Do men hold African-American and Caucasian women to different standards of beauty?✩
Abstract
Racial differences in men’s preferences for African-American and Caucasian women’s body size and shape were examined. As expected, there was a trend for African-American men to choose ideal figures with a lower waist-to-hip ratio (WHR), which is associated with a more curvaceous figure. Contrary to expectations, however, African-American men did not choose heavier female figures as ideal. In fact, both groups chose underweight and normal weight figures as ideal. The results from this study suggest that while preferences for WHR may continue to be associated with cultural factors, African-American and Caucasian men may have become more similar than different in their preferences for female weight. Also, the results suggest that within the African-American sample, there were two subsamples with regard to WHR preferences, with one subgroup endorsing the same ideal WHR as their Caucasian counterparts. The results are discussed in terms of possible changes to cultural values that may be reflected in a change in what is considered attractive.Keywords: African-Americans, Caucasians, Waist-to-hip ratio
1. Women to different standards of beauty
Recent estimates are that over 73% of African-American women are overweight or obese compared to 63% of Caucasian women (Center for Disease Control and Prevention, NHANES, National Center for Health Statistics, 2005). On the other hand, Caucasian women, especially those of higher Socioeconomic Status (SES), have been thought to be disproportionately represented among those with anorexia nervosa and other types of unhealthy, restrictive eating patterns (Crago, Shisslak, & Estes 1996; Striegel-Moore et al., 2003). Other research alternatively suggests that the rates of eating disorders among ethnic minority women may be comparable to a similar sample of Caucasian women (Mulholland & Mintz, 2001). Similarly, Striegel-Moore and Cachelin (2001)concluded that the rates of eating disorders in ethnic minority women are underreported due to the lack of participation of ethnic minority women in treatment studies. The investigations based upon research showing differential rates between groups have suggested that different beauty ideals between the two cultures may contribute to the differences seen in eating pathology and weight, especially among women (Abrams, Allen, & Gray, 1993; Harris, 1994; Greenberg & LaPorte, 1996; Parker et al., 1995; Wilfley et al., 1996).2. Body size
Research exploring differences between African-American and Caucasian male preferences for female body size over the past two decades has generally found African-American men to be more accepting of larger body sizes for women than Caucasian men (e.g., Cohn & Adler, 1992; Fallon & Rozin, 1985; Greenberg & LaPorte, 1996; Thompson, Sargent, & Kemper, 1996). For example, Thompson et al. (1996), examined adolescent males’ perceptions of ideal female body size. They found that African-American males preferred a larger female size than Caucasian males. Furthermore, when subjects without girlfriends were asked to estimate the height and weight they would desire for a girlfriend, the calculated Body Mass Index (BMI) was significantly different between ethnicities, with African-American males desiring a BMI in the appropriate weight category for 15 year old females and Caucasian males desiring a BMI value that fell into the underweight category (Thompson et al., 1996). Thompson et al. (1996) also assessed social norms and found that Caucasian males surmised that their parents, male friends, and female friends would prefer an ideal female size smaller than what African-American subjects felt their parents, male friends, and female friends would prefer.When the authors of this study from the “male perspective” compared their findings to other studies from the “female perspective,” using the same silhouettes, they found that their Caucasian male participants selected an ideal female size larger than that chosen by Caucasian females of similar age in the previous studies (e.g. Fallon & Rozin, 1985; Kemper, Sargent, Drane, Valois & Hussey, 1994). Additionally, their African-American male participants selected as ideal female figures that were larger than those chosen by African-American females of similar age in other studies (Kemper et al., 1994; Cohn et al., 1987). One limitation of this study lies in the failure of investigators to assess what participants believed members of the other ethnic group would select as ideal. This limitation is addressed in the current study.
