The following is the first original empirical research project that I completed for my psychology research class at UW-Madison as an undergrad. I took the class at summer and went around asking men to participate in the study about how gender schema impacts perception of the female form in advertisements, making it my first audience reception study as well. It’s rough, but I’ve kept it for over 20 years as a reminder of where I started and how far I’ve come.
was designed to test for the effects of a man’s gender schema and a woman’s
somatotype on how attractive men rate women and thus how much they recall of
stimuli surrounding the women.
Participants were 10 men, equally divided between being either gender
schematic or gender aschematic, found on the University of Wisconsin-Madison
campus who received candy as a compensation.
No significant main effects or interactions were found for either
attractiveness or recall, but several strong trends for subgroups of overall
attractiveness ratings and strong correlations were calculated. The results indicate that a man’s gender
schema and a woman’s somatotype has no influence on how attractive she is
rated, that recall might be due to an unpredicted cognitive phenomenon, and
that an attractive personality is just as important as an attractive face. Implications are discussed for marketing
strategies and women’s appraisals of their attractiveness to the opposite sex.
Gender Schematic Men and the Somatotypes They Perceive: Can One Influence the Other?
People want to understand who they are. Social cognition psychologists say a method of self-realization is through developing a self-concept that categorizes one’s personality and invokes predictions of one’s behavior. One theory to explain how a person thinks of oneself envisions a complex network of cognitive structures known as self-schemas. According to Markus (1977), self-schemas are self-generalizations created from past experiences and from both personal and social evaluations of behavior. These generalizations are networks of trait associations that combine to create a sense of identity and self. Generalizations are created through the organization of past experiences and evaluations to develop a cognitive pattern which can be used for future predictions about the self. A person who realizes he or she has been responding to similar situations in a consistent fashion will form a schema defining his or her personality along that dimension of behavior, and this schema will in turn influence her future behavior. Self-schemas can be formed for a variety of personality traits although most schema research is done on bipolar dimension traits (e.g. Fong & Markus, 1982; Markus, 1977; Renn & Calvert, 1993).
One such dimension often studied is gender and the associated gender schema, although it may not be appropriate to label gender in such a fashion. Bem (1981), in describing gender schemas, explained schemas as methods for categorizing and imposing structure on the vast amount of external stimuli an individual must interpret and understand everyday. Bem saw this process beginning early in childhood for gender schemas through sex-typing. Through socialization methods such as the media and adults, children learn the appropriate behaviors and attributes for the sexes and these lessons form the basis of the child’s gender schema; they become sex-typed to be either feminine or masculine. Yet because every person has a different past, there exists an array of gender schemas between the two polar extremes of pure masculine and pure feminine (Markus, 1977). Social stereotypes presented by the media and biological differences between the sexes serve to counterbalance personal difference so that in general boys will develop more masculine schemas than girls, and girls will develop more feminine schemas than boys.
According to Bem’s gender schema theory (1981), a person with a high gender schema (i.e., a masculine man) will centralize this schema, assigning a high level of importance to gender-related traits in their self-concept. A schema in such a person will influence not only how they distribute importance to gender related traits but how they view gender in general. This characteristic of a centralized self-schema has been found even outside of gender studies in more dichotomous traits. Fong and Markus (1982) found that people high on certain schema sought out the same self-relevant trait in others while those low on that trait looked at more neutral traits. Fong and Markus attributed this result to the idea that a self-schema can influence the perception of others by constraining what information is sought in another person to self-relevant information. In other words, people evaluate others by their schemas.
This logic of Fong and Markus’s can be applied to gender schemas. People with high gender schemas are masculine men and feminine women. There also exist people with low gender schemas who do not define themselves with either polar end of the dimension. Someone who has traits from both schemas are labeled “androgynous” (Bem, 1974). Not to be confused with physiological androgyny, psychological androgyny means that the man or woman does not define him or herself strictly along the lines of masculinity or femininity. As they do not have a polarized gender schema, androgynous people are often labeled aschematic. Since Bem first attempted categorizing aschematics, other researchers have found multiple levels of psychological androgyny (Frable & Bem, 1985; Markus, Crane, Bernstein, & Siladi, 1982), but the basic definition is still the same. Because androgynous people have less well-defined gender schemas, it is assumed that gender holds less importance to them and that they do not perceive the world in the same gender-related way as schematics (Renn & Calvert, 1993). Gender aschematics recall more counterstereotypical information than gender schematics (Renn & Calvert) and refer to neutral traits more as a source of information about another person (Fong & Markus, 1982).
