Abstract
Purpose
This paper aims to investigate cross-gender effects of gender stereotypes in advertising. More specifically, it proposes that the negative effects found in studies of women’s reactions to stereotyped female portrayals should hold across gender portrayal and target audience gender.
Design/methodology/approach
In two experimental studies, the effects of stereotyped portrayals (vs non-stereotyped portrayals) across gender are compared.
Findings
The results show that advertising portrayals of women and men have a presumed negative influence on others, leading to higher levels of ad reactance, which has a negative impact on brand-related effects across model and participant gender, and for gender stereotypes in terms of physical characteristics and roles.
Research limitations/implications
Whereas previous studies have focused on reactions of women to female stereotypes, the current paper suggests that women and men alike react negatively to stereotyped portrayals of other genders.
Practical implications
The results indicate that marketers can benefit from adapting a more mindful approach to the portrayals of gender used in advertising.
Originality/value
The addition of a cross-gender perspective to the literature on gender stereotypes in advertising is a key contribution to this literature.
Keywords
Citation
Åkestam, N., Rosengren, S., Dahlén, M., Liljedal, K.T. and Berg, H. (2021), "Gender stereotypes in advertising have negative cross-gender effects", European Journal of Marketing, Vol. 55 No. 13, pp. 63-93. https://doi.org/10.1108/EJM-02-2019-0125
Publisher
:Emerald Publishing Limited
Copyright © 2021, Nina Åkestam, Sara Rosengren, Micael Dahlén, Karina T. Liljedal and Hanna Berg.
License
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial & non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
Introduction
This research investigates consumer reactions to gender-stereotyped advertising portrayals. Differing from most studies in this field of research (Eisend, 2019), it specifically focuses on how consumers of a different gender than that of the portrayed models respond to gender-stereotyped advertising. By studying the cross-gender effects of gender-stereotyped portrayals in advertising (i.e. the effects on consumers of a different gender to the persons depicted in the advertising), this research highlights the need to consider reactions across gender, regardless of target audience. This investigation of cross-gender effects is warranted because the discourse on gender issues in advertising research has changed considerably in the past few years. The study of gender roles in advertising is evolving into a research field closely aligned with the current societal and cultural changes, where a clear distinction between gender and sex is prevalent (Eisend, 2019; Tuncay Zayer et al., 2020). Gender stereotypes, sometimes simply referred to as gender roles, have a long history in advertising, where they are still very frequently used (Eisend, 2010). In recent years, the potentially negative effects of using stereotypes in advertising have been highlighted by both advertising research (Eisend, 2019) and practice (Windels, 2016). This is evidenced, for example, by the growing industry use of so-called femvertising (Åkestam et al., 2017a) and dadvertising (Bukszpan, 2016), as well as in the increasing number of research articles exploring gender issues in advertising (Tuncay Zayer et al., 2020; Liljedal et al., 2020; Azar et al., 2018). There is also an increasing awareness in the advertising industry of the potentially negative effects that stereotyped gender portrayals can have. A 2017 report by the British Advertising Standards Authority (ASA) concluded that “gender stereotypes have the potential to cause harm by inviting assumptions about adults and children that might negatively restrict how they see themselves and how others see them” (Advertising Standards Authority, 2018, p. 3) and therefore called for regulations of the use of gender stereotypes in advertising. Based on the above reasoning, the purpose of the present research is to extend the research on gender-stereotyped portrayals in advertising to include a cross-gender perspective, thereby increasing the relevance of the advertising research literature in the contemporary development of how gender and stereotypes are considered in advertising research and practice.
Cross-gender reactions to gender-stereotyped portrayals in advertising should be of great interest to marketers. First of all, as the discourse concerning the concept of gender becomes more nuanced in many societies today, some leading advertising researchers, like Eisend (2019) argue that the gender concept in advertising research should follow suit. According to his criticism, research about gender stereotypes has often been limited to binary definitions of gender and to studying the effect of gender stereotypes on consumers of the same (binary) gender as the depicted advertising models. Eisend (ibid.) also notes that few studies have examined how men respond to non-stereotyped gender depictions in advertising.
Still, marketers wanting to reduce the use of gender stereotypes by exploring non-stereotyped gender portrayals (i.e. portrayals showing a person in a way that does not correspond with the gender stereotype) have little guidance in terms of what to expect from such efforts. For example, there is considerably more research about how women react negatively to stereotyped portrayals of women in advertising and the media (Åkestam et al., 2017a; Bower, 2001; Halliwell and Dittmar; 2004; Janssen and Paas, 2014), than about how men react to stereotyped portrayals of men in advertising and the media (Tuncay Zayer et al., 2020; Ricciardelli et al., 2010; Branchik and Chowdhury, 2012). Since targeted ads are by no means safe from further dissemination (for example through social media), ads targeting one gender will likely end up also being displayed to other genders (Huhmann and Limbu, 2016). In most cases, cross-gender targets will even be a desired audience, since customers of different genders will likely be interested in the same products (either buying for themselves or as a gift) and constitute an exposure audience (e.g. for giving advice or forming ideals). As such, marketers urgently need guidance on which cross-gender effects to expect from using stereotyped versus non-stereotyped gender portrayals in advertising. Advertising research, however, does not yet have much to offer in this particular field (Gulas and McKeage, 2000, for an exception).
At a societal level, advertising is part of the cultural expressions that reflect and impact on how consumers perceive themselves and others (Lorenzen et al., 2004; Palan, 2001; Rutledge Shields and Heinecken, 2001; Steele et al., 2002). Both advertising practitioners and researchers should thus be aware of and understand the impact of the visual displays of advertising. If we ignore the effects of gender-stereotyped advertising across the population regardless of gender, we also limit our understanding of our present society and how advertisers can take a more responsible part in it. Consequently, by examining the cross-gender effects of gender-stereotyped portrayals, the present research offers a more holistic understanding of gender stereotypes in advertising. Building on research showing that women’s reactions to stereotyped portrayals of women in advertising are influenced by the presumed influence of the portrayals on others (Åkestam, 2018; Choi et al., 2008; Wan et al., 2003) and psychological reactance to the advertising (Åkestam et al., 2017a), this research shows that the negative effects found in studies of women’s reactions to female portrayals hold across gender portrayals and target audiences. More specifically, given the social nature of advertising exposure, and that consumers view advertising portrayals as self-relevant (Dahlén et al., 2014), we argue that both women and men will react negatively to stereotyped portrayals of a gender other than their own. A wider contribution of this research is demonstrating the need for research to consider the effects of gender-stereotyped portrayals across genders if it is to continue the important and ongoing investigation in advertising research of consumer reactions to stereotyped portrayals of gender in advertising. More broadly speaking, this research also highlights the social effects of advertising, because consumers react to portrayals of people even though they may be dissimilar to the consumers themselves (Häfner, 2003).
Gender stereotypes in advertising
Stereotypes are generally defined as sets of beliefs about the characteristics, attributes, and behaviors of members of certain groups (Ashmore and Del Boca, 1981). Gender stereotypes are consequently based on the beliefs that certain characteristics, attributes and behaviors differentiate the different genders (Eisend, 2010). Historically, most research on gender roles in advertising has focused on a sex-based and binary (male/female) concept of gender and while this approach does reflect the stereotyped portrayals of women and men so often found in advertising, it also fails to encompass the true diversity of gender identities (Eisend, 2019). A binary concept of gender is also central to the gender stereotypes found in advertising, which are characterized by portrayals of women and men (and those two genders only) that conform to the stereotypes for their respective genders (Grau and Zotos, 2016). Gender stereotypes are characterized in four dimensions: gender stereotypes concerning physical characteristics, role behaviors, occupational status and personality traits (Deaux and Lewis, 1984). Non-stereotyped gender portrayals in advertising thus differ from stereotyped portrayals by not portraying gender-stereotyped physical characteristics, role behaviors, occupational status, or personality traits (e.g. by showing less thin women or less muscular men). In advertising, gender stereotypes are often portrayed as a combination of different aspects of the stereotype, for example in an ad where the physical characteristics of a model aligns with the gender stereotype in terms of body size, pose and facial expression (Åkestam, 2018; Deaux and Lewis, 1984). Conversely, non-stereotyped portrayals of gender in advertising should avoid portraying several aspects of a gender stereotype in order to challenge it (Åkestam et al., 2017a).
Like other stereotypes, gender stereotypes function as a type of heuristics, in that they facilitate information processing and thus require less cognitive effort to process (Pratto and Bargh, 1991). This, however, has to be weighed against the negative effects of exposure to gender stereotypes. During the past decades, many research studies have been able to demonstrate negative effects of social comparison (Festinger, 1954) with decorative models in advertising on the appearance satisfaction and body-focused anxiety of women and girls (Halliwell and Dittmar, 2004; Richins, 1991; Lavine et al., 1999; Martin and Gentry, 1997; Hogg et al., 1999). While stereotyped depictions of women tend to portray slim women, the stereotyped depictions of men rather emphasize athleticism and muscularity, but exposure to stereotyped depictions of men has similarly negative effects on men’s body satisfaction, body esteem, and self–esteem (Lorenzen et al., 2004; Barlett et al., 2008). Another negative effect of using gender stereotypes in advertising is the career disadvantages of not considering some occupations due to the prolonged exposure to occupational gender role stereotypes (Eisend, 2010). A number of studies have also showed positive effects of using non-stereotyped depictions of women in advertisements (Liljedal et al., 2020; Bian and Wang, 2015; Janssen and Paas, 2014; Antioco et al., 2012; Richins, 1991).
Although much of the literature to date has focused on women’s reactions toward gender stereotypes, the literature also includes negative effects for men, where for example gender role attitudes are seen as dynamic entities that can be at least momentarily shifted due to gender portrayals in advertising (Garst and Bodenhausen, 1997). However, there is still a perception among advertising practitioners that men are immune to the negative consequences of gender stereotypes in advertising (Tuncay Zayer and Coleman, 2015). This reasoning is in line with objectification theory (Fredrickson and Roberts, 1997), which focuses on describing the effects of sociocultural contexts where women experience sexual objectification, especially in terms of the resulting self-objectification in women. In contradiction to this, however, exposure to sexualized advertising can lead to increased self-objectification in both men and women (Karsay et al., 2018).
Like all stereotypes, gender stereotypes are also culturally dependent and change over time, meaning that responses to gender stereotypes in advertising can vary between different cultures and across different periods of time (Eisend, 2010, 2019). Stereotyped portrayals of women and men have a long history in advertising and are still very frequently used in advertisements (Grau and Zotos, 2016; Eisend, 2019; Knoll et al., 2011). Interestingly, stereotyped images of men predominantly focused on facial photos (and sports action shots) until the 1990s when the male body became more prevalent in line with other culturally normative and influential ideals of masculinity (i.e. hegemonic masculinity, Connell, 1987) (Tuncay Zayer et al., 2020; Ricciardelli et al., 2010; Davis, 2002). A recent study among advertising practitioners by Windels (2016) found that advertisers are aware of using gender stereotypes (although reluctant to admit this), and view them as useful tools, that have quick effects and are at least partly based on truths, in that they represent the societal view of gender roles. Furthermore, the advertising practitioners perceived that consumers find stereotypes attractive, and that stereotypes communicate quickly and simplify ad processing, which in turn helps prevent distraction and encourages emotional rather than rational responses. There is also some evidence that gender-stereotyped advertisements for gendered products may have better effects on consumers with a traditional gender role congruence (“masculine men” and “feminine women”) than for consumers with a more non-traditional gender role congruence (Morrison and Shaffer, 2003).
The question is, however, whether the advertising industry’s use of gender stereotypes in advertising is about to change. There are many signs that both marketers and consumers are moving away from gender-stereotyped depictions, such as the inclusion of more plus-size models and non-retouched images in advertisements (for brands such as Aerie), and that social media now makes it possible for consumers to voice their dissatisfaction with gender-stereotyped depictions in public (Pounders, 2018). Gender stereotypes will likely be affected by changing societal views of the gender concept, including the increased understanding of concepts such as gender fluidity and the limitations of using binary (male/female) definitions of gender (Eisend, 2019). As mentioned, stereotypes are culturally and temporally contingent, and as time passes and cultures change, so will the gender stereotypes.