3. Body shape
Other studies have identified body shape, or waist-to-hip ratio, as an important feature of female attractiveness for which Caucasian and African-American men may have different preferences. Low WHR, typically defined as .68–.80, has been suggested to be an indicator of reproductive health and fertility (Henss, 1995; Singh, 1993, 1994a,b,c) and also a signal of the absence of major diseases (Bjorntorp, 1988, 1991; Leibel, Edens, & Fried, 1989). Thompson et al. (1996) found African-American males were 1.9 times more likely to select a larger ideal female hip/buttocks size and 1.7 times more likely to choose a larger ideal female thigh size than Caucasian subjects. Evidence from other investigations demonstrated that African-Americans indicated a greater attraction to heavier figures than did Caucasian men, including a higher ideal weight and a preference for larger buttocks (Cunningham, Roberts, Barbee, Druen, & Wu, 1995). Similarly, in a recent study, Freedman, Carter, Sbrocco, and Gray (2004) found that African-American men often preferred WHRs in the range of .55–.65. They found that in addition to being more accepting of larger body sizes for women, African-American men preferred a WHR that was lower than the WHR preferred by Caucasian men. Such findings lend support to the notion that men’s weight and shape preferences constitute cultural variants that may contribute to the maintenance of unhealthy weights among African-American women. However, not all studies have supported these conclusions. Although Singh (1994a,b,c) found that African-American men did not prefer heavier women, the range of WHRs presented to participants was restricted (.70 through 1.0). Additionally, participants were only presented with figures that were Caucasian in appearance.4. Acculturative factors and ethnic identification
Acculturation to mainstream culture may impact ethnic differences in women’s body image and beauty ideals, as well as men’s perceptions of female beauty. Several studies have suggested that African-American women who ascribe to the values of mainstream Caucasian culture are more at risk for eating disorders and maladaptive eating habits (Abrams et al., 1993; Crago et al., 1996; Osvold and Sodowsky, 1993; Rucker & Cash, 1992). Abrams and colleagues (1993), for example, demonstrated that African-American women who embrace Caucasian culture also endorse eating disorder-related attitudes. Furthermore, scores on the pre-encounter subscale of the Racial Identity Attitudes Scale (Helms, 1990), a measure of stage of racial identity, were significantly positively correlated with measures of eating pathology including Restrain, Fear of Fat, and Drive for Thinness subscales. It should be noted African-Americans in the pre-encounter stage of identity development typically exhibit a desire to be more like Caucasians culturally. As such, they may behave in a manner consistent with Caucasian culture and believe more in traditionally Caucasian values (e.g., individuality, competitiveness). Therefore, it is not surprising that such African-Americans would also adhere to Caucasian attitudes regarding eating and beauty. Conversely, it would seem likely, then, that African-American men who ascribe to pro-Black viewpoints and who are less acculturated will select heavier ideal figures due to a rejection of Caucasian culture, however this has not been successfully investigated to date. Additionally, one’s level of acculturation may impact an individual’s beliefs about what a member of the other ethnic group would choose as ideal. For example, a Caucasian man living in an integrated community might begin to become aware of differences between what he and what African-American men find attractive about women’s bodies. This awareness may or may not influence his own preferences and ideals of beauty.5. Inter-racial preferences
Greater acceptance of a variety of body sizes and shapes and even idealization of heavier body sizes seem to be factors that serve to buffer African-American women from restrictive eating and body image pathology (Cunningham et al., 1995; Thompson et al., 1996). However, such cultural variants may also be factors that serve to maintain unhealthy attitudes and practices that contribute to the higher rates of overweight and obesity in African-American women (Hebl & Heatherton, 1998; Steele, 1992). Alternately, African-American women’s awareness that African-American men may be more accepting of heavier body sizes (Freedman et al., 2004) may contribute to the perception that there are few incentives to lose weight. The role of cultural variants, however, becomes even more complex, when the prevalence of inter-racial relationships is considered.Level of acculturation may also differentially affect the ideal figure chosen by those African-Americans who are and are not willing to date women outside of their own racial group. Acculturation may either indirectly impact choice of the ideal figure for the other ethnic group, via one’s beliefs about inter-racial dating, or it may directly impact the selection of an ideal female figure, due to level of exposure to individuals of the other ethnic group. For example, Jackson and McGill (1996), in their investigation of African-American and Caucasian men’s body type preferences for women, found that when looking at same-race females, African-American men preferred a larger female body type than did Caucasians. In addition, African-American men associated fewer unfavorable characteristics and more favorable characteristics with obese same-race females than did Caucasian males.
