ABSTRACT
Stereotyped occupational gender-role portrayals still are used frequently in advertising. Despite this, no previous research has examined their effects. Tentative findings from the advertising industry and some recent research suggest that other types of nonstereotyped gender-role portrayals in advertising can have a number of positive effects. This article corroborates these findings in three empirical studies that demonstrate the positive brand-related and social effects of nonstereotyped occupational gender-role portrayals in advertising on respondents, regardless of their gender. The findings also reveal that the positive effects of nonstereotyped occupational gender-role portrayals can be explained by signaling mechanisms. This article thus contributes to the literature on stereotyped gender-role portrayals as well as to the wider application of signaling theory and the growing research interest in how social and brand-related effects are connected.
MANAGEMENT SLANT
Nonstereotyped occupational gender-role portrayals in advertising have positive effects on attitudes toward the advertisement and the brand, perceived product quality, and social connectedness.
The findings suggest that advertisers may use nonstereotyped occupational gender-role portrayals to achieve signaling effects, in line with how consumers make inferences about brands and their offerings on the basis of how they advertise.
Social connectedness, in turn, has a positive impact on attitudes toward the advertisement and the brand, which lends support to the growing interest in how advertising simultaneously may have a positive impact on both society and advertisers.
INTRODUCTION
One of the most common and prevailing gender-stereotype components in advertising is the stereotyped depiction of occupational gender roles (Eisend, 2010). This may be about to change, however. The advertising industry lately has recognized the need for further investigation of stereotyped gender portrayals in advertisements and has begun to take action.
A number of initiatives have been taken to reduce the use of gender stereotypes in advertisements, such as the #SeeHer initiative and the Cannes Glass Lion award. The Cannes Lions organization created the Glass Lion award in 2015 to promote work in advertising that challenges gender biases in media and marketing (Toure, 2015). The #SeeHer campaign was founded in 2016 by the Association of National Advertisers (ANA) with the goal to achieve a 20 percent increase in the accurate portrayal of girls and women in media by 2020 (Neff, 2016). The #SeeHer organization recently reported that it already has reached its stated goal for improving gender portrayals in advertising. In light of this development, it is imperative for advertising research to contribute to the process by providing the advertising industry with more research about the effects of less-stereotyped gender portrayals in advertisements and the mechanisms that explain them.
The use of gender stereotypes in advertising has been studied extensively since the 1970s, although most research studies to date have focused on documenting the prevalence of gender stereotypes rather than on examining their effects on consumer responses to the advertisements featuring them (Åkestam, Rosengren, and Dahlen, 2017a; Eisend, 2010; Venkatesan and Losco, 1975). Some recent findings from both the advertising industry and research indicate positive effects of advertisements featuring less-stereotyped gender portrayals (Åkestam, 2017; Åkestam et al., 2017a). These positive effects counter both earlier research (Eisend, Plagemann, and Sollwedel, 2014; Windels, 2016) and what still is believed to be the overarching practice of the advertising industry's inclusion of gender-based stereotypes in advertising (Eisend, 2019). Stereotyped portrayals of women and men are still very common in advertising, especially if compared with the equality standard, which posits that men and women should be represented equally in advertising with regard to, for example, occupational roles (Eisend, 2010, 2019; Windels, 2016).
Now, when more advertisers are attempting to replace stereotyped gender portrayals with nonstereotyped ones, research on the latter urgently is needed to understand when and why these portrayals may have positive effects. The present article adds to the new research stream by focusing on the stereotype category that is most prevalent in advertising—namely, occupational gender-role stereotypes (Eisend, 2010). To the authors' knowledge, no previous research has focused on the effects of using stereotyped occupational gender-role portrayals in advertising. Some recent advertising research, however, has been able to demonstrate positive effects for advertisements, brands, and products using other types of nonstereotyped gender portrayals (Åkestam, 2017; Åkestam et al., 2017a; Eisend et al., 2014; Feiereisen, Broderick, and Douglas, 2009; Huhmann and Limbu, 2016).
The present article therefore aimed to examine the effects of nonstereotyped occupational gender-role portrayals on attitudes toward advertisements, brand attitudes, and perceptions of product quality. More specifically, in line with the advertising-industry initiatives to decrease the stereotyped representation of women in advertising (e.g., #SeeHer, 2019), this research focused on nonstereotyped depictions of women in occupational roles that are stereotyped as masculine.
The authors proposed that signaling theory might be relevant to investigate as a means of explaining the effects of nonstereotyped gender portrayals, in line with the previous research in advertising on signaling effects (e.g., Dahlén, Rosengren, and Törn, 2008; Modig, Dahlén, and Colliander, 2014; Rosengren, Dahlén, and Modig, 2013). Signaling theory typically is used to explain how consumers infer information about brands, products, and services (Kirmani and Rao, 2000). The information available to consumers often is a brand's advertising, which is why signaling theory is particularly suitable to advertising research. The current study examined whether nonstereotyped occupational gender-role portrayals can be considered as a signal in advertising (Dahlén et al., 2008).
The authors also suggested that nonstereotyped occupational gender-role portrayals may contribute positively, not only toward advertisement and brand-related effects, but also toward social effects. Social effects of advertising are another understudied topic; only about 1 percent of studies in advertising research include social effects (Eisend, 2019; Kyongseok, Hayes, Avant, and Reid, 2014), although these have been assumed to have an impact on brand-related effects (Eisend, 2016). Only a few studies have investigated this relationship, but so far they seem to provide evidence that this may very well be the case (Åkestam, Rosengren, and Dahlen, 2017b).
The present article extends signaling theory and places it in the context of gender stereotypes in advertising. The authors propose that the signaling of brand effort in using nonstereotyped occupational gender-role portrayals in advertising increases viewers' social connection with the people portrayed in the advertisements, leading to advertisement and brand-related effects, in line with signaling theory.
Additionally, the findings in this article hold for respondents regardless of their gender and thereby contradict some previous research that has demonstrated differing effects of stereotyped advertising for women and men (e.g., Bower, 2001; Eisend et al., 2014; Whipple and Courtney, 1985). This is an important contribution that, together with the findings on social effects, points toward the societal impact of advertising, where nonstereotyped gender portrayals in advertising may benefit both advertisers and consumers, regardless of their gender. The findings in this article are also in conflict with the continued prevalence of stereotyped gender depictions in advertising (Eisend, 2010, 2019; Knoll, Eisend, and Steinhagen, 2011; Middleton, Turnbull, and de Oliveira, 2019; Windels, 2016).
In three studies, the authors demonstrate positive signaling effects of nonstereotyped occupational gender-role portrayals in advertisements and show that these results held for both men and women. These signaling effects, in turn, had positive effects on attitudes toward the advertisement, perceptions of product quality, and attitude toward the brand. Additionally, positive effects of nonstereotyped occupational gender-role portrayals on social connectedness were indicated, which extends the signaling perspective of brands' use of nonstereotyped gender roles.
The studies in this article also show that social effects and brand-related effects seem to be connected, given that social connectedness was found to affect consumers' attitudes toward both advertisement and brand. This article thus contributes to the research on stereotyped gender-role portrayals by adding research on the effects of nonstereotyped occupational gender-role portrayals in advertising. The article also offers signaling theory as the explanatory mechanism for the results, through which social effects also are included.
LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
Stereotyped Gender-Role Portrayals in Advertising
Stereotypes can be described as a set of generalized beliefs about the characteristics and attributes of a specific group (Bodenhausen and Richeson, 2010). Gender stereotypes thus are the beliefs that certain characteristics and attributes differentiate men and women (Eisend, 2010). Gender stereotypes consist of a number of separate components that each have a masculine and a feminine version—namely, physical appearance, traits, role behaviors, and occupations (Deaux and Lewis, 1984). In advertising, a stream of research based on content analyses has documented stereotyped gender portrayals in terms of these components (Eisend, 2010; Grau and Zotos, 2016; Knoll et al., 2011).
Although advertisements portraying women in decorative and sexualized roles (i.e., using physical-appearance stereotypes) have received much criticism throughout the years, the most commonly occurring gender-stereotype component in advertising actually relates to occupational roles (Eisend, 2010). The existence of gender stereotypes in most societies regarding occupational roles is well documented (Helwig, 1998). Even young children display advanced knowledge of gendered occupational-role stereotypes—for example, that men more likely are medical doctors than nurses (Wilbourn and Kee, 2010).
The negative effects on women of being exposed to stereotyped portrayals in advertisements are well documented, such as the negative effects of self-comparison with decorative models on the body image of young women (Dittmar and Howard, 2004). Occupational gender-role stereotypes obviously also can have negative effects on women, especially in terms of career disadvantages (Eisend, 2010). Occupational gender-role stereotypes also have negative effects on men. Men and boys tend to consider fewer occupational choices to be available to them, because they only will consider occupations that align with the male occupational-role stereotype, whereas women and girls will consider a wider range of occupations (Helwig, 1998; Wilbourn and Kee, 2010).
Despite the criticism, advertising containing gender stereotypes has prevailed, possibly because people implicitly assume that it still can improve brand evaluations and sales (e.g., Eisend, 2019; Middleton et al., 2019; Windels, 2016). It has been argued, for example, that men may favor more traditional stereotypes in advertising, whereas women prefer more nonstereotyped depictions (Eisend et al., 2014; Whipple and Courtney, 1985). The naïve beliefs in the generally positive effects of stereotypes in advertising have persevered even though negative brand effects have been documented often. Female consumers in particular sometimes display reactance against stereotyped portrayals of women in advertising (Bower, 2001; Bower and Landreth, 2001). A recent stream of advertising research, conversely, has indicated positive effects for advertisements, brands, and products using types of nonstereotyped gender portrayals other than occupational, such as stereotyped depictions related to physical appearances (Åkestam, 2017; Åkestam et al., 2017a; Feiereisen et al., 2009; Huhmann and Limbu, 2016) and role behaviors (De Meulenaer, Dens, De Pelsmacker, and Eisend, 2018; Eisend et al., 2014).
There also have been recent indications from the advertising industry that nonstereotyped gender portrayals in advertisements can have positive effects on consumer responses to the advertisements, products, and brands. Together with the ANA, the previously mentioned #SeeHer organization has developed a data-tracking methodology called the Gender Equality Measure to measure perceptions of how women are portrayed in advertising and programming. Their data analyses show that television advertisements with positive GEM scores are liked more and have more positive effects on purchase intentions and brand reputation (#SeeHer, 2019).
This article adds to the literature by studying the effects of reversing the most commonly occurring gender stereotype: stereotyped occupational gender-role portrayals. In line with the current initiatives in the advertising industry to improve how women are portrayed in advertising, such as #SeeHer, this article focuses on nonstereotyped depictions of women in occupational roles that are stereotyped as masculine. Following the positive results for using other types of nonstereotyped depictions and the findings from the advertising industry, the authors of the current article expected positive effects from breaking the occupational gender-role stereotype.
To explain these positive effects, the authors turn to signaling theory. One might expect that using nonstereotyped gender would have the same effects as using creative advertising, in that it increases consumer processing of the advertising (Rosengren et al., 2013). In other words, nonstereotyped advertising is not itself equal to creative advertising, but it may signal information to recipients in much the same way.
From a cognitive perspective, stereotypes can be described as a type of heuristic, in that they facilitate the processing of information, which then requires less cognitive effort (Macrae, Milne, and Bodenhausen, 1994; Sherman, Macrae, and Bodenhausen, 2000). This applies also to gender stereotypes, such as the occupational gender-role stereotype that men more likely will be medical doctors (Pratto and Bargh, 1991; White and White, 2006). Nonstereotyped occupational gender-role portrayals in advertisements consequently are processed more extensively, because the stereotype cannot be used to facilitate the information processing. The authors of the current article thus argue that nonstereotyped gender portrayals in advertisements will have much the same effect as creative advertising, which also is processed more, and that employing them could be another way to cut through the clutter and reach the intended audience for an advertisement (Danaher, Bonfrer, and Dhar, 2008; Rosengren et al., 2013).
Following the increased processing of advertisements featuring nonstereotyped gender portrayals of occupational roles and in line with the research on signaling effects, the current authors proposed that there would be positive effects on attitudes toward the advertisements, the perceived quality of products featured in them, and attitudes for the brands. This reasoning is in line with findings from the industry and the previous research that indicated positive effects of advertisement with other types of nonstereotyped gender portrayals (e.g., Åkestam, 2017; Åkestam et al., 2017a; Eisend et al., 2014). The authors of the current article expected that advertisements, which included nonstereotyped occupational gender-role portrayals, would result in more favorable effects on advertisement attitudes than advertisements that feature more stereotyped gender-role portrayals.
H1: Attitudes toward advertisements portraying nonstereotyped occupational gender roles will be more positive than attitudes for advertisements portraying stereotyped occupational gender roles.
In line with this reasoning, the authors also expected that advertisements featuring nonstereotyped occupational gender-role portrayals would have more favorable effects on perceptions of product quality and brand attitudes.
H2: Products in advertisements portraying nonstereotyped occupational gender roles will be perceived to be of higher quality than products in advertisements portraying stereotyped occupational gender roles.
H3: Attitudes toward brands whose advertisements portray nonstereotyped occupational gender roles will be more positive than attitudes toward brands whose advertisements portray stereotyped occupational gender roles.
As an explanation for the positive effects of portraying nonstereotyped occupational gender roles in advertising hypothesized in H1–H3, the authors of the present article proposed signaling theory.
Advertising as a Signal of Brand Effort and Ability
Signaling theory typically is used to describe how consumers infer information about brands, products, and services that they have not experienced themselves (Kirmani and Rao, 2000). The underlying rationale is that there is an information asymmetry between the advertising brand and its potential customers, and different types of signals may help the potential customer in shaping his or her view of the brand and its offering. Early research investigated different types of marketing signals that guide consumer evaluations and choice, such as warranties and price (Kirmani and Rao, 2000), as well as perceived advertising expense, in terms of, for example, advertising formats and repetition (Kirmani, 1990, 1997; Kirmani and Wright, 1989) and found that these positively affected attitudes toward the brand and perceptions of product quality. Later, advertising creativity also was categorized as a marketing signal that influences consumers' perceptions of the brand (Dahlén et al., 2008). Signaling theory lately has been demonstrated to be rather versatile, in that it also can help explain similar effects in business-to-business, recruitment, and investor relations (Dahlen, Rosengren, and Karsberg, 2018).