The research literature on gender stereotypes in advertising is scattered across different academic fields such as marketing/advertising (Tuncay Zayer et al., 2020; Huhmann and Limbu, 2016; Grau and Zotos, 2016; Antioco et al., 2012; Wolin, 2003), psychology (Lockwood and Kunda, 1997; Lavine et al., 1999; Halliwell and Dittmar, 2004; Zawisza and Cinnirella, 2010; Meyers-Levy and Loken, 2015), consumer research (Richins, 1991) and gender studies (Kaufman, 1999; Morrison and Shaffer, 2003; Lorenzen et al., 2004; White and White, 2006; Ricciardelli et al., 2010). It also suffers from a lack of common terminology. While studies of gender role behaviors, occupational status, and personality traits tend to explicitly use a stereotype framework (Bakir et al., 2008; Eisend, 2010; Knoll et al., 2011), studies of physical characteristics tend to apply other concepts, such as idealization (Richins, 1991), body-size (Bian and Wang, 2015), and attractiveness (Bower, 2001). For portrayals of men in media and advertising, masculine hegemony is also used similarly to the stereotype framework, as it can be found in male roles, status set, perspective, behavior and personal characteristic (Ricciardelli et al., 2010; Connell, 1987). For the purpose of the present research, the literature on the physical characteristics of advertising models will be interpreted using a stereotype framework (Eisend, 2010). Table 1 provides an overview of previous experiment studies of gender stereotypes in advertising since 1990.
As seen in Table 1, most of the studies have investigated women’s reactions to female gender stereotypes and the effects found have predominantly been negative. What is more, most studies have investigated the social and brand-related effects for same-gender consumers only. Cross-gender reactions to stereotypes have thus been largely overlooked in the existing literature.
Of interest in this research is cross-gender reactions, meaning how for example consumers who identify as men react to portrayals of women, or how consumers who identify as women react to portrayals of men. Although women are often thought of as the only target audience of ads featuring women (and likewise for other genders), this may not always be the case. Consumers of other genders may also be appropriate targets of the advertising in their role as consumers, purchasers of gifts, or as influencers in the purchase process. For example, retailers have reported that half of their men’s underwear are in fact bought by women (Ferrier, 2018). What is more, advertising often reaches consumers who are not the intended target audience (cf. Huhmann and Limbu, 2016; Dahlén et al., 2013). In line with the reasoning of Åkestam et al. (2017a) and Dahlén et al. (2014), consumers of any gender may thereby use advertising in forming perceptions of themselves and others. Understanding the reactions of men to stereotyped portrayals of women and of women to stereotyped portrayals of men is thus of great importance.
The presumed influence of gender stereotypes
The present research hypothesizes that the negative reactions to stereotyped portrayals of women found in previous research on female consumers will hold across gender for both female and male consumers (i.e. for male consumers exposed to stereotyped portrayals of women and for female consumers exposed to stereotyped portrayals of men). This is in line with recent research demonstrating that consumers regardless of gender respond more positively toward non-stereotyped gender portrayals in advertising (Liljedal et al., 2020; Baxter et al., 2016). The proposed theoretical model is summarized in Figure 1 and is further discussed in the following sections.
Previous research shows that one reason why women react negatively to stereotyped advertising portrayals of women is that they believe that others are negatively affected by such portrayals (Åkestam, 2018; Choi et al., 2008; Milkie, 1999; Wan et al., 2003). This influence of presumed influence (IPI) of gender stereotypes (Gunther and Storey, 2003), adds to a compelling body of research showing that the expected impact of media or advertising on other people mediates consumer reactions to a wide range of stimuli (Eisend, 2017; Sun et al., 2008). A common explanation for this mechanism is motivational: people tend to feel better about themselves when they perceive others to be more susceptible to potentially harmful persuasion attempts, such as advertising for alcohol (Noguti and Russell, 2014), gambling (Youn et al., 2000), and offensive advertising messages (Dahlén et al., 2013). Women presume female gender-stereotyped portrayals to constitute such harmful persuasion attempts (Choi et al., 2008) and expect “others” to be more influenced by exposure to stereotyped portrayals of women than they themselves are (Choi et al., 2008; Milkie, 1999; Wan et al., 2003). This suggest that a male (female) consumer viewing a female (male) portrayal in advertising would perceive women (men) to be affected by it and thus we would expect the effects of IPI found in previous studies (Åkestam, 2018) to hold up across gender. It is therefore hypothesized:
Men presume stereotyped (vs non-stereotyped) portrayals of women in advertising to generate higher (vs lower) levels of negative influence on women.
Women presume stereotyped (vs non-stereotyped) portrayals of men in advertising to generate higher (vs lower) levels of negative influence on men.
Another reason that women react negatively to stereotyped female portrayals, is that the narrow set of gender stereotypes frequently used to portray women in advertising (Eisend, 2010; Zimmerman and Dahlberg, 2008) puts pressure on them to behave in a manner that is consistent with these stereotypes, which in turn leads to psychological reactance (Åkestam et al., 2017a). The same effect could be expected for portrayals of other genders, as indicated by recent research on how male informants (de)legitimize messages of masculinity in advertising in different ways (Tuncay Zayer et al., 2020).
The theory of psychological reactance (Brehm, 1966; Thorbjørnsen and Dahlén, 2011) proposes that people have a pre-disposition to preserve and restore their personal freedom. When personal freedom is reduced, eliminated, or threatened with elimination, people will experience a state of arousal (reactance) that induces attempts to recover or re-establish the lost or threatened behavior. These attempts can include opposing or protesting the threat or turning against the source of the threat. For gender stereotypes, the influence that the stereotyped portrayal is expected to have on others (H1) would also lead to psychological reactance in consumers of another gender than that portrayed in the ad. If consumers assume that gender-stereotyped portrayals have a negative influence on others, so that “others” are more likely to a) conform to the stereotype, and b) treat others in line with the stereotype, this will affect the reactions of those consumers as well, regardless of their gender.
Non-stereotyped portrayals, however, move away from simplistic gender stereotypes toward more complex and varied portrayals (Åkestam et al., 2017a). This puts less strain on consumers of all genders to comply with a specific stereotype, and thus leads to more possibilities to relate to the portrayals used. As non-stereotyped portrayals are more open to consumers creatively deconstructing meanings (Puntoni et al., 2010), reactance would be less likely to occur in reaction to for example a woman (vs a man) being portrayed as strong. Overall, this suggests that stereotyped portrayals of a gender that is different from the consumer’s own gender will be more likely to lead to ad reactance than non-stereotyped portrayals. It is thus hypothesized:
Stereotyped (vs non-stereotyped) portrayals of women in advertising generate higher (vs lower) levels of ad reactance for men.
Stereotyped (vs non-stereotyped) portrayals of men in advertising generate higher (vs lower) levels of ad reactance for women.
When reactance occurs, people are motivated to recover the lost or threatened behavior (Brehm, 1966; Thorbjørnsen and Dahlén, 2011). A person experiencing reactance thus becomes more likely to resist persuasion. In an advertising context, this would, in turn, lead to lower levels of brand-related effects, as consumers defend themselves by concluding that “I’m not wrong, they are” (Obermiller et al., 2005). As argued above, both women and men are expected to experience reactance after exposure to gender-stereotyped advertising portrayals of a gender other than their own. This would, in turn, lead to lower brand-related effects mediated by presumed negative influence on others and reactance. It is thus hypothesized:
For men, stereotyped (vs non-stereotyped) portrayals of women in advertising generate lower (vs higher) levels of brand-related effects in terms of a) ad attitudes, b) brand attitudes and c) purchase intentions.
For women, stereotyped (vs non-stereotyped) portrayals of men in advertising generate lower (vs higher) levels of brand-related effects in terms of a) ad attitudes, b) brand attitudes and c) purchase intentions.
The above hypotheses can also be described as a chain of effects. Stereotyped gender portrayals in advertising are proposed to make consumers of other genders react negatively compared to non-stereotyped portrayals; first by creating the assumption that others are negatively affected by such portrayals (the influence of presumed influence; Åkestam, 2018; Gunther and Storey, 2003), which in turn is believed to create a negative reaction among the viewers of the ads (psychological reactance; Brehm, 1966; Thorbjørnsen and Dahlén, 2011, Åkestam et al., 2017a), which then finally is believed to spill over on to the ad and brand evaluations and intentions of these consumers, such as ad attitudes, brand attitudes, and purchase intentions.
Taken together, the underlying theoretical explanation for the above proposed effects can thus be described in terms of a serial mediation, where the effects of gender-stereotyped portrayals on ad attitudes, brand attitudes and purchase intentions (H3) are mediated by the presumed negative influence on women/men (H1) and ad reactance (H2). Stated more formally:
For men, presumed negative influence on women and ad reactance will mediate the negative effects of gender-stereotyped portrayals of women on a) ad attitudes, b) brand attitudes and c) purchase intentions.
For women, presumed negative influence on men and ad reactance will mediate the negative effects of gender-stereotyped portrayals of men on a) ad attitudes, b) brand attitudes and c) purchase intentions.
Study 1
Study 1 tested the hypotheses using portrayals of gender stereotypes in terms of physical characteristics in a 2 (stereotyped vs non-stereotyped portrayal) × 2 (female vs male model gender) × 2 (female vs male participant gender) between-subjects, full factorial experiment. The stereotype component of physical characteristics (Deaux and Lewis, 1984) was selected as it is frequently used in advertising featuring both stereotyped and non-stereotyped portrayals of gender (Sobande et al., 2020; Bian and Wang, 2015; Janssen and Paas, 2014; Antioco et al., 2012; Richins, 1991), and should therefore promote ecological validity in all experimental conditions.
Stimulus development
Four print ads for underwear brands were used as stimuli. Underwear ads naturally show the bodies of models, making them suitable for studying gender stereotypes in terms of physical characteristics, without potentially confounding elements of sexualization that can appear if bodies are exposed in other contexts (Mayer and Peev, 2017). Using gendered products, such as underwear marketed toward men and women ensures that the product is perceived as gendered in line with the model’s perceived gender (i.e. reducing the risk of potentially confounding effects). Many products still tend to be perceived as gendered by consumers, whether they are specifically marketed as gendered or not. For example, beer, athletic shoes and dog food are perceived as masculine products and wine, facial tissue and cat food as feminine products (Fugate and Phillips, 2010; Morrison and Shaffer, 2003). For all conditions, stereotypicality was operationalized in line with so called femvertising advertisements that challenge several aspects of female portrayals simultaneously (Åkestam et al., 2017a). In other words, several different elements of gender stereotypes in terms of physical characteristics often found in advertising were included to create the stereotyped ads. In the present context, the gender stereotypes (or lack thereof in the non-stereotyped ads) are thus not operationalized in terms of one single aspect such as body-size, because changing the size of the model per se does not automatically change a stereotyped ad portrayal into a non-stereotyped ad portrayal. For this particular study, portrayals predominantly differed with regards to body-size (Bian and Wang, 2015; Åkestam et al., 2017a), pose (Jones, 1991; cf. Wiles et al., 1995), and mood (Åkestam et al., 2017a). For the stereotyped female portrayal condition, a real ad featuring a slim woman in underwear posing on a bed, with the message “Reclaim Your Shape” was used. For the non-stereotyped female portrayal ad, the image in the ad was replaced by an image from another real campaign, featuring another (less slim) woman. In terms of posing, the stereotyped female portrayal depicts a woman standing on all fours in bed, in line with Jones’ (1991) summary of subtle stereotyping in advertising, as opposed to the non-stereotyped portrayal, which shows a woman sitting in a relaxed position in a chair. For the male portrayal condition, the advertisement image was replaced by a slim, athletic man in a muscle pose (stereotyped condition), and a heavier, less athletic man in a more relaxed pose (non-stereotyped condition). This manipulation is in line with research showing that stereotyped depictions of men tend to emphasize muscularity (Barlett et al., 2008). The ad portrayals also differed in terms of other aspects related to gender stereotypes in terms of physical characteristics, such as grooming, make-up and photo retouching. For the male portrayal condition, the brand used in the female portrayal ads (Triumph) was replaced by a men’s underwear brand (Jockey). This rendered four ads that were similar in terms of message and layout, but different in terms of the gender-stereotyped depiction of physical characteristics and the model gender. In line with previous research that has advocated that real ads enhance external validity (Fox et al., 1998; Morgan and Reichert, 1999; Moorman et al., 2002; Feiereisen et al., 2009), all images used were sourced from real advertising campaigns that had not run on the market of the study specifically. A pre-test (n = 275, 60% woman, mean age= 30.55; participants similar to, but not included in the main study) showed that the female stereotyped ad was perceived as significantly more stereotyped than the non-stereotyped ad (Mstereotyped = 6.69; Mnon-stereotyped = 3.20 vs, p < 0.01, items: “The ad is stereotyped” and “The ad shows a stereotyped image of women/men”, measured on seven-point Likert scales, Cronbach’s alpha = 0.952), as was the male stereotyped ad (Mstereotyped = 5.95; Mnon-stereotyped = 2.23 vs, p < 0.01). This suggests, in line with the findings of Åkestam et al. (2017a), that consumers view certain portrayals of models’ physical characteristics (such as body weight and pose) as stereotyped. It is to be noted that the overall objective in selecting these ads was to create a set of stereotyped and non-stereotyped ads for portrayals of both women and men. These ads thus differ somewhat since the beauty ideals differ between men and women, for example in terms of body-size and pose.