6. The present study: rationale and purpose
To date, there have been no studies investigating men’s body size and shape preferences for women of different ethnic backgrounds. Using silhouette drawings of Caucasian and African-American female figures, Caucasian and African-American men’s preferences were assessed for both sets of silhouettes. In doing so, this study attempted to clarify whether men apply differential standards to African-American and Caucasian women, in terms of overall body weight and waist-to-hip ratio. In addition, men’s levels of acculturation were assessed in order to discern the influence of acculturation on dating preferences and subsequent female body size and shape preferences.Three primary hypotheses were put forth. First, African-American men were expected to prefer a heavier body size and a lower WHR than their Caucasian counterparts. Furthermore, African-American men who were more acculturated to Caucasian culture were expected to show preferences more aligned with those of Caucasian men. Second, men who date inter-racially were expected to hold all women to standards of beauty similar to those of their ethnic group. That is, Caucasian men who date inter-racially would choose women with thinner, more tubular figures as ideal for both groups, while African-American men who date inter-racially would choose heavier, more curvaceous figures as ideal for both groups. Third, when asked about their beliefs about the preferences of the other ethnic group, participants were expected to cite an ideal female figure that aligned with cultural stereotypes for the other ethnic group. For example, African-American men were expected to report that Caucasian men prefer a thin, tubular figure. Conversely, Caucasian men were expected to report that African-American men prefer a heavier, curvaceous figure.
7. Method
7.1. Participants
The participants were 100 non-Hispanic males (50 Caucasian; 50 African-American) between 18 and 58 years of age. Participants were recruited from the Washington, DC and surrounding communities through flyers and newspaper advertisements. All participants self-identified their ethnic group membership.7.2. Measures
7.2.1. Demographic questionnaire
The demographic information form assessed age, ethnicity, current height and weight, marital status, employment, and level of education. It also assessed level of occupation and education for participants’ parents. The answers were used to determine participants’ SES according to a derivation of the Hollingshead Four-Factor Scale of Socioeconomic Status (Hollingshead, in press). Socioeconomic scores were computed using the formula (Occupation × 5)+ (Education × 3). Participants were placed into one of five social class categories ranging from 1 = lowest to 5 = highest. The cut-off scores that were used were the ones advocated by Hollingshead (in press).7.2.2. Dating preferences questionnaire
Participants were asked questions about their current and past romantic and sexual relationships in order to assess whether or not they had flexible dating practices with regard to ethnicity. In addition, if participants did not date individuals of ethnic and racial groups other than their own, they were asked to respond in an open-ended fashion about their reasons for not doing so. Since the vast majority of participants (93%) endorsed flexible dating practices, the open-ended responses of the few remaining participants were not analyzed.7.2.3. African-American Acculturation Scale (AAAS-33)
The short-form of the African-American Acculturation Scale (Landrine & Klonoff, 1995) measures the extent to which African-American individuals participate in the cultural traditions, beliefs, assumptions, and practices of the dominant Caucasian society vs. remaining immersed in their own cultural traditions. This version correlates well with the original 74-item long form (r=.94) and has good concurrent and group differences validity. The total score is computed by summing across the 33 items and can range from 33–231 with lower scores indicating greater acculturation to Caucasian culture.7.2.4. Height and weight
Body weight and height were self-reported by participants. Body mass index (BMI) was calculated as weight (kg)/height (m)2. Participants were classified as underweight, normal, weight, overweight or obese based on the NHLBI standards (NIH, 1998).7.2.5. Silhouette stimuli
The silhouette stimuli used were based on those used by Freedman et al. (2004). In the Freedman et al. (2004) study, participants indicated their preferences for female body size and shape using both modified Singh (1993, 1994a,b,c) stimuli and modified Tassinary and Hansen (1998) stimuli. Singh’s (1994a) original stimuli depicted women with a height of five feet four inches. The original stimuli varied along 3 levels of body weight (underweight, normal weight, and overweight) and four levels of WHR (.70, .80, .90, 1.0). This range of WHRs encompasses some of the types of figures that occur in the general population, but does not account for females with very curvaceous figures (WHRs in the .50 to .70 range). The revised Singh stimuli used in theFreedman et al. (2004) study used 11 levels of WHR (from .50 to 1.0 with increments of .05) with the same 3 levels of body weight as used in the Singh studies. These weight categories, however, were not representative of actual women, and were skewed toward the lower end of the Body Mass Index scale. Specifically, the heaviest silhouette depicted a 68.2 kg woman, a BMI of 25.7 kg/m2, which falls in the very low end of the overweight BMI range, which is 25 kg/m2 (NIH, 1998). These silhouettes are not representative of today’s U.S. population in which over 60% of adults and over 70% of African-American women are overweight (BMI between 25 kg/m2 and less than 30 kg/m2). Similarly, 30.5% of adults and almost 50% of African-American women are obese, with a BMI equal to or greater than 30 kg/m2.The present study developed new figures, based upon the original Singh (1994a) figures and the Freedman et al. (2004) modifications. The WHR modifications made by Freedman et al. (2004)were maintained. The two major modifications included the addition of heavier silhouettes and shading of the figures to represent African-American and Caucasian women. These figures are depicted in Figs. 1 and and22.