In this article, the current authors suggest that brands' use of nonstereotyped occupational gender-role portrayals in their advertisements will evoke similar reactions as when brands use, for example, creative or expensive advertising. Both creative and expensive advertising signal perceived brand effort (Ambler and Hollier, 2004; Dahlen et al., 2008; Modig et al., 2014). That is, the higher the perceived level of creativity is or the more expensive the advertisement is perceived to be, the more effort the brand is thought to have put behind the advertisement. The reasoning goes that the brand would not have put the extra effort in unless it were worth it, because the brand would not have invested time and resources that could not be offset by revenues following the advertising campaign (Ambler and Hollier, 2004; Dahlen et al., 2008; Kirmani, 1997; Kirmani and Wright, 1989).
The academic interest in perceived brand effort lies in its effects. It has been found to mediate the effect of advertising creativity on more favorable advertising attitudes (Dahlen et al., 2018) as well as brand attitudes (Modig et al., 2014). Connecting with the earlier reasoning, the authors suggest that a brand's usage of nonstereotyped occupational gender-role portrayals in advertising can have similar effects as creative advertising, because in both cases they break with the “normal,” more stereotyped advertising. This would lead to the perception that the brand put more effort into the advertisement—for example, by not routinely adding a man in an occupation perceived as stereotypically masculine but instead portraying a woman in such a role. Advertisements that include nonstereotyped occupational gender-role portrayals will lead to higher levels of perceived brand effort than advertisements that display more stereotyped gender roles. In turn, it is proposed that the higher perceived brand effort will explain the more positive advertisement attitudes hypothesized in H1.
H4: Perceived brand effort will be higher for advertisements portraying nonstereotyped occupational gender roles than for advertisements portraying stereotyped occupational gender roles.
Advertisers that are perceived to have invested more effort than usual in their advertising also are perceived to have higher brand ability (Dahlén et al., 2008, 2018). On the basis of the perceived brand effort (stemming from, e.g., a highly creative advertisement), consumers likely will infer that the brand is smarter and more innovative and thus likely will put this ability to use not only to produce creative advertisements but also to create better, higher quality products (Ambler and Hollier, 2004; Dahlén et al., 2008; Kirmani and Rao, 2000; Kirmani and Wright, 1989). It thus seems plausible that the inclusion of nonstereotyped occupational gender-role portrayals in advertisements will signal not only brand effort but also brand ability. Advertisements that include nonstereotyped occupational gender-role portrayals will lead to higher levels of perceived brand ability than advertisements that display more stereotyped occupational gender roles.
H5: Perceived brand ability will be higher for advertisements portraying nonstereotyped occupational gender roles than for advertisements portraying stereotyped occupational gender roles.
This increase in perceived brand ability, in turn, should affect consumer perceptions of product quality positively with regard to the advertised product, offering an explanation for the effects hypothesized in H2. Positive signals stemming from creative or expensive advertising typically are evaluated by consumer attitudes toward the advertising brand (Dahlen et al., 2018; Modig et al., 2014). In line with the reasoning above, the underlying mechanisms for this are that advertisements that send a positive signal increase consumers' perceptions of brand effort, followed by perceived brand ability, which, in turn, affects the perceived product quality, which affects consumer attitudes toward the brand (Dahlen et al., 2018). These mechanisms are also present for nonstereotyped occupational gender-role portrayals in advertising, offering an explanation for the effects hypothesized in H3.
H6: Brand effort, brand ability, and perceived product quality will mediate the positive effects of advertisements portraying nonstereotyped occupational gender roles on attitudes toward brands.
Social Connectedness in Advertising
Social connectedness is one of several social effects of advertising (Åkestam et al., 2017b). Defined as the extent to which a person feels connected to other people (Hutcherson, Seppala, and Gross, 2008; Lee and Robbins, 1995), it can be directed toward both people in general (Lee and Robbins, 1995) and, as in the case of this article, people portrayed in a specific advertisement (Hutcherson et al., 2008). This type of social effect can be a form of unintended, or extended, effect of advertising, because its effect is not focused primarily on the persuasive element of advertising (Dahlen and Rosengren, 2016).
As mentioned earlier, although negative social effects of stereotypes in advertising have been found (Dittmar and Howard, 2004; Eisend, 2010; Wilbourn and Kee, 2010), social effects tend to be desirable to marketers nevertheless. This partly is due to marketers increasingly wanting to take responsibility for negative effects of advertising and to promote positive effects (Dahlen and Rosengren, 2016) and partly because social effects also are assumed to affect traditional advertising metrics, such as consumer evaluations of the brand (Åkestam et al., 2017b; Eisend, 2016). Recent research predominantly has highlighted examples of positive social effects, such as healthy consumption (Yoon, 2015), empathy (Åkestam et al., 2017b; Escalas and Stern, 2003), creativity (Fitzsimons, Chartrand, and Fitzsimons, 2008; Rosengren et al., 2013), benevolent behaviors (Defever, Pandelaere, and Roe, 2011), and impact on body ideals (Bissell and Rask, 2010).
Returning to social connectedness, this article investigates whether nonstereotyped occupational gender-role portrayals in advertising also can influence consumers' social connectedness with the people portrayed in advertisements. As mentioned, nonstereotyped portrayals require more extensive processing, because a stereotype cannot be used as a heuristic to facilitate the social categorization of the people who are portrayed (Macrae et al., 1994; Sherman et al., 2000). Consumers have to think more about the people in the advertisements.
Nonstereotyped gender portrayals in advertising also can produce thoughts about the fact that they are unusual (Brumbaugh and Grier, 2006; Grier and Brumbaugh, 1999). There are indications that an increase in social thoughts about people in advertisements can have positive effects on consumers' perceptions of social connectedness with them. Other portrayals of people in advertising that challenge advertising stereotypes, such as the portrayal of homosexuality in advertising, create a sense of shared reality by priming the viewer to relate to others as well as themselves, which has positive effects on social connectedness (Åkestam et al., 2017b).
H7: Social connectedness (with the people portrayed in the advertisements) will be higher for advertisements featuring nonstereotyped occupational gender roles compared with advertisements with stereotyped occupational gender roles.
It has been indicated that although social effects can be desirable in their own right (Dahlen and Rosengren, 2016), they also are thought to be a means for brand-related effects (Eisend, 2016). Very few studies, however, have investigated the potential connection between social effects and brand-related effects (Åkestam et al., 2017b; Eisend, 2016). In this article, the authors suggest that social connectedness with the people portrayed in the advertisements can be described as a result of the underlying process of how consumers respond to advertising according to signaling theory.
It therefore is plausible that the perceived brand effort of unusual advertising, such as nonstereotyped portrayals, spurs further thoughts on both the brand and the people portrayed in the advertisement. The fact that the brand has taken the effort to include unusual portrayals in its advertising should evoke more processing on the portrayals, which increases the social connectedness with the portrayed people. This perceived connection improves the attitudes toward the advertisement and, in turn, brand attitudes (Brown and Stayman, 1992) because of feelings of connectedness. The brand is rewarded for its effort in including nonstereotyped portrayals, because of increased processing.
