Sex Roles, Sex, and Stereotyping in Advertising: More Questions Than Answers

Mark I. Alpert, The University of Texas at Austin
[ to cite ]:
Mark I. Alpert (1979) ,"Sex Roles, Sex, and Stereotyping in Advertising: More Questions Than Answers", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 73-77.

Advances in Consumer Research Volume 6, 1979      Pages 73-77

SEX ROLES, SEX, AND STEREOTYPING IN ADVERTISING: MORE QUESTIONS THAN ANSWERS

Mark I. Alpert, The University of Texas at Austin

INTRODUCTION

Recently there has been an increase in attention focused on sex role portrayal and sexuality in advertising. As this session on "Sex in Advertising" illustrates, this attention has come not only from critics of advertising, but from practitioners, regulators, and academicians, all with a variety of perspectives. The papers in this session range from a very specific study of aspects of sex in advertising, to a general theoretical discussion and literature review of sex in advertising, to a more general discussion of issues in role portrayal and stereotyping of women in advertising. Their orientations vary from pragmatic and managerial issues of what attitudinal effects result from sex and nudity in ads, to questions of the empirical validity of various normative statements concerning how advertisers "should" deal with stereotypes in a changing social milieu. Taken as a whole, these three papers provide a useful review of what is thought to be true, suggest issues where questions remain substantial, and propose a number of approaches to studying the phenomena in question.

Let us proceed to examine some contributions from each of these papers, as well as questions that might be raised in regard to each. In closing, we shall discuss issues in the general area of sex, sex roles, and stereotyping in advertising.

First Paper

The first paper, "The Role of Sexually-Oriented Stimuli in Advertising: Theory and Literature Review," by Wilson and Moore, (1979), provides a useful statement of theoretical considerations underlying the differential effects of degrees of advertisement sexuality on varying products and audiences. This paper is particularly useful, given the validity of the authors' assertion that most marketing research into this topic has lacked theoretical bases for generating and testing hypotheses. Their theoretical discussion is brief, and space in these papers is limited. However, it illuminates several important elements in the process by which sexually-oriented ads may interact with particular products and receivers to produce particular effects such as attention, retention, recall of brand vs. ad, and attitudes towards the ad, product, and company.

The discussion of arousal is especially interesting, as the authors argue that erotic ads may create disequilibrium, then psychological tension which may cause increased cognitive activity directed towards the ad and/ or the advertisement product. Depending on how this advanced level of information processing interacts with the consumer's value system, positive or negative effects may be obtained.

The authors also claim that arousal theory explains the negative impact of sexy advertising, where the inclusion of a female model is not logically related to the product. They reason that since arousal inhibits the impact of irrelevant information, gratuitous use of sexy models may depress perception of advertising messages. However, the logic here is not clear, since it would seem that the irrelevant information that would be depressed would pertain to the sexy stimulus, but not to the advertising messages, assuming they are comparable to those used in the non-sexy ads.

Since descriptive studies rarely have controlled for message content, comparisons of attitudes towards sexy vs. non-sexy hardware or computer ads may have confounded message with degree of erotic content. The authors' explanations for this phenomenon that are based on self-concept theory seem more valid. Here they argue that females who identify positively with a sexy model will like sexy ads, while those who do not want to be seen as sex objects will react negatively to such ads, particularly when the irrelevance of the model to the product enhances perceptions that she is solely used as a sex object.

Discussion of selective perception and absolute limits of sensory stimulation is also effectively illustrated with examples relating how older women may not respond to sexual innuendo in some advertising, due to non-familiarity with the verbal idioms (and lower sensitivity to sexuality). It would be useful to consider the concept of perceptual threshold in the light of subliminal sexuality in advertising. It might be argued that erotic stimuli may not be consciously perceived as sexy by segments of the audience who may nonetheless be positively (or in some cases, negatively) influenced by the subconsciously perceived sexual content. Certainly many advertisers and critics of advertising operate on this premise. Research questions might be formulated in which one could contrast the perceived level of erotic content with that measured by physiological means (e.g. pupil dilation, heart-rate, etc.) as stimulus intensity is varied, and the resulting impact on attitudes and behaviors towards the ads and products, and their co-variation with audience variables such as education, values, sex, age, and marital status.

It is worth noting that the theoretical discussion is written mainly from the standpoint of males' perceptions of sexy women in advertisements, with little emphasis on women's perceptions of sexy males. The generality of the phenomenon should be kept in mind, however, and the authors' discussion of empirical findings in terms of both sexes' responses to stimuli of the opposite and/or same sex show that they are aware of this generality.