Procedure
Two hundred and fifty-four people (age 18–79, mean age = 43.7) participated in the study. Participants were asked which gender they perceived themselves to belong to with the options “Woman”, “Man”, and “Other”. In all, 51% responded woman, 49% man and 0% other. The participants were randomly allocated to experiment groups and each participant answered one of the four versions of the questionnaire. Of the participants identifying as women, 32 viewed the male non-stereotyped ad, 30 viewed the male stereotyped ad, 33 viewed the female non-stereotyped ad, and 32 viewed the female stereotyped ad. Among those participants who identify as men, 29 viewed the female non-stereotyped ad, 30 viewed the female stereotyped ad, 32 viewed the male non-stereotyped ad, and 36 viewed the male stereotyped ad.
All participants were recruited by intercept at major train stations and in shopping malls in four Swedish large and midsized cities during the end of 2017. This method for recruiting participants was employed in order to achieve a heterogeneous sample that should reflect a general consumer audience in terms of demographic variables, such as age, gender, and ethnicity (Dahlén et al., 2009). Participants were told that researchers interested in a novel advertising concept created by a company conducted the study. This information was the same for all conditions. Each participant was randomly assigned a booklet with one stimulus ad, followed by a questionnaire featuring several filler tasks, as well as measures of the dependent variables. Participants were instructed to look at the ad for as long as they wanted, and then fill out the questionnaire (as recommended by e.g. Dahlén et al., 2008; Moorman et al., 2002). All questionnaires were printed out on paper and filled in by the respondents using a pen while the field worker was present nearby, thereby ensuring that all distributed questionnaires were also returned. All field workers (predominantly women, mean age 25 years old) had been trained before the data collection started. After the questionnaires had been collected, participants were debriefed and dismissed. Participants were not compensated for their participation.
Measures
As the stimuli featured real brands, brand familiarity was assessed before exposure to the ad by asking: “What is your current relation to brand X?”, followed by three semantic differential scales (don’t know at all/know very well, not familiar with/very familiar with, have no prior experience with/have extensive prior experience with, Cronbach’s alpha = 0.929).
Following brand familiarity, the three dependent variables followed. Ad attitudes were measured using the items bad/good, dislike/like, negative opinion/positive opinion, rated on seven-point semantic differential scales, in response to the question “What is your opinion on the ad?” (Dahlén et al., 2009, Cronbach’s alpha = 0.985). Brand attitudes were measured on identical scales in response to the question “What is your opinion on the brand?” (Cronbach’s alpha = 0.988).
Purchase intentions were measured using the items not at all probable/very probable and not at all likely/very likely, rated on seven-point semantic differential scales, in response to the question “If you were to buy underwear as a gift for someone else, how likely is it that you would choose something from Brand X?” (Dahlén et al., 2008; Conbach’s alpha = 0.988).
Mediating variables followed with presumed negative influence on women/men was measured on seven-point Likert scales with the item “I believe that other women/men are negatively affected by this ad” (Eisend, 2008, 2015; Gunther and Storey, 2003), with the gender in the question adapted to fit with model gender.
Ad reactance was measured on seven-point Likert scales with four items: “The ad makes me want to be the exact opposite”, “I do not approve of how the ad tries to affect me”, “The ad portrays an ideal that annoys me” and “The message in this ad limits my freedom of choice” (Cronbach’s alpha = 0.869). This measure was adapted from Hong (1992) to track situational reactance caused by the ad (Fitzimons and Lehmann, 2004; Thorbjørnsen and Dahlén, 2011).
Finally, the questionnaire included the two-item measure of perceived stereotypicality used in the pre-test as a manipulation check (Cronbach’s alpha = 0.958).
Results
Manipulation checks confirmed that participants perceived the stereotyped ads as significantly more stereotyped than the non-stereotyped ad (female model: Mnon-stereotyped = 2.82 vs Mstereotyped = 5.77, t(122) = −9,81, p < 0.01, male model: Mnon-stereotyped = 2.57 vs Mstereotyped = 5.06, t(128) = −8,11, p < 0.01). As real brands were used in the stimulus ads, all hypotheses were tested while controlling for brand familiarity.
Assessing H1, a MANCOVA analysis using brand familiarity as covariate (thus neutralizing any potential effect it may have) showed a direct effect of stereotypicality on presumed negative influence on women and men (Wilks Lambda = 0.877; F = 17.07, p < 0.01, η2 = 0.12). In this case brand familiarity was not significant as a covariate, (p = 0.763). Further, there was a direct effect of model gender (Wilks Lambda = 0.925; F = 9.82, p < 0.01, η2 = 0.08), and a significant interaction effect between stereotypicality and model gender (Wilks Lambda = 0.899; F = 13.73, p < 0.01, η2 = 0.10). The analysis showed no significant direct effect of participant gender (p = 0.879), but a significant interaction effect between stereotypicality and participant gender (Wilks Lambda = 0.971; F = 3.69, p < 0.05, η2 = 0.03). In support of H1a, planned comparisons revealed that male participants presumed the stereotyped female portrayal to have a higher level of negative impact on women (Mnon-stereotyped = 3.17 vs Mstereotyped = 4.90, t(57) = −3.97, p < 0.01). Female participants did however not presume the stereotyped male portrayal to have a higher level of negative of impact on men (Mnon-stereotyped = 3.53 vs Mstereotyped = 4.20, t(60) = −1.50, p = 0.136). H1b is thus rejected. Within-gender planned contrasts were then performed, revealing that male participants did not presume the stereotyped male portrayal to have a higher level of negative impact on men (Mnon-stereotyped = 3.56 vs Mstereotyped = 3.94, t(66) = −0.97, p = 0.333). Female participants did, however, presume the stereotyped female portrayal to have a higher level of negative of impact on women (Mnon-stereotyped = 2.36 vs Mstereotyped = 5.34, t(63) = −6.87, p < 0.01).
Testing H2, an ANCOVA analysis showed a direct impact of stereotypicality (F = 51.24, p < 0.01, η2 = 0.17), and of participant gender (F = 6.13, p < 0.05, η2 = 0.02) on ad reactance. There was no significant direct effect of model gender (p = 0.358), nor was there an effect of the covariate brand familiarity (p = 0.902). However, there were significant interaction effects between stereotypicality and model gender (F = 10.75, p < 0.01, η2 = 0.04), and between stereotypicality and participant gender (F = 5.89, p < 0.05, η2 = 0.02). In support of H2a, planned comparisons revealed that male participants reported higher levels of reactance after exposure to the stereotyped female portrayal (Mnon-stereotyped = 2.04 vs Mstereotyped = 3.60, t(57) = −3.50, p < 0.01). Similarly, female participants reported higher levels of ad reactance after exposure to male stereotyped portrayals (Mnon-stereotyped = 2.89 vs Mstereotyped = 4.11, t(60) = −2.92, p < 0.01), thus supporting H2b. Within-gender planned contrasts also show that male participants did not report higher levels of reactance after exposure to the stereotyped male portrayal (Mnon-stereotyped = 2.71 vs Mstereotyped = 2.99, t(66) = −0.81, p = 0.420). In contrast, female participants reported higher levels of ad reactance after exposure to female stereotyped portrayals (Mnon-stereotyped = 1.90 vs Mstereotyped = 4.39, t(63) = −7.18, p < 0.01).
To assess H3, a MANCOVA analysis showed a direct effect of stereotypicality on presumed ad attitudes, brand attitudes, and purchase intentions (Wilks Lambda = 0.747; F = 27.36, p < 0.01, η2 = 0.25). There were no significant direct effects of model gender (p = 0.374) or participant gender (p = 0.176), but the covariate brand familiarity was significant (Wilks Lambda = 0.892; F = 9.79, p < 0.01, η2 = 0.11). The analysis showed significant interaction effects between stereotypicality and model gender (Wilks Lambda = 0.955; F = 3.83, p < 0.01, η2 = 0.05) and stereotypicality and participant gender (Wilks Lambda = 0.918; F = 7.28, p < 0.01, η2 = 0.08). Planned comparisons revealed that for men, the stereotyped female ad generated lower levels of ad attitudes (Mnon-stereotyped = 4.90 vs Mstereotyped = 2.88, t(57) = 4.93, p < 0.01), brand attitudes (Mnon-stereotyped = 4.61 vs Mstereotyped = 2.99, t(57) = 3.75, p < 0.01), and purchase intentions (Mnon-stereotyped = 4.38 vs Mstereotyped = 2.95, t(57) = 3.03, p < 0.01). For women, the pattern was similar for the male ads in terms of ad attitudes (Mnon-stereotyped = 3.82 vs Mstereotyped = 2.52, t(60) = 3.19, p < 0.01) and brand attitudes (Mnon-stereotyped = 4.00 vs Mstereotyped = 2.83, t(60) = 3.17, p < 0.01), but the effects did not carry over to purchase intentions (Mnon-stereotyped = 3.34 vs Mstereotyped = 3.00, t(60) = 0.72, p = 0.476). H3a is thus fully supported, while H3b is partially supported. The results for H1-3 are summarized in Table 2. Within-gender planned comparisons also reveal that for men, the stereotyped male ad generated overall lower levels of ad attitudes (Mnon-stereotyped = 4.29 vs Mstereotyped = 3.25, t(66) = 2.74, p < 0.01), brand attitudes (Mnon-stereotyped = 4.33 vs Mstereotyped = 3.65, t(66) = 2.01, p < 0.05), and purchase intentions (Mnon-stereotyped = 3.14 vs Mstereotyped = 2.71, t(66) = 1.11, p = 0.270). For women, the pattern was similar for the female ads in terms of ad attitudes (Mnon-stereotyped = 5.36 vs Mstereotyped = 2.23, t(63) = 7.13, p < 0.01) and brand attitudes (Mnon-stereotyped = 5.48 vs Mstereotyped = 2.97, t(60) = 6.26, p < 0.01), but the effects did not carry over to purchase intentions (Mnon-stereotyped = 4.18 vs Mstereotyped = 2.80, t(60) = 2.92, p < 0.01).