By way of increased perceived brand effort, therefore, nonstereotyped occupational gender-role portrayals will affect consumers' social connectedness with the people portrayed in the advertisements. This, in turn, will affect their attitudes toward the advertisements, followed by their attitudes toward the advertising brands.
H8: Brand effort, social connectedness with the people portrayed in the advertisements, and attitudes toward the advertisements will mediate the positive effects of advertisements portraying nonstereotyped occupational gender roles on attitudes toward brands.
METHODOLOGY
The signaling effects of nonstereotyped occupational gender-role portrayals were investigated in three experimental studies. All studies tested the hypothesized effects on attitudes toward the advertisements (H1), perceived product quality (H2), and attitudes toward the brands (H3) as well as the underlying mechanisms of perceived brand effort (H4) and perceived brand ability (H5). The authors thereby could test for the theoretical mechanisms behind these hypothesized effects, in line with signaling theory. H6 tested the mediating effects of perceived brand effort, perceived brand ability, and perceived product quality. The mechanism for H6 is illustrated by the underlying theoretical path of effects (See Figure 1).
Theoretical Framework for Hypotheses 2–6 in Line with Signaling Theory, as Tested in Study 1, Study 2, and Study 3
Theoretical Framework for Hypotheses 1, 3, 4, 7, and 8, as Tested in Study 2 and Study 3
Studies 2 and 3 additionally included social connectedness (H7), with the aim to extend the signaling perspective of brands' use of nonstereotyped occupational gender roles. The authors thus also tested for the mediating effects of social connectedness (H8 in Study 2 and Study 3; See Figure 2).
STUDY 1
The study used an experimental design, whereby stereotyped versus nonstereotyped occupational gender-role portrayals in advertising were tested. Study 1 employed a convenience sample from MTurk, consisting of 224 participants (61 percent men; aged 24–72 years; mean age = 39 years) after 28 respondents failed one of the three control questions: “Did the ad include pictures of people?” “How many persons were displayed in the ad?” and “Which profession or professions did the ad display?”
As recommended by previous researchers (Peer, Vosgerau, and Acquisti, 2014), all participants had an MTurk approval rate of 98 percent, and attention-check questions (i.e., the control questions) were used. The participants each answered one of four versions of an online questionnaire, in which one of the four advertisements was displayed. The participants were assigned randomly to the four questionnaire versions. Participants each were compensated with $1.
Method
Stimuli. Four print advertisements were created for the purpose of this study: Two versions included a woman in a stereotypically masculine occupational role, and two versions included a man in the same role. Two occupations were included: military (soldier) and medical (doctor).
All four versions consisted of a large photo portraying two or three people, in which the focal person was a man or woman depicted as either a doctor or a soldier. In line with many other experimental studies in advertising research, the brand was made anonymous (in this case, blurred) to avoid confounding effects from previous brand experience (Adil, Lacoste-Badie, and Droulers, 2018; Choi, Taylor, and Lee, 2017; Dahlén, Rosengren, and Smit, 2014; Liljedal, 2016; Peter and Ponzi, 2018). The advertising stimuli are available from the authors of the current study on request.
The degree to which these advertisements were perceived as including nonstereotyped or stereotyped occupational gender-role portrayals was tested in a separate prestudy with a convenience sample (n = 36; 50 percent female; mean age = 31 years) using the items “This is a stereotypical portrayal of gender roles” and “This is a stereotypical portrayal of gender roles in advertising” (1 = “do not agree at all,” 10 = “agree completely”; Cronbach's α = 0.81). The results were satisfactory (MStereotyped = 7.30, MNonstereotyped = 2.88, p < 0.01).
A second prestudy, using a separate MTurk sample (n = 30; 33 percent female; mean age = 37 years) tested attitudes toward the people depicted in the advertisements, explicit occupational gender stereotypes, and perceptions of occupations in the pictures. Attitudes toward the depicted people were measured with the question “What do you think about the people in this picture?” and three items answered on a 10-point semantic differential scale (“bad” [1] to “good” [10], “do not like” [1] to “like” [10], “negative impression” [1] to “good impression” [10]). There were no differences in attitudes toward the depicted people in the pictures depicting either doctors (MFemaleDoctor = 1.58, MMaleDoctor = 2.19, p = 0.08) or soldiers (MFemaleSoldier = 1.83, MMaleSoldier = 1.69, p = 0.63).
Explicit occupational gender stereotypes were measured for the occupations doctor and soldier with the question “On the following scale, where would you place this occupation?” (1 = “masculine,” 10 = “feminine”). Similar scales were used in previous research (White and White, 2006). One-sample t tests indicated that means for both doctor (MDoctor = 4.47, p < 0.01) and soldier (MSoldier = 2.13, p < 0.01) were lower than the midpoint of the scale (5.5), indicating that the occupations were perceived as more masculine than feminine.
Perceptions of occupations in the pictures were measured with the open question “What would you guess are the occupations of the people in the picture?” Responses to the questions indicated that respondents perceived the people in the pictures as doctors and soldiers, respectively, just as intended.
Measures. Five measures were included in the main study. Perceived product quality consisted of three items on a 10-point semantic differential scale:
“low” (1) to “high” (10),
“lower than average” (1) to “higher than average” (10), and
“worse than competing brands” (1) to “better than competing brands” (10) (Dahlén et al., 2008); Cronbach's α = 0.94.
The question included a sentence asking the respondents to “state your spontaneous opinion, even if the product information is limited.”
Attitude toward the advertisement was also measured with three items on a 10-point scale:
“bad” (1) to “good” (10),
“dislike” (1) to “like” (10),
“unpleasant” (1) to “pleasant” (10) (Holbrook, 1987); Cronbach's α = 0.95.
Attitude toward the brand was similarly measured with three items on a 10-point scale:
“bad” (1) to “good” (10),
“dislike” (1) to “like” (10),
“unpleasant” (1) to “pleasant” (10) (Holbrook, 1987); Cronbach's α = 0.96.
For this measure, the respondents also were asked to “state your spontaneous opinion even if the brand is anonymous.”
Finally, to investigate the theoretical reasoning on signaling theory as a potential way to explain the results of this experiment, two items were included. Perceived effort on behalf of the advertiser (Modig et al., 2014) was measured on a 10-point Likert scale (“disagree” (1) to “agree” (10); Cronbach's α = 0.92) with three items:
“I feel that the brand has put a lot of time behind the ad,”
“I feel that the brand has put a lot of effort behind the ad,” and
“I feel that the brand has put a lot of thought behind the ad.”
Perceived ability of the brand was similarly measured with the three items adapted from previous research (Dahlén et al., 2008); Cronbach's α = 0.90:
“The brand is leading,”
“The brand is successful,” and
“The brand is modern.”
Results
To ensure that the stimuli sampling (medical and military occupations) did not create any unexpected interaction effects, the authors conducted a multivariate analysis. Gender was included as a covariate because it may affect how advertising is processed and perceived (Meyers-Levy and Maheswaran, 1991), including advertisements featuring nonstereotyped gender portrayals (Eisend et al., 2014). On the basis of the three dependent variables of the study, gender did not act as a covariate, F(3, 223) = 0.10, ns, and there was no interaction effect between occupation (i.e., military versus medical) and the non-/stereotyped gender portrayal (i.e., man versus woman), F(3, 223) = 1.79, ns. The authors ran a second multivariate analysis to ensure that there was no interaction effect between the respondents' gender and non-/stereotyped gender portrayals specifically, F(3, 223) = 0.27, ns. The complete sample thus was analyzed jointly.
