Literary Analysis paper

ORIGINAL ARTICLE

The Empowering (Super) Heroine? The Effects of Sexualized Female Characters in Superhero Films on Women

Hillary Pennell & Elizabeth Behm-Morawitz

Published online: 11 March 2015 # Springer Science+Business Media New York 2015

Abstract The present study was conducted with female un- dergraduates in the Midwestern region of U.S. to examine the potential positive and negative influences of the gendered de- pictions of women in superhero films. This study utilized so- cial cognitive and objectification theory frameworks to exper- imentally examine the short-term effects of exposure to sexu- alized female characters in superhero films on 83 female viewers’ gender role beliefs, body esteem, and self-objectifi- cation. Results show that exposure to the sexualized-victim images of women in superhero films decreased egalitarian gender role beliefs. Exposure to the sexualized-heroine im- ages resulted in lower body esteem. Additionally, a positive effect emerged with a greater belief in the importance of body competence to the self-concept for women who were exposed to the superheroine characters. This study demonstrates short- term effects from viewing sexualized images of women in superhero films and provides a significant understanding of how sexualized female representations may impact gender related beliefs as well as perceptions of one’s self-esteem and body objectification.

Keywords Superhero . Sexualization . Objectification .

Gender . Stereotyping

Introduction

This research explores the short-term effects of exposure to gendered representations in superhero films on female audi-

ences. This study was conducted with female undergraduates in the U.S. to examine the potential positive and negative influences of the representations of women in superhero films. According to the U.S. Movie Market Summary, in 2012, 1.3 billion tickets were purchased, and the U.S. film industry earned an impressive $10.8 billion. The popularity of film has held steady over the years, but certain genres of film have become more recognizable than ever. Action films, more spe- cifically comic book/superhero films, have become increas- ingly popular in the since the beginning of the new millenni- um (McCallister et al. 2006).

Although U.S. comic book-based films have greatly con- tributed to the U.S. film industry since the 1930s (McCallisteret al. 2006), with new technology, today’s comic book films have become Bblockbusters^ and sure successes for movie studios. These blockbuster comic book films are characterized by big action sequences coupled with computer generated image (CGI) special effects technology; they are fast-paced, entertaining, and have become cultural phenome- na. Comic book movies, specifically those films done by the comic power-house of Marvel Entertainment, the largest U.S. comic publisher, with approximately 37 % of the comic book market (McCallisteret al. 2006), have taken over the box of- fice with movies like Spider-Man (http://marvel.com/movies/ movie/6/spider-man), Blade (http://marvel.com/movies/ movie/10/blade), X-Men (http://marvel.com/movies/movie/8/ x-men), The Incredible Hulk (http://marvel.com/movies/ movie/23/the_incredible_hulk), and Iron-Man (http://marvel. com/movies/movie/19/iron_man). These films have become well-liked and successful due to the extensive fan base already created by the comic books themselves (Meslow 2012). They attract the youthful movie-going demographic, appeal to the nostalgic older audiences, and offer well-established charac- ters and storylines with sequel potential. They reach wide audiences across the world on a variety of cultural levels due to the films worldwide distribution (Richwine 2014).

H. Pennell (*): E. Behm-Morawitz Department of Communication, University of Missouri, 315 Switzler Hall, Columbia, MO 65211, USA e-mail: [email protected]

Sex Roles (2015) 72:211–220 DOI 10.1007/s11199-015-0455-3

Research suggests (all studies referenced in this article are based on U.S. samples unless otherwise noted) that cultural messages about gender within popular media can become normative and influence real-world perceptions. Media effects research provides evidence that entertainment media mes- sages impact individuals’ stereotypical gender role beliefs and expectations (Behm-Morawitz and Mastro 2008, 2009; Lafky and Duffy 1996; Signorielli 1989, 1993). Specifically, images and messages within these popular films can reinforce as well as create gender stereotypical expectations (Behm- Morawitz and Pennell 2013). Because images of women in these films are so limited in comparison to men (Carroll 1990), it is especially important to examine these character- izations that are highly memorable, in part due to their novelty. U.S. psychology research suggests that novel stimulus, like superheroines and their stories, interest audiences because of their newness (Musen and Treisman 1990). Thus it can be reasoned that superhero sagas have unique appeal in that they offer an escape into fantastical worlds rich with novel beings and places, that present audiences with stories that are cen- tered on stimuli unique from real life.

Also it is important to understand from where our concepts of gender and standards of physical appearance based on gen- der arise and critique the images and messages we receive through mediums like film. Gender is particularly important to consider in the present study because this can help explain an individuals’ ability to learn about gendered norms through observations of media models like film. Bussey and Bandura (1999) suggest that media character portrayals, for instance women in superhero films, may be used to help build an indi- vidual’s social standards about gender related-roles, attitudes, appearance and behaviors.

In the present research, an experimental study was conducted with an undergraduate female population to examine the potential positive and negative effects of the gendered depictions of women in superhero films. This study was guided by social cognitive (Bussey and Bandura 1999) and objectification theory frameworks (Fredrickson and Roberts 1997) and experimentally ex- amined the short-term effects of viewing sexualized im- ages of women in superhero films. It was predicted that viewing sexualized images of women in these films would result in less egalitarian gender role beliefs about women in the real world, lower body esteem, and pro- duce higher degrees of self-objectification.

Portrayals of Women in Action Films

U.S. research indicates that women in U.S. action films are often exploited, using only their sexuality as the focus of their characters; and, those characters are not offered complexity in the film’s storyline (Clover 1999; Mulvey 1999; Tasker 1993). Not unlike other media, film offers a narrow range of roles for

female characters and typically those roles focus on their sex appeal. Even in U.S. G-rated films, women are often hypersexualized with unrealistically small waists, unusually large chests, and unattainable hourglass figures (Herbozoet al. 2004). Women are usually the love interest of the hero and function as a spectacle, typically playing victims being rescued by the hero, therefore not narrating the plot (Clover 1999; Mulvey 1999; Tasker 1993).