The final theoretical link is to aggression theory, and the authors direct most of their attention to whether sexy ads provoke aggressive behavior. From a social and public policy, this is a key issue and it warrants concern, but the managerial significance of the sex-aggression link seems slighted. A number of advertisements seem to combine an appeal to sex and aggression drives linked to product characteristics: witness advertising for "Cougar" automobiles featuring Farrah Fawcett-Majors. If research were to establish that those who are most apt to perceive (via selective perception) these ads were also seeking a manifestation for aggressive motivations (or at least tend towards aggressiveness) managers could see relevance to matching erotic ads to "aggressive" products and customers.

The literature review relates the theoretical rationale(s) to empirical findings from psychology and marketing. Considering the vast body of psychological literature to be summarized, the authors did a creditable job. It would have been more useful had they been able to provide the same degree of critical evaluation to this literature that they did for the marketing studies. For example, it would be easier to generalize from particular studies of differential effects of erotic content on males vs. females if the psychological literature review were to note which, if any, used adult samples. The possibility exists that part of the reason why Kinsey found females "less responsive" to erotic content than did some later studies may be that he used adults, and several later studies used college students, whose females may be more "liberated" and less sexually inhibited than the adult females to whom their responses are perhaps being generalized. It is difficult to comment in depth on a vast literature, but an attempt to account for the effect of sample biases would aid interpretation of generalizations that might be made.

In a related vein the authors should be credited with making the useful distinction between the degree of explicitness in sexual stimuli in psychological experiments (high) vs. that for advertising (low to moderate). The extent to which these differences temper the applicability of psychologically-based studies on advertising predictions is effectively noted.

Six published studies of sexual stimuli in advertisements are reviewed in detail, and an effective table summarizes the variables studied, context, sample, and other relevant considerations. In general, the authors seem on sure ground in this review, and they comment effectively concerning what each study shows and what it does not. Space constrains me from summarizing key findings that are in the paper; rather, I would like to point out areas for further consideration, with the understanding that these comments should not imply that considerable insight has not already been shown by the authors.

In reviewing Alexander and Judd (1978), they point out that nudity and even the "human element" in an ad may distract subjects from learning brand names (confirming distraction hypotheses); yet in a realistic setting they state that net learning may be enhanced due to higher attention to the ads. This is a useful assertion regarding the external validity limitations of showing people ads vs. measuring actual attention, learning, and so forth, in a natural setting (ala Schwerin or K-Tel). One should note that in this study, respondents were only allowed to look for 8 seconds or 15 seconds, with this controlled on a slide projector. It would also be useful to note the necessity of exploring the generality of the nudity - recall relationships over other products and subjects, since Alexander and Judd used college students.

The review of Peterson and Kerin (1977) notes the effective use of experimental design to explore relationships among product type, degree of model nudity, and respondent sex on perceptions of product, ad, and company. Overlooked in their critique of this study (and perhaps relevant to others) is that attitudes towards the "topless" model might have been lower than the less-nude poses because of an interaction with the attractiveness (or lack of) her unclothed figure. Considerable literature suggests that attractiveness affects communication, liking, and so forth; hence erotic stimuli studies may be confounded by the attractiveness of available models. Ratings of this variable might be obtained and its impact on key variables could be measured.

The authors close with some research recommendations, including the use of focus groups for hypothesis generation, multiple measures of advertising effectiveness, and controls for realistic advertisements, without confounding lighting and other factors. Additional research recommendations would make the paper a bit more useful, particularly regarding the hypotheses that should be tested to validate the various theoretical positions that the authors summarized. Suggestions for operationalizing and testing appear in the paper, and their quality is such that even more would be welcome. The paper has synthesized considerable literature, and yet one has the sense that there is not much that can be generalized at this point. To the extent that more of the theoretical links that are initiated in their paper begin to influence future research, future summaries might become more definitive and more satisfying.

Second Paper

The second paper, "Demographic and Cognitive Factors Influencing Viewers Evaluations of 'Sexy' Advertisements," by Sciglimpaglia, Belch, and Cain, (1979), provides empirical evidence concerning some of the conceptual questions raised in the first paper. The authors studied the impact of variations in nudity, "suggestiveness'', and sex of the model(s) on perceptions of simulated ads given by male and female students, with varying sexual values and role perceptions. Their findings provide some insight into the relationships among respondent variables and perceptions of "sexy" advertising.

