Non‐traditional male gender portrayal as a persuasion tool in advertising

We would like to thank Professor Anthony R. Pratkanis, Dr. Patrick Leman, and two anonymous reviewers for their insightful and useful comments on the manuscript.

Abstract

The outcomes of an experiment investigating the effectiveness of gendered advertisements that use traditional or non‐traditional male portrayals are reported. The effectiveness of the ads was tested in relation to perceivers' attitudes to male gender roles on a Polish student sample. The experiment followed a 2 (ad type)×2 (gender attitude) mixed factorial design. The main finding was that the non‐traditional ad strategies were more effective than the traditional ones, supporting the predictions of research on message elaboration and the pique technique. Moreover, attitudes to male gender roles played a moderating role where the predictions of the gender attitude/ad strategy match hypothesis were supported. Theoretical and practical implications are discussed.

We would like to thank Professor Anthony R. Pratkanis, Dr. Patrick Leman, and two anonymous reviewers for their insightful and useful comments on the manuscript.

Keywords:

• advertising
• gender roles
• gender stereotypes
• persuasion

Although the topic of gender stereotypy in advertising is not new, most of the existing research focuses exclusively on female roles, female target groups, and the role of attitudes to women, in determining ad effectiveness (Wolin, 2003). Surprisingly little is known about male gender stereotypes in advertising and the effects of breaking or following them. What are the effects of the use of men in non‐traditional gender roles on persuasion? Could ads that use such portrayals be effective? What is the role of initial attitudes to male gender roles in the effectiveness of such ads? This paper attempts to investigate these questions by reviewing the literature on gender stereotypy in advertising and by applying a social influence perspective (e.g., research on message elaboration and the pique technique) to analyzing the problem.

MALE GENDER ROLES AND THEIR PORTRAYAL IN ADVERTISING

As early as 1979, Kerin, Lundstrom, and Sciglimpaglia predicted that future advertising would become more egalitarian and break traditional gender stereotypes. However, it has been shown that, so far, advertisers have instead continued to use gendered advertising strategies, apparently believing in their effectiveness (Goffman, 1976; Klassen, Jasper, & Schwartz, 1993; Mullany, 2004; Signorielli & McLeod, 1994). Recent studies of gender portrayals in the media have reported little change: although both genders are equally often portrayed as authorities (at least in the UK), males are still more likely to do voiceovers, are less often depicted performing dependent roles, are less often presented in the home and with children, and are much more likely to be the central figures in ads, than women. Additionally, ads that reverse gender‐role stereotypes are rare, while those few that do, often depict men in a comic (if not ironic) way (Furnham & Skae, 1997; Kaufman, 1999). Despite the fact that gender stereotyping in printed media is becoming more subtle, it still persists (Thomas & Treiber, 2000). Here, two types of male gender roles are distinguished and tested: traditional, associated with work (e.g., Businessman), and non‐traditional, associated with “the home” (e.g., Househusband), as this domain is traditionally associated with women and not men (Diekman & Eagly, 2000; Eagly, Mladanic, & Otto, 1991; Eckes, 1994).

PREVIOUS RESEARCH ON ADVERTISING, GENDER ROLES, AND GENDER ATTITUDES

Although there is a wide body of research exploring the typology of stereotypical gender portrayals used in ads (Barthel, 1992; Mandal, 1998; Wolin, 2003), there is still a shortage of research that addresses the problem of the effectiveness of ads that break or follow traditional male gender roles. The majority of studies addressing the question have focused mainly on female depictions and reported varied findings. For example, preference for non‐traditional advertisements (egalitarian and “super‐women” portrayals) over traditional ones was reported by Jaffe and Berger (1994). Bellizzi and Milner (1991) found that the non‐traditional positioning of an automobile repair service (with female voiceover) was favored by women but not by men, who preferred a traditional positioning strategy (with male voiceover) in radio advertising. However, Whipple and Courtney (1980) demonstrated equal effectiveness of traditional and progressive ad strategies. Duker and Tucker (1977) on the other hand, found that traditional housewife ads were more effective than any of the three non‐traditional portrayals (working mother, modern woman, and professional) tested.