Anderson and Bem (1981) found that the different levels of gender schemas also attend to attractive stimuli differently. They showed gender schematics and aschematics a picture of the person they were about to interact with over a phone. The pictures showed either an attractive or unattractive person of the same of opposite sex. Schematics were more likely to behave towards perceived attractive individuals as dictated by the cultural stereotypes of attractiveness; that is, schematics were more responsive to their partner when they believed their partner was an attractive member of the opposite sex while this difference was not seen for aschematics. These results indicate that gender schematic men would attend more towards women whose physical appearance matches the cultural definitions of attractiveness.
To explain these results, Bem and Anderson (1981) used Bem’s theory of the heterosexuality subschema (1981). The heterosexuality subschema states that the heterosexual attraction fosters the development of gender schematic related processing in young people by creating the generalization that the two sexes are biologically different which in turn may create a general readiness to encode the opposite sex in terms of sexual attraction. This readiness could be a result of sex-typing, with the stimuli being the need to find a mate of the opposite sex and gender. However formed, it seems an object perceived as attractive will activate the heterosexuality subschema and focus attention onto this object because it is perceived as a possible mate, thus excluding any other stimuli that may surround it.
Adding to Bem and Anderson’s (1981) work indirectly, Roskos-Ewoldsen and Fazio (1992) found that certain attitudes can affect attention to visual stimuli in a field of visual stimuli. They discovered that if a presented object evoked a readily accessible attitude or an attitude with strong associations, then attention would become focused on that object, leaving other objects to become lost along the way to recall. The way Roskos-Ewoldsen and Fazio discussed the orientating value of attitude is similar to the way Bem (1981) discusses gender schemas. Roskos-Ewoldsen and Fazio explained that because humans have to cope with an incredible amount of visual stimuli each day, they have to create a functional method for deciding which stimuli to attend to. Attitudes, basically structures that define what the self likes and dislikes, could then serve to direct attention to stimuli on the basis of “good” or “bad.” Substitute “schemas” for “attitudes,” and the idea is the same.
Schemas are based on the same dichotomous system, the self is “this” but not “that,” so it seems likely that schemas could influence attention in the same manner. Individuals who are gender schematic have been shown to view the world in gender-related terms and stereotypes. It has been shown that gender schematic men attend more to the women they perceive as attractive. Roskos-Ewoldsen and Fazio (1992) might say this is due to an attitude gender schematic men have towards women: they like attractive women and thus focus their attention towards them. Yet the study performed by Anderson and Bem (1981) was in the absence of visual stimuli after the initial photograph. The results suggest that the attitude of gender schematic men can be primed, and that this priming has subsequent effects on behavior. Priming with a visual stimuli was found to activate the attitude prior to seeing the stimuli in a field of stimuli (Roskos-Ewoldsen & Fazio). So what would happen if after the initial priming a gender schematic male both saw and interacted with a member of the opposite sex?
A study conducted by Mckenzie-Mohr and Zanna (1990) investigating the relationship of sexism, pornography, and gender schemas, begins to answer this question. College men were selected based on their gender schemas from a larger pool and randomly assigned to one of two conditions. Either the student watched a nonviolent pornographic episode or they viewed an episode of a session of Canada’s House of Commons. They were told this was the first of three unrelated studies. In the second part, a woman interviewed each student about college life. The third part was a computer program designed to test their recall and to measure their response latency to a number of questions about the interview. Mckenzie-Mohr and Zanna found that gender schematic men who had been exposed to the pornography behaved more sexually towards the woman as rated by the woman and recalled more information with a quicker response rate about her physical appearance than the room’s appearance. Mckenzie-Mohr and Zanna (1990) attributed this result to the characteristic of gender schematics to selectively attend to items relevant to their schemas. In other words, because the gender schematic men who viewed the pornography were already primed to respond in terms of gender, they acted towards the woman in these terms. Roskos-Ewoldsen and Fazio (1992) would explain these results with different words, but the same meaning. It would appear that the gender schematic attitude towards attractiveness does affect attention.
But under what conditions must this attractiveness be apparent for the schema to have an effect on attention? Mckenzie-Mohr and Zanna (1990) used an erotic stimuli in which a woman was already acting in sexual terms. It was her actions that really primed the attitude. Unless a man is in the Red Light District late at night, seeing women acting sexually in public is not that common. Perhaps something as simple and common as a woman’s body shape can suffice in eliciting this attitude.