Stabile (2009) specifically examined the superhero film genre. She observed that according to U.S. cultural norms, the role of someone is need of protection is invariably female or feminized. The active roles of the protector are masculin- ized, having strength and power as the foundation of mascu- linity; therefore, femininity is seen as vulnerable and weak. Consequently, superheroes are men because men are under- stood as protectors in our culture and women more often ap- pear as victims.

In particular, it is the sexualization through clothing and body type (Herbozo et al. 2004) and victimization through passivity and submissiveness (Clover 1999; Stabile 2009) of female characters in film that reflect traditional roles for wom- en. Because cognitions and behavior can be shaped through the observations of models presented by the media (Bussey and Bandura 1999) exposure to sexualized/victimized (i.e., damsel in distress) and sexualized/empowered (i.e., superher- oine) characterizations of women in superhero films would be expected to negatively influence short-term real world percep- tions of women’s traditional gender roles, body esteem, and self-objectification. However it can be argued that women in these superhero films can have positive, rather than negative effects on audiences. Many of the superheroines like X-Men’s Storm or Mystique can be viewed as having empowering characteristics. These characters are physically strong, athlet- ic, proficient, confident, and intelligent, characteristics that are counter-stereotypical for female characters (Behm-Morawitz and Pennell 2013; Stabile 2009). Although their sex appeal may be seen as objectification of the female body, it may also be constructed as empowering by showing distain for tradi- tional feminine modesty. Indeed, these characters represent a departure from the traditional role of the damsel in distress for women in superhero films. However, U.S. research (e.g., Behm-Morawitz and Mastro 2009) suggests that the sexualization of these characters serves to reinforce traditional gender role beliefs in female audience members, rather than challenge them. Although female participants may identify with and report feeling empowered by superheroines due to the female characters’ commanding cognitive and physical abilities (Behm-Morawitz and Pennell 2013), the sexualization of these female characters may supersede these empowering attributes and produce stereotypical gender- related perceptions. It is the dichotomous representation of superheroines as being extraordinary, yet sexualized that muddies the potential for these characters to be empowering

212 Sex Roles (2015) 72:211–220

versus stereotypically gendered (Behm-Morawitz and Pennell 2013).

Social Cognitive Theory, Gender, and Learning from Superhero Films

A better understanding of how individuals process film im- ages can be obtained by drawing upon social cognition. People often make judgment about gender, race, sexuality, and appearance through the use of cognitive mechanisms. Bussey and Bandura (1999) offer a framework based in social cognitive theory that helps understand how exposure to medi- ated models like those in superhero films can influence atti- tudes and beliefs specifically about gender. According to Bussey and Bandura (1999), media messages serve as a source for Bthe development of gender-linked knowledge and competencies^ (p. 686). Accordingly, sexualized por- trayals of women may be used to aid in developing an indi- vidual’s standards about gender-appropriate appearance, gender-based conduct, and normative gender roles.

U.S.-based research reveals that there is a significant relation- ship between media exposure to gender stereotypes and individ- uals’ stereotypical gender role beliefs and expectations (Lafky and Duffy 1996; Signorielli 1989, 1993). Studies have found that U.S. mainstream media, such as blockbuster superhero films, often hold on to traditional gender stereotypes in their presentation of women and men (Behm-Morawitz and Mastro 2008; Signorielli 1989). In a more recent study, Coyne et al. (2014) demonstrated that exposure to superheroes in television and a film was related to gender stereotypic play among chil- dren. This study provides empirical evidence between super- hero media depictions and gender stereotyping among viewers.

These gender stereotypes are usually defined as broad gen- eralizations about what it means to be male and female based on some knowledge of limited aspects of certain members of that sex. For women, they are stereotypically judged to be less intellectual, more emotional, and less aggressive than men; physically, women are expected to be youthful, beautiful, soft, and voluptuous (large chest and round buttocks) but thin (Deaux and Lewis 1983). In regard to role behaviors, women are supposed to be more nurturing and motherly. Finally, for women, their occupations are to hold jobs that involve care- taking, low status (e.g., secretarial positions), and jobs that do not involve manual labor. As further evidence from experi- mental work, video game play with sexualized heroines (sim- ilar in characterization to the superheroines of film) can unfa- vorably influence people’s beliefs about women in the real world (Behm-Morawitz and Mastro 2009).

While U.S. media effects research has demonstrated the short term effects of exposure to gendered depictions in media on individuals’ unfavorable beliefs and attitudes about people in the real world (Behm-Morawitz and Mastro 2009; Coyne et al. 2014; Signorielli 1989; &Signorielli 1993), media

effects research has also demonstrated recurrent or regular viewing can impact viewers’ perceptions. For example, a re- cent study has shown effects of regular viewing of MTV’s 16 and Pregnant on female adolescents’ perceptions (Aubrey et al. 2014). In this study, the effects of regular viewing on beliefs related to pregnancy, relationships, and parenting were measured. Aubrey et al. (2014) found that regular viewers of the show had more favorable attitudes toward teen pregnancy, perceived lower personal risk for getting pregnant, and accept- ed more myths about teen pregnancy than did their peers who were non-regular viewers of the series. Similarly, we might expect that regular viewing of superhero-based films, for the present study, may influence participants’ perceptions of gender-related beliefs. Although the primary aim of this study is to conduct a controlled test of the short-term effects of gendered representations of women in superhero films, it was also important to control for the potential relationship between regular viewing and the dependent variables.

Therefore, from the social cognitive theory of gender de- velopment and differentiation perspective, it is expected that viewing sexualized-victim and sexualized-heroine portrayals of women will result in increased belief in traditional, stereo- typical roles for women in the real world. It is expected that exposure to these sexualized representations of women will result in less favorable perceptions in comparison to women with no exposure to these representations. Thus the following hypothesis was predicted:

H1 Viewing the sexualized heroine condition and sexual- ized victim condition is associated with decreased egal- itarian gender role beliefs in comparison to the no stimuli condition.