To test H4a and H4b, the underlying premise that the stereotyped portrayals increase presumed negative impact on women/men (H1), which in turn increases ad reactance (H2), which consequently has a negative impact on ad attitudes, brand attitudes, and purchase intentions (H3), mediation tests were conducted using the Preacher-Hayes approach (Model 6, 5000 bootstrapping samples, 95% confidence interval, as recommended by Hayes et al. (2011), Zhao et al. (2010). The first analysis focused on male participants (H4a). Female ad condition (non-stereotyped vs stereotyped ad) was used as the independent variable, presumed negative influence on women as the first mediator, ad reactance as the second mediator, and ad attitudes, brand attitudes, and purchase intentions as the dependent variables (in three separate mediation models). Significant mean indirect effects from the bootstrap analyses were found for ad attitudes (−0.28, 5,000 bootstrap samples; 95% CI: −0.68 to −0.06) brand attitudes (−0.31, 5,000 bootstrap samples; 95% CI: −0.75 to −0.08), and purchase intentions (−0.22, 5,000 bootstrap samples; 95% CI: −0.60 to −0.01). The tests also showed a significant effect of stereotypicality on presumed negative influence on women (1.67, 95% CI: 0.77 – 2.56), and of presumed influence on ad reactance (0.43, 95% CI: 0.19 – 0.67). Further, ad reactance had a significant negative effect on ad attitudes (−0.31, 95% CI: −0.51 – −0.11) and brand attitudes (−0.33, 95% CI: −0.54 – −0.13), but not on purchase intentions (−0.15, 95% CI: −0.39 – 0.10). Thus, for men, H4a was accepted for ad attitudes and brand attitudes, whereas it was rejected for purchase intentions.
The second analysis focused on female participants and male portrayal ads, thereby testing for H4b. Male ad condition (non-stereotyped vs stereotyped ad) was used as the independent variable in the mediation models, presumed negative influence on men as the first mediator, ad reactance as the second mediator, and ad attitudes, brand attitudes, and purchase intentions as the dependent variables. No significant mean indirect effects from the bootstrap analyses were found for the moderation models predicting ad attitudes (−0.07, 5,000 bootstrap samples; 95% CI: −0.26 to 0.02), brand attitudes (−0.04, 5,000 bootstrap samples; 95% CI: −0.18 to 0.01), or purchase intentions (−0.04, 5,000 bootstrap samples; 95% CI: −0.17 to 0.02). The tests also showed no significant effect of stereotypicality on presumed negative influence on men (0.78, 95% CI: −0.09 – 1.76), but a significant effect of presumed influence on ad reactance (0.25, 95% CI: 0.01 – 0.49). Further, ad reactance had a significant negative effect on ad attitudes (−0.46, 95% CI: −0.70 – −0.23), brand attitudes (−0.29, 95% CI: −0.50 – −0.07), and purchase intentions (−0.32, 95% CI: −0.61 – −0.03). H4b was thus rejected due to the lack of significant effect of stereotypicality on presumed negative influence on men for female participants.
Overall, the results of the mediation analyses (summarized in Figures 2 and 3) support the proposed theoretical reasoning for men exposed to female ad stereotypes (with the exception of purchase intentions) but offers only partial support for women exposed to male stereotyped ads as indicated by an insignificant effect on the perceived negative of influence on men.
Discussion
Study 1 suggest that stereotyped portrayals of gender can lead to lower levels of ad attitudes, brand attitudes, and in some cases purchase intentions across gender compared to non-stereotyped portrayals. Mediation analyses indicate that this is partly due to stereotyped portrayals leading to higher levels of presumed negative influence on others (women or men), which in turn leads to ad reactance. While the proposed relationships are significant for men exposed to ads featuring women, there is no significant effect on perceived influence on men for women exposed to ad featuring men. Such ads, however, still generate higher levels of reactance with women, which has a negative impact on brand-related effects.
Comparing the cross-gender effects with the within-gender effects, it seems that the effects stipulated for cross-gender typically also holds for within-gender analyses of women’s reactions, much in line with previous research on how women respond to stereotyped portrayals of women (Åkestam et al., 2017a; Åkestam, 2018; Antioco et al., 2012; Bower, 2001; Halliwell and Dittmar; 2004; Janssen and Paas, 2014). Our data does, however, not indicate that men respond to portrayals of men in the same way. In other words, the results indicate that the effects are based on a combination of factors, rather than gender in isolation.
The results are subject to several limitations. Most notably, Study 1 only addressed stereotypes in terms of physical characteristics for one product category. It also used different brands for men and women, making it possible that the differences between men and women could be attributed to the different brands used. What is more, it did not take previous product category use and buying behaviors into account. Although all images used was sourced from a real underwear campaigns, perceived realism of the stimuli could also be a concern in the male non-stereotype ad condition. Finally, although a sample of 254 respondents is a considerable number of respondents to have been intercepted, it only results in approximately 30 respondents per cell used to test the hypotheses. Study 2 was designed with all of these limitations in mind.
Study 2
Study 2 sets out to remedy the limitations discussed above using ads featuring stereotyped portrayals in terms of gender role behavior, rather than physical characteristics. The experimental design was similar to Study 1.
Stimulus development
Four print ads for a real deodorant brand (Dove) were created as stimuli. Deodorant was chosen as the product category as it is frequently bought and used by women as well as men. Dove was selected as the brand because it has the advantage of being a gender-neutral brand (since gendered brands may elicit different effects, cf. Azar et al., 2018; Ulrich, 2013). The brand logos and pack-shots were adapted from real Dove ads to enhance the external validity, but the stimuli ads were created for the experiment by a professional art director to avoid confounding effects from previous exposure (as recommended by Moorman et al., 2002). For the stereotyped female portrayal condition, the ad portraying stereotyped gender role behavior featured a woman and a child cleaning. This is in line with stereotypes of women, where women are often portrayed in the home and together with children (Wiles et al., 1995). The non-stereotyped female ad featured a woman lifting heavy weights at the gym. For the male portrayal condition, the roles were reversed; the ad portraying stereotyped gender role behavior featured a strong-looking man lifting heavy weights at the gym with his eyes focused purposefully (Wiles et al., 1995), and the non-stereotyped ad featured a man and a child cleaning. All ads featured the message “Real women/men do the heavy lifting”. This rendered four ads that were similar in terms of message and layout, but different in terms of gender role behavior stereotypicality and model gender. A pre-test (n = 122, 57% women, mean age = 34.70; participants similar to, but not included in the main study) showed that the stereotyped ads were perceived as significantly more stereotyped than the non-stereotyped ads (female model: Mstereotyped = 6.03; Mnon-stereotyped = 2.47 vs, p < 0.01, male model: Mstereotyped = 5.84; Mnon-stereotyped = 3.07 vs, p < 0.01, same items as Study 1, Cronbach’s alpha = 0.872).
Procedure and measures
Three hundred and sixty-six people participated in the study. The procedure and measures were identical to those of Study 1 and the data collection took place in February 2018. Purchase intentions were measured using “skincare product from Dove” rather than a deodorant specifically. Participants who had never bought deodorant or did rarely use deodorant were removed from the sample. This rendered data from a total of 330 participants (age 18–77, mean age = 33.7) that was used in the analysis. As in Study 1, participants were asked for their gender with the resulting distribution 53.3% women, 44.8% men, and 1.8% other. The 6 individuals who answered “other” were subsequently removed from analysis since this group is too small for any meaningful statistical analyses. The participants were randomly allocated to experiment groups and each participant answered one of the four versions of the questionnaire. Of the participants identifying as women, 46 viewed the male non-stereotyped ad, 34 viewed the male stereotyped ad, 52 viewed the female non-stereotyped ad, and 44 viewed the female stereotyped ad. Among those participants who identify as men, 30 viewed the female non-stereotyped ad, 40 viewed the female stereotyped ad, 36 viewed the male non-stereotyped ad, and 42 viewed the male stereotyped ad.
Results
A manipulation check confirmed that the manipulation was successful (female model: Mnon-stereotyped = 3.46 vs Mstereotyped = 5.58, p < 0.01, male model: Mnon-stereotyped = 3.65 vs Mstereotyped = 5.75, p <. 01). Both the female ads (Mnon-stereotyped = 3.71 vs Mstereotyped = 3.40, p = 0.474) and male ads (Mnon-stereotyped = 3.86 vs Mstereotyped = 3.64, p = 0.485) were considered as equally realistic across conditions. As the male non-stereotyped model was perceived as significantly older (38.9 vs 30.2, p < 0.01), all analyses were conducted while controlling for model age. As in Study 1, all hypotheses were tested while controlling for brand familiarity.
Assessing H1, a MANCOVA analysis showed a direct effect of stereotypicality on presumed negative influence on women/men (Wilks Lambda = 0.914; F = 14.00, p < 0.01, η2 = 0.09). The analysis showed no significant direct effect of participant gender (p = 0.122), but a significant interaction effect between stereotypicality and participant gender (Wilks Lambda = 0.980; F = 3.04, p < 0.05, η2 = 0.02). Further, there was direct effect for model gender (Wilks Lambda = 0.954; F = 7.10, p < 0.01, η2 = 0.05), but no significant interaction effect between stereotypicality and model gender (p = 0.242). Neither of the covariates were significant; brand familiarity (p = 0.878) and model age (p = 0.148). Planned comparisons revealed that male participants presumed the stereotyped female portrayal to have a higher level of negative impact on women (Mnon-stereotyped = 3.53 vs Mstereotyped = 4.55, t(68) = −2.84, p < 0.01). Female participants similarly presumed the stereotyped male portrayal to have a higher level of negative of impact on men (Mnon-stereotyped = 3.59 vs Mstereotyped = 4.76, t(76) = −2.23, p < 0.05). H1a and H1 b are thus supported. Within-gender planned contrasts were then performed, revealing no significant difference in how male participants presume that the stereotyped and non-stereotyped male portrayals may affect the level of negative impact on men (Mnon-stereotyped = 3.44 vs Mstereotyped = 3.95, t(76) = −1.49, p = 0.140). Female participants did however presume the stereotyped female portrayal to have a higher level of negative of impact on women (Mnon-stereotyped = 3.85 vs Mstereotyped = 5.43, t(92) = −4.79, p < 0.01).
In assessment of H2, an ANCOVA analysis showed a direct impact of stereotypicality on ad reactance (F = 38.76, p < 0.01, η2 = 0.11). There was also a significant direct effect of participant gender (F = 19.16, p < 0.01, η2 = 0.06), and a significant direct effect of model gender (F = 5.80, p < 0.05, η2 = 0.02). The interaction effect between stereotypicality and participant gender was not significant (p = 0.157) and neither was the interaction effect between stereotypicality and model gender (p = 0.551). The covariate for brand familiarity was also significant (F = 6.30, p < 0.05, η2 = 0.02), but not for model age (p = 0.520). Further, planned comparisons revealed that male participants reported higher levels of reactance after exposure to the stereotyped female portrayal (Mnon-stereotyped = 2.96 vs Mstereotyped = 3.72, t(68) = −1.98, p < 0.05). Similarly, female participants reported higher levels of reactance after exposure to male stereotyped portrayals (Mnon-stereotyped = 2.91 vs Mstereotyped = 3.98, t(78) = −3.00, p < . 01). This supports H2a and H2b. Within-gender planned contrasts show that male participants did in fact report higher levels of reactance after exposure to the stereotyped male portrayal (Mnon-stereotyped = 2.61 vs Mstereotyped = 3.65, t(76) = −3.42, p < 0.01), just as female participants also did after exposure to female stereotyped portrayals (Mnon-stereotyped = 3.37 vs Mstereotyped = 4.98, t(94) = −5.81, p < 0.01).