The authors subsequently performed independent-samples t tests to test whether nonstereotyped portrayals of occupational gender roles in advertising produced more positive results than did stereotyped portrayals:
Attitude toward the advertisement (MStereotyped = 6.93, SD = 2.35; MNonstereotyped = 7.77, SD = 2.27), t(222) = −2.71, p < 0.01. H1 thus is supported.
Perceived product quality (MStereotyped = 6.44, SD = 2.12; MNonstereotyped = 6.88, SD = 1.94), t(223) = −1.61, p = 0.05. H2 thus is supported.
Attitude toward the brand (MStereotyped = 6.91, SD = 7.46; MNonstereotyped = 7.46, SD = 2.13), t(223) = −1.90, p < 0.05. H3 thus is supported.
Theoretical Framework for Understanding the Signaling Effects of Nonstereotyped Occupational Gender Role Portrayals in Advertisements on Brand Attitudes: All Hypotheses as Tested in Study 2 and Study 3
Testing whether these results can be explained by signaling theory, the authors ran independent-samples t tests for perceived brand effort and perceived brand ability, respectively, followed by a serial mediation to confirm that the effects of non-/stereotyped occupational gender-role portrayals in advertising were mediated by effort and ability (cf. Dahlen et al., 2018). First, perceived brand effort and ability were shown to be higher for brands using nonstereotyped gender roles in their advertising:
Perceived brand effort (MStereotyped = 5.63, SD = 2.77; MNonstereotyped = 6.64, SD = 2.64), t(223) = −2.79, p < 0.001. H4 thus is supported.
Perceived brand ability (MStereotyped = 6.31, SD = 2.42; MNonstereotyped = 7.32, SD = 2.05), t(223) = −3.39, p < 0.001. H5 thus is supported.
Second, testing the hypothesis that stipulated the underlying theoretical explanations for the direct effects, the authors performed two serial mediation analyses using bootstrap analyses (Preacher and Hayes, 2008) in PROCESS Version 3.0. Testing for H6, which addresses the underlying explanation for the effects found for brand attitude in H3, the authors used Model 6 to test the mediation non-/stereotyped portrayal → perceived brand effort → perceived brand ability → perceived product quality → brand attitude (cf. Dahlen et al., 2018). The authors first tested age and gender as covariates for the mediation, but none were significant, and the mediation analysis thus was run without covariates. The subsequent mediation analysis showed a significant mean indirect effect from the bootstrap analysis of 0.20 (5,000 bootstrap samples, 95 percent confidence interval [CI]: 0.05 to 0.40), which lends support to the explanatory model using signaling theory.
Discussion
The results of the first study show positive effects of using nonstereotyped occupational gender-role depictions in advertising on advertisement attitudes, brand attitudes, and perceived product quality. Study 1 also indicates that the signaling theory provides relevant support to explain the effects found for nonstereotyped occupational gender portrayals in advertising. These portrayals signal that the brand has made an effort, which is rewarded by the viewers in terms of more positive attitudes toward the brand as well as the advertisement itself.
Study 2 was designed to replicate these findings and to include social connectedness with the persons portrayed in the advertisements. It was expected that the perceived increased brand effort would induce the viewers to process the advertisement more, which would increase social connectedness with the person portrayed in the advertisement (H7). This, in turn, was expected to spill over to more favorable attitudes toward the advertisement as well as the brand (H8), as a summary of the proposed hypotheses demonstrates (See Figure 3).
STUDY 2
Method
Study 2 was conducted in the same manner as Study 1, with an experimental design and the same four stimuli advertisements. The convenience sample from Sweden consisted of 151 respondents (58 percent female; 20–71 years old; mean age = 33 years) who were recruited via online forums and messaging tools. Although a convenience sample was used, care was taken to achieve a sample that would reflect the Swedish market in terms of age and gender. No rewards were paid to the respondents.
The same dependent variables were included as in Study 1 to allow for direct replication of the initial findings, and indices were created (every Cronbach's α > 0.86). Additionally, one social effect was measured to investigate whether the stimuli would evoke different levels of feelings of belongingness with the people portrayed in the advertisement. Social connectedness was measured on a 10-point Likert scale (“disagree” [1] to “agree” [10]), with three items adapted from previous research (Hutcherson et al., 2008; Cronbach's α = 0.86):
“I feel like I belong with the people in the ad,”
“I am similar to the people in the ad,” and
“I feel positive towards the people in the ad.”
Results
Once again, because stimuli sampling was used in Study 2, the authors conducted a multivariate analysis to verify that no interaction effect was present between occupation and the non-/stereotyped portrayal and that gender did not operate as a covariate. As in Study 1, testing for gender as a covariate did not produce any significant results, F(4, 143) = 0.75, ns, nor were any interaction effects from the stimuli sampling present, F(4, 143) = 0.58, ns. The total sample thus could be analyzed. Additionally, as in Study 1, the authors ran a second multivariate analysis to ensure that there was no interaction effect between the respondents' gender and non-/stereotyped gender portrayals, F(3, 144) = 0.031, ns.
As expected, independent-samples t tests confirmed that nonstereotyped portrayals of occupational gender roles in advertising produced more positive results than did stereotyped portrayals:
Attitude toward the advertisement (MStereotyped = 4.42, SD = 2.19; MNonstereotyped = 6.01, SD = 2.30), t(150) = −4.39, p < 0.001. H1 thus is supported.
Perceived product quality (MStereotyped = 4.79, SD = 1.95; MNonstereotyped = 5.70, SD = 1.98), t(150) = −2.86, p < 0.01. H2 thus is supported.
Attitude toward the brand (MStereotyped = 4.74, SD = 2.29; MNonstereotyped = 5.80, SD = 2.27), t(150) = −2.88, p < 0.01. H3 thus is supported.
Social connectedness (MStereotyped = 3.91, SD = 2.10; MNonstereotyped = 5.03, SD = 2.47), t(150) = −3.02, p < 0.01. H7 thus also is supported.
Testing whether these results can be explained by signaling theory, the authors ran independent-samples t tests for perceived brand effort and perceived brand ability, followed by mediation analyses to confirm that the effects of non-/stereotyped gender-role portrayals in advertising were mediated by effort and ability (cf. Dahlen et al., 2018). First, replicating the results found in Study 1, perceived brand effort and ability were shown to be higher for brands using nonstereotyped gender roles in their advertising:
Perceived brand effort (MStereotyped = 3.36, SD = 1.90; MNonstereotyped = 5.16, SD = 2.61), t(137) = 4.87, p < 0.001. H4 thus also is supported.
Perceived brand ability (MStereotyped = 3.87, SD = 2.00; MNonstereotyped = 5.36, SD = 2.37), t(150) = −4.18, p < 0.001. H5 thus also is supported.