Fig. 1
Depiction of African-American figures.
Fig. 2
Depiction of White American figures.
7.2.5.1. Weight categorizations
Four new weight categories were introduced to replace the former, skewed weight categories. The five weight categories in the present study were “underweight,” “normal weight,” “overweight,” and “obese,” and “very obese,” corresponding to the following BMI values (in kg/m2): 17, 22, 27, 32.5, and 37.5, respectively.7.2.5.2. Shading
The Singh line drawn figures were shaded and colored by a graphic artist to represent African-American and Caucasian figures. The hairstyle of the figures is similar. Facial features were not added. Aside from shading, the figure sets are identical. The purpose of this addition was to clarify any differential standards of beauty that men apply to women of the two racial groups. The two sets of figures displayed identical weight and WHR levels.8. Procedure
Participants were recruited to participate in a study examining dating preferences. Participants were first asked to complete the packet of self-report measures. Each set of figures was shown to each participant. African-American participants were shown the African-American figures first while the Caucasian participants were shown the Caucasian figures first. The cover story states that “For each of the following sets of figures, you will be asked to select your ideal female figure. There are no right or wrong answers. Please look at all figures and select just one.” For each set, participants were asked the same question, ”Which figure represents your ideal female figure?” Then, participants were shown both sets of figures simultaneously and asked to choose the figure that represents the ideal female figure chosen by a member of the other racial group. Following completion of the study, all participants were debriefed. As an incentive for participating, each individual received $25.9. Results
Group differences on continuous measures were examined using multiple analysis of variance (MANOVA) and analysis of variance (ANOVA). Group differences on categorical measures were examined using chi-square and Mann–Whitney tests. All analyses were examined using an alpha of .05 and 95% confidence intervals.9.1. Demographic data
Demographic and descriptive data for age, height, weight, Body Mass Index scores, acculturation scores, and SES levels are presented in Table 1. A MANOVA, examining the variables age, height, and weight, indicated significant group differences F(5, 90)=3.346, p<.01. Follow-up ANOVAs indicated significant differences between groups on age, F(1, 94)=15.696, p<.01 with African-American participants being significantly older than Caucasian participants. No other significant differences were found. (see Table 1).Table 1
Descriptive data for African-American and Caucasian participants | African-American | Caucasian | ||
|---|---|---|---|---|
| | | ||
| X | (SD) | X | (SD) |
Age | 33.60 | (9.24)a | 26.10 | (9.30)a |
Height (in.) | 71.13 | (2.74) | 71.00 | (3.18) |
Weight (lb) | 190.85 | (33.61) | 185.61 | (30.48) |
BMI (kg/m2) | 26.52 | (4.01) | 25.72 | (4.41) |
Acculturation | 124.62 | (28.36) | N/A | |
SES | 39.62 | (11.04)b | 48.06 | (8.67)b |
Note. X = Mean, SD = standard deviation, N = number of subjects, BMI = Body Mass Index difference, SES = Socioeconomic Status, at p<.01,
astatistically significant,
bstatistically significant difference at p<.05.
The range of SES scores was 12–64. For Caucasian participants, the mean SES score was 48.06, SD=8.67 (category=4), indicating an upper middle class position. For African-American participants, the mean SES score was 39.62, SD=11.04 (category=3), indicating a middle class position. Overall, the mean SES score was 43.80, SD=10.76 (category=3), indicating a middle class position. Groups differed in their placement in the five SES categories (chi-square (4)=11.548, p<.05). Contrary to expectations, BMI and SES were not significantly correlated as might be expected by the fact that individuals with heavier body weights are over-represented in the lower SES levels. This is likely due to a restricted range on SES (with relatively few represented in the lower class).