In assessment of H3, a MANCOVA analysis showed a direct effect of stereotypicality on presumed ad attitudes, brand attitudes, and purchase intentions (Wilks Lambda = 0.858; F = 16.38, p < 0.01, η2 = 0.14). There were also significant direct effects of model gender (Wilks Lambda = 0.967; F = 3.41, p <. 05, η2 = 0.03) and participant gender (Wilks Lambda = 0.941; F = 3.07, p < 0.01, η2 = 0.03). The analysis further showed a significant three-way interaction effect between stereotypicality, model gender and participant gender (Wilks Lambda = 0.969; F = 3.19, p < 0.05, η2 = 0.03). Neither brand familiarity (p = 0.868), nor model age (p = 0.571) were significant as covariates. Planned comparisons revealed that for men, the stereotyped female ad generated lower levels of ad attitudes (Mnon-stereotyped = 4.27 vs Mstereotyped = 2.73, t(64) = 4.06, p < 0.01), brand attitudes (Mnon-stereotyped = 4.24 vs Mstereotyped = 3.47, t(68) = 2.35, p < 0.05), but no significant difference for purchase intentions (Mnon-stereotyped = 2.73 vs Mstereotyped = 3.48, t(68) = −1,69, p = 0.095). For women, the pattern was similar for the male ads in terms of ad attitudes (Mnon-stereotyped = 4.38 vs Mstereotyped = 3.10, t(78) = 2.92, p < 0.01) and brand attitudes (Mnon-stereotyped = 4.42 vs Mstereotyped = 3.69, t(78) = 1.79, p < 0.05), but the effects did not carry over to purchase intentions (Mnon-stereotyped = 3.15 vs Mstereotyped = 2.85, t(78) = 0.75, p = 0.465). H3a and H3b are thus both partially supported. The results are summarized in Table 3. Within-gender planned comparisons reveal that for men, the stereotyped male ad generated overall lower levels of ad attitudes (Mnon-stereotyped = 4.67 vs Mstereotyped = 3.78, t(76) = 2.91, p < 0.01) and brand attitudes (Mnon-stereotyped = 4.59 vs Mstereotyped = 3.75, t(76) = 2.98, p < 0.01), but the effects did not carry over to purchase intentions (Mnon-stereotyped = 3.92 vs Mstereotyped = 3.79, t(76) = 0.39, p = 0.702). For women, the pattern was similar for the female ads in terms of ad attitudes (Mnon-stereotyped = 3.85 vs Mstereotyped = 2.26, t(94) = 5.43, p < 0.01) and brand attitudes (Mnon-stereotyped = 4.23 vs Mstereotyped = 2.77, t(94) = 4.78, p < 0.01), but the effects also carried over to purchase intentions (Mnon-stereotyped = 3.77 vs Mstereotyped = 3.10, t(92) = 2.08, p < 0.05).
Mediation tests for H4a and H4b were conducted using the same approach as in Study 1. The analyses for male participants reacting to female ads showed significant mean indirect effects from the bootstrap analyses for the moderation models predicting ad attitudes (−0.23, 5,000 bootstrap samples; 95% CI: −0.52 to −0.03) and brand attitudes (−0.27, 5,000 bootstrap samples; 95% CI: −0.56 to −0.05), but not for purchase intentions (−0.19, 5,000 bootstrap samples; 95% CI: −0.53 to 0.12). The tests for male participants reacting to female ads also showed a significant effect of stereotypicality on presumed negative influence on women (1.31, 95% CI: 0.52 – 2.09), and of presumed influence on ad reactance (0.69, 95% CI: 0.48 – 0.69). Further, ad reactance had a significant negative effect on ad attitudes (−0.26, 95% CI: −0.52 – −0.01) and brand attitudes (−0.38, 95% CI: −0.60 – −0.16), but not on purchase intentions (−0.25, 95% CI: −0.63 – 0.14). Mirroring Study 1, H4 is accepted for ad attitudes and brand attitudes, but rejected for purchase intentions.
The corresponding analyses for female participants reacting to male ads showed no significant mean indirect effects from the bootstrap analyses for the moderation models predicting ad attitudes (−0.10, 5,000 bootstrap samples; 95% CI: −0.33 to 0.03), brand attitudes (−0.02, 5,000 bootstrap samples; 95% CI: −0.20 to 0.11), or purchase intentions (−0.10, 5,000 bootstrap samples; 95% CI: −0.03 to 0.34). The tests for female participants reacting to male ads also showed a significant effect of stereotypicality on presumed negative influence on men (1.24, 95% CI: 0.15 – 2.63), and a significant effect of presumed influence on ad reactance (1.32, 95% CI: 0.46 – 2.18). Further, ad reactance had a significant negative effect on ad attitudes (−0.52, 95% CI: −0.82 – −0.22), brand attitudes (−0.42, 95% CI: −0.72 – −0.13), but not on purchase intentions (−0.03, 95% CI: −0.25 – 0.32). In this study, these results replicate the findings of H4a rather than the findings of H4b in Study 1 since H4b is here accepted for ad attitudes and brand attitudes but not for purchase intentions.
Overall, the results of the mediation analyses (summarized in Figures 4 and 5) support the theoretical reasoning for ad and brand attitudes, but not for purchase intentions.
Discussion
Study 2 set out to remedy some of the limitations of Study 1, using advertising portrayals of gender role behavior stereotypes. For this study, H1 and H2 were fully supported, and H3 was supported with the exception of purchase intentions. This is also visible in the mediation analyses, where H4a and H4b are accepted for ad and brand attitudes but rejected when it comes to purchase intentions. This largely replicates the findings of Study 1, with one important exception: for gender role behavior stereotypes, the proposed theoretical reasoning is supported for men exposed to portrayals of women, as well as for women exposed to portrayals of men. As previously, within-gender analyses of female respondents’ reactions are in line with extant research on how women respond to stereotyped portrayals of women (Åkestam et al., 2017a; Åkestam, 2018; Antioco et al., 2012; Bower, 2001; Halliwell and Dittmar; 2004; Janssen and Paas, 2014). Similar to the cross-gender analyses of interest, we also find a little stronger support for the theoretical framework among men responding to portrayals of men. As in Study 1, however, the responses still seem to be dependent on a combination of factors as opposed to merely gender. Taken together, however, as the studies in this research also differed in terms of product category, brands, and potentially realism of the stimuli, more studies are needed to confirm if this is indeed the case.
General discussion
The results in this research suggest that gender-stereotyped portrayals in advertising have a negative impact across genders, for both women and men. Men presume stereotyped portrayals of women to have a negative influence on women, both in terms of physical characteristics and role behavior, which in turn leads to ad reactance. For female consumers, reactions to male portrayals are similar for stereotypes in terms of gender role behavior. For stereotyped portrayals of physical characteristics, however, there is no difference in the presumed influence on men between the two portrayals. This could be due to another common gender stereotype: the notion that men are less concerned with their looks than women are (Van Hellemont and Van Den Bulck, 2012). Our within-gender analyses also point in this direction. Another potential explanation is that female stereotyping has been more frequent (Eisend, 2010; Fredrickson and Roberts, 1997; Wolin, 2003) and is more intensely discussed in society (Åkestam et al., 2017a), making its impact on women a more salient concern than for men (as our within-gender analyses also suggest). If this explanation holds, the fact that consumers adjust their perception of influence on others over time (Eisend, 2017), suggests that as male stereotypes related to physical characteristics in advertising are increasingly discussed and challenged (by brands such as Axe and Gillette), the presumed influence on men might come to increase (cf. Tuncay Zayer et al., 2020; Ricciardelli et al., 2010; Davis, 2002).
Research implications
This research contributes with a much-needed cross-gender perspective to the research literature on gender stereotypes in advertising which, to date, has mostly been concerned with women’s reactions to female stereotypes (Antioco et al., 2012; Eisend, 2019), and to a lesser degree with male reactions to male stereotypes (Tuncay Zayer et al., 2020; Lorenzen et al., 2004). The findings indicate that consumers react negatively to stereotyping across genders; a perspective that has received considerably less research attention (with exceptions such as Liljedal et al., 2020; Baxter et al., 2016). These findings also suggest that the negative effects of gender-stereotyped portrayals are more far-reaching than previous studies have suggested, because they extend to exposure audiences regardless of gender. Although much targeted advertising supposedly addresses a single-gender audience, consumers of other genders will still play important roles as purchasers of gifts, influencers in the decisions-making process, or – of course – as product users. This research indicates that consumers take stereotypes into account when evaluating advertising in these roles as well. This resonates with the understudied but growing literature on the social effects of advertising (Eisend, 2019; Kim et al., 2014; Gulas and McKeage, 2000; Häfner, 2003), where several studies have found that consumers can respond in terms of for example empathy (Åkestam et al., 2017b; Escalas and Stern, 2003) and social connectedness (Lee and Robins, 1995) to portrayals of people who are superficially dissimilar to themselves.
Another contribution of this research is that it proposes that the model of the influence of presumed influence (Eisend, 2017; Gunther and Storey, 2003) can help explain consumer reactions to gender-stereotyped portrayals in advertising. To date, the literature on gender stereotypes has mainly been concerned with self-relevant reactions, thus overlooking the social, cultural and societal contexts of these effects (Eisend, 2019). By connecting presumed influence with the theory of ad reactance (Brehm, 1966; Thorbjørnsen and Dahlén, 2011) our findings suggest that presumed influence on others can also lead to reactance and that gender stereotypes can be perceived as limiting consumers’ freedom in general. Thus, this research interprets the literature on the physical characteristics and role behavior of advertising models using a stereotype framework and thereby ties it closer to the literature on gender stereotypes in advertising. In other words, this research connects studies previously scattered across different theoretical fields such as advertising (Tuncay Zayer et al., 2020; Grau and Zotos, 2016; Antioco et al., 2012; Wolin, 2003), psychology (Lockwood and Kunda, 1997; Lavine et al., 1999; Halliwell and Dittmar, 2004; Zawisza and Cinnirella, 2010; Meyers-Levy and Loken, 2015), consumer research (Richins, 1991) and gender studies (Kaufman, 1999; Morrison and Shaffer, 2003; White and White, 2006; Ricciardelli et al., 2010).
Implications for marketers
The present research joins the growing body of work on the negative effects of gender stereotypes in advertising (Eisend, 2019) urging advertising and marketing professionals to reconsider their continued use of gender stereotypes for creating easily recognizable narratives is thought to connect with desirable brand values (Windels, 2016; Tuncay Zayer and Coleman, 2015; Middleton et al., 2019). That is, the argument that divergence from normative elements of stereotypes risk being evaluated negatively (Zawisza and Cinnirella, 2010) should no longer be perceived as valid in practice. The findings in this research, that non-stereotyped portrayals of both women and men can lead to positive ad and brand attitudes from exposure audiences, complement previous studies of stereotype effects and indicate that marketers should continue to explore non-stereotyped gender portrayals in their advertising. This suggests that marketers have much to gain, both in terms of sales and goodwill, from pro-actively challenging gender stereotypes, rather than waiting for regulatory efforts (such as those of the ASA) or further guidance from self-regulatory bodies (such as the European Advertising Standards Alliance and the ICC Code on Advertising Practice). Non-stereotyped gender portrayals also benefit marketers in the long run, since gender stereotypes have been found to harm consumers’ attitude toward advertising in general (Huhmann and Limbu, 2016).
The results also indicate that marketers should consider reactions of target as well as exposure audiences when designing and evaluating campaigns. This is partly because consumers of all genders may play important roles both as purchasers of a product, as influencers in the decision-making process and also as users of a product originally marketed toward another gender. It is also partly due to advertising’s role as culture and as a mechanism for an ongoing discourse on societal values, such as those pertaining to gender and stereotypes (Tuncay Zayer et al., 2020; Schroeder and Borgerson, 1998; Steele et al., 2002; O’Donohoe, 1994). In this regard, to include cross-gender reactions is a step to increase our overall understanding in this area. Further, although advertising can influence consumers to (re)negotiate gender in their daily lives (Tuncay Zayer et al., 2020), we know that consumers tend to behave in congruence with advertised values (Defever et al., 2011) and may even alter their own gender role attitudes following exposure to advertising (Tuncay Zayer et al., 2020; Garst and Bodenhausen, 1997). These types of findings place further responsibility on advertisers to explore more modern, non-stereotyped gender portrayals to their advertising. As society changes, we are likely to see this change take place. This would in turn further strengthen the mirror argument over the mold argument (Pollay, 1986) as proposed by Eisend (2010).
Finally, the findings suggest that using stereotyped portrayals in advertising to attract persons of other genders gender is a flawed strategy. The stereotyped portrayals are not only harmful for people of the same gender, but as consumers take the presumed impact on others into account when evaluating ads, such executions may be evaluated less favorably, regardless of the consumer’s own gender and sexual preferences.