Previous research has also investigated the role of various gender‐related variables (e.g., gender identity, ideology, attitudes, career orientation) in the effectiveness of advertising, again with mixed results (see Wolin, 2003; Zawisza, 2006, for overviews). For example, Ford and LaTour (1993) found that women who belong to feminist organizations preferred progressive female‐role depictions more than a general sample. Similarly, Barry, Gilly, and Doran (1985) have shown that both career‐ and “home‐maker”‐oriented women preferred ads that match their orientations. However, other researchers did not find evidence to support the match hypothesis for variables such as feminism (Duker & Tucker, 1977; Whipple & Courtney, 1980), or career versus homemaker orientation (Bellizzi & Milner, 1991).

The authors are aware of only two experimental investigations that have looked into male gender roles in advertising. Debevec and Iyer (1986a) showed that breaking male stereotypes in (radio) advertising resulted in better evaluation and higher levels of purchase interest in the advertised female product (washing‐up liquid) irrespective of the participants' gender. Garst and Bodenhousen (1997) investigated mock magazine advertisements for a cup of coffee and a personal computer. Although they did not find an ad type effect on ad effectiveness (as measured by purchase intent and brand‐related reactions), the androgynous models were perceived more favorably than the traditional masculine ones, and this main effect was not qualified by the initial attitudes to male gender roles. Debevec and Iyer (1986a), although they tested the mediating effects of a slightly different concept, gender identity (as measured by BSRI), similarly did not find it predictive of the effectiveness of the (non)traditional ads. Both the investigations indicated, however, greater effectiveness of the non‐traditional ad strategy. Why should the non‐traditional ad strategies be more effective? Should gender attitude play any role in the effectiveness of the (non)traditional advertisements? Research on race and persuasion provide some answers to these questions.

NON‐TRADITIONAL MALE SOURCE AS A PERSUASION TOOL

Debevec and Iyer (1986a) explained the preference for non‐traditional positioning by referring to the novelty of such an advertising message. However, they did not test this possibility explicitly. Literature on persuasion and the non‐stereotypical use of African‐Americans in messages provides more explicit support for this explanation. For example, Aronson and Golden (1962) found that a non‐stereotypical Afro‐American source of a persuasive message on arithmetic (Black engineer) was equally effective as a stereotypical White source (White engineer) and more effective than a stereotypical Afro‐American source (Black dishwasher). However, unprejudiced individuals were persuaded by the non‐stereotypical African‐American to a greater extent than by a White communicator of an appeal on arithmetic. White and Harkins (1994) also found that the Black, but not the White, source motivated White individuals to elaborate the message on exam policy despite low involving conditions. In fact, the motivation was so high that only exceptionally strong distraction neutralized this effect. The authors demonstrated that this enhanced elaboration was due to prejudice rather than to mere violation of expectations. More precisely, their findings supported Gaertner and Dovidio's (1986) aversive racism theory according to which most Whites are strongly motivated to appear non‐racist and therefore they scrutinize messages from stigmatized sources to a greater extent. Since there are many analogies between racism and sexism (Swim, Aikin, Hall, & Hunter, 1995) these findings showing the greater persuasiveness of a novel, non‐stereotypical, Black source could, in theory, be applied to a novel, non‐traditional male source, such as a “Househusband”. Indeed, non‐traditional males, likewise Blacks, are stigmatised (Brescoll & Uhlmann, 2005; Rost, 2002) and in the case of sexism, too, egalitarian norms are in operation (Glick & Fiske, 1996, 1999; Swim et al., 1995; Tougas, Brown, Beaton, & Joly, 1995). Therefore, we could expect that non‐traditional male ad strategies would result in greater ad effectiveness than the traditional one, especially in non‐prejudiced (here “Liberal”) individuals who, due to the egalitarian norms, would be motivated to respond positively to it.

On a more global level the pique technique (Santos, Leve, & Pratkanis, 1994), also suggests the greater effectiveness of a non‐traditional, therefore novel, male source. The technique involves asking a strange or unusual question, which piques the target's interest and thus disrupts the mindless refusal script. This leads to more positive thoughts and greater compliance. Santos et al. (1994) demonstrated that passers‐by in the strange question condition (e.g., Can you spare 17 cents?) were almost 60% more likely to give money than those in the usual question condition (e.g., Can you spare a quarter?). The authors showed that this tactic is effective under three conditions: (1) the situation needs to involve mindless refusal; (2) the persuasive message needs to be strange (to attract attention); (3) it should disrupt counter‐arguing (negative thoughts) and provoke positive thoughts by providing an unambiguous positive statement. When applied to the advertising situation investigated here it could be argued that: (1) advertising in principle involves mindless refusal (Dolinski, 2003), (2) the Househusband ad strategy can be regarded as strange and unusual (Furnham & Skae, 1997); and (3) it has the potential of disrupting negative thoughts and provoking positive ones (as the advertising appeal contains positive statements), especially among those with a liberal compared to traditional attitudes.