9.1.1. Hypothesis: do African-Americans prefer heavier body sizes and a lower WHR?
The preferences for participants’ ideal and least favorite weights were tested using chi-square. The preferences for participants’ ideal and least favorite WHR were tested using a Mann–Whitney, because with 6 levels of WHR, one can assume approximate continuity. The distribution of these preferences was not normal, as indicated by a Shapiro–Wilk test of normality. Follow-up t-tests for proportions were used to clarify the significant differences between groups for each WHR level. African-Americans were expected to prefer a heavier body size and lower WHR than their Caucasian counterparts. This hypothesis was partially supported. There were no group differences for ideal and least favorite weight. As depicted in Table 2, there was moderate support for African-Americans preferring a lower WHR (see Table 2).Table 2
African-American and Caucasian mens’ ratings of ideal and least favorite weight and WHR for female silhouettes | African-American silhouettes | Caucasian silhouettes | ||
|---|---|---|---|---|
| | | ||
| % AA raters | % CA raters | % AA raters | %CA raters |
| | | | |
| (n=50) | (n=50) | (n=50) | (n=50) |
Ideal weight | | | | |
Underweight | 52.0 | 56.0 | 40.0 | 52.0 |
Normal | 32.0 | 36.0 | 42.0 | 44.0 |
Overweight | 12.0 | 8.0 | 14.0 | 4.0 |
Obese | 4.0 | 0.0 | 2.0 | 0.0 |
Very obese | 0.0 | 0.0 | 2.0 | 0.0 |
Least favorite weight | | | | |
Underweight | 18.0 | 12.0 | 18.0 | 10.0 |
Normal | 4.0 | 0.0 | 0.0 | 2.0 |
Overweight | 0.0 | 0.0 | 4.0 | 0.0 |
Obese | 4.0 | 6.0 | 4.0 | 2.0 |
Very obese | 74.0 | 82.0 | 74.0 | 86.0 |
Ideal WHR | | | | |
Low | 36.0 * | 18.0 * | 40.0 | 32.0 |
Moderate | 58.0 * | 68.0 * | 48.0 | 58.0 |
High | 6.0 * | 14.0 * | 12.0 | 10.0 |
Least favorite WHR * | | | | |
Low | 46.0 | 62.0 | 44.0 * | 64.0 * |
Moderate | 16.0 | 12.0 | 28.0 * | 20.0 * |
High | 38.0 ** | 26.0 | 28.0 * | 16.0 * ** |
AA = African-American, CA = Caucasian American.
*Significant Mann–Whitney, p<.05.
**Significant follow-up t-test for proportions, p-value<critical p-value=.05.
9.1.2. Preferences for same-race figures
Did African-Americans’ preferences for African-American figures differ significantly from Caucasians’ preferences for Caucasian figures? There were no significant differences between the groups for either ideal weight for same-race figure (chi-square (3)=4.947, p=.176) or least favorite weight for (chi-square (3)=2.260, p=.520) same-race figures.Based on a Mann–Whitney test, there were no significant differences between groups for ideal WHR for same-race figures (U=1171.000, p=.534). For least favorite WHR for same-race figures, there were significant differences between groups (U=962.000, p=.027). Here, 64.0% of Caucasians chose a low WHR as least favorite while only 46.0% of African-Americans did for same-race figures. In addition, more African-Americans disliked a high WHR than did Caucasians for same-race figures (38.0% vs. 16.0%). Follow-up t-tests for proportions showed this comparison to be significant, t(99)=2.48, p=.0149 (<critical p=.016). Thus, there appears to be greater group consensus on which WHR constitutes an unattractive figure than what constitutes an attractive one.
9.1.3. Preferences for other-race figures
When asked to choose an ideal weight and a least favorite weight of the opposite ethnic/racial group the choices of African-American men and Caucasian men did not significantly differ for either the ideal weight (chi-square (4)=4.382, p=.357) or for the least favorite weight (chi-square (3)=3.005, p=.391).For ideal WHR for other-race figures, there were significant differences between groups, (Mann–Whitney test U=993.000, p=.045). Here, more African-Americans than Caucasians chose a low WHR as ideal for other-race figures (40.0% vs. 18.0%). This comparison approached significance in a follow-up t-test for proportions, t (99)=2.42, p=.0172 (<critical p=.016). Also, more Caucasians chose a moderate WHR as ideal for other-race figures than did African-Americans (68.0% vs. 48.0%). Here, there was more group consensus on what was an attractive WHR for other-race figures amongst Caucasian men than amongst African-American men.