Limitations and suggestions for further research
There are several limitations of this research that should be considered in developing the line of research further. First, more research studies are needed to assess whether the differences in reactions observed for male ads stereotyped in terms of physical characteristics and role behavior are indeed a result of the different stereotype components, or due to a confounding element of Study 1, such as realism or product category. Future research should further assess whether product categories with multi-gender target audiences that are perceived as androgynous rather than gendered in themselves (cf. Fugate and Phillips, 2010; van Tilburg et al., 2015), such as grocery stores or soft drinks, would render similar results. Our two empirical studies both used real brands, which enhances the ecological validity of the studies but could potentially also have had a confounding effect on the results.
While ad attitudes, brand attitudes, and purchase intentions are certainly of interest to most marketers, they do not provide a full understanding of the potential effects of stereotyped and non-stereotyped advertising portrayals. Previous research has suggested that non-stereotyped advertising portrayals can work as a priming cue to generate social effects, such as social connectedness and empathy (Åkestam et al., 2017b). Future studies should include social effects in parallel to brand-related effects when studying non-stereotyped portrayals of gender, to investigate the presumed connection between social effects and advertising metrics (Dahlén and Rosengren, 2016; Eisend, 2019). In this context, effects of self-objectification (Fredrickson and Roberts, 1997) could provide a useful path of further exploring cross-gender effects of stereotyped gender portrayals. Hopefully the present research will open up for more such investigations, thereby contributing to a deeper understanding of the effects of stereotyped portrayals of gender in advertising.
In line with the above discussion, it is also interesting to ponder if – and in that case how – the context of the data collection might have had an impact on the results. That is, can using intercept at major train stations and shopping malls to recruit participants for our studies have primed the participants to consider the role of advertising in society and retail? Compared to other methods of data collection, such as online panels, the method afforded us a certain degree of control over the context in which the participants were exposed to the ads. Compared with other sampling methods, such as classroom experiments, the present methodology also made it possible to recruit a wide and heterogenous set of respondents, many of whom are not participating in online panels and similar. Future research may want to explore if advertising research concerning societal aspects and issues differ depending on where and how the data is collected. Further, this study was unable to investigate responses in terms a broader spectrum of gender (Fischer, 2015; Eisend, 2019), since relatively few of the respondents in our studies self-categorized as other genders than male or female and this group was thus too small to analyze separately. Researchers may want to explore other ways of measuring gender identities (a combination of culture and biological sex; Palan, 2001) through for example the BEM scale (Bem, 1974; or Barak and Stern, 1986 for a shortened version) to allow for categorization of respondents across two separate dimensions – one feminine and one masculine. Alternatively, the Gender Attitude Inventory (GAI: Ashmore et al., 1995) can also help researchers in assessing how respondents’ acceptance of traditional stereotypes influence their advertising responses (cf. Ulrich, 2013). These types of categorizations might further help to uncover why and how consumers across genders respond to gender stereotypes.
In social sciences, gender is rapidly moving beyond the binary definitions of only a male and a female gender and advertising research is no exception (Eisend, 2019). A valid critique toward this research is thus that although it explores the cross-gendered effects of gender-based stereotypes, it is still concerned with only the male and female gender. This is also a weakness of the chosen methodology. Intercepts are unfortunately problematic if one wants to recruit minorities. Out of six hundred and twenty respondents, only six individuals opted for “other” when indicating their gender. A logical next step for future research would therefore be to use other forms of recruitment, such as online panels that segment respondents on gender, to better investigate the full spectrum how consumers of different genders respond to gender-based stereotyped advertising. Finally, gender stereotypes understandably vary with culture since what it means to identify with any gender identity is heavily influenced by culture (Palan, 2001). Our empirical studies were both carried out in Sweden, the most feminine country as per Hofstede’s categorization of national cultures (Hofstede, 2001). Although stereotypes were pre-tested in the Swedish market, we do not know how other culturally relevant stereotypes may affect consumers in other cultures, for example more masculine countries such as Japan (Hofstede, 2001). Studies have repeatedly shown that countries vary in the degree to which gender stereotypes are used, how they are used, and which stereotypes are most frequently used (An and Kim, 2007; Wiles et al., 1995). Investigating if the present results can be replicated in other cultures and if the underlying theoretical reasoning also holds for other cultures would therefore be an interesting next step in further developing the research on cross-gender responses to gender stereotypes in advertising.
Figures
Experiment studies of gender stereotypes in advertising since 1990
Name | Stereotype Component | Participants | Study performed in (country) | Findings |
---|---|---|---|---|
Richins (1991) | Female physical | Women | Stereotyped portrayals lead to lower levels of self-satisfaction | |
Martin and Kennedy (1993) | Female physical | Women (preadolescent, adolescent) | USA | Stereotyped portrayals raise comparison standards, but do not affect self-perceptions |
Jaffe and Berger (1994) | Female role | Women | USA | An egalitarian positioning of female portrayals in advertising is favored over a superwoman or traditional positioning in terms of affect and purchase inteterest |
Garst and Bodenhausen (1997) | Male role | Men | Less traditional men’s gender role attitudes are more traditional following exposure to traditionally masculine models, but remain unchanged after exposure to androgynous models | |
Martin and Gentry (1997) | Female physical | Women (preadolescent, adolescent) | USA | Through social comparison, girls’ self- perceptions and self-esteem may be affected. Motive for social comparison moderates results. |
Lavine et al. (1999) | Male/female physical, role | Men and women | USA | Ad attitudes are higher for nonsexist (nonstereotyped) ads. Attitude toward feminism moderate results. Both men and women report larger body size discrepancies following exposure to sexist ads |
Gulas and McKeage (2000) | Male/female physical, role | Men (students) | USA | Stereotyped portrayals of men and women have a negative impact on men’s self-evaluations |
Bower (2001) | Female physical | Women | Stereotyped portrayals lead to lower levels of model and product evaluation | |
Morrison and Shaffer (2003) | Male/female Role (endorser) | Men and women (students) | USA | Traditional participants respond more favorably to stereotyped ads; non- traditional participants respond somewhat more favorably to nontraditional ads (ad attitude, product attitude, purchase intention). Self-referencing moderates the results |
Lorenzen et al. (2004) | Male, physical | Men (students) | USA | Men’s self-rated body satisfaction decreased (did not change) after viewing images of muscular (average) men |
Halliwell and Dittmar (2004) | Female physical | Women | Stereotyped portrayals lead to higher levels of body-focused anxiety (moderated by internalization of thin ideals) but has no impact on advertising effectiveness | |
Orth and Holancova (2004) | Male/female, occupation | Men and women | Czech Republic | Men and women react differently (e.g. ad attitude, brand attitude, purchase intention) to non-/stereotyped portrayals, with the most favorable responses to exclusive portrayals of their own gender. Responses are affected by consumers' a priori attitude toward the sex role portrayal issue |
Choi et al. (2008) | Female physical | Women (students) | Stereotyped portrayals lead to higher levels of presumed influence on others | |
Harper and Tiggemann (2008) | Female physical | Women (students) | Australia | Stereotyped portrayals lead to higher levels of self-objectification and body dissatisfaction |
Zawisza and Cinnirella (2010) | Male/female role, traits | Men and women respectively (2 studies) (students) | United Kingdom | For both male and female versions of the ads, the paternalistic ad strategies are more effective than envious ones, supporting the predictions of the stereotype content model |
Buunk and Dijkstra (2011) | Female physical | Women (students) | The Netherlands | Stereotyped portrayals lead to lower levels of product attitudes. Moderated by gender-priming |
Antioco et al. (2012) | Female physical | Women | France | Stereotyped portrayals lead to lower levels of self-esteem, model trustworthiness, brand attitude and purchase intention |
Eisend et al. (2014) | Male/female role | Men and women (mostly students) | Germany | Ad attitude and brand attitude are increased for humorous stereotyped portrayals, especially non-stereotyped (compared to non-humorous). Women evaluate gender portrayals more favorably in humorous ads, especially for nontraditional stereotyping |
Janssen and Paas (2014) | Female physical | Women | The Netherlands | Stereotyped portrayals lead to lower levels of ad attitude, brand attitude, and purchase intention. Consumers’ fashion leadership may moderate results |
Lee (2014) | Male/Female Traits | Men and women | Taiwan | In political advertising, congruity between candidate gender and campaign theme enhances attitudes toward the political candidate and voting intentions in positive advertising, whereas complementarity enhanced persuasion in negative advertising |
Baxter et al. (2016) | Male/Female role | Men and Women (parents) | Australia | Non-stereotyped portrayals of males lead to higher levels of advertising attitude, especially by consumers who hold a non-traditional (egalitarian) gender role ideology |
Bian and Wang (2015) | Female physical | Women | Stereotyped portrayals lead to lower levels of model attractiveness. Moderated by brand and self-esteem | |
Chu et al. (2016) | Male/female, role (endorser) | Men and women | Korea | Consumers’ ad attitude (toward non-stereotyped ads) depends on self-construal and need for uniqueness, with perceived novelty and cognitive resistance as mediators |
Åkestam et al. (2017a) | Female physical, role | Women | Stereotyped portrayals lead to lower levels of ad attitude and brand attitude | |
Åkestam (2018) | Female physical, role | Women | Sweden | Stereotyped portrayals lead to lower levels of ad attitude and brand attitude |
Liljedal et al. (2020) | Male/female, occupation | Men and women | USA, Sweden, UK | Nonstereotyped occupational gender-role portrayals in advertising generate positive brand-related and social effects on respondents, regardless of their gender |
Mean values of planned comparisons, study 1
Men reacting to female ads | |||
---|---|---|---|
Variable | Stereotyped mean (SD) | Non-stereotyped mean (SD) | p |
PI women | 4.92 (1.58) | 3.18 (1.75) | <0.01 |
Reactance | 3.60 (1.91) | 2.04 (1.47) | <0.01 |
AdAtt | 2.88 (1.52) | 4.90 (1.63) | <0.01 |
BAtt | 2.99 (1.67) | 4.61 (1.65) | <0.01 |
PI | 2.95 (1.88) | 4.38 (1.75) | <0.01 |
Women reacting to male ads | |||
Variable | Stereotyped mean (SD) | Non-stereotyped mean (SD) | p |
PI men | 4.20 (1.51) | 3.53 (1.95) | 0.136 |
Reactance | 4.11 (1.60) | 2.89 (1.68) | <0.01 |
AdAtt | 2.52 (1.28) | 3.82 (1.89) | <0.01 |
BAtt | 2.83 (1.25) | 4.00 (1.61) | <0.01 |
PI | 3.00 (1.78) | 3.34 (1.99) | 0.476 |
Mean values of planned comparisons, study 2
Men reacting to female ads | |||
---|---|---|---|
Variable | Stereotyped mean (SD) | Non-stereotyped mean (SD) | p |
PI women | 4.55 (1.55) | 3.53 (1.38) | <0.01 |
Reactance | 3.72 (1.80) | 2.96 (1.25) | <0.05 |
AdAtt | 2.73 (1.60) | 4.27 (1.45) | <0.01 |
BAtt | 3.47 (1.46) | 4.24 (1.25) | <0.05 |
PI | 3.48 (1.85) | 2.73 (1.76) | 0.095 |
Women reacting to male ads | |||
Variable | Stereotyped | Non-stereotyped | p |
PI men | 4.76 (1.72) | 3.59 (2.67) | <0.05 |
Reactance | 3.98 (1.46) | 2.91 (1.65) | <0.01 |
AdAtt | 3.10 (1.96) | 4.38 (1.91) | <0.01 |
BAtt | 3.69 (1.90) | 4.42 (1.76) | <0.05 |
PI | 2.85 (1.89) | 3.15 (1.67) | 0.465 |
References
Advertising Standards Authority (2018), “Depictions, perceptions, and harm. A report on gender stereotypes in advertising”, Retrieved March 19 2018 from ASA’s website, available at: www.asa.org.uk/genderresearch.html
Åkestam, N. (2018), “Caring for her: the influence of presumed influence on female consumers’ attitudes toward advertising featuring gender-stereotyped portrayals”, International Journal of Advertising, Vol. 37 No. 6, pp. 871-892.