The Sexual Objectification of Women in Superhero Films

Objectification theory offers a framework for understanding the consequences of being female in a culture that sexually objec- tifies women, the media being a wide-reaching source of con- structing this culture. Thus, objectification theory by Fredrickson and Roberts (1997) posits that the media encour- age women to self-objectify through the emphasis of the sexual objectification of women’s bodies in media such as magazines, television, and film, which associate the value of women large- ly with their appearance. Fredrickson and Roberts (1997), de- fine sexual objectification as Boccur[ing] whenever a woman’s body, body parts, or sexual functions are separated out from her person, reduced to the status of mere instruments, or regarded as if they were capable of representing her^ (p. 175). In the case of women in superhero films, the sexual objectification of the body is central to the identity of the victim and the superhero- ine. Many times when women are included in the superhero narrative, they tend to play victims in need of rescue by the

Sex Roles (2015) 72:211–220 213

male superheroes. Their abilities are stripped from them and their bodies are the primary focus. According to Behm- Morawitz and Pennell (2013), B…superhero stories take a fa- miliar gender narrative and transform it into a supergendered narrative, with hypermuscular male heroes and hypersexualized female victims, glorifying traditional gender roles^ (p. 75). The images of the female victim fit with stereo- typical conventions of femininity (i.e., weak and sexy) and serve to support normative messages of masculinity (i.e., in control and capable). For the superheroine, although they offer a much more physically capable picture of a woman in com- parison to the female victim, they are also presented as hypersexualized in order to balance their masculine traits of physical capabilities (Behm-Morawitz and Pennell 2013). The superheroine characters are often portrayed wearing tight, body-revealing clothing and use their sexuality to get what they want. Thus the superheroines can appeal to audiences in the way that they do not deviate too far from traditional gender norms by maintaining a focus on their body and sexuality.

U.S.-based research by Aubrey et al. (2009) showed that women who were exposed to images of female models with high skin exposure used more negative words to describe their own appearance than participants assigned to view control im- ages. In a similar U.S. study Aubrey et al. (2011) examined the effects of sexual objectification of female artists in U.S. music videos on male undergraduates’ sexual beliefs. The study re- vealed that participants who viewed music videos of highly objectified female artists reported more adversarial sexual be- liefs, acceptance of interpersonal violence, and negative atti- tudes about sexual harassment than participants assigned to low-sexual objectifying music videos by the same female artists.

Accordingly, our goal of this study is to extend our aware- ness and understanding of how visual representations of women’s bodies in superhero films may affect women’s levels of self-objectification. Our prediction is that superhero films’ objectification of the female body will result in women objec- tifying their own bodies by privileging their concern about how they look over the competence of their body. It is expect- ed that female participants who view the superhero films will report greater perceived importance of appearance and lower perceived importance of body competence (e.g., strength and health) to the self-concept in comparison to those who do not view the clips. Additionally, while self-objectification entails the viewing of the body for the pleasure of others, body es- teem relates to level of satisfaction with one’s own body (Franzoi and Shields 1984).

In terms of the media’s role in affecting female body esteem, much research has examined the effects of exposure to the thin (or curvaceously thin) female body ideal so prevalent in main- stream U.S. media. Developmental U.S. research suggests that for women, in particular, the image of the body often becomes very important to their identity (Ussher 1989). In other words, research suggests that the appearance and functionality/

competence of the female body becomes an important compo- nent of a female’s personal identity, wherein she may look to her body as being central to her self-worth and self-esteem. Moreover, research indicates media become a significant source of information about the normative standards of appear- ance of the female body and contribute to women’s feelings about their bodies (Murnen, et al. 2003; Ussher 1989) thus affecting their body esteem. The images of the female form portrayed by the media are used as a source of comparison, through which women and girls look to determine their own worth and men and boys may use to form expectations of females. The awareness of one’s own body in comparison to media images of the female body begins to be a source of self- concept struggle for females beginning during the adolescent period, and this struggle is thought to continue throughout life (Ussher 1989). U.S. empirical research has demonstrated that exposure to such idealized media images of females typically lowers girls’ and women’s esteem with their own bodies (Bissell and Zhou 2004; Murnen, et al. 2003; Ussher 1989).

Though these two concepts are distinct, the representations of sexualized and idealized images of the female body in su- perhero films would be expected to not only heighten self- objectification post viewing, but to also decrease body esteem among women. Thus, in line with social cognitive theory of gender development and differentiation, observation of these media models would be linked to problematic perceptions of the body, whether through others’ perceptions or one’s own perceptions. It is expected that exposure to sexualized repre- sentations of women in superhero films will result in less favorable perceptions of the body in comparison to women with no exposure to these representations. The following hy- potheses were predicted:

H2 Viewing the sexualized heroine condition and sexual- ized victim condition is associated with decreased body esteem in comparison to the no stimuli condition.

H3a Viewing the sexualized heroine condition and sexual- ized victim condition is associated with increased focus on the importance of appearance to one’s self-concept in comparison to the no stimuli condition.

H3b Viewing the sexualized heroine condition and sexual- ized victim condition is associated with decreased fo- cus on the importance of body competence to one’s self-concept in comparison to the no stimuli condition.

Method

Stimuli

The Spider-man and X-Men films were chosen as stimuli be- cause of exemplification of the sexualized female victim and

214 Sex Roles (2015) 72:211–220

sexualized superheroine roles, respectively. In the Spider-man series, the character Mary Jane is shown as weak and in need of rescue by her male counterpart. Conversely, in the X-Men series the female characters are strong, intelligent, and physi- cally capable. However, the female characters in both series are sexualized through their appearance (the characters are all thin, busty, and wear tight/revealing clothing). Thus the fe- male character in Spider-man represented the sexualized- victim stimuli and the female characters in X-Men represent the sexualized-heroine stimuli.