Recent research into somatotype stereotypes may explain what might cause a gender schematic male to attend towards one body shape more than another. Spillman and Everington (1989), in investigating the changes in these stereotypes, have found that the somatotype given the highest attractiveness rating has changed. Where classically the muscular mesomorph was always rated the most attractive, college men and women now rate the thin ectomorph the ideal. Ectomorph women were rated the most sexually appealing and the most likely to have dates than the other body types (Spillman & Everington, 1989). It would then appear that the thin body has become the social standard of attractiveness. Because gender schematic men are the ones who have been sex-typed by society on how to see themselves and others, they would be those men who have been most influenced with the social standards of beauty. Thus, the gender schematic men of today would find the ectomorph the most attractive. Following Anderson and Bem’s (1981) and Roskos-Ewoldsen and Fazio’s (1992) theories, seeing an ectomorph may cause a gender schematic man to focus on her and, based on the research of Mckenzie-Mohr and Zanna (1990), to act more sexually towards her. However, the results of Spillman and Everington (1990) say that the ectomorph is the most “physically” attractive. Mesomorphs were assigned the most attractive personality traits. Gender aschematic men were not as affected by the physical attractiveness of the opposite sex (Anderson & Bem, 1981) and relied less on gender related traits in general when assessing others. Thus aschematics may find the mesomorphs the most attractive overall and attend the most to them.
This study was designed to explore the issue of gender schema attentiveness towards attractive visual stimuli by using somatotypes. To test if gender schematics and aschematics rate the attractiveness of somatotypes differently and thus attend to the somatotypes differently, the study incorporated a medium in the hot seat of the thin ideal debate: magazine ads. By varying the women’s body shapes while keeping the background stimuli constant, the effect of attention on recall can be measured. If the attention theory of Roskos-Ewoldsen and Fazio (1992) is true, then there should be a negative correlation between the attractiveness rating and the recall of surrounding information. According to the somatotype research (Spillman & Everington, 1989), there should be positive correlations for physical attractiveness and overall attractiveness for the ectomorph body type and for personality attractiveness and overall attractiveness for the mesomorph body type.
If the research on the heterosexuality subschema (Anderson & Bem, 1981) and somatotypes (Spillman & Everington, 1989) are sound, then a gender schematic man should rate an ectomorph the most attractive while a gender aschematic man should find the mesomorph the most attractive. This can be shown by breaking the attractiveness rating into two parts, physical attributes and personality traits. A gender schematic man should rate the physical attributes of an ectomorph the highest while a gender aschematic man should rate the personality traits of the mesomorph the highest. Also, a gender schematic man should recall less of the information about the ad when the ad features the ectomorph, and an aschematic man should recall less of the ad when a mesomorph is featured. Since the obese endomorph is still stereotypically labeled the most unattractive, it will probably be given the lowest attractiveness rating by both schemas, although aschematics will probably rate endomorphs higher based on personality traits. Since it will be rated the least attractive and thus causing the least attention, the most recall of the ad should be found for the one featuring endomorphs. The average body shape that has no cultural stereotypes associated with it will probably fall within the range of the other body types across the schemas for both attractiveness and information recall.
Ten men from the University of Wisconsin-Madison campus participated in this study. They ranged from 18 years old to 45 years old with none in their 30’s. One aschematic man was Asian-American with no affiliation to the University. Half of the men never looked through a man’s fashion magazine, and over two-thirds never looked through a woman’s fashion magazine. Only two people refused to accept the candy offered upon completing the study, and according to their personal statements none were aware of the study’s goal before debriefing. Random assignment was only possible with regard to the order of the somatotypes seen.
The study contained three parts: a personality questionnaire, the magazine ads and the questionnaires associated with these ads.
The Personality Questionnaire. As seen in Appendix A, the first questionnaire consisted of 30 personality traits on a 7-point scale, with one being “Never true” and seven being “Always true.” The instructions were for the man to rate himself on these traits as they apply to him on a day-to-day basis. The questionnaire was a modified version of the questionnaire developed by Bem to test for psychological androgyny. Bem’s (1974) questionnaire had 60 traits on the same scale; twenty traits were masculine, 20 were feminine, and 20 were neutral. However, Bem’s sex-type questionnaire was meant to screen participants in advance of the study. Such a selection luxury was not available for this study, so a less daunting questionnaire of 15 masculine traits (e.g., aggressive) and 15 feminine traits (e.g., affectionate) were selected and randomly organized. This modification also allowed for scanning the results to a form a general idea of the man’s gender schema.