The study is an ambitious one in several respects. The authors randomly ordered 8 dummy ads (with sunsets, children, etc.) and 9 treatment ads that were largely drawn from actual print media advertising examples of varying degree of nudity and suggestiveness. Attitudes towards these advertising excerpts might have more external validity than the usual studies employing "made-up" advertisements. Unfortunately, this realism had to be achieved at the cost of confounding degree of nudity and suggestiveness with other factors that varied nonsystematically across the treatments. Different models were used in each ad, while ideally the same male and female would have been used in a controlled experiment. There are also differences in background that may in some cases have influenced the findings. For example, how much of the female respondents' greater antipathy towards the fully nude female "ad" than the male nude "ad" might be due to the former model being shown in a bedroom setting and the male standing in water and viewed from slightly below the waist? Attractiveness of the models, facial expression, and other confounding variables present additional problems, especially with variations in the models. In that their conclusions generally support those of other studies, this confounding may not be a major problem, but there remain questions of what was causing the effects that were measured.

Another useful extension in this paper is the use of measures of sexual attitude, advertising role portrayals, and role orientations to predict attitudes towards particular ad treatments. For the males in the study, a number of intuitively reasonable and statistically significant correlations were found, particularly those relating conservatism towards sexual values with disapproval of nudity or suggestiveness in ads involving one or both sexes. Incidentally, the correlations in Table 2 would be easier to interpret if scales were scored so that the same sign of a correlation coefficient would have the same meaning across scales. For example, role orientation scores are reversed, such that a negative correlation means that conservative attitude toward social roles is associated with negative statements about the ads, while the opposite is meant for negative correlations with the sexual value scale.

Apparently, these attitude scales were less successful in "explaining" female advertising perceptions, for as the authors noted, few significant correlations were shown. Lest anyone attempt to generalize from the female half of Table 2, it should be noted that the 10 correlations significant at the .05-level of alpha may not be significantly more than the 5 (5% of 108 "tests") that would be expected due to type-I errors, and the variance would be higher due to inter-correlations among variables and ratings of multiple ads per subject.

Response effects may also be present in this study, given that subjects were asked to provide paper-and-pencil statements about advertising that was clearly related to sex and nudity. This is a problem in any subjective statements about advertising, and attitudes towards the products would be better, particularly if obtained less obtrusively. In this case the purpose may have been particularly obvious, given the presence of 7 out of 17 ads that were sexual, and 3 with outright nudity. In the Peterson/Kerin (1977) study, only one nudity treatment per subject was used, possibly making the research objective less obvious, particularly for the less blatant ads (for the "topless" treatment, they too may have this problem). In the present study, response set bias may be amplified by administering the (sensitizing) sexual attitude questions prior to the ads. According to Bem (1972) and Reingen (1978) self-perception theory may predict that having described oneself as sexually liberal, one may respond more positively to evaluation of sexy ads, and negatively for self-described conservatives. Less reactive dependent variables and/or separate measurement of sexual and role attitudes would improve the study.

In summary, the study suggests that prior research showing females less tolerant of sexuality in advertising may be given additional support by their findings that extend the literature to look at suggestiveness as well as nudity. Caution is needed in generalizing the re-suits, due in part to the use of Southern California college students, and also to confounds and other threats to internal validity. The scales that were used may be studied further and their reliabilities and validities strengthened. Further, one might temper statements to the effect that evaluations are "strongly" related to men's personal sexual orientation, since statistical significance may differ from practical significance. Even for nude portrayals, the correlations that were "highly significant" would have r2s of .048 to .142 (without adjusting for response-effect's upward bias). From a conceptual standpoint, this study might be integrated with the theoretical positions taken in the Wilson/Moore paper, and it would support some of the conclusions from the literature. No doubt, injecting realism into advertising research in this area can be worthwhile, and future research can build upon the exploratory work here presented. Replications under conditions in which non-sexual aspects of the stimuli are more controlled may add validation to what is here, and careful manipulation of other advertising elements may also be undertaken.

Third Paper

The third paper, "How Should Women Be Portrayed In Advertisements?--A Call For Research," by Roberts and Koggan (1979) applies a basically case-study approach to current advertising in order to generate some hypotheses about women's role portrayals and their impact on advertising messages. It would have been helpful, in the context of this session, had the authors chosen to deal more extensively with the issue of nudity in advertising, so that one could contrast their views with those of the two preceding papers. From their discussion of the first "problem," one may infer that advertising using women as sex-objects should be replaced by those appealing "to a woman's sense of well being for her own sake." (Roberts and Koggan, 1979). However, here the authors are directing their messages to advertisers aiming at women customers, and it would be interesting to get their views on the ethics and pragmatics of advertisements aimed at males.