Second, two mediation analysis were conducted with bootstrap analyses (Preacher and Hayes, 2008) in PROCESS Version 3.0. The first mediation analysis replicated the results from Study 1. To test H6 and address the underlying explanation for the effects found for brand attitude in H3, the authors once again used Model 6 to test the mediation non-/stereotyped portrayal → perceived brand effort → perceived brand ability → perceived product quality → brand attitude (Dahlen et al., 2018). Again, age and gender were added first as covariates, but results indicated that they did not affect the results. They subsequently were removed, and the mediation analysis produced a significant mean indirect effect from the bootstrap analysis of 0.24 (5,000 bootstrap samples; 95 percent CI: 0.09 to 0.45), once again lending support to the explanatory model using signaling theory.
Next, the authors ran one more mediation analysis to test the reasoning that social connectedness may affect attitudes toward both the advertisement and the brand. The authors proposed that the perceptions of brand effort involved in nonstereotyped occupational gender portrayals in advertising may affect social connection toward the people portrayed in the advertisement. This, in turn, should affect positively attitudes toward both the advertisements and the brands.
Using Model 6, the authors thus tested the mediation non-/stereotyped portrayal → perceived brand effort → social connectedness → attitude toward the advertisements → attitude toward the brands. Again, neither gender nor age was needed as a covariate, and thus they were removed. As expected, the mediation analysis showed a significant mean indirect effect from the bootstrap analysis of 0.33 (5,000 bootstrap samples; 95 percent CI: 0.17 to 0.52), thus confirming H8.
Discussion
Study 2 replicates the results from Study 1, in terms of providing results for the positive effects found for nonstereotyped occupational gender-role portrayals in advertising. Indeed, the perceived sense of brand effort and ability underlies the effects on consumer attitudes toward the brand. Social connectedness with the people displayed in the advertisements also responded to the perceived brand effort and subsequently affected attitudes toward both the advertisements and the brands. These findings lend support for the connection between social and brand-related effects. The logic presented in signaling theory thus applies well to explain the positive effects found for brand attitudes, through two different paths of causality: one “traditional” via effort and ability, and one “social” via social connectedness.
The results do not indicate any differences depending on the respondents' gender. This may be because the usage of a nonstereotyped portrayal stands out and is processed more thoroughly, regardless of gender. This increased processing of the portrayed person in the advertisement subsequently may result in more favorable evaluations of the advertisement and brand, as discussed in the reasoning leading up to the hypotheses. Study 3 aimed to investigate whether this is a plausible explanation for the effects, by including a thought protocol.
Study 3 further aimed to replicate the results while also employing a different set of stimuli advertisements, featuring another occupation and only one person in each advertisement. In the advertisements used in Study 1 and Study 2, several people were depicted. This could have led to confusion in terms of the relationships and interactions, such as emotional charge and hierarchical relationship, between the depicted people.
STUDY 3
Study 3 also was conducted with an experimental design but with two different stimuli advertisements. These advertisements (see the Appendix) included a firefighter as an example of a stereotyped occupation. The two stimuli versions furthermore each included only one person in a similar standing position to avoid any impact that might arise from, for example, posturing or composition of the photograph. These stimuli also included a photo of the product to help the respondents make the connection with the perceived product quality. The brand, however, was blurred to avoid confounding effects from previous experience.
Method
Stimuli and Pretests. Two print advertisements were created for the purpose of this study. The authors selected firefighters as a stereotyped occupation, partly because of the ongoing debate surrounding female firefighters (Ward, 2019). The authors also performed a separate pretest using a U.K.-based Prolific sample (N = 30; 60 percent female, 3 percent other; mean age = 31 years) to examine explicit occupational gender stereotypes for firefighter as an occupation. The same measurements were used as in Study 1, and a one-sample t test indicated that the mean rating for firefighter (MFirefighter = 3.43, p < 0.01) was lower than the midpoint of the scale (5.5).
The advertisements in Study 3 featured photos of people and products from the stock photography provider Shutterstock. The photos were pretested in two separate studies, and a license for the selected pictures was purchased from Shutterstock. First, the authors performed a pretest among other advertising researchers (n = 14) to test the similarity between the pictures of men and women. The authors selected nine picture pairs, all consisting of one photo of a male firefighter and one photo of a female firefighter. Picture similarity was measured with the question “How similar or dissimilar are these two pictures?” (1 = “dissimilar,” 10 = “similar”). The final two pictures were selected on the basis of the ratings (MSimilarity = 8.08).
Second, the authors performed another prestudy with a separate U.K.-based Prolific sample (n = 30; 60 percent female; mean age = 32 years) to examine attitudes toward the people depicted in the photos. Attitudes were measured with the same measurements as in Study 1. There were no differences between attitudes toward the depicted people in the two selected photos of firefighters (MFemaleFirefighter = 8.23, MMaleFirefighter = 8.13, p = 0.61). Perceptions of occupations in the pictures were measured with the same measurements as in Study 1. Responses indicated that, as expected, respondents mostly perceived the people in the pictures as firefighters.
The authors then created the stimuli advertisements by combining the selected photographs with a picture of a water bottle from Shutterstock on a plain white background (see the Appendix for the stimuli advertisements).
Respondents and Measures. Like the previous studies, the third study used an experimental design. The sample consisted of 504 participants (50 percent men, 0.2 percent other, 49.8 percent women; aged 18–75 years; mean age = 37 years) recruited from Prolific's U.K.-based pool of nationally representative respondents. The respondents were compensated with £1 each for their participation. All respondents were included in the analysis because none failed the three control questions: “Did the ad include pictures of people?” “How many persons were displayed in the ad?” and “Which profession or professions did the ad display?”
The participants were assigned randomly to answer one of the two versions of the online questionnaire. The same dependent variables were included as in Study 2 to allow for direct replication of the initial findings, and indices were created (every Cronbach's α > 0.77).
Finally, a thought protocol was included in the questionnaire. In the discussion leading up to the hypotheses earlier, it was suggested that the positive effects from nonstereotyped occupational gender roles may stem from increased processing. To test for this, the authors included a thought protocol by asking the respondents simply to write down their thoughts, as many as they liked, after having seen the stimuli. This procedure has been used previously (Åkestam et al., 2017b).
Two of the authors of the current article then coded the listed thoughts independently. For each respondent, they counted the number of thoughts pertaining to the person depicted in the stimuli. This included observations on the occupation (e.g., “He is a firefighter”) and the physical characteristics (e.g., “She has blue eyes”) of the people in the pictures, as well as comments on objects in relation to the depicted person (e.g., “His coat is too big”). Thoughts that related to the occupation generally (e.g., “Firefighting is a tough job”) and objects that were mentioned without relating to the person (e.g., “Red hat”) were not included. An index then was computed on the basis of the coding of both researchers (Pearson r = 0.972, p < 0.01).
Results
First, the authors conducted a multivariate analysis to verify that no interaction effect was present between the non-/stereotyped portrayal and the gender of the respondents. Results confirmed this was the case, F(3, 497) = 1.77, ns.
Replicating the results in Study 1 and Study 2, independent-samples t tests confirmed that nonstereotyped portrayals of gender roles in advertising produced more positive results than did stereotyped portrayals:
Attitude toward the advertisement (MStereotyped = 4.84, SD = 2.09; MNonstereotyped = 5.38, SD = 2.21), t(502) = −2.85, p < 0.01. H1 thus is supported.
Perceived product quality (MStereotyped = 4.35, SD = 2.15; MNonstereotyped = 4.80, SD = 2.16), t(502) = −2.38, p < 0.01. H2 thus is supported.