The Magazine Ads. The four representative magazine ads (see Appendices B1-B4) were created by hand by combining elements of real ads found in women’s fashion magazines. Full length figures of women were copied from various ads and were then modified into silhouettes to fit one of the four body shapes being studied. Each ad contained information and pictures from various real ads to create an ad to sell make-up for Cover Girl. The information was constant in each ad. The placement and the silhouettes of the women changed from each ad so that the ads appeared different but actually had the same content. One particular body type was represented in each ad by three silhouettes.
The Pictures Questionnaire. Each picture had a corresponding questionnaire (see Appendices C1-C4). The first part asked the men to rate the women in the ad they just saw on 10 character traits borrowed from Spillman and Everginton’s research (1989, e.g. sexually appealing) using a 7-point scale. These traits were consistent for all the ads. The second part had 13 True or False questions (e.g., “There are 11 flowers in the ad”). Since the information was consistent across the ads, 12 questions were the same but in different random orders for each ad. The thirteenth question varied for each picture because it asked something specific about the location of a woman in respect to a picture of a product. This was done to avoid complete repetition. Attached to these questionnaires was a short questionnaire to ascertain demographic information (see Appendix D).
The ads and their questionnaires formed packets whose orders were arranged using a Latin square counterbalance to diminish order effects. The packets were then randomly shuffled and passed out by picking the top one first.
The study was conducted using a 2 x 4 mixed design. The four levels of the somatotypes served as the independent variables for the within-subjects measure. These were ectomorph, mesomorph, endomorph and average. The two applicable gender schemas of schematic and aschematic served as the independent variables for the between-subjects measure. Gender schemas were determined from the personality questionnaire by using the median scores established by Mckenzie-Mohr and Zanna (1990) who used a 300 person pool in which to determine the cutoff values. A man with median scores above 4.97 on the masculine traits and below 4.26 on the feminine traits was classified as gender schematic. A man whose median scores were above both of these cutoffs was classified as gender aschematic. One of the dependent variables was overall attractiveness of the somatotypes derived from the trait ratings for the women and broken down into the subgroups of personality attractiveness and physical attractiveness. The other dependent variable was recall of the information in the ad obtained from the True/False section. For all the dependent variables, the higher the score indicated the higher the attribute.
Before approaching any man, the experimenter checked the top packet for the order and arranged the pictures to correspond with the packet.
Men who were sitting alone were approached at various locales across the campus. The experimenter introduced herself as a researcher for the University who was conducting an experiment on how men’s personality influences their perceptions of women’s images found in magazines. They were informed that their participation was not expected to exceed ten minutes, and that it was the man who would set the pace.
When the men agreed to participate, they signed the consent form and filled out the personality questionnaire. The experimenter than said:
What I’m going to do is show you a series of pictures, one at a time and four in all. I’m only going to show you them for fifteen seconds each, so I don’t expect you to memorize them or anything, just look them over. After each picture there’s a quick questionnaire, and when we’re done I just need to collect some demographic data.
Each picture was then shown according to the packet, and the men were timed while looking at it. After 15 seconds, the experimenter removed the picture and handed the men the corresponding questionnaire. After everything was completed, the men were debriefed and presented with three lollipops as compensation.
The dependent variable of overall attractiveness was obtained by finding the mean of the 10 traits. Four of the traits were reversed to allow equal representation for the three main somatotype stereotypes as found by Spillman and Everington (1989). Personality attractiveness and physical attractiveness each had five traits calculated like overall attractiveness to derive their ratings. Four scores were reversed to ensure equal representation across the stereotypes. These reversals were: vain, lazy, sloppy, conforming. Recall was measured by tallying how many correct answers were obtained in the True and False section. Any crossed out, blank or otherwise indecisive answers were treated as incorrect responses.
When the scores for the character traits were tallied for an overall attractiveness rating, there were no significant effects found. As seen in Table 1, the means for the somatotypes collapsed over the gender schemas were essentially the same. Because Mauchley’s sphericity test was violated, a calculated Pillai’s Trace demonstrates this lack of difference, F (3, 6) = .094, p = .961. Also by comparing the means for the two gender schemas for each somatotype, as seen in Table 2, it is clear that gender schematic men did not differ from gender aschematic men, shown by Pillai’s Trace, F (3, 6) = .109, p = .952. The overall rating of attractiveness did not seem to be effected by the different somatotypes or an interaction of somatotypes and gender schema.