Åkestam, N., Rosengren, S. and Dahlén, M. (2017a), “Advertising ‘like a girl’: toward a better understanding of ‘femvertising’ and its effects”, Psychology and Marketing, Vol. 34 No. 8, pp. 795-806.
Åkestam, N., Rosengren, S. and Dahlén, M. (2017b), “Think about it – can portrayals of homosexuality in advertising prime consumer-perceived social connectedness and empathy?”, European Journal of Marketing, Vol. 51 No. 1, pp. 82-98.
An, D. and Kim, S. (2007), “Relating Hofstede’s masculinity dimension to gender role portrayals in advertising: a cross-cultural comparison of web advertisements”, International Marketing Review, Vol. 24 No. 2, pp. 181-207.
Antioco, M., Smeesters, D. and Le Boedec, A. (2012), “Take your pick: Kate Moss or the girl next door?”, Journal of Advertising Research, Vol. 52 No. 1, pp. 15-30.
Ashmore, R.D. and Del Boca, F.K. (1981), “Conceptual approaches to stereotypes and stereotyping”, in Hamilton, D.L. (Ed.), Cognitive Processes in Stereotyping and Intergroup Behavior, Psychology Press, New York, NY, pp. 1-35.
Ashmore, R.D., Del Boca, F.K. and Bilder, S.M. (1995), “Construction and validation of the gender attitude inventory, a structured inventory to assess multiple dimensions of gender attitudes”, Sex Roles, Vol. 32 Nos 11/12, pp. 753-785.
Azar, S.L., Aimé, I. and Ulrich, I. (2018), “Brand gender-bending. The impact of an endorsed brand strategy on consumers’ evaluation of gendered mixed-target brands”, European Journal of Advertising, Vol. 52 Nos 7/8, pp. 1598-1624.
Bakir, A., Blodgett, J.G. and Rose, G.M. (2008), “Children's responses to gender-role stereotyped advertisements”, Journal of Advertising Research, Vol. 48 No. 2, pp. 255-266.
Barak, B. and Stern, B.B. (1986), “Sex-linked trait indexes among baby-boomers and pre-boomers”, in Lutz, R. (Ed.), Advances in Consumer Research, Vol. 14, Association for Consumer Research, Provo, UT, pp. 204-209.
Barlett, C.P., Vowels, C.L. and Saucier, D.A. (2008), “Meta-analyses of the effects of media images on men’s body-image concerns”, Journal of Social and Clinical Psychology, Vol. 27 No. 3, pp. 279-310.
Baxter, S.M., Kulczynski, A. and Illicic, J. (2016), “Ads aimed at dads: exploring consumers’ reactions toward advertising that conforms and challenges traditional gender role ideologies”, International Journal of Advertising, Vol. 35 No. 6, pp. 970-982.
Bem, S.L. (1974), “The measurement of psychological androgyny”, Journal of Consulting and Clinical Psychology, Vol. 42 No. 2, pp. 155-162.
Bian, X. and Wang, K. (2015), “Are size-zero female models always more effective than average-sized ones? Depends on brand and self-esteem!”, European Journal of Marketing, Vol., Vol. 49 Nos 7/8, pp. 1184-1206.
Bower, A.B. (2001), “Highly attractive models in advertising and the women who loathe them: the implications of negative affect for spokesperson effectiveness”, Journal of Advertising, Vol. 30 No. 3, pp. 51-63.
Branchik, B.J. and Chowdhury, T.G. (2012), “Self-oriented masculinity: advertisements and the changing culture of the male market”, Journal of Macromarketing, Vol. 33 No. 2, pp. 160-171.
Brehm, J.W. (1966), A Theory of Psychological Reactance, Academic Press, Oxford.
Bukszpan, D. (2016), “DADvertising: how realistic images of dads took over TV ads”, Retrieved October 4, 2017, from Fortune’s website, available at: http://fortune.com/2016/06/19/dadvertising-commercials-fathers-day-ads/
Buunk, A.P. and Dijkstra, P. (2011), “Does attractiveness sell? Women's attitude toward a product as a function of model attractiveness, gender priming, and social comparison orientation”, Psychology and Marketing, Vol. 28 No. 9, pp. 958-973.
Choi, Y.Y., Leshner, G. and Choi, J. (2008), “Third-person effects of idealized body image in magazine advertisements”, American Behavioral Scientist, Vol. 52 No. 2, pp. 147-164.
Chu, K., Lee, D.-H. and Kim, J.Y. (2016), “The effect of non-stereotypical gender role advertising on consumer evaluation”, International Journal of Advertising, Vol. 35 No. 1, pp. 106-134.
Connell, R.W. (1987), Gender and Power: Society, the Person and Sexual Politics, Stanford University, CA.
Dahlén, M., Granlund, A. and Grenros, M. (2009), “The consumer-perceived value of non-traditional media: effects of brand reputation, appropriateness and expense”, Journal of Consumer Marketing, Vol. 26 No. 3, pp. 155-163.
Dahlén, M., Rosengren, S. and Smit, E. (2014), “Why the marketer's view matters as much as the message”, Journal of Advertising Research, Vol. 54 No. 3, pp. 304-312.
Dahlén, M., Rosengren, S., Törn, F. and Öhman, N. (2008), “Could placing ads wrong be right?: advertising effects of thematic incongruence”, Journal of Advertising, Vol. 37 No. 3, pp. 57-67.
Dahlén, M., Sjödin, H., Thorbjørnsen, H., Hansen, H., Linander, J. and Thunell, C. (2013), “‘What will ‘they’ think?’ marketing leakage to undesired audiences and the third-person effect”, European Journal of Marketing, Vol. 47 Nos 11/12, pp. 1825-1840.
Dahlén, M. and Rosengren, S. (2016), “If advertising won’t die, what will it be? Toward a working definition of advertising”, Journal of Advertising, Vol. 45 No. 3, pp. 334-345.
Davis, K. (2002), “‘A dubious equality’: men, women and cosmetic surgery”, Body and Society, Vol. 8 No. 1, pp. 49-65.
Deaux, K. and Lewis, L.L. (1984), “Structure of gender stereotypes: interrelationships among components and gender label”, Journal of Personality and Social Psychology, Vol. 46 No. 5, pp. 991-1004.
Defever, C., Pandelaere, M. and Roe, K. (2011), “Inducing value-congruent behavior through advertising and the moderating role of attitudes toward advertising”, Journal of Advertising, Vol. 40 No. 2, pp. 25-37.
Eisend, M. (2008), “Explaining the impact of scarcity appeals in advertising: the mediating role of perceptions of susceptibility”, Journal of Advertising, Vol. 37 No. 3, pp. 33-40.
Eisend, M. (2010), “A meta-analysis of gender roles in advertising”, Journal of the Academy of Marketing Science, Vol. 38 No. 4, pp. 418-440.
Eisend, M. (2015), “Persuasion knowledge and third-person perceptions in advertising: the moderating effect of regulatory competence”, International Journal of Advertising, Vol. 34 No. 1, pp. 54-69.
Eisend, M. (2017), “The third-person effect in advertising: a meta-analysis”, Journal of Advertising, Vol. 46 No. 3, pp. 377-394.
Eisend, M. (2019), “Gender roles”, Journal of Advertising, Vol. 48 No. 1, pp. 72-80.
Eisend, M., Plagemann, J. and Sollwedel, J. (2014), “Gender roles and humor in advertising: the occurance of stereotyping in humorous and nonhumorous advertising and its consequences for advertising effectiveness”, Journal of Advertising, Vol. 43 No. 3, pp. 256-273.
Elliott, R. and Elliott, C. (2005), “Idealized images of the male body in advertising: a reader‐response exploration”, Journal of Marketing Communications, Vol. 11 No. 1, pp. 3-19.
Gentry, J. and Harrison, R. (2010), “Is advertising a barrier to male movement toward gender change?”, Marketing Theory, Vol. 10 No. 1, pp. 74-96.
Escalas, J.E. and Stern, B.B. (2003), “Sympathy and empathy: emotional responses to advertising dramas”, Journal of Consumer Research, Vol. 29 No. 4, pp. 566-578.
Feiereisen, S., Broderick, A.J. and Douglas, S.P. (2009), “The effect and moderation of gender identity congruity: utilizing ‘real women’ advertising images”, Psychology and Marketing, Vol. 26 No. 9, pp. 813-843.
Ferrier, M. (2018), “Why women are buying men’s underwear”, Retrieved 19 March 2018 from the Guardian’s website, available at: www.theguardian.com/fashion/2014/oct/29/is-this-the-end-of-gender-specific-underwear-or-just-a-load-of-pants
Festinger, L. (1954), “A theory of social comparison processes”, Human Relations, Vol. 7 No. 2, pp. 117-140.
Fischer, E. (2015), “Toward more marketing research on gender inequality”, Journal of Marketing Management, Vol. 31 Nos 15/16, pp. 1718-1722.
Fitzimons, G.J. and Lehmann, D.R. (2004), “Reactance to recommendations: when unsolicited advice yields contrary responses”, Marketing Science, Vol. 23 No. 1, pp. 82-94.
Fox, R.J., Krugman, D.M., Fletcher, J.E. and Fischer, P.M. (1998), “Adolescents' attention to beer and cigarette print ads and associated product warnings”, Journal of Advertising, Vol. 27 No. 3, pp. 57-68.
Fredrickson, B.L. and Roberts, T.A. (1997), “Objectification theory: toward understanding women's lived experiences and mental health risks”, Psychology of Women Quarterly, Vol. 21 No. 2, pp. 173-206.
Fugate, D.L. and Phillips, J. (2010), “Product gender perceptions and antecedents of product gender congruence”, Journal of Consumer Marketing, Vol. 27 No. 3, pp. 251-261.
Garst, J. and Bodenhausen, G.V. (1997), “Advertising's effects on men's gender role attitudes”, Sex Roles, Vol. 36 Nos 9/10, pp. 551-572.
Grau, S.L. and Zotos, Y.C. (2016), “Gender stereotypes in advertising: a review of current research”, International Journal of Advertising, Vol. 35 No. 5, pp. 761-770.
Gulas, C.S. and McKeage, K. (2000), “Extending social comparison: an examination of the unintended consequences of idealized advertising imagery”, Journal of Advertising, Vol. 29 No. 2, pp. 17-28.
Gunther, A.C. and Storey, J.D. (2003), “The influence of presumed influence”, Journal of Communication, Vol. 53 No. 2, pp. 199-215.
Häfner, M. (2003), “How dissimilar others may still resemble the self: assimilation and contrast after social comparison”, Journal of Consumer Psychology, Vol. 14 Nos 1/2, pp. 187-196.
Halliwell, E. and 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, Vol. 23 No. 1, pp. 104-122.
Harper, B. and Tiggemann, M. (2008), “The effect of thin ideal media images on women’s self-objectification, mood, and body image”, Sex Roles, Vol. 58 Nos 9/10, pp. 649-657.
Hayes, A.F., Preacher, K.J. and Myers, T.A. (2011), “Mediation and the estimation of indirect effects in political communication research”, Sourcebook for Political Communication Research: Methods, Measures, and Analytical Techniques, Vol. 23, pp. 434-465.
Henderson-King, D., Henderson-King, E. and Hoffmann, L. (2001), “Media images and women’s self-evaluations: social context and importance of attractiveness as moderators”, Personality and Social Psychology Bulletin, Vol. 27 No. 11, pp. 1407-1416.
Hofstede, G. (2001), Culture's Consequences: Comparing Values, Behaviors, Institutions and Organizations across Nations, Sage, Thousand Oaks, CA.
Hogg, M.K., Bruce, M. and Hough, K. (1999), “Female images in advertising: the implications of social comparison for marketing”, International Journal of Advertising, Vol. 18 No. 4, pp. 445-473.
Hong, S. (1992), “Hong's psychological reactance scale: a further factor analytic validation”, Psychological Reports, Vol. 70 No. 2, pp. 512-514.