These films were also selected because of their popularity, adding to the external validity of the study design. The first X- Men film earned $296 million worldwide, X2 earned $408 mil- lion, and the third film in the series, X-Men: The Last Stand, earned an impressive $459 million. Spider-man, brought in $404 million in the United States alone and $822 million world- wide in 2002 (Spider-man 2013). On its opening day, Spider-man grossed more than $40 million dollars, breaking Hollywood’s previous record for Harry Potter and the Sorcerer’s Stone (Buncombe 2002). A little over a decade after the film’s debut, Sony Pictures rebooted the Spider-Man franchise by releasing the first installment of another trilogy entitled The Amazing Spider- man, which grossed more than $700 million worldwide (The Amazing Spider-Man 2012). Spider-man 3 brought in the highest grossing Marvel comic to date with an unbelievable $891 million worldwide (US Movie Market Summary for 2012 2013).

Each stimulus used in the experimental conditions contained similar plotlines. The plots centered around superheroes battling to overthrow a male villain and save innocent people. For ex- ample, the scenes from the Spider-man series, which represented the sexualized-victim condition, centered on Spider-man saving his female love interests Mary Jane and Gwen Stacy as well as ordinary citizens from a male mutant villain. Similar to the Spider-man stimuli, the X-Men plotline also centered on superheroes/superheroines battling to save innocent citizens from mutant villains. However, the X-Men stimuli, representing the sexualized-heroine condition, did have more female charac- ters than the Spider-man stimuli. The addition of more women in the sexualized-heroine stimuli was chosen in order to repre- sent female superheroines equal to that of male superhero char- acters. Thus the X-Men series was chosen due to the more bal- anced representation of both male and female superheroes.

Pretesting of Stimuli

Prior to the experiment, a pretest of the stimuli was conducted with undergraduate students outside of the experimental sam- ple (N=36) to test the validity of the sexualized female char- acter images. Of the participants in the Spider-man viewing condition (n=18), 40 % (n=6) were female and 60 % (n=9) were male. The average age of the participants in the Spider- man viewing condition was 20.00 years old (SD=1.46). In X- men viewing condition (n=18), 57 % (n=12) were female and

43 % (n=9) were male. The average age of the participants in the X-men viewing condition was 19.67 years old (SD=0.97).

Participants were assigned to watch a 13 min long montage of video clips of female characters from the film series Spider-man (sexualized-victim) or X-Men (sexualized-heroine). The sexualized-victim female character was shown with large breasts and a small waist; her skin was exposed in the stomach, breast, and thigh area accompanied by tight clothing. This female char- acter was also shown in harmful or dangerous situations where she was rescued by her male counterpart. The sexualized-heroine female characters similarly were shown with large breasts, small waists, and tight clothing. However, these characters were shown in situations where they were physically strong, intelligent, and capable compared to their male-counterparts. A montage of clips was used in order to include prototypical representations of wom- en in superhero films from multiple films rather than just one film.

After viewing film clips containing these characterizations, participants filled out a questionnaire asking about the female characterizations and their level of enjoyment of the clips. Overall, the sexualized-victim and sexualized-heroine video montages were rated to be similar in terms of enjoyment, F (1, 36)=2.69, p=0.11, and entertainment, F (1, 36)=0.03, p=0.88. The sexualized-victim female character was judged to be less aggressive, F (1, 36)=201.03, p<0.01; physically strong, F (1, 36)=175.52, p<0.01; violent, F (1, 36)=148.36, p<0.01; skilled, F (1, 36)=84.30, p<0.01; competent, F (1, 36)=23.97, p<0.01; and empowered, F (1, 36)=448.27, p<0.0, than the sexualized-heroines. Participants rated both types of sexualized characters similarly in terms of attractiveness, F (1, 36)=0.03, p=0.86, and degree of sexualization, F (1, 36)=0.04, p=0.85. Thus, both video montages were judged to be representative of the sexualized representations of women in superhero films. Analysis of variance tests were run and found participant gender had no significant effect on the ratings of these characterizations and there were no significant interactions between participant gender and conditions. See Table 1 for descriptive statistics.

Experiment

Participants

A total of 82 female undergraduate students from a large, U.S. Midwestern University participated. Participants were assigned to condition using the block randomization method to ensure the conditions were relatively equal in size. Of the participants, 35 % (n=29) participated in the control condi- tion, 37 % (n=30) in the sexualized-victim condition, and 28 % (n=23) in the sexualized-heroine condition. See Table 2 for participant descriptives. The participants were re- cruited from a basic communication course that draws majors from across the university. An all-female population was appro- priate for this study because research shows individuals typi- cally look to and identify with the same sex or those they

Sex Roles (2015) 72:211–220 215

Table 1 Descriptive Statistics for Pretesting of Stimuli

Variable N Female Participant Mean (SD)

Male Participant Mean (SD)

Overall Mean (SD)

Entertaining Spider-man 15 5.83 (1.17) 5.92 (0.669) 6.07 (0.884)

X-Men 21 6.22 (0.667) 6.22 (0.667) 6.05 (0.669)

Enjoyment Spider-man 15 6.33 (0.516) 6.00 (0.866) 6.13 (0.743)

X-Men 21 5.75 (0.996) 5.80 (0.833) 5.87 (0.913)

Attractiveness Spider-man 15 1.67 (0.516) 1.33 (0.500) 1.47 (0.516)

X-Men 21 1.42 (0.515) 1.60 (1.00) 1.52 (0.750)

*Physical Strength Spider-man 15 4.17 (0.753) 4.00 (1.00) 4.07a (0.884)

X-Men 21 1.08 (0.289) 1.22 (0.441) 1.14b (0.359)

*Aggressiveness Spider-man 15 4.00 (0.632) 4.11 (0.928) 4.07a (0.799)