When the five traits created to measure physical attractiveness were considered separately, there was a strong trend to indicate there was a difference between the somatotypes, but again there was no effect of gender schemas. As seen in Table 2, the mean scores for gender schemas are almost identical, a lack of interaction proven by Pillai’s Trace, F (3, 6) = .609, p = .633. Table 1 shows the means for the somatotypes, and there appears to be a clear difference between average at the high end and endomorph at the low end. However, this difference is only a trend with Pillai’s Trace, F (3, 6) = 3.636, p = .084.
The five traits assigned to assess the attractiveness of the women’s personalities shows a very strong trend for somatotypes and no difference for the gender schema hypotheses. The means in Table 2 would seem to show an effect of gender schema on personality attractiveness, but this belief is not supported statistically, F (3, 24) = .543, p = .657. Likewise, the means in Table 1 seem to indicate a significant difference between the four somatotypes, but only a very strong trend was detected with F (3, 24) = 2.978, p = .052.
For all the somatotypes, scores for the two subgroups of attractiveness, using the Pillai’s Trace, were significant between each other for only mesomorph ( F (1, 8) = 5.369, p= .049). Average and ectomorph only showed strong trends ( F (1, 8) = 4.805, p = .060 and F (1, 8) = 5.256, p = .051, respectively) while endomorph showed no difference ( F (1, 8) = 1.72, p = .689). These differences along with the trends for the subgroups can be seen in Figure 1.
The number of correct responses did not differ within the somatotypes or in the interaction. The median scores in Table 2 for the gender schemas show the narrow variability, a lack of difference highlighted by Pillai’s Trace, F (3, 6) = .053, p = .983. Table 1 shows the means for the somatotypes and shows the same tightness, Pillai’s Trace at F (3, 6) = 3.057, p = .113. While this result may become stronger with more participants, the value is too weak to be anything but a possible trend.
Due to the nature of the codependence of the dependent variables, various correlations were run to either clarify or strengthen results.
It was predicted that a negative correlation would exist between attractiveness and recall. While this may be true to ectomorph and endomorph ( r (8) = -.037, p = .918 and r (8) = -.147, p = .686, respectively), these results are not strong enough to infer such a correlation. In fact, a strong positive correlation for mesomorph ( r (8) = .907, p < .001) is the exact opposite of what was predicted. The average had with an r (8) = .086, p = .813 had a positive slope but was not strong enough to infer correlation.
Breaking apart the overall attractiveness rating paints a more colorful picture. Once again there is no significant correlation for the two ratings for ectomorphs, but now only physical attractiveness is negatively correlated with recall ( r (8) = -.362, p= .303) while personality attractiveness is quite the opposite ( r (8) = .265, p = .460). Average, which was so insignificant overall, remains insignificant for personality attractiveness ( r (8) = .244, r = .497) but develops a very weak negative trend for physical attractiveness ( r (8) = -.443, p = .200). Endomorph, which was seen to have a negative slope overall, develops weak positive slopes for both personality and physical attractiveness ( r (8) = .191, p = .597 and r (8) = .401, p = .251, respectively). Mesomorph maintains its nearly perfect relationship when broken down, with physical attractiveness ( r (8) = .906, p < .001) slightly more predictive of recall than personality attractiveness ( r (8) = .818, p < .001).
Also of interest was to see how good the subgroups were as predictors of overall attractiveness. Physical attractiveness is a good predictor for ectomorph’s overall attractiveness ( r (8) = .811, p = .004) but so is personality attractiveness ( r (8) = .849, p = .002). The same findings for physical ( r (8) = .968, p < .001) and personality attractiveness ( r (8) = .905, p < .001) apply to mesomorph, only stronger. Personality is the only one strong enough to be considered for predicting overall attractiveness for average ( r (8) = .874, p = .001). Strong trends were found for both personality and physical for endomorph ( r (8) = .570, p = .085 and r (8) = .602, p = .065).
Testing to see if the two attractiveness subgroups were correlated achieved no significant findings for ectomorph and average ( r (8) = .389, p = .267 and r (8) = .147, p = .686, respectively). Quite the contrary, the results indicate a strong positive correlation for both endomorph and mesomorph ( r (8) = .747, p = .013 and r (8) = .821, p = .004, respectively). The positive correlations indicate that as one rating went up, so did the other.