Attitude toward the brand (MStereotyped = 5.35, SD = 2.14; MNonstereotyped = 5.68, SD = 2.26), t(502) = −1.68, p < 0.05. H3 thus is supported.
Social connectedness (MStereotyped = 3.88, SD = 1.83; MNonstereotyped = 4.28, SD = 1.82), t(502) = −2.42, p < 0.01. H7 thus also is supported.
As above, the authors also ran independent-samples t tests for perceived brand effort and perceived brand ability, followed by mediation analyses to further strengthen the argument that signaling theory can help explain the positive effects. First, replicating the results found in Studies 1 and 2, perceived brand effort and ability were shown to be higher for brands using nonstereotyped occupational gender roles in their advertising:
Perceived brand effort (MStereotyped = 2.99, SD = 2.08; MNonstereotyped = 3.47, SD = 2.36), t(493) = −2.43, p < 0.01. H4 thus also is supported.
Perceived brand ability (MStereotyped = 3.83, SD = 2.08; MNonstereotyped = 4.35, SD = 2.45), t(491) = −2.53, p < 0.01. H5 thus also is supported.
Second, the authors conducted two mediation analyses using bootstrap analyses (Preacher and Hayes, 2008), all of which were run with PROCESS Version 3.0. Both mediation analyses replicated the results from Studies 1 and 2. To address the underlying explanation for the effects found for brand attitude in H3 and test for H6, the authors once again used Model 6 to test the mediation non-/stereotyped portrayal → perceived brand effort → perceived brand ability → perceived product quality → brand attitude (Dahlen et al., 2018).
Age and gender, as well as basic demographics such as occupation and level of education, were included as covariates. Of these, only gender was indicated as a covariate (p < 0.05; 95 percent CI: 0.05 to 0.53). The subsequent mediation analysis thus included gender as a covariate, thereby removing its effect on the mediation. A significant mean indirect effect from the bootstrap analysis of 0.11 (5,000 bootstrap samples; 95 percent CI: 0.02 to 0.21) was found.
Next, the authors ran mediation analyses to replicate further the results from Study 2 concerning H8, which indicates that social connectedness affects attitudes toward the brand. Replicating the final mediation from Study 2, the authors used Model 6 to try the mediation non-/stereotyped portrayal → perceived brand effort → social connectedness → attitude toward the advertisement → attitude toward the brand. Potential covariates (gender, age, occupation, and educational level) were included first, but none produced significant results. Mirroring the previous results and thereby confirming H8, the subsequent mediation analysis showed a significant mean indirect effect from the bootstrap analysis of 0.06 (5,000 bootstrap samples; 95 percent CI: 0.01 to 0.11).
After that, the authors performed an analysis of the data from the thought protocols to examine whether the positive effects from nonstereotyped occupational gender roles might stem from increased processing. An independent-samples t test indicated that there indeed was a difference in the amount of person-related processing stemming from the stimuli featuring stereotyped and nonstereotyped depictions. Results indicate that the non stereotyped occupational gender-role depictions produced more thoughts about the person in the advertisement (MFemaleFirefighter = 1.51, SD = 1.30; MMaleFirefighter = 1.16, SD = 1.18), t(502) = −3.24, p < 0.01.
Discussion
Study 3 further replicates the results from Studies 1 and 2, this time using a different set of stimuli featuring another occupation, including a product display and another photo composition, and using a larger sample to further strengthen the validity of the results. The findings further lend support for the connection between social and brand-related effects as well as the reasoning that nonstereotyped occupational gender roles induce more processing than do stereotyped occupational gender roles. Study 3 strengthens the assumption that there is a connection between social connectedness with the person in the advertisements and attitudes toward the advertisement and brand.
GENERAL DISCUSSION
Results
The findings reported in this article demonstrate several positive effects of turning away from the routine use of a common gender stereotype that is used widely in advertising (Eisend, 2010, 2019): stereotyped occupational gender-role depictions. When the stereotype was reversed in the studies of this article, so that women were seen in occupational roles that were stereotyped as masculine, in the populations in which the studies were carried out, the effects were overwhelmingly positive. In three studies, positive effects on brand and advertisement attitudes were demonstrated and explained with (positive) signaling effects and social effects of the advertisements.
These findings are in line with the recent research findings of positive effects from using other types of nonstereotyped gender depictions in advertising (Åkestam, 2017; Eisend et al., 2014) as well as with the preliminary findings from the advertising industry concerning positive effects of the improved representation of women (Neff, 2016). This article thus adds another component of gender stereotypes to the literature and corroborates previous research and industry findings about the positive effects of nonstereotyped depictions of gender in advertising.
This article also adds to the research demonstrating that when brands do something beyond what is expected, positive results emerge because of consumers' perceptions of brand effort and brand ability (Ambler and Hollier, 2004; Dahlén et al., 2008, 2018; Kirmani and Wright, 1989; Rosengren et al., 2013). Nonstereotyped occupational gender-role portrayals thus join other marketing signals, such as advertising creativity and expense, that advertisers can employ to signal effort. An additional contribution of this article is the connection between social and brand-related effects and the finding that these can be explained with signaling theory.
Previous research already has pointed toward several beneficial effects of social connectedness whereby it is regarded to contribute toward a more aggregated society (Åkestam et al., 2017b) and correlates with both physical and mental health (Hutcherson et al., 2008). When the perceived effort on behalf of the brand was “awarded” more attention, the nonstereotyped occupational gender-role portrayals were processed more and evoked more social thoughts, which translated into both social connectedness and, consequently, more positive attitudes toward the advertisement and the brand. Thinking a bit more about and feeling a little closer to the people portrayed in the advertisement made consumers like the advertisement and the brand a little more.
Combined with the effects found in this article, further research into the social effects of advertising looks promising for advertisers, consumers, and society. It is interesting that both proposed paths of causality found support in the studies. The positive effects of nonstereotyped occupational gender roles in advertising on attitudes toward the brands stemmed from both the previously proposed chain of effects in signaling theory (perceived effort → perceived ability → perceived product quality; Dahlen et al., 2018) and the currently proposed path, which includes social effects (perceived effort → social connectedness → attitudes toward the advertisements). This further validates signaling theory as the underlying explanation for the demonstrated effects.
Managerial Implications
The findings presented in this article should be seen as a strong encouragement to advertisers who consider using nonstereotyped occupational gender-role depictions in their advertising. To many advertisers, the use of such stereotyped occupational gender-role depictions as were included in this article may not seem like such a problem. Occupational gender-role stereotypes have been used in marketing for a long time and are still very common (Eisend, 2010, 2019; Grau and Zotos, 2016; Knoll et al., 2011; Middleton et al., 2019). Positive effects of using occupational gender-role stereotypes in advertising also have been demonstrated, at least for some groups of consumers (Eisend et al., 2014; Windels, 2016).
Given all this, advertisers easily could remain in the belief that the continued use of stereotyped occupational gender-role depictions is a tried and tested method with a proven effectiveness. The authors of the current article hope that the findings from these studies, together with industry initiatives to decrease the use of gender stereotypes in advertising, can support advertisers in their decisions to use less-stereotyped depictions.
