount of body-focused anxiety experienced by women, which is not in line with previous research that has consistently shown an effect on body-focused anxiety (Dittmar & Howard, 2004; Halliwell, & Dittmar, 2004; Halliwell, Dittmar, & Howe, 2005; Tiggemann & McGill, 2004). One possible explanation might be rooted in the sample size, because the findings were in the expected direction. That is, there was a slightly higher mean of body-focused anxiety for women who were exposed to a thin model compared to women who were exposed to an average sized model. Perhaps, with a larger sample size, this difference in means might have been significant. A second possible reason is that the participants did not engage in social comparison. As was stated in the theoretical framework, self-evaluation through upward comparisons with thin-idealized or unrealistic models in advertising can result in negative body-image and increased body-focused anxiety (Sohn & Youn, 2013). This reasoning is in line with previous studies that have pointed at the moderating (Dittmar & Howard, 2004) and the mediating effect (Tiggemann & McGill, 2004) of social comparison tendency on body-focused anxiety. Consequently, if this study sample did not engage in social comparison, the model’s body size would thus also not impact body-focused anxiety. Perhaps the individuals did not engage in social comparison because they did not identify with the model, or because they did not process the advertisement long enough. If the advertisement was processed more peripherally, rather than reflectively, because of the brief amount of time that they were asked to observe the advertisement, then the participants might

(33)

not have engaged in social comparison. Finally, perhaps no significant effect of model’s body size was found on body-focused anxiety because the sample mainly consisted out of young students who possibly resembled the body size of the thin model, and were thinner than the average sized model. If that was the case then the thin model might not necessarily cause significantly more body focused anxiety than the average sized model.

There was also no significant main effect of model’s body size on mood, which is not in line with the limited previous research, which have shown exposure to thin models to result in a more negative mood compared to average sized models (Pinhas et al, 1999; Tiggemann & McGill, 2004). Again, this might be explained by the possibility that the participants did not associate themselves with, or compared themselves to, the model in the advertisement, which subsequently could not affect their moods. A methodological

explanation could be that perhaps the mood scale was not the best reflection of mood, since it had items like “confidence” and “anger”. Perhaps a scale measuring “sadness” and

“happiness” would have been a better measure of mood and would have given significant results, as this scale has been proven to be reliable in a previous study (Anschutz, Engels, Becker, & Van Strien, 2009).Additionally, subjective body size evaluation was analyzed as a moderator of the relationship between model’s body size and body-focused anxiety. No interaction effect was found, which can be attributed to the aforementioned reason of the participants not comparing themselves with the models in the advertisement.

There was a significant interaction effect between the model’s body size and subjective body size evaluation on mood, but it was not in the expected direction. It was hypothesized that women would have a more negative mood when exposed to a thin model compared to when exposed to an average sized model, and that this discrepancy would be greater for women with a more negative subjective body size evaluation compared to women with a more positive body size evaluation. While women with a more negative body size

(34)

evaluation had a more negative mood when exposed to a thin model compared to an average sized model, women with a more positive body size evaluation actually experienced a more positive mood in general, compared to women with a more negative body size evaluation. This result can be explained by the previous finding that negative body image predicts negative mood in both women and men (Holsen, Kraft, & Roysamb, 2001). Results thus indicate that although body size of the model did not have a main effect on mood, the use of average sized models could diminish the negative mood experienced among women with a negative body size evaluation.

Serial mediation of model’s body size, advertisement attitude, brand attitude, and purchase intention

The serial mediation was partially confirmed. Advertisement attitude did significantly mediate the relationship between model’s body size and brand attitude, and between model’s body size and purchase intention. This means that a positive attitude towards the

advertisement had a spillover effect on the brand attitude and purchase intention, which is in line with the existing literature (Balachander and Ghose, 2003). However, the fact that attitude towards the brand did not mediate the relationship between attitude towards the advertisement and purchase intention is not in line with the TBP (Ajzen, 1991). A possible explanation for why ad attitude, but not brand attitude positively influenced purchase intention, is that the participants that the reason that the participants want to buy the product might be because they want to support a company with a body positive message, and not necessarily because they liked the actual product.

(35)

Limitations and suggestions for future research

One limitation of the present study is that there was only one stimulus for the thin model condition and one stimulus for the average sized model condition, which can lead to case-by-category confound. This means that we cannot be entirely confident whether any observed difference in outcome variables between the two conditions was due to the categories (thin vs. average sized model) or due to two individual cases. This possible limitation was eliminated as much as possible by creating an identical advertisement with only the size of the models differing between the two conditions, but future research should repeat the study and create multiple stimuli per category.

Another limitation might have been a language barrier. For the majority of the

participants, English was not their first language, which may have led to not understanding or misinterpreting some of the questions, and subsequently not answering correctly. Future research should ensure to only recruit participants who completely master the language of the survey.