While this study did not support any of the predictions as they were hypothesized, the results are nonetheless interesting and important in their implications to the real world and future research. A man’s gender schema appears to have no effect on how attractive he perceives a woman based on her body shape, and thus there was no difference on how well he remembered the content of the ads. While there were no clear-cut differences, there were several correlations that suggest explanations not predicted.
A gender aschematic man is just as likely as a gender schematic man to rate an ectomorph or average body shape as attractive overall. In addition, a woman’s body shape does not affect how attractive she is perceived. An underweight woman appears to be no more attractive than an overweight woman. It also appears that no matter how masculine a man defines himself and no matter what shape a woman’s body is, these men are attracted to women. Breaking down the overall attractiveness into its components of perceived physical and personality attractiveness found trends towards the body types being different from each other along these dimensions, indicating that with a larger sample these differences have a good chance of becoming significant. Because the trends were strong but not significant enough to justify performing a post hoc examination, the means were examined to understand where the trends were leading. The average body shape was perceived to have the most attractive physical characteristics and mesomorphs was perceived to have the most attractive personality traits. Endomorphs were perceived to be the lowest on both these scales. While these results were predicted to occur for endomorphs and mesomorphs, they were predicted to occur within the gender schemas. This was not the case, and these trends were not sufficient to affect the overall attractiveness rating each somatotype received, as was predicted to occur. What appears to have happened is that the subgroups averaged themselves out so that all somatotypes were rated similarly overall despite the differences in their subgroups, which is apparent because the ratings across the somatotypes are moderate, staying close to the middle range of possible scores.
Perhaps the men did not feel comfortable rating a woman based solely on the shape of her body. Several men after completely the study voiced a desire to have been able to see more than just the silhouette on which to base their judgments. This discomfort seems to indicate that these men may hold off making judgments about a woman until more than just her surface features are known. For women, this hesitation means that bringing out one’s personality more in first impression situations with a man would influence how attractive one is perceived more than just displaying physical attributes. However, when all that is available is the superficial body shape, being thin is no more a surefire method for being perceived as more attractive than being voluptuous or overweight is.
In fact, these results seem to indicate that both gender schematics and gender aschematics have some version of the heterosexuality subschema. As explained earlier, the heterosexuality subschema generalizes how a person encodes the traits of a member of the opposite sex along a sexual attraction dimension. The belief is that a gender schematic man who attributes more importance to gender-related traits would be affected by a woman’s gender- and sex-related traits to encode her overall characteristics in terms of sexual attractiveness (Bem, 1981). However, perceived attractiveness for the women did not depend on how the man view gender-related traits. It appears that the heterosexuality subschema is not simply a subschema for men with well-defined, masculine traits because they are the typical “heterosexual male.” It may not be a subschema at all. Any heterosexual may have the tendency to perceive members of the opposite sex in terms of sexual attractiveness, an idea that seems well-suited to be answered by cultural and sociobiological research.
The heterosexuality subschema and the research done on attitude-evoking stimuli (Roskos-Ewoldsen & Fazio, 1992) had predicted that the more attractive a man perceived a woman, the less he would remember of the stimuli surrounding her. The amount of information a man is able to recall does not appear affected by either the body shape of the woman, and thus his perceived attraction, and the man’s gender schema. The idea was that viewing a certain somatotype would activate in the gender schema an attitude of like or dislike on the attraction dimension and thus focus attention unto a favorable somatotype while averting attention from an unfavorable somatotype. However, the correlations show the opposite occurring, but only for mesomorphs. The more attractive a man rates the mesomorph, the more he remembers the content of the ad in which she is featured. This result was found not only for overall perceptions of attractiveness, but with the two subgroups as well, suggesting that the ads invoked a different type of attention. Unlike the field of unpaired visual stimuli used in the Roskos-Ewoldsen and Fazio study, the field of visual stimuli used in this study had associations of the women and the product they were selling. Perhaps these associations became more attended to as the perceived attraction increased. Men see a woman they like, see she is selling something, and thus pay more attention to what she is selling. This explanation could also be used to explain the marketing strategy of draping attractive women on sports cars or surrounding alcoholic drinks with willing women. Yet these ads use thin women normally. This study suggests using voluptuous women may be a better marketing strategy.