X-Men 21 1.08 (0.289) 1.22 (0.441) 1.14b (0.359)

*Violence Spider-man 15 4.00 (1.10) 4.33 (0.707) 4.20a (0.862)

X-Men 21 1.33 (0.492) 1.33 (0.500) 1.33b (0.483)

Sexualization Spider-man 15 4.67 (0.516) 4.11 (1.05) 4.33 (0.900)

X-Men 21 4.36 (0.888) 4.36 (0.707) 4.33 (0.796)

*Skilled Spider-man 15 3.67 (0.816) 3.00 (0.707) 3.27a (0.799)

X-Men 21 1.17 (0.389) 1.44 (0.726) 1.29b (0.561)

*Empowerment Spider-man 15 4.83 (0.408) 4.56 (0.527) 4.67a (0.488)

X-Men 21 1.17 (0.389) 1.44 (0.527) 1.29b (0.463)

*Competence Spider-man 15 3.00 (0.894) 3.00 (0.707) 3.00a (0.756)

X-Men 21 1.58 (0.669) 1.56 (1.13) 1.57b (0.870)

*Values with differing subscripts were significantly different from one another at p<0.05. Entertainment and enjoyment were scored on a 1–7 scale, and all other variables were scored on a 1–5 scale. No significant differences emerged between the male and female participants ratings of the stimuli

Table 2 Participant Demographics Condition Race n (%) Age Range (M, SD) Covariates:

Action is Favorite Movie Genre n (%)

Prior Movie Exposure M (Median, SD)

Sexualized Victim (Spider-man)

White=19 (63) 17–24 (20.47, 1.36) Yes=5 (16) 10 (5, 18.23) Black=7 (23)

Asian=2 (7) No=25 (84) Latino=0 (0)

Other=2 (7)

Sexualized Heroine (X-Men)

White=17 (74) 18–23 (20.74, 1.39) Yes=1 (5) 9 (8, 4.87) Black=2 (9)

Asian=2 (9) No=22 (95) Latino=2 (9)

Other=0 (0)

Control (No stimuli) White=24 (83) 17–23 (20.62, 1.18) Yes=2 (7) 11 (6, 19.78) Black=3 (10)

Asian=1 (3) No=27 (93) Latino=0 (0)

Other=1 (3)

216 Sex Roles (2015) 72:211–220

perceive to be similar to themselves (Hoffner 1996). Thus be- cause this study focuses on the representation of women in the superhero film genre, women are more likely to look to these media models when making gender-related perceptions.

Procedure

Participants were assigned to one of three conditions: sexualized- victim condition (clips from the Spider-Man films), sexualized- heroine condition (clips from the X-Men films), and the control (no stimulus) condition. Participants were told that the study was inquiring about why people like and identify with superhero films. When assigned to the film viewing conditions, each par- ticipant watched a 13 min video montage of scenes featuring female characters from the Spider-Man series or the X-Men se- ries. After watching the video, participants were instructed to take a survey asking about traditional gender role stereotypes, body image, and self-appearance. Additional items about movie-going habits, enjoyment of the film clips, and non gender-related atti- tudes and beliefs scales were included to help disguise the pur- pose of the study. The participants assigned to the control condi- tion did not watch the film montage and simply were asked to complete the questionnaire.

Dependent Measures

To measure gender stereotyping (α=0.86, M=2.93, SD=0.72), an adaptation of the Attitudes toward Women Scale (AWS) created by Behm-Morawitz and Mastro (2009) was used. To construct their scale, Behm-Morawitz and Mastro used the AWS and additional items derived from existing gender- related scales (i.e., Personal Attributes Questionnaire, Sex-Role Egalitarianism Scale, and Gender Attitude Inventory) to create a stronger measure of gendered attitudes and beliefs. The four dimensions of the scale included: career and domestic labor, appearance, cognitive capabilities, and physical capabilities (Behm-Morawitz and Mastro 2009). Participants responded to statements about men’s and women’s responsibilities related to household and career duties, women’s attire and appearance in public, men’s and women’s rationality and problem-solving skills, and men’s and women’s physical strength and ability to handle physical challenges. The scales included items such as, BAwoman’s children should come before her career^, BMen and women should share household work equally ,̂ and BMen are better at taking on mental challenges than women^. The items were measured on a 7-point scale ranging from Bstrongly disagree^ (1) to Bstrongly agree^ (7), with low numbers dem- onstrating egalitarian attitudes toward women and high numbers demonstrating more traditional, less egalitarian attitudes toward women. One composite score, which included all four dimen- sions, was calculated for each participant.

Body esteem (α=0.91, M=4.69, SD=1.12) was measured using the Body Esteem Scale (Franzoi and Shields 1984). The

scale required participants to rate, from 1 (very dissatisfied) to 7 (very satisfied), their satisfaction with their general appear- ance and specific body parts (e.g., face, stomach, chest, and thighs). A sum score was created for each participant, with lower numbers indicating lower body satisfaction.

Self-objectification was measured using the Self- Objectification Questionnaire (Noll and Fredrickson 1998), which consists of body attributes that participants must rank in terms of their importance to their self-concept. Computing Chronbach’s alpha is not appropriate for this rank order scale, as the scale does not allow for participants to select consistent responses for like items. Participants are forced to assign only one value per item on the scale, such that they rank order the importance of the attributes to their self-concept. Six items relate to body appearance and six items relate to body competence. A sum score for appearance (Median=28.00, M=28.29, SD=7.56) and a sum score for competence (Median=28.00, M=27.75, SD=7.05) was created for each participant. Participant scores could range between 6 and 42 on each sub-scale. Lower scores indicate greater importance of attribute type (appearance and competence) to the participant’s self-concept. The calculation of a score for body appearance and a score for body competence (rather than one composite self-objectification score) allows for the examination of the potentially unique effects of images of women in superhero films – some of which portray a highly competent and strong female body – on participants’ view of the body as an object of appearance and of competence/health.