Given their scope, the authors provide some useful evidence and arguments on the topic of stereotyping females in advertising. Their treatment of female role portrayal relationships to significant others as depicted in ads, and relationships to various product categories reflects concern with broader and more fundamental issues than the question of sexy advertising (which is a special case of role portrayal). Their literature review helps summarize the studies of the extent to which women's stereotyping in advertising is increasing or decreasing. Unfortunately, the results of the studies are mixed, and it is not clear whether the "gain" in portrayal in professional roles is offset by an increase in sex-object roles. Studies of attitudes towards advertisements were summarized and found to relate to respondents' attitudes towards women, and socio-economic characteristics. Limitations in the external validity of these studies were noted in a general sense. The literature is well summarized by the authors, who point out that many proposed suggestions for changing the traditional female stereotypes do not provide substantial guidance for advertising strategy. They noted in their introduction that well-meaning attempts to portray women in a positive role may have stray connotations that disturb some women's groups. Others may attempt to place women in a highly professional light but miss the target market of housewives, as when a Ph.D. female biochemist endorses a breakfast drink.

Accordingly, their call for research to test propositions concerning the impact of stereotyped female role portrayal on marketplace behavior is well taken. The balance of their paper is given to discussion of a series of "problems" in stereotyping, examples of advertisements that may be part of the "problem" and/or part of the "solution," and hypotheses concerning how role portrayals should be done. Many of these are interesting examples and speak to alleged problems in the area. It would be useful to know how representative are the ads that have been abstracted for the paper, although a detailed and random sampling of advertisements is probably beyond the scope of the paper.

A more substantial limitation to the usefulness of the paper is the lack of a tight theoretical basis for describing how stereotyping works, why advertisers use it, and how attitudes towards stereotyped females may be likely to vary according to particular sociodemographic, cultural, and other relevant attributes of the (male and female) perceivers of advertising messages. A number of hypotheses are drawn from an implicit framework that would be more effective if made more explicit.

Another area where the paper could become more useful would be through the addition of empirical guides for operationalizing and testing at least some of the specific hypotheses that are generated from the advertising examples presented. For example, if the authors believe that ads for health and beauty products should appeal to a woman's sense of well being rather than her status as a sex object, it would be useful to present a conceptual argument that says for what kinds of women this might be true, and where it might not be. In addition, they might propose a study in which advertisements could be scored according to objective criteria and rated for "sense of well being", "sex object", and so forth, and exposed to subjects who are measured according to their buyer characteristics. Measuring instruments for key advertising effectiveness variables should be discussed, both verbal and more direct measures of attention, interest, and persuasion. Analytical procedures such as analysis-at-variance and covariance could be presented in the context of a proposed study. Some of these measures are treated in the Wilson/Moore paper, and a classic study by Evans (1957) illustrates how ads could be scored for congruence with strategies suggested by motivation research and this measure correlated with readership scores. As far as the present paper is concerned, some of the hypotheses could be tested with Starch scores for males and females.

Although ideally all elements of the comparison ads would be the same except for the role portrayal utilized, this is hard to achieve in an ex post design with current advertisements. However, this paper's exploratory research status might support such analyses for one or two hypotheses, and it would be stronger than the pure case study approach used alone.

Most of the examples and hypotheses consist of providing a rationale for why an ad will be effective, or will not be, in terms of how stereotyping is handled in the ad. The logic is often attractive, but the contribution would be more substantial if we had a stronger conceptual framework, and either more empirical support for the hypotheses, or a proposed methodology for operationalizing and testing the hypotheses that are put forth. A number of the authors' hypotheses are reasonable and interesting; others may be more viewed as controversial, depending on one's perspective. It is hoped that the authors and others may be motivated to continue the work presented in this paper, and thereby provide the empirical validation that is needed to improve our understanding of this area.

DISCUSSION

The papers in this session are certainly of topical interest. Beyond the matter of sexy advertising may be more fundamental questions of managerial strategy and responsibility. How can advertisers creatively respond to the challenges of forsaking traditional sex-role stereotypes, and what impact will this have on advertising to customers whose sex and sex-role orientations may vary? Much work remains to be done to provide theoretical and empirical understanding of the extent to which advertising uses stereotypes vs. the extent to which it creates or at least reinforces them. Most of the papers have focused on female role-typing in advertising, but male role-typing is at least implied. The societal and managerial perspectives on what advertisers should do in response to changes in sex role orientations and sensitivities to them will continue to be debated for some time to come. Further, if Tucker (1976) is correct in the assertion that verbal attitudes towards these issues may be discrepant from behavior, research in this area will be more challenging and potentially more misleading than in many other problem areas.