For least favorite WHR for other-race figures, there were no significant differences between groups (U=1074.000, p=.181). The majority of participants in both groups chose a low WHR as least favorite for other-race figures (44.0% of African-Americans, and 62.0% of Caucasians).
9.1.4. Group differences by figure set
Finally, chi-square and Mann–Whitney tests were used to compare African-Americans’ and Caucasians’ preferences for weight and WHR for each set of figures. For ideal weight, there were no significant differences between groups for either African-American figures (chi-square (3)=2.592, p=.459) or for Caucasian figures (chi-square (4)=5.584,p=.232). There were no significant group differences for least favorite weight for African-American figures (chi-square (3)=3.005, p=.391) or for Caucasian figures (chi-square (4)=4.926, p=.295).For ideal WHR for African-American figures, there were statistically significant differences between groups in the expected direction (U=974.500, p=.026). For African-Americans, 36.0% chose a low WHR as ideal compared to only 18.0% of Caucasians. Another 58.0% of African-Americans chose a moderate WHR as ideal compared to 68.0% of Caucasians. Six percent of African-Americans chose a high WHR as ideal compared to 14.0% of Caucasians. Follow-up t-tests for proportions found that the comparison between Caucasian and African-Americans for low WHR approached significance, t(99)=2.03, p=.045, (<critical p=.016).
For least favorite WHR for African-American figures, the differences between groups approached significance (U=1045.000, p=.116) because there were considerably more Caucasians who disliked a lower WHR (62.0% vs. 46.0%) for African-American figures and considerably more African-American participants (38.0% vs. 26.0%) who disliked a higher WHR for African-American figures. There were about equal percentages of participants who disliked a moderate WHR (16.0% of African-Americans vs. 12.0% of Caucasians).
For ideal WHR for Caucasian figures, there were no statistically significant differences between groups (U=1177.000, p=.574). For least favorite WHR for Caucasian figures, the numbers roughly mirrored participants’ choices for least favorite WHR for African-American figures. There were statistically significant differences between groups (U=986.000, p=.044). Similar to participants’ choices for least favorite WHR for African-American figures, more African-American participants (28% vs. 16%) disliked a high WHR and more Caucasian participants disliked a low WHR (64.0% vs. 44.0%). However, follow-up t-tests for proportions showed no significant comparisons between groups. The comparison between groups for those who chose a low WHR as least favorite approached significance, t(99)=−2.01, p= .048>critical p= .016. See Figs. 3 and Figs. 4
Fig. 3
Participants’ ideal and least favorite weight.
Fig. 4
Participants’ ideal and least favorite whr.
Overall, for WHR preferences for African-American figures, moderate WHRs were chosen as ideal and low WHRs as least favorite. There was a trend for slightly more African-Americans preferring a lower WHR than their Caucasian counterparts. For Caucasian figures, moderate WHRs were preferred, while low WHRs were considered least attractive. Again, there was a trend for more African-Americans to prefer a low WHR and for more Caucasians to dislike a low WHR for these figures. Although the majority of all participants preferred a moderate WHR, for those who did prefer a low WHR, there were more African-American men in this category. Such results lend some support to the stated hypothesis in that some African-American men would prefer a lower WHR than their Caucasian counterparts.