Covariates

Prior movie exposure (Median=6.0, SD=16.22) was used as a covariate in analyses comparing the effects of viewing condi- tion. Participants reported the estimated number of movies they view per month. See Table 2 for descriptive statistics.

Favorite action movie genre was also used as a covariate in analyses comparing the effects of viewing condition. Participants responded to a one item question. The item asked participants, BIs the superhero action genre of film your favor- ite film genre?^ Participants could respond (1) BYes^ or (2) BNo^. See Table 2 for descriptive statistics.

Results

Analysis of variance tests were used to examine differences between conditions. Movie viewing variables were included as covariates in the models. Specifically, participants were asked if the superhero action genre was their favorite genre of film. This was measured to understand participants’ prefer- ence for this film genre. Only 13 % (n=11) selected the su- perhero action genre as their favorite genre and 87 % (n=71) did not. Also participants were asked about their prior movie exposure to specifically superhero films. On average participants

Sex Roles (2015) 72:211–220 217

reported watching 10 superhero movies per month (Median=6, SD=16.22).

Hypothesis 1 predicted that exposure to the sexualized images of women in superhero films would be associated with decreased egalitarian gender role beliefs in comparison to those that do not view any stimuli (see Table 3 for statistics). An analysis of var- iance test, using condition as the independent variable, gender role beliefs as the dependent variable, and movie viewing as covariates revealed significant effects of condition on partici- pants’ gender role beliefs, F (2, 80)=3.43, p=0.04, η2=0.08. Specifically, viewing the clips featuring the sexualized-victim female character resulted in less egalitarian beliefs about women’s roles in society in comparison to the control condition (i.e., viewing no stimuli). However, no significant differences were found between viewing the sexualized-heroine female char- acters and the sexualized-victim characters or no stimuli. Results suggest that the sexualization of the female victim in superhero films may have different effects on gender-related beliefs than the depiction of the sexualized (super)heroine.

Next, hypothesis 2 predicted that exposure to the sexualized images of women in superhero films would be associated with decreased body esteem. An analysis of variance test, using con- dition as the independent variable, body esteem as the dependent variable, and movie viewing as covariates revealed the effect of condition on participant body esteem was significant at the p= 0.10 level, F (2, 80)=2.86, p=0.06, η2=0.07, thus the results were marginal at the p=0.05 level. However, no significant group differences were detected using the Tukey post hoc anal- ysis. To further explore possible effects, independent t tests were conducted. Viewing the sexualized-heroine clips predicted lower body esteem in comparison to viewing no stimuli, t (50)=−2.46, p=0.018. There were no significant differences detected when comparing the sexualized-victim condition with the control con- dition, t (57)=−1.85, p=0.069. Additionally, no differences were found between the sexualized-victim and the sexualized-heroine viewing groups, t (51)=0.45, p=0.652. Thus, results cautiously suggest that viewing superheroines in superhero films may, at least temporarily, lower body esteem for female viewers.

Hypothesis 3a predicted that viewing sexualized images of women in superhero films would be associated with increased perception of the importance of appearance to one’s self-con- cept. An analysis of variance test was run with condition as the independent variable, importance of appearance as the depen- dent variable, and the movie viewing variables as covariates. The test revealed no support was found for this hypothesis, F (2, 80)=2.15, p=0.12. Condition did not significantly predict importance of appearance to self-concept.

For hypothesis 3b, which predicted that viewing sexualized images of women in superhero films would be associated with decreased focus on the importance of body competence, a sig- nificant effect of condition emerged but not in the direction ex- pected. An analysis of variance test was run with condition as the independent variable, importance of body competence as the dependent variable, and the movie viewing variables as covari- ates. The test indicated that condition significantly predicted im- portance of body competence, F (2, 79)=4.28, p=0.02, η2=0.10, such that viewing the sexualized-heroine clips re- sulted in higher ranking of importance of body competence in comparison to viewing no stimuli. No significant differences were found between viewing the sexualized-victim female characters and the sexualized-heroine characters or no stimuli.

Discussion

This study examined the short-term effects of exposure to sexualized images of women in superhero films on women’s gender-related beliefs and perceptions of the body. The gender dynamics of character representations in superhero films gar- ner special attention, yet little U.S. research has empirically examined the influence of the superhero saga on gender- related cognitions and perceptions of the self. What makes superhero stories unique is their reliance on highly sexualized images of women as well as their contrasting representations of women as victims (i.e., damsels in distress) and heroes (i.e., superheroines). The results of this study demonstrate differen- tial effects of the portrayal of the sexualized victim and the sexualized heroine images of women in superhero films.

Specifically, viewing of the sexualized-victim female char- acters in superhero films resulted in less egalitarian beliefs about women’s roles in society in comparison to the control condition (i.e., no exposure to superhero films). Exposure to the sexualized-heroine characters, however, was not signifi- cantly linked to gender role beliefs. This suggests that the damsel in distress representation may influence female audi- ences to endorse more traditional gender beliefs. Additionally, though superheroines may be constructed as empowering, ex- posure to these powerful characters did not result in more egalitarian views of women in society. However, the empowering nature of these characterizations may have tem- pered the negative effects of the sexualization of these

Table 3 Descriptive Statistics

Condition Gender Roles M (SD)

Body Esteem M (SD)

Self-Objectification M (SD)

Body Comp Appearance

Sexualized Victim

3.12 (0.80) a 4.53 (1.27) b* 28.79 (5.95) 27.24 (8.11)

Sexualized Heroine

3.02 (0.58) 4.43 (1.02)c* 24.23 (7.78) d 31.23 (7.01)

Control 2.67 (0.69) a 5.08 (0.98) bc* 29.86 (6.71) d 26.79 (7.11)

Note. Mean scores were created for gender roles and body esteem mea- sures and were ranked on a 1–7 item scale. A sum score was created for self-objectification for each participant with possible scores that ranged from 11 to 77. Cells with the same subscripts are significantly different at p<0.05. *Cells are significantly different at p<0.10 level

218 Sex Roles (2015) 72:211–220

characters on women’s gender-related beliefs that were evi- dent with the sexualized-victim female characters.