The unexpected results of the relationships between the two subgroups have implications for advertising and women in general. For both endomorphs and mesomorphs, the more attractive a man finds her personality, the more attractive he finds her physical characteristics. This relationship did not occur for ectomorphs. Moreover, for these two somatotypes, as either of the subgroups increased, overall attractiveness increased. These results add to the importance of a strong, attractive personality, and to the notion that physical characteristics are no more predictive of who will be labeled attractive than personality traits. For women, this again implies the necessity for worrying just as much about one’s personality flaws as one’s physical flaws. When advertisers want to use women other than the thin ideal, a wise strategy would be to highlight an attractive personality as much as superficial attractive features, especially if the target audience is mostly men. Yet all of these results about what is most important for attractiveness must be taken with a grain of salt. The men used in this study were educated men, men who most likely treat women more than objects. Testing more of a general populace with different levels of education and different socioeconomic levels may find different results, but these results imply the possibility that the more educated a man is, the less he views women solely on their appearances. Also, it must be remembered that many correlations were run, increasing the chance of mistakenly believing something these results when the opposite are correct. The truth about the men’s attitudes and perceptions could in reality be quite different.
Another factor to consider when relating these results to the real world is the ages of the participants. Not all the men were in their 20s, and it seems likely that those gender schematic men in their 40s might have a different sex-type in their mind about what type of body is the most attractive. As cited earlier (Spillman & Everington, 1989), mesomorphs were classically seen as the most attractive. Perhaps those men of the older generation had been sex-typed to accept this body type as the cultural standard for beauty. A difference in generations might have caused the lack of difference between body types.
The experimenter also might have skewed these results. For one thing, she was not totally blind to the experiment, and subtle body language might have influenced the men. Also, because she was a woman, they might not have wanted to look like “pigs;” that is, they might have wanted to appear like gentlemen and thus not be so judgmental about the women. It would be interesting, then, to have the experimenter be a man and see how the results come out.
One of the major problems was the method in which the men were solicited. Problems with the sample as being representative start with how the men were classified into a gender schema. Unlike the Mackenzie-Mohr and Zanna (1990) study, men could not be randomly selected from a larger pool where their scores on the gender schema dimensions were already calculated. Men were chosen out of convenience with the hope that they fell into either of the categories. Due to this sampling procedure, four of the men classified as gender schematic had scores that barely met the median criteria set by Mckenzie-Mohr and Zanna. It is possible that by adding more masculine and feminine traits, the difference between schematics and aschematics might have been more pronounced, but the opposite is also possible. The best way to ensure proper representation of the two gender schemas would to have sampled a large pool and drawn from it what was needed.
Comparing the procedural differences of this study and the Mckenzie-Mohr and Zanna (1990) study also may explain why there failed to be any significant difference between the recall scores. In the Mckenzie-Mohr and Zanna study, recall was measured by response latency and the difference between remembering the background and remembering the woman, but in this study recall was measured by correctness to questions only about the background. This study did not really test to see if the interaction of gender schema and the priming of somatotypes affected remembering the women over remembering the background information. This comparison was not possible in the design of this study, so it is possible recall was not tested for properly. Also, because each packet contained all four ads and the information did not change, carryover effects might have been high. While counterbalancing hopefully accounted for most of this problem, men, once they were aware of the pattern of the study, did attend to the background information more, as verified by their comments to the experimenter at the end. The best way to alleviate these concerns would be to structure the study more like the Mckenzie-Mohr and Zanna study. Most importantly, a between-subjects design where the subject is primed by one body type, interacts with a woman, and is then tested for recall of the woman and the background. Using this design would make it possible to vary the body shape of the woman whom which the men interact, thus testing to see if being primed by one type of body shape affects social interaction with the same or different body types.
The lack of differences of attractiveness between the four body shapes might be do to the short list of traits on which the attractiveness was measured. It would be interesting to see if more traits increase or decrease the perceived attraction of the somatotypes. However, the results of the perceptions are reassuring because they contradict much of the thin ideal as being the most attractive figure. The unexpected results of the attractiveness ratings in and of themselves are worthy of further study. Are the somatotypes truly seen as equally attractive? Under what conditions is this true? Is this only true in the absence of other more pronounce physical or personality traits? Can it be manipulated by surrounding stimuli?
Also worthy of further testing is the explanation for why higher attractiveness caused higher recall. If the associations between the women and the product are the cause for higher recall, then removing this association might cause the attraction-attention (Roskos-Ewoldsen & Fazio, 1992). This could be tested by either substituting a feminine product with something more neutral (i.e. food, aspirin, real estate) or cross-gender (cars, beer, tools) or by placing the silhouette in a field of visual stimuli with no uniting pattern. With the lack of associations, a woman’s figure which elicits attraction might then elicit attention over the rest of the stimuli. While the associations have implications for marketing, placing a woman amongst a field of non- or neo-associative stimuli would have implications for the real world. Just as it is not common to see women acting sexually in public, save for bars and parties, women are not always surrounded by feminine stimuli because they are not normally walking billboards. Women are increasingly found in cross-trait stimuli, like a business environment. Would a man interacting with a woman in these pretexts focus on a woman he finds attractive and block of the surrounding stimuli? Would this interaction be affected by the woman’s body type, by the man’s gender schema, by both factors, or by neither factor?