A number of implications for research are covered in or may be stimulated by the papers discussed in this session. Wilson and Moore have presented a first step in building a conceptual framework for the analysis of sexually-oriented advertising. The broader sex-role issues may require integration of recent contributions from sociology as well.

A number of research issues have been noted above by the session authors and in these comments. Among them is the choice of dependent variables to measure advertising effectiveness. Most of the marketing literature has used statements towards the advertisements. Less reactive measures that utilize other indices of interest, attention, attitude change, and behavioral intentions should be employed in future research and validity-generalizations may be attempted from these multiple measures. Further, respondent sensitivity to this topic calls for care in securing non-biased samples, asking questions (Blair, Sudman, Bradburn, and Stocking, 1977). and in debriefing.

The stimuli variables to be operationalized and explored also need research under controlled conditions. Among these are the number and sex of persons used as models in the ad, advertising medium, degree of nudity, degree of attractiveness of the model(s) and/or their clothing, matching of sex of model with that of ad respondent (or non-matching), background, product or service advertised brand "name", perceived relevance of model to advertising situation, and numerous interactions among these related stimulus variables. For example, consider the effect of partial nudity on the meaning of brand names such as "Close-Up" vs. "Crest" toothpaste.

Similar questions remain concerning the differential effects of stimulus variables on receivers of varying sex, role-orientation, sexual attitudes, involvement with the product, and a variety of psychographic variables. Situational variables also may affect the communication process, with variations induced by who else is present during the advertising exposure, individual vs. group decision-making, and so forth. In addition, there are a number of ethical issues in conducting such research, as well as questions concerning the legitimate use of such findings in a commercial context. No matter what perspective from which one approaches the area, improved knowledge of the process by which sex roles are communicated and to what effect is a prerequisite to effective actions to influence advertising strategies. At this point, most of what is believed is fragmented and anecdotal. If future research is done with the care suggested by a careful study of the existing literature, our knowledge may be considerably strengthened by those efforts

REFERENCES

M. Wayne Alexander and Ben Judd, Jr., "Do Nudes in Advertisements Enhance Brand Recall?", Journal of Advertising Research, 18(February 1978), 47-50.

Donald J. Bem, "Self-Perception Theory," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 6 (New York: Academic Press, 1972)

Ed Blair, Seymour Sudman, Norman M. Bradburn, and Carol Stocking, "How to Ask Questions About Drinking and Sex: Response Effects in Measuring Consumer Behavior," Journal of Marketing Research, 14(August 1977), pp. 316-321.

Franklin B. Evans, "Motivation Research and Advertising Readership," Journal of Business, 30(April 1957), 141-146.

Robert Peterson and Roger Kerin," The Female Role in Advertisements: Some Empirical Evidence," Journal of Marketing, 41(October 1977), pp. 59-63.

Peter H. Reingen, "On Inducing Compliance With Requests," Journal of Consumer Research, 5(September 1978), 96-102.

Mary Lou Roberts and Perri B. Koggan, "How Should Women Be Portrayed in Advertisements?--A Call for Research," in Advances in Consumer Research, 6(Chicago: Association for Consumer Research, 1979).

Horace S. Schwerin, "Why Television Commercials Succeeds" in Robert Ferber and Hugh G. Wales (eds.), Motivation and Market Behavior, (Homewood, Illinois: Richard D. Irwin, Inc., 1958), pp. 321-333.

Donald Sciglimpaglia, Michael A. Belch, and Richard F. Cain, Jr., "Demographic and Cognitive Factors Influencing Viewers Evaluations of 'Sexy' Advertisements," in Advances in Consumer Research, 6(Chicago: Association for Consumer Research, 1979).

W. T. Tucker, "A Long Day of Discrepant Behavior," in Marketing: 1776-1976 and Beyond, edited by Kenneth L. Bernhardt, (Chicago: American Marketing Association, 1976), 351-353.

R. Dale Wilson and Noreen K. Moore, "The Role of Sexually-Oriented Stimuli in Advertising: Theory and Literature Review," in Advances in Consumer Research, 6 (Chicago: Association for Consumer Research, 1979).

----------------------------------------