The effects of the sexualization of women in superhero films on women’s body satisfaction and self-objectification were also examined. Results marginally suggest that viewing sexualized- heroine superhero characters may lower women’s body esteem. Here, the heroine rather than victim representation had an im- pact on level of body satisfaction, negatively impacting women’s perceptions of their bodies. Though both types of superhero film portrayals represent idealized images of the fe- male body, only the heroine depiction had an impact on body esteem. The power and status of the superheroine characters may illicit wishful identification among female viewers, in which case women may use these characters as a point of comparison for judging the self, a comparison that they feel they cannot live up to thus lowering one’s body esteem. Future studies should examine the role of character identifica- tion as well as the process of social comparison in the effects of superhero films on female audience members’ body esteem.

Viewing of sexualized superheroines was also associated with increased importance of body competence to one’s identity. In contrast to what was predicted, exposure to sexualized images of women in superhero films did not increase self-objectification. Past research suggested that the mere sexualization of these pow- erful characters would increase women’s tendency to view their bodies as objects of desire for others, with a heightened focus on the importance of physical appearance rather than competence (Aubrey, et al. 2009; Aubrey et al. 2011). Thus, a significant effect of condition emerged but not in the direction expected. Condition significantly predicted importance of body compe- tence such that the health and performance of one’s body was deemed to be more important after viewing the superheroines. Viewing the sexualized-heroine and the sexualized-victim por- trayals had no significant effect on perception of importance of physical appearance to the self. These results suggest that super- heroines in superhero sagas may indeed be empowering in en- couraging female audiences to focus on the strength and health of their bodies.

However, taking together the effects of the superheroine on body esteem and body competence, a contradictory picture emerges. On the one hand, the superheroine may have deleteri- ous effects on women’s body satisfaction. On the other hand, the superheroine may influence women to focus on body compe- tence rather than appearance. Future studies should work to fur- ther explore the relationship between body esteem and self- objectification in relation to the effects of the depiction of a sexualized female victim versus sexualized female hero on fe- male audiences. Examination of media identification and social comparison processes may help to unpack these complex relationships.

Additionally, one possible limitation of the present study is the use of video montages rather than full films for viewing conditions. Though this afforded the researchers the use of

multiple films for the experiment, the viewing may not entirely replicate the experience of watching a full superhero film. Future studies may want to consider screening full superhero films and examining effects. Also, the control condition could have used a superhero narrative that did not contain sexualized images of females as opposed to a no-stimuli control condition. This study only utilized a non-stimulus control condition as it was thought watching a superhero narrative might prime or conjure up im- ages of the superheroes’ female counterparts (which is typical in superhero narratives). However, future studies might utilize both a non-stimuli condition as well as additional control condition containing no images of sexualized females to explore potential differences in the effects. Furthermore, as previously mentioned, more work needs to be done in terms of understanding the cog- nitive and emotional processes through which these effects may occur. Social comparison, character identification, and emotional responses to superhero films may lend further insight into the effects of the sexualization of women in superhero films on women’s gender-related beliefs and perceptions of the body.

Conclusions and Further Considerations

Superhero texts do not just simply entertain, but have real world implications for how we construct our gender-related identities, attitudes, and beliefs. Female characters both resist and conform to familiar gender conventions. The results of this study illustrate the implications of this dichotomy. A po- tential negative influence of representations of women in su- perhero films emerged with exposure to sexualized female characters in these films resulting in more traditional gender stereotypes and lower body esteem for female participants. In contract, potential positive impacts were observed with expo- sure to sexualized-heroine characters increasing focus on body competence for women.

Future research should also consider psychological expla- nations for why viewing superhero films may impact women’s gender-related beliefs. We do not know from the present re- search if level of affinity for the characters in the superhero films, for example, influences subsequent effects of the films. It is also unknown whether audience investment would influ- ence the relationship between affinity and superhero film ef- fects, although, we might suspect that invested audiences would report greater affinity.

Lastly, although college students are an appropriate sample for the present study based on their life stage as well as media habits, the findings from this study might not be generalizable to other viewing populations. Similarly, this study was con- ducted in the context of U.S. culture, thus findings from this study cannot be generalized to other cultural populations. Because the target audience of the superhero genre is still men and boys, future studies should examine the unique in- fluence of superhero sagas on male audiences as well as on individuals from other cultures.

Sex Roles (2015) 72:211–220 219

References

Aubrey, J., Hensen, J., Hopper, K., & Smith, S. (2009). A picture is worth twenty words (about the self): Testing the priming influence of vi- sual sexual objectification on women’s self-objectification. Communication Research Reports, 26, 271–284. doi:10.1080/ 08824090903293551.

Aubrey, J., Hopper, K., & Mbure, W. (2011). Check that body! The effects of sexually objectifying music videos on college men’s sex- ual beliefs. Journal of Broadcasting and Electronic Media, 55, 360– 379. doi:10.1080/08838151.2011.597469.

Aubrey, J., Behm-Morawitz, E., & Kim, K. (2014). Understanding the ef- fects of MTV’s 16 & Pregnant on adolescent girls’ beliefs, attitudes and behavioral intentions toward teen pregnancy. Journal of Health Communication, 19, 1145–1160. doi:10.1080/10810730.2013.872721.

Behm-Morawitz, E., & Mastro, D. (2008). Mean girls? The influence of gender portrayals in teen movies on emerging adults’ gender-based attitudes and beliefs. Journalism and Mass Communication Quarterly, 85, 131–146. doi:10.1177/107769900808500109.