While with these questions it appears the study has traveled in a circle, the study has not been for naught. Knowledge about how men perceive women on levels of attractiveness are important to research in a time of heightened concern about appearance where thinner is considered the ideal and young women feel pressured to meet this ideal. While the questions proposed were not specifically answered, unexpected findings were made and through them new questions were raised to hopefully help find the answers to the main question of men, women, and cross-sex interactions. A circle was made, but the circle became a little more complex.
Anderson, S.M. & Bem, S.L. (1981). Sex-typing and androgyny in dyadic interaction: Individual differences in responsiveness to physical attractiveness. Journal of Personality and Social Psychology, 41, 74-88.
Bem, S.L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155-162.
Bem, S.L. (1981). Gender schema theory: A cognitive account of sex-typing. Psychological Review, 88, 354-364.
Fong, G.T. & Markus, H. (1982). Self-schemas and judgments about others. Social Cognition, 1, 191-204.
Frable, D.E.S & Bem, S.L. (1985). If you are gender schematic, all members of the opposite sex look alike. Journal of Personality and Social Psychology, 49, 459-468.
Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35, 63-78.
Markus, H., Crane, M., Bernstein, S. & Siladi, M. (1982). Self-schemas and gender. Journal of Personality and Social Psychology, 42, 38-50.
Mckenzie-Mohr, D. & Zanna, M.P. (1990). Treating women as sexual objects: Look to the (gender schematic) male who has viewed pornography. Personality and Social Psychology Bulletin, 16, 296-308.
Renn, J.A. & Calvert, S.L. (1993). The relation between gender schemas and adults recall of stereotyped and counterstereotyped televised information. Sex Roles, 28, 449-459.
Roskos-Ewoldsen, D.R. & Fazio, R.H. (1992). On the orienting value of attitudes: Attitude accessibility as a determinant of an object’s attraction of visual attention. Journal of Personality and Social Psychology, 63, 198-211.
Spillman, D.M. & Everington, C. (1989). Somatotypes revisited: Have the media changed our perception of the female body image? Psychological Reports, 64, 887-890.
The purpose of this questionnaire is to assess your general personality. Answer the following questions as they relate to you on a day-to-day time frame. The questions are personality traits, and for each you need to circle the number on the 1 to 7 scale that best fits your general personality, so answer impulsively.
Never true Always true
1 2 3 4 5 6 7 1) Leader
1 2 3 4 5 6 7 2) Affectionate
1 2 3 4 5 6 7 3) Aggressive
1 2 3 4 5 6 7 4) Compassionate
1 2 3 4 5 6 7 5) Assertive
1 2 3 4 5 6 7 6) Sensitive
1 2 3 4 5 6 7 7) Self-reliant
1 2 3 4 5 6 7 8) Soft-spoken
1 2 3 4 5 6 7 9) Independent
1 2 3 4 5 6 7 10) Childlike
1 2 3 4 5 6 7 11) Warm
1 2 3 4 5 6 7 12) Forceful
1 2 3 4 5 6 7 13) Gentle
1 2 3 4 5 6 7 14) Cheerful
1 2 3 4 5 6 7 15) Risk taker
1 2 3 4 5 6 7 16) Tender
1 2 3 4 5 6 7 17) Individualistic
1 2 3 4 5 6 7 18) Sympathetic
1 2 3 4 5 6 7 19) Self-sufficient
1 2 3 4 5 6 7 20) Yielding
1 2 3 4 5 6 7 21) Decision-maker
1 2 3 4 5 6 7 22) Dominant
1 2 3 4 5 6 7 23) Ambitious
1 2 3 4 5 6 7 24) Shy
1 2 3 4 5 6 7 25) Analytical
1 2 3 4 5 6 7 26) Athletic
1 2 3 4 5 6 7 27) Understanding
1 2 3 4 5 6 7 28) Rigid
1 2 3 4 5 6 7 29) Gullible
1 2 3 4 5 6 7 30) Loyal