Huhmann, B.A. and Limbu, Y.B. (2016), “Influence of gender stereotypes on advertising offensiveness and attitude toward advertising in general”, International Journal of Advertising, Vol. 35 No. 5, pp. 846-863.
Jaffe, L.J. and Berger, P.D. (1994), “The effect of modern female sex role portrayals on advertising effectiveness”, Journal of Advertising Research, Vol. 34 No. 4, pp. 32-42.
Janssen, D.M. and Paas, L.J. (2014), “Moderately thin advertising models are optimal, most of the time: moderating the quadratic effect of model body size on ad attitude by fashion leadership”, Marketing Letters, Vol. 25 No. 2, pp. 167-177.
Johnson, G.D. and Grier, S.A. (2012), “What about the intended consequences?”, Journal of Advertising, Vol. 41 No. 3, pp. 91-106.
Jones, M. (1991), “Gender stereotyping in advertisements”, Teaching of Psychology, Vol. 18 No. 4, pp. 231-233.
Karsay, K., Knoll, J. and Matthes, J. (2018), “Sexualizing media use and self-objectification: a meta-analysis”, Psychology of Women Quarterly, Vol. 42 No. 1, pp. 9-28.
Kaufman, G. (1999), “The portrayal of men’s family roles in television commercials”, Sex Roles, Vol. 41 Nos 5/6, pp. 439-458.
Kim, K., Hayes, J.L., Avant, J.A. and Reid, L.N. (2014), “Trends in advertising research: a longitudinal analysis of leading advertising, marketing, and communication journals. 1980 to 2010”, Journal of Advertising, Vol. 43 No. 3, pp. 296-316.
Knoll, S., Eisend, M. and Steinhagen, J. (2011), “Gender roles in advertising: measuring and comparing gender stereotyping on public and private TV channels in Germany”, International Journal of Advertising, Vol. 30 No. 5, pp. 867-888.
Lavine, H., Sweeney, D. and Wagner, S.H. (1999), “Depicting women as sex objects in television advertising: effects on body dissatisfaction”, PSPB, Vol. 25 No. 8, pp. 1049-1058.
Lee, Y.K. (2014), “Gender stereotypes as a double-edged sword in political advertising: persuasion effects of campaign theme and advertising style”, International Journal of Advertising, Vol. 33 No. 2, pp. 203-234.
Lee, R.M. and Robins, S.B. (1995), “Measuring belongingness: the social connectedness and the social assurance scales”, Journal of Counseling Psychology, Vol. 42 No. 2, pp. 232-241.
Liljedal, K.T., Berg, H. and Dahlén, M. (2020), “Effects of nonstereotyped occupational gender role portrayal in advertising. How showing women in male-stereotyped job roles sends positive signals about brands”, Journal of Advertising Research, Vol. 60 No. 2, pp. 179-196.
Lockwood, P. and Kunda, Z. (1997), “Superstars and me: predicting the impact of role models on the self”, Journal of Personality and Social Psychology, Vol. 73 No. 1, pp. 91-103.
Lorenzen, L.A., Grieve, F.G. and Thomas, A. (2004), “Brief report. Exposure to muscular male models decreases men’s body satisfaction”, Sex Roles, Vol. 51 Nos 11/12, pp. 743-748.
Martin, M.C. and Gentry, J.W. (1997), “Stuck in the model trap: the effects of beautiful models in ads on female pre-adolescents and adolescents”, Journal of Advertising, Vol. 26 No. 2, pp. 19-33.
Martin, M.C. and Kennedy, P.F. (1993), “Advertising and social comparison: consequences for female preadolescents and adolescents”, Psychology and Marketing, Vol. 10 No. 6, pp. 513-530.
Mayer, J.M. and Peev, P. (2017), “Do sexy mouthwash ads leave a bad aftertaste? The interaction of sexual self-schema and brand positioning fit on female-targeted sexual advertising efficacy”, Marketing Letters, Vol. 28 No. 2, pp. 321-333.
Meyers-Levy, J. and Loken, B. (2015), “Revisiting gender differences: what we know and what lies ahead”, Journal of Consumer Psychology, Vol. 25 No. 1, pp. 129-149.
Middleton, K., Turnbull, S. and de Oliveira, M.J. (2019), “Female portrayals in Brazilian advertising: are outdated cultural stereotypes preventing change?”, International Journal of Advertising, Vol. 39 No. 5, doi: 10.1080/02650487.2019.1658428.
Milkie, M.A. (1999), “Social comparisons, reflected appraisals, and mass media: the impact of pervasive beauty images on Black and White girls' self-concepts”, Social Psychology Quarterly, Vol. 62 No. 2, pp. 190-210.
Moorman, M.P., Neijens, P.C. and Smit, E.G. (2002), “The effects of magazine-induced psychological responses and thematic congruence on memory and attitude toward the ad in a real-life setting”, Journal of Advertising, Vol. 31 No. 4, pp. 27-40.
Morgan, S.E. and Reichert, T. (1999), “The message is in the metaphor: assessing the comprehension of metaphors in advertisements”, Journal of Advertising, Vol. 28 No. 4, pp. 1-12.
Morrison, M.M. and Shaffer, D.R. (2003), “Gender-role congruence and self-referencing as determinants of advertising effectiveness”, Sex Roles, Vol. 49 Nos 5/6, pp. 265-275.
Noguti, V. and Russell, C.A. (2014), “Normative influences on product placement effects: alcohol brands in television series and the influence of presumed influence”, Journal of Advertising, Vol. 43 No. 1, pp. 46-62.
O’Donohoe, S. (1994), “Advertising uses and gratifications”, European Journal of Marketing, Vol. 28 Nos 8/9, pp. 52-75.
Obermiller, C., Spangenberg, E. and MacLachlan, D.L. (2005), “Ad skepticism: the consequences of disbelief”, Journal of Advertising, Vol. 34 No. 3, pp. 7-17.
Orth, U.R. and Holancova, D. (2004), “ Men’s and women’s responses to sex role portrayals in advertisements”, International Journal of Research in Marketing, Vol. 21 No. 1, pp. 77-88.
Palan, K.M. (2001), “Gender identity in consumer behavior research: a literature review and research agenda”, Academy of Marketing Science Review, Vol. 10 No. 2001, pp. 1-31.
Plakoyiannaki, E. and Zotos, Y. (2009), “Female role stereotypes in print advertising. Identifying associations with magazine and product categories”, European Journal of Advertising, Vol. 43 Nos 11/12, pp. 1411-1434.
Pollay, R.W. (1986), “The distorted mirror: reflections on the unintended consequences of advertising”, Journal of Marketing, Vol. 50 No. 2, pp. 18-36.
Pounders, K. (2018), “Are portrayals of female beauty in advertising finally changing?”, Journal of Advertising Research, Vol. 58 No. 2, pp. 133-137.
Pratto, F. and Bargh, J.A. (1991), “Stereotyping based on apparently individuating information: trait and global components of sex stereotypes under attention overload”, Journal of Experimental Social Psychology, Vol. 27 No. 1, pp. 26-47.
Puntoni, S., Schroeder, J.E. and Ritson, M. (2010), “Meaning matters”, Journal of Advertising, Vol. 39 No. 2, pp. 51-64.
Ricciardelli, R., Clow, K.A. and White, P. (2010), “Investigating hegemonic masculinity: portrayals of masculinity in men’s lifestyle magazines”, Sex Roles, Vol. 63 Nos 1/2, pp. 64-78.
Richins, M.L. (1991), “Social comparison and the idealized images of advertising”, Journal of Consumer Research, Vol. 18 No. 1, pp. 71-83.
Rutledge Shields, V. and Heinecken, D. (2001), Measuring up. How Advertising Affects Self-Image, University of PA Press, Philadelphia, PA.
Schroeder, J.E. and Borgerson, J.L. (1998), “Marketing images of gender: a visual analysis”, Consumption Markets and Culture, Vol. 2 No. 2, pp. 161-201.
Sobande, F., Mimoun, L. and Torres, L.T. (2020), “Soldiers and superheroes needed! Masculine archetypes and constrained bodily commodification in the sperm donation market”, Marketing Theory, Vol. 20 No. 1, pp. 65-84.
Steele, C.M., Spencer, S.J. and Aronson, J. (2002), “Contending with group image: the psychology of stereotype and social identity threat”, Advances in Experimental Social Psychology, Vol. 34 No. 2002, pp. 379-440.
Sun, Y., Pan, Z. and Shen, L. (2008), “Understanding the third‐person perception: evidence from a meta‐analysis”, Journal of Communication, Vol. 58 No. 2, pp. 280-300.
Taylor, C.R. and Stern, B.B. (1997), “Asian-Americans: television advertising and the ‘model minority’ stereotype”, Journal of Advertising, Vol. 26 No. 2, pp. 47-61.
Thorbjørnsen, H. and Dahlén, M. (2011), “Customer reactions to acquirer-dominant mergers and acquisitions”, International Journal of Research in Marketing, Vol. 28 No. 4, pp. 332-341.
Tuncay Zayer, L. and Coleman, C.A. (2015), “Advertising professionals’ perceptions of the impact of gender portrayals on men and women: a question of ethics?”, Journal of Advertising, Vol. 44 No. 3, pp. 1-12.
Tuncay Zayer, L., McGrath, M.A. and Castro-González, P. (2020), “Men and masculinities in a changing world: (de)legitimizing gender ideals in advertising”, European Journal of Marketing, Vol. 54 No. 1, pp. 238-260.
Ulrich, I. (2013), “The effect of consumer multifactorial gender and biological sex on the evaluation of cross-gender brand extensions”, Psychology and Marketing, Vol. 30 No. 9, pp. 794-810.
Van Hellemont, C. and Van Den Bulck, H. (2012), “Impacts of advertisements that are unfriendly to women and men”, International Journal of Advertising, Vol. 31 No. 3, pp. 623-656.
van Tilburg, M., Lieven, T., Herrmann, A. and Townsend, C. (2015), “Beyond ‘pink it and shrink it’. Perceived product gender, aesthetics, and product evaluation”, Psychology and Marketing, Vol. 32 No. 4, pp. 422-437.
Wan, F., Faber, R.J. and Fung, A. (2003), “Perceived impact of thin female models in advertising: a cross-cultural examination of third person perception and its impact on behaviors”, Asia Pacific Journal of Marketing and Logistics, Vol. 15 Nos 1/2, pp. 51-73.
White, M.J. and White, G.B. (2006), “Implicit and explicit occupational gender stereotypes”, Sex Roles, Vol. 55 Nos 3/4, pp. 259-266.
Wiles, J.A., Wiles, C.R. and Tjernlund, A. (1995), “A comparison of gender role portrayals in magazine advertising”, European Journal of Marketing, Vol. 29 No. 11, pp. 35-49.
Windels, K. (2016), “Stereotypical or just typical: how do US practitioners view the role and function of gender stereotypes in advertisements?”, International Journal of Advertising, Vol. 35 No. 5, pp. 864-887.
Wolin, L.D. (2003), “Gender issues in advertising: an oversight synthesis of research 1970-2002”, Journal of Advertising Research, Vol. 43 No. 1, pp. 111-129.
Youn, S., Faber, R.J. and Shah, D.V. (2000), “Restricting gambling advertising and the third-person effect”, Psychology and Marketing, Vol. 17 No. 7, pp. 633-649.
Zawisza, M. and Cinnirella, M. (2010), “What matters more – breaking tradition or stereotype content? Envious and paternalistic gender stereotypes and advertising effectiveness”, Journal of Applied Social Psychology, Vol. 40 No. 7, pp. 1767-1797.
Zhao, X., Lynch, J.G. and Chen, Q. (2010), “Reconsidering Baron and Kenny: myths and truths about mediation analysis”, Journal of Consumer Research, Vol. 37 No. 2, pp. 197-206.
Zimmerman, A. and Dahlberg, J. (2008), “The sexual objectification of women in advertising: a contemporary cultural perspective”, Journal of Advertising Research, Vol. 48 No. 1, pp. 71-79.
Acknowledgements
Jan Wallanders och Tom Hedelius stiftelse W2016-0498:1