Behm-Morawitz, E., & Mastro, D. (2009). The effects of the sexualization of female video game characters on gender stereotyping and female self-concept. Sex Roles, 61, 808–823. doi: 10.1007/s11199-009-9683-8.

Behm-Morawitz, E., & Pennell, H. (2013). The effects of superhero sagas on our gendered selves. In R. Rosenberg (Ed.), Our superheroes, ourselves (pp. 73–93). New York, NY: Oxford University Press.

Bissell, K. L., & Zhou, P. (2004). Must-see TVor ESPN: Entertainment and sports media exposure and body-image distortion in college women. Journal of Communication, 54, 5–21. doi:10.1111/j.1460- 2466.2004.tb02610.x.

Buncombe, A. (2002). Spiderman swings its way to box office record. The Independent, p. 4.

Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychological Review, 106, 676– 713. doi:10.1037/0033-295X.106.4.676.

Carroll, N. (1990). The image of women in film: A defense of a paradigm. The Journal of Aesthetics and Art Criticism, 48, 349–360. doi:10. 2307/431572.

Clover, C. (1999). Her body, himself: Gender in the slasher film. In S. Thornham (Ed.), Feminist film theory: A reader (pp. 234–247). New York: New York University Press.

Coyne, S., Linder, J., Rasmussen, E., Nelson, D., & Collier, K. (2014). It’s a bird! It’s a plane! It’s a gender stereotype!: Longitudinal asso- ciations between superhero viewing and gender stereotyped play. Sex Roles, 70, 416–430. doi:10.1007/s11199-014-0374-8.

Deaux, K., & Lewis, L. (1983). Structure of gender-stereotypes: Interrelationships among components and gender label. Journal of Personality and Social Psychology, 46, 991–1004. doi:10.1037/ 0022-3514.46.5.991.

Franzoi, S. L., & Shields, S. A. (1984). The body-esteem scale: Multidimensional structure and sex differences in a college popula- tion. Journal of Personality Assessment, 48, 173–178. doi:10.1207/ s15327752jpa4802_12.

Fredrickson, B. L., & Roberts, T.-A. (1997). Objectification theory: Toward understanding women’s lived experiences and mental health risks. Psychology of Women Quarterly, 21, 173–206. doi:10.1111/j. 1471-6402.1997.tb00108.x.

Herbozo, S., Tantleff-Dunn, S., Gokee-Larose, J., & Thompson, J. K. (2004). Beauty and thinness messages in children’s media: A con- tent analysis. Eating Disorders, 12, 21–34. doi:10.1080/ 10640260490267742.

Hoffner, C. (1996). Children’s wishful identification and parasocial inter- action with favorite television characters. Journal of Broadcasting & Electronic Media, 40, 389–402. doi:10.1080/08838159609364360.

Lafky, S., & Duffy, M. (1996). Looking through gendered lenses: Female stereotyping in advertisements and gender role expectations. Journalism and Mass Communication Quarterly, 76, 379–388. doi:10.1177/107769909607300209.

McCallister, M. P., Gordon, I., & Jancovich, M. (2006). Blockbuster meets superhero comic, or art house meets graphic novel? The contradicto- ry relationship between film and comic art. Journal of Popular Film and Television, 34, 108–115. doi:10.3200/JPFT. 34.3.108-115.

Meslow, S. (2012). The revenge of the second-tier superhero film. The Atlantic. Retrieved from http://www.theatlantic.com/entertainment/ archive/2012/02/the-revenge-of-the-second-tier-superhero-film/253243/.

Mulvey, L. (1999). Visual pleasure and narrative cinema. In S. Thornham (Ed.), Feminist film theory: A reader (pp. 58–69). New York: New York University Press.

Murnen, S. K., Smolak, L., Mills, J. A., & Good, L. (2003). Thin, sexy women and strong, muscular men: Grade-school children’s re- sponses to objectified images of women and men. Sex Roles, 49, 427–437. doi:10.1023/A:1025868320206.

Musen, G., & Treisman, A. (1990). Implicit and explicit memory for visual patterns. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 127–137.

Noll, S. M., & Fredrickson, B. L. (1998). A mediational model linking self-objectification, body shame, and disordered eating. Psychology of Women Quarterly, 22, 623–636. doi:10.1111/j.1471-6402.1998. tb00181.x.

Richwine, L. (2014). X-Men sequel tops Avatar as Fox’s biggest global debut. Retrieved from http://www.reuters.com/article/2014/05/26/ us-boxoffice-idUSKBN0E61QZ201405.

Signorielli, N. (1989). Television and concepts about sex roles: Maintaining conventionality and the status quo. Sex Roles, 21, 337–356. doi:10.1007/BF00289596.

Signorielli, N. (1993). Television and adolscenet’s perceptions about work. Youth and Society, 24, 314–341. doi:10.1177/ 0044118X93024003004.

Spider-Man. (2013). The numbers: Box office data [Fact Sheet]. Retrieved from http://www.the-numbers.com/movies/2002/SPIDR.php.

Stabile, C. A. (2009). Sweetheart, this ain’t gender studies: Sexism and superheroes. Communication and Critical/Cultural Studies, 6, 86– 92. doi:10.1080/14791420802663686.

Tasker, Y. (1993). Action heroines in the 1980s: The limits of ‘musculinity’. In D. Morley (Ed.), Spectacular bodies: Gender, genre and the action cinema (pp. 132–152). London: Routledge. doi:10.4324/9780203221846.

The Amazing Spider-Man. (2012). The numbers: Box office data [Fact Sheet]. Retrieved from http://www.the-numbers.com/movies/2012/ SPID4.php.

US movie market summary for 2012. (2013). The numbers: Box office data [Fact Sheet]. Retrieved from http://www.the-numbers.com/ market/2012.php.

Ussher, J. M. (1989). They psychology of the female body. New York: Routledge.

220 Sex Roles (2015) 72:211–220