As the data in this study was all self-reported, social desirability bias needs to be considered. It could be argued that, since there is a recent social trend in shaming campaigns that use very thin models, and praising campaigns using average sized models (Kemp, 2018; Mabry-Flynn & Champlin, 2018), that the questions asking about one’s attitude about the advertisement might have been prone to social desirability. Participants could possibly have been unconsciously influenced to respond to the questions in line with the current social trend, thus indicating a more positive attitude for the advertisements featuring an average sized model compared to those featuring a thin model. This could then have contributed to the significant difference found between the two model stimuli on advertising attitude. In real life, these same participants would not be prone to social desirability since no one is

(36)

monitoring their response to an ad or product, and thus might not actually prefer the ad featuring the average sized model.

Finally, the study measured intentions, but not behavior. Although participants might have indicated a higher intention to purchase the body scrub when exposed to the

advertisement featuring the averaged model, compared to the one featuring the thin model, they might not actually purchase the product, a phenomenon known as the intention-behavior gap (Sheeran & Webb, 2016). In the end the purpose of an advertising campaign is to

increase sales, however through this study we cannot conclude that the use of an average sized model would actually lead to an increase in sale, since only attitude and purchase intention was measured. To further develop the findings, actual purchasing behavior should be investigated as well. This could be done by placing a stand with the printed advertisement and the body scrub in a store and observing people’s purchasing behavior of the scrub.

Furthermore, it would be interesting to further investigate the impact of models’ body sizes on body-focused anxiety, as the present study’s lack of significant results was

unexpected. It would, for example, be fruitful to investigate underlying moderating and mediating processes between model’s body size and body-focused anxiety. The present study could be replicated while measuring social comparison processes to investigate whether the effect of model’s body size on body-focused anxiety and mood can indeed be ascribed to the (lack of) social comparison.

Conclusion

` The thin idealized female body that is communicated in the media sets largely

unattainable standards for women and consequently can have detrimental effects on women’s physical and mental health. The thin beauty norm has been shown countless times to have aversive effects on women’s body image and body-focused anxiety and has been associated

(37)

with disordered eating and depression. Although this study’s results indicated no difference in body-focused anxiety between women who were exposed to the thin model and the average sized model, this is not in line with the extensive amount of previous studies and warrants further research. While some companies have made the step to use more average sized models in their advertising campaigns (Bazilian, 2016), the majority continue to use thin models, with the idea that thin models will have a positive effect on their advertising effectiveness. The current study adds to the existing literature by showing that advertisements displaying average sized models are in fact more effective, as they result in a more positive attitude towards the advertisement and brand, as well as greater purchase intention of the product. Finally, the study attempted to close a gap in literature by investigating women’s subjective body size evaluation as a moderator on the aforementioned relationships, instead of objective measures of body size, such as BMI. Subjective body size evaluation only proved to be a moderator between the relationship between the model’s body size and mood, as women with a negative evaluation of their body displayed a more negative mood when exposed to a thin model compared to an average sized model, while women with a positive evaluation of their body had an overall more positive mood. The current study thus indicates that the use of averaged sized models would be most beneficial in avoiding causing a

negative mood in women with a negative subjective body size evaluation, as well as in creating greater advertising effectiveness. The old fashioned “thinness sells” argument is contradicted in this study as women appear to appreciate when advertisements defy old fashioned beauty norms, which positively spill over to the brand and purchase intention. We thus urge companies to use models with more realistic body sizes in order to increase sales while simultaneously combatting possible aversive effects on women’s mental health.

(38)

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211. doi: 10.1016/0749-5978(91)90020-T

Arlidge, J. (2000, June 25). Now bigger is the ideal figure. Observer, p. 15. London: Guardian Newspapers Ltd.

Andersen, K., & Paas, L. J. (2014). Extremely thin models in print ads: The dark sides. Journal of Marketing Communications, 20(6), 447-464. doi:

10.1080/13527266.2012.723027

Anschutz, D. J., Engels, R. C., Becker, E. S., & Van Strien, T. (2009). The effects of TV commercials using less thin models on young women's mood, body image and actual food intake. Body Image, 6(4), 270-276. doi: 10.1016/j.bodyim.2009.07.007

Baker, M. J., & Churchill Jr, G. A. (1977). The impact of physically attractive models on advertising evaluations. Journal of Marketing Research, 14(4), 538-555. doi:

10.1177/002224377701400411

Balachander, S., & Ghose, S. (2003). Reciprocal spillover effects: A strategic benefit of brand extensions. Journal of Marketing, 67(1), 4-13. doi: 10.1509/jmkg.67.1.4.18594 Bandura, A., & Walters, R. H. (1977). Social Learning Theory (Vol. 1). Englewood Cliffs,

NJ: Prentice-hall.