Given the above, in the present experiment we should expect the non‐traditional male depiction (Househusband) to be more effective than the traditional (Businessman) one (our first hypothesis). However, the processes involved in the pique technique could be moderated by the perceiver's initial attitudes to male gender roles. The pique technique may not work for Traditionals because the stimulus does not induce positive thoughts in them (as it would in Liberals). In other words, the non‐traditional male will trigger interest, and disrupt the counter‐arguing directed at the advertising, but the valence of the response to the ads will be either positive or negative according to the person's gender attitudes. Therefore, we may expect that the ad strategy/gender attitude match will result in greater ad effectiveness than a mismatch, with Traditionals responding better to the Businessman ad than the Househusband one, and Liberals responding better to the Househusband ad than to the Businessman one. This would therefore be demonstrated by a significant ANOVA interaction between ad type and gender attitudes (our second hypothesis).

METHOD

Design

The experiment followed a 2 (ad type: traditional vs non‐traditional) within‐subjects×2 (gender attitude: traditional vs liberal) between‐subjects, mixed factorial design, with ad effectiveness as the key dependent variable.

Participants

A total of 82 volunteer first‐year undergraduate students, (42 men and 40 women) from the University of Gdansk (Poland), aged 18–25 were tested. Participants were recruited from psychology, pedagogy, and international trade and politics departments.

Procedure

Participants took part in the experiment in groups of 20. They were given a set of questionnaires consisting of: (1) a measure of male gender role attitude; (2) exposure to two printed ads (one traditional and one non‐traditional; order counterbalanced); (3) evaluation of the two ads on an Ad Effectiveness Scale developed for the experiment. Participants were told that the experiment tested the effectiveness of various ads. The experimental sessions took around 15 minutes. At the end participants were thanked and debriefed.

Ad type

Printed ads created specifically for the purpose of the study were used where atypical characters performed roles typically assigned to the opposite sex in accordance with Kerin et al.'s (1979) role‐switching strategy. They were carefully pre‐selected so that the traditional Businessman (Bm) portrayal was perceived as independent and dominant while the non‐traditional Househusband (Hh) portrayal was perceived as dependant and submissive. Both the portrayals also met the requirement of evoking positive affect. The Bm ad depicted a smartly dressed man in his early 30s in an office setting, standing next to his desk with a laptop on it, holding a newspaper, and talking on a mobile phone. The Hh ad portrayed a different casually dressed man in his mid 30s in a kitchen setting, standing next to a kitchen table with bowls of ingredients on it, slicing vegetables. Other elements of the ads were kept the same: in both cases the men were depicted from their waist up, they were smiling and looking at the laptop or at the sliced vegetables depending on the photo, the headline at the top read “Vitamin,” the product was placed at the bottom left corner and a headline next to it read “Vitamin – a blend of vitamins for everyone” all presented on a blue background in both cases.

Product

Vitamins were chosen as the advertised product, which was assumed to be unisex and low‐involvement.1 “Vitamin” was selected as a brand name that was generally liked and unknown.2

Test of initial attitudes toward gender roles

An evaluation task was used as the method of testing initial attitudes to male roles. Participants were provided with descriptions of two men (see Appendix 1): one reflecting and the other one departing from the traditional gender role stereotypes,3 and were asked to answer a question: “To which of these two persons described above would you like to be similar/would you like your life partner to be similar to?”, depending on the participant's gender. This procedure tapped into ideal self and partner image concepts and therefore was supposed to secure greater attitude–behavior consistency (Mehta, 1999). The expressed preference was treated here as an indicator of positive attitude to a particular traditional or modern male gender role.

Measure of ad effectiveness

An Advertising Effectiveness Scale (AES) was created as an operationalisation method for ad effectiveness. Out of the initial pool of 81 items, 15 were selected on the basis of the magnitude of the inter‐judge congruity coefficient obtained from the scores of five independent judges, and made up the final version of the Ad Effectiveness Scale (see Appendix 2). A forced factor analysis showed that the items loaded on one factor (loadings between .447 and .701) that explained 31.13% of the variance. The magnitude of Cronbach's Alpha for the whole scale was 0.80. The analyses were conducted on the scores from the whole scale after the scales for the negative items had been reversed.