The stereotyped occupational gender-role depictions included in this article well may seem like harmless advertisement pictures to many advertisers. After all, most firefighters and soldiers are still men, so what is the harm of showing men in such occupational roles in advertising? Advertisers need to be aware that the use of such stereotyped depictions risks perpetuating occupational gender-role stereotypes that are harmful to both women and men (Helwig, 1998; Wilbourn and Kee, 2010).
Although occupational gender-role stereotypes are diminishing in many cultures, they still persistently remain (White and White, 2006). It well may be the time for the advertising industry to do what it can to reduce the harmful gender stereotypes. As demonstrated in this article, advertisers even can be rewarded for their efforts.
This article also suggests that practitioners may do right in turning their attention toward the social effects of advertising. The results indicate that instead of spending heavily on media budgets to accomplish advertiser effort (Kirmani, 1997; Kirmani and Wright, 1989; Modig et al., 2014), advertisers instead may use nonstereotyped gender portrayals to signal brand effort and ability, with positive consequences for the brand. As advertising avoidance (Baek and Morimoto, 2012; Dix and Phau, 2017), including the use of advertising blockers (McDonald, 2018), increases, being able to find ways for consumers to connect with advertising should be a focus for advertisers and advertising agencies. Studies 2 and 3 in this article suggest that a way to break through the clutter (Danaher et al., 2008) may be the usage of nonstereo typed portrayals.
It remains to be seen for how long nonstereotyped occupational gender-role portrayals in advertising will have this potential to stand out in contemporary advertising. On the one hand, women's advancement in the workplace seems to have stalled (Thomas, Cooper, Konar, Rooney, et al., 2018). This development indicates that the managerial recommendations should last. On the other hand, much is pointing toward a general direction of more balanced portrayals of people in advertising.
As mentioned earlier, there is currently a movement in the United States toward a more fair representation of women from advertising and media (Neff, 2016), running in parallel with the advertising industry's own promotion of advertising that promotes gender equality, as described earlier in this article. The American Marketing Association (2018) also has highlighted the importance of further research on the effects of marketing on society at large. Interesting times lie ahead, when researchers and practitioners may discover for how long the effects found in this study will remain. Once nonstereotyped portrayals are no longer unusual, they may no longer evoke increased processing among their receivers.
Limitations and Further Research
Because of the methodology used in this article, the authors attempted to study the topic at hand, occupational gender-role stereotypes in advertising, in as much isolation as possible. The stimuli advertisements deliberately were designed (and pretested) to differ only in the gender of the people depicted in the occupational roles. Although this methodological decision concerning the experimental conditions helped isolate the studied phenomenon, it also prevented the examination of interactions between the gender stereotypes and other factors, such as other stereotypes. When research uses pictures of people in advertising, there is always an inherently intersectional component to the analysis. The people in the pictures are perceived not only in terms of their gender but also by other characteristics, such as their age, ethnicity, body type, physical ability, sexual orientation, and attractiveness.
The studies presented here used a very limited selection of traditional stock photos that were selected and pretested to be as similar to one another as possible across experimental conditions. The advertisements thus ended up representing men and women who had many similarities; they were all seemingly young, attractive, white, cis-gendered, and able-bodied. A recommendation for future research thus is to examine the joint effects of using occupational gender-role stereotypes combined with other stereotypes in advertising.
For many of the same reasons, the studies in this article also used rather narrow and very simplified definitions of both gender and stereotypes. The studies used a binary division of the people depicted in the advertisements (into men and women). The studies also dichotomized the depictions into either “stereotyped” or “nonstereotyped,” not taking into account the degree of stereotype in each picture.
This type of simplification, in terms of both a binary gender concept and stereotypes, has been criticized (Eisend, 2019), and another recommendation for future research is to study both gender and stereotyped depictions in a less-dichotomized fashion. The studies in this article also only examined depictions of women in occupational roles stereotyped as masculine, not men in occupational roles stereotyped as feminine. Because occupational gender-role stereotypes can be as detrimental to men as to women (Helwig, 1998; Wilbourn and Kee, 2010), another recommendation for future research is to examine occupational roles stereotyped as feminine.
Overall, the studies in this article did not find any differences between the effects on different groups of consumers. Previous research, however, often has found different effects of using gender stereotypes for different groups of consumers, such as gender and age differences (Eisend, 2019; Eisend et al., 2014; Whipple and Courtney, 1985). The authors of this article recommend future research to examine differences in effects among consumer groups.
That the studies presented here did not show any differences should not be taken as an indication that there are no such differences. Studies using other samples and other research methodologies well may find differences among groups. Although the main effects on the populations the authors studied were positive, it is very possible that there are groups for which the effects of using occupational gender-role stereotypes in advertising would even be negative.
It is, in fact, very likely that the effects of using occupational gender-role stereotypes in advertising vary among different cultures. Stereotypes are social constructs, and gender stereotypes are sets of generalized beliefs about the characteristics and attributes of men and women (Bodenhausen and Richeson, 2010). Gender stereotypes are not static; they evolve over time, which is why some occupations that previously were stereotyped as masculine (e.g., accountant) now are seen as more gender neutral (White and White, 2006). Previous research has demonstrated that responses to gender stereotypes in advertising can vary among different cultures (De Meulenaer et al., 2018; Eisend, 2010), which is why the studies in this article pretested the occupational gender-role stereotypes on the same samples that were used in the main studies.
A recommendation for future research, therefore, is to investigate cultural differences in how consumers respond to stereotyped occupational gender-role depictions in advertising. It should be noted that the studies tested occupations that are stereotyped as masculine but that women are allowed to have. For a woman to be a firefighter or soldier in any of the countries of the studies' convenience samples (i.e., United States, Sweden, United Kingdom) is a transgression neither of cultural norms nor of legal boundaries. There is, however, as demonstrated in the prestudies, still a stereotype that the occupational roles are masculine. It also should be noted that if the stereotypes change, the effects shown in this article should change, too.
ABOUT THE AUTHORS
Karina T. Liljedal is an assistant professor at the Center for Consumer Marketing and a research fellow at the Center for Retailing at Stockholm School of Economics. Her research interests are consumer behavior, advertising, and retail. Liljedal's work has been published in journals such as Journal of Advertising Research, European Journal of Marketing, Journal of Marketing Communications, and International Review of Retail, Distribution and Consumer Research.
Hanna Berg is an assistant professor at the Center for Consumer Marketing, Stockholm School of Economics. Her main research interests are consumer behavior, visual marketing, advertising, and online retailing. Berg's work has been published in journals such as Journal of Business Research, Psychology & Marketing, Journal of Service Management, and Journal of Retailing and Consumer Services.
Micael Dahlen is professor at the Stockholm School of Economics, Center for Consumer Marketing. His research interests are consumer behavior, social psychology, and communications. Dahlen's work can be found in Journal of Advertising Research, International Journal of Research in Marketing, and Psychology & Marketing, among other journals.
Appendix Stimulus Advertisements Used in Study 3
Stimuli Advertisement Used in Study 3 Showing A Female Firefighter
(Mock advertisement with photo from Shutterstock)
Stimuli Advertisement Used in Study 3 Showing A Male Firefighter
(Mock advertisement with photo from Shutterstock)
- Received January 30, 2019.
- Received (in revised form) September 10, 2019.
- Accepted October 18, 2019.
- Copyright© 2020 ARF. All rights reserved.
REFERENCES
Vol 60 Issue 2