Bazilian, E. (2016, April 25). Why More Brands Are Embracing Plus-Size Models. Retrieved from https://www.adweek.com/brand-marketing/why-more-brands-are-embracing-plus-size-models-170984/

Borau, S., & Bonnefon, J. F. (2017). The advertising performance of non-ideal female models as a function of viewers' body mass index: A moderated mediation analysis of two competing affective pathways. International Journal of Advertising, 36(3), 457-476. doi: 10.1080/02650487.2015.1135773

Bower, A. B., & Landreth, S. (2001). Is beauty best? Highly versus normally attractive models in advertising. Journal of Advertising, 30(1), 1-12. doi:

10.1080/00913367.2001.10673627

Brausch, A. M., & Gutierrez, P. M. (2009). The role of body image and disordered eating as risk factors for depression and suicidal ideation in adolescents. Suicide and

Life-Threatening Behavior, 39(1), 58-71.

Brown, C., & Laland, K. N. (2002). Social learning of a novel avoidance task in the guppy: Conformity and social release. Animal Behaviour, 64(1), 41-47. doi:

10.1006/anbe.2002.3021

Clayton, R. B., Ridgway, J. L., & Hendrickse, J. (2017). Is plus size equal? The positive impact of average and plus-sized media fashion models on women’s cognitive resource allocation, social comparisons, and body satisfaction. Communication Monographs, 84(3), 406-422. doi: 10.1080/03637751.2017.1332770

Cash, T. F., & Deagle III, E. A. (1997). The nature and extent of body image disturbances in anorexia nervosa and bulimia nervosa: A meta‐analysis. International Journal of Eating Disorders, 22(2), 107-126. doi: 10.1002/(SICI)1098-108X(199709)22:2<107::AID-EAT1>3.0.CO;2-J

(39)

Diedrichs, P. C., & Lee, C. (2011). Waif goodbye! Average-size female models promote positive body image and appeal to consumers. Psychology & Health, 26(10), 1273-1291. doi: 10.1080/08870446.2010.515308

Dittmar, H., & Howard, S. (2004). Thin-ideal internalization and social comparison tendency as moderators of media models' impact on women's body-focused anxiety. Journal of Social and Clinical Psychology, 23(6), 768-791. doi: 10.1521/jscp.23.6.768.54799 Etikan, I., Musa, S. A., & Alkassim, R. S. (2016). Comparison of convenience sampling and

purposive sampling. American Journal of Theoretical and Applied Statistics, 5(1), 1-4. doi: 10.11648/j.ajtas.20160501.11

Festinger, L. (1954). A theory of social comparison processes. Human relations, 7(2), 117-140. doi: 10.1177/001872675400700202

Fitts, W. H., & Roid, G. H. (1964). Tennessee self concept scale. Nashville, TN: Counselor Recordings and Tests.

Furnham, A., Badmin, N., & Sneade, I. (2002). Body image dissatisfaction: Gender differences in eating attitudes, self-esteem, and reasons for exercise. The Journal of Psychology, 136(6), 581-596. doi: 10.1080/00223980209604820

Gecas, V. (1982). The self-concept. Annual Review of Sociology, 8(1), 1-33. doi: 10.1146/annurev.so.08.080182.000245

Gerbner, G., Gross, L., Morgan, M., Signorielli, N., & Shanahan, J. (2002). Growing up with television: Cultivation processes. In Media effects (pp. 53-78). Routledge.

Gillian, A. (2000, May 31). Skinny models ‘send unhealthy message’. Guardian (p. 7). London: Guardian Newspapers Ltd.

Grabe, S., Ward, L. M., & Hyde, J. S. (2008). The role of the media in body image concerns among women: A meta-analysis of experimental and correlational studies. Psychological Bulletin, 134(3), 460-476. doi: 10.1037/0033-2909.134.3.460

Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect of experimental presentation of thin media images on body satisfaction: A meta‐analytic review. International Journal of Eating Disorders, 31(1), 1-16. doi: 10.1002/eat.10005

Häfner, M., & Trampe, D. (2009). When thinking is beneficial and when it is not: The effects of thin and round advertising models. Journal of Consumer Psychology, 19(4), 619-628. doi: 10.1016/j.jcps.2009.06.004

Halliwell, E., & Dittmar, H. (2004). Does size matter? The impact of model's body size on women's body-focused anxiety and advertising effectiveness. Journal of Social and Clinical Psychology, 23(1), 104-122. doi: 10.1521/jscp.23.1.104.26989

Halliwell, E., Dittmar, H., & Howe, J. (2005). The impact of advertisements featuring ultra-thin or average size models on women with a history of eating disorders. Journal of Community & Applied Social Psychology, 15(5), 406-413. doi: 10.1002/casp.831 Hargreaves, D. A., & Tiggemann, M. (2004). Idealized media images and adolescent body

image: “Comparing” boys and girls. Body Image, 1(4), 351-361. doi: 10.1016/j.bodyim.2004.10.002

Heinberg, L. J., Thompson, J. K., & Stormer, S. (1995). Development and validation of the sociocultural attitudes towards appearance questionnaire. International Journal of Eating

Defying Beauty Norms: its Effect on Advertising Effectiveness and Women’s Mental Health The Effect of Advertising Models’ Body Size on Advertising Effectiveness and Women’s Body-Focused Anxiety and Mood