RESULTS

To test the predictions a 2 (ad type: traditional vs non‐traditional) within×2 (gender attitudes: traditional vs liberal) between factor mixed design ANOVA analysis was computed using the GLM command within the SPSS^TM software package, where ad effectiveness served as the dependent variable. Since the gender of the participant did not affect the perception of the ad effectiveness for any of the tested ads—independent two‐tailed t(80) = −0.78, p = .44 for Househusband and t(80) = −1.19; p = .24 for Businessman—this variable was excluded from further analyses. This decision was further supported by the fact that participants' gender did not interact with either Ad Type, F(1, 75) = .756, p = .38, or Gender Attitude, F(1, 75)  =  .621; p  =  .43. Also the three‐way Ad Type × Gender Attitude×Gender interaction was not significant, F(1, 75) = 1.165; p = .28). The statistical analyses were computed on means of scores from the AES with the scores from negatively worded items reversed.

The analyses revealed a significant main Ad Type effect, Λ = .935; F(1, 77) = 5.35, p = .02, indicating that generally the Househusband ad was perceived as more effective (M = 3.43) than the Businessman ad (M = 3.28) (see table 1). Thus, the prediction that the Househusband ad will be more effective than the Businessman one was supported. The expectation that a match between attitude to men and ad strategy will result in higher ad effectiveness, was also supported: a significant Ad Type×Gender Attitude interaction effect was found, Λ = .833; F(1, 77) = 15.5; p = .000, indicating that the responses to the two ad types differed according to the attitude to men. A follow up analysis with t tests (one‐tailed) showed that Traditionals perceived the Businessman ad as more effective (M = 3.41) than the Househusband ad (M = 2.30); t(58) = −1.704, p = .047. Also, a significant difference was found within Liberals, who perceived the Househusband ad as more effective (M = 3.56) than the Businessman ad (M = 3.14); t(19) = 3.156; p = .003.

TABLE 1. Means and SD for two gender attitudes groups and two ad types

Ad type	Traditionals (N = 20)	Liberals (N = 59)	Marginal means for ad type
Businessman (traditional)
M	3.41	3.14	3.27
SD	0.46	0.42	0.06
Househusband (non‐traditional)
M	3.30	3.56	3.43
SD	0.54	0.44	0.067
Marginal means for attitude to men	3.35	3.35
	SD = .054	SD = .092

DISCUSSION

Support for both hypotheses was obtained. The non‐traditional advertising message was generally more effective than the one based on traditional gender roles. Thus, it turned out that although men seem to be stigmatized more when breaking gender roles (Brescoll & Uhlmann, 2005; Rost, 2002), the novel picture of a househusband was still better received than the one depicting a traditional businessman. This finding corresponds with the predictions of the pique technique according to which the novelty of the message disrupts the mindless refusal script (reluctance to advertising) and therefore increases the compliance rates. It is also in line with Debevec and Iyern's (1986a) finding that breaking male stereotypes resulted in higher effectiveness with radio ads, irrespective of participants' gender. The second (attitude match) hypothesis was also fully supported: individuals holding more progressive attitudes to male gender roles responded better to the non‐traditional portrayal of a man, while those with more conservative beliefs found the traditional businessman ad more appealing. This finding is, again, in line with the predictions of the pique technique. It seems that the cognitive response that follows the disruption of the reluctance script is more positive in Liberal than in Traditional participants. This finding also fits well with the body of empirical research that predicted and found a more general match between gender attitudes and advertising strategy to be effective (Barry, Gilly, & Doran, 1985; Ford & Latour, 1993; Jaffe, 1991; Jaffe & Berger, 1994).

What processes are responsible for the greater effectiveness of the non‐traditional male character in advertising remains to be investigated. On the one hand it could be due to egalitarian norms and the self‐presentation motives predicted by Gaertner and Dovidio's (1986) aversive racism theory (in this case extended to sexism/perceptions about gender). On the other hand it could be due to the mere novelty of this particular portrayal as predicted by the pique technique (and also suggested by Debevec & Iyer, 1986a). Alternatively, both processes might have a cumulative effect on persuasion. Also, if the attention‐getting mechanisms similar to those postulated by the pique technique were involved here, it would be necessary to determine whether Traditionals responded less positively to the Hh portrayal because the disruption of the mindless refusal script was followed by a character they liked less, or because the disruption also affected the cognitive resources needed to control the automatic negative responses suggested by White and Harkins (1994), or both.

Unlike the Aronson and Golden (1962) experiment, the non‐stereotypical source turned out to be more effective than the stereotypical one, irrespective of prejudice levels. However, as in their case, non‐prejudiced individuals (Liberals here) were indeed persuaded by the Hh ad type to a greater extent than were prejudiced individuals (Traditionals here). These differences could be due to a culturally related difference in the strength of egalitarian norms between the USA and Poland (Furnham & Saar, 2005; Robila & Krishnakumar, 2004). It is possible that in the Polish sample Traditionals were less motivated to appear non‐prejudiced. Alternatively, it is possible that the egalitarian norms directed at Afro‐Americans are stronger than those directed at non‐traditional men (as suggested by Czopp & Monteith, 2003; Rodin, Price, Bryson, & Sanchez, 1990). More experiments involving cross‐cultural comparisons and other non‐stereotypical sources are needed to gain greater insight into the nature of the persuasive effects discovered here. Moreover, the experiment here focused entirely on ad/gender attitude match, while it has also been demonstrated that the match between the product gender and the spokesperson's gender may also influence ad effectiveness (Bellizzi & Milner, 1991; Debevec & Iyer, 1986a, 1986b; Wolin, 2003). It is possible that the effects of match and mismatch reported here would not hold for strongly masculine or feminine products.

Overall, the experiment here provided support for the generalisability of the pique technique over unusual pictorial stimuli and the persuasive effects on non‐traditional male sources on non‐prejudiced individuals. It also demonstrated the potential effectiveness of a non‐traditional male source, something as yet very rare in the domain of advertising. By focusing on breaking the male (rather than female) gender role stereotype, by applying a social influence perspective, and by testing a Polish sample, this experiment has extended the existing literature on the effectiveness of gendered advertisements in new empirical and theoretical directions. Further work is needed to answer the remaining questions regarding the exact cognitive processes behind these findings as well as to determine possible limitations of using such an ad strategy.

Notes

We would like to thank Professor Anthony R. Pratkanis, Dr. Patrick Leman, and two anonymous reviewers for their insightful and useful comments on the manuscript.

1. These features of the product were tested post‐factum in a study in which 18 participants were asked to indicate (using 7‐point semantic differential scales) how feminine vs masculine (−3, 3) and requiring little thought vs a lot of thought (1–7) the product was. The vitamins were perceived as moderately involving (M = 4.27) and only slightly feminine (M = −1.27).

2. Note that the brand name differs from the Polish word witamina, which translates into English “vitamin.”

3. A previously conducted Internet poll was the basis for the contents of the descriptions used in the task. Respondents were asked to define the appearance, behavior, and traits of traditional and atypical men. These characteristics were then evaluated by five independent judges on three dimensions: domination vs submission, independence vs dependence, and positive vs negative affect evoked by the described representatives of the two categories (typical man – Businessman; atypical man – Househusband). The outcomes were satisfactory—each of the descriptions was evaluated as was expected and then used in the evaluation task (that is, dominant and independent for traditional man and submissive and dependent for counter‐stereotypical man, with both described characters rated positively).

APPENDIX 1

Description 1 (Businessman)

Andy is a director of a big corporation. He achieved this position thanks to his ambition, dynamism, and determination in attaining his goals. Thanks to his sharp intellect and wide knowledge he became a highly respected person. He is always up‐to‐date with the market situation and his competitors and makes appropriate decisions in time.

Andy maintains his wife and two children. He built the house for them and bought a holiday house at the lake, where he tries to go as often as possible to spend quality time with his family. Recently, he bought a car as a gift for his wife's birthday.

As an energetic and health‐conscious individual Andy tries to lead a healthy lifestyle—he goes to a swimming pool and to a gym at least once a week and treats his family with relaxing holidays abroad every year.

Andy is happy with his life‐style and his work gives him a lot of satisfaction.

Description 2 (Househusband)

Tony is a husband and a father to three children. He likes the quiet of the house and therefore he agreed to take care of the household, and thus to help his wife in developing her professional career. He spends the day doing housework and cooking for the children. He also maintains a garden which he is proud of. Sometimes he repairs other people's devices to earn some money.

Tony made many friends among his neighbors, with whom he likes to play cards or chess in the evenings. Also, he often plays basketball or football with his sons.

During weekends Tony likes going for bike trips with his family or fishing with friends.

He believes he is a good father (the kids love him) and a good husband—he supports his wife by doing the house chores and thus enables her to develop her professional career.

APPENDIX 2

Ad Effectiveness Scale (Chronbach's alpha = .80)