9.2. The impact of acculturation
The hypothesis that the most acculturated African-Americans would show preferences for figures most like those of Caucasian men was not supported. The range for the total AAAS-33 score was 62–183. The range possible is somewhat wider: participants can score between 33 and 231, indicating that this sample was restricted in range. The mean score was 124.62, SD=28.36 and the median score was 124.50. The mean score for all African-Americans in the Landrine and Klonoff (1995)combined sample (original and new) was 147.80, SD=28.26, which is somewhat higher than the mean obtained here, suggesting that the present study consisted of a sample of African-Americans that was more acculturated to Caucasian culture than the original sample. There were no differences in acculturation between older and younger African-American participants, t(48)=.797, p=.429). The mean AAAS score for African-American participants younger than 30 years of age (N=22) was 121.00, SD= 27.38 while the mean for those 30 years of age and older was 127.46 (N=28), SD=29.28.In order to assess whether scoring high or low on the acculturation measure impacted the expressed preferences of African-American participants, a median split was implemented to select out participants who scored low vs. high on the measure. There were 25 participants scoring below the median and 25 scoring above the median. For African-American silhouettes, there were no significant differences between the two groups on expressed preferences for ideal weight (chi-square (3)=2.154, p=.541), for least favorite weight (chi-square (3)=3.027, p=.387), for ideal WHR (U=289.500, p=.640), or least favorite WHRs (U=287.500, p=.618). Similarly, for Caucasian silhouettes, there were no significant differences high and low acculturation African-Americans for ideal weight (chi-square(4)=3.714, p=.446), for least favorite weight (chi-square(3)=3.027, p= .387), for ideal WHR (U=302.500, p=.842), or for least favorite WHR (U=245.000, p=.182).
9.3. Dating practices and beauty ideals
We expected that men who date-interracially would hold women of both ethnic groups to the same standards of beauty. Because almost all participants, regardless of race, endorsed having flexible dating practices (92.8%), (as evidenced by either currently dating someone outside one’s own racial group, or by expressing a willingness to date someone outside one’s own racial group), the analysis of any resulting differences on the basis of dating practices was hindered. When comparing those who do have flexible dating practices versus those who do not, there were no differences between racial groups (chi-square (2)=.031, p=.758). Of the 47 African-American participants who answered this question, 93.6% had flexible dating practices, while 6.4% did not. Similarly, 92.0% of Caucasian participants had flexible dating practices, while 8.0% did not.9.4. Cross-ethnic beauty preferences
Regardless of dating practices, did participants hold women to the same standards of beauty? This question was explored using Fisher’s exact test to compare differences in the frequency with which Caucasian versus African-American participants applied the same standards of beauty across figure sets. A total of 14 participants (14%) applied the exact same standards of beauty (same categories of weight and WHR for ideal and least favorite figures) to African-American and Caucasian figures. Out of these 14 participants, 4 were African-American and 10 were Caucasian. This difference was not significant (p=.148).Some participants who did not hold women to the exact same standards of beauty did apply the same standards for either weight or WHR but not both. Seventy-two (72%) participants applied the same standards to African-American and Caucasian figures for ideal weight. Thirty-two of these participants were African-American and 40 were Caucasian. This difference was not significant (p=.118). For least ideal weight, 84 (84.0%) participants applied the same standards to both sets of figures, 40 of whom were African-American and 44 of whom were Caucasian. This difference was not significant (p=.414). For ideal WHR, 36 (36.0%) participants applied the same standards to both sets of figures, 20 of whom were African-American and 16 of whom were Caucasian. This difference was not significant (p=.532). Finally, for least favorite WHR, 56 (56.0%) participants applied the same standards to both sets of figures, 25 of whom were African-American and 31 of whom were Caucasian. This difference did not reach statistical significance (p=.314).
9.5. Beauty ideals for the other racial group
Caucasian men’s expectations for African-American men. To assess cultural stereotypes, each participant was asked which figure he believed a member of the other racial group would choose as ideal (see Table 3). For Caucasian men, 34.0% believed an African-American man would choose an underweight figure as ideal. Another 46.8% believed an African-American man would choose a normal weight figure as ideal. Only 19.2% of Caucasian men believed African-American men would choose overweight and obese figures as ideal, contrary to expectations that most Caucasian men would believe that African-American men liked overweight female figures. In reality, most African-American participants chose underweight and normal weight figures as ideal, in accordance with Caucasian participants’ expectations.Table 3
What did participants believe a member of the other racial group would choose as ideal? | % Caucasians (N=47) | % African-Americans (N=49) |
|---|---|---|
| | |
| For African-Americans | For Caucasians |
Ideal weight | | |
Underweight | 34.0 | 73.5 |
Normal | 46.8 | 22.4 |
Overweight | 12.8 | 4.1 |
Obese | 6.4 | 0.0 |
Very obese | 0.0 | 0.0 |
Ideal WHR | | |
Low | 55.1 | 22.4 |
Moderate | 36.7 | 49.0 |
High | 8.2 | 28.6 |
For WHR, 55.1% believed African-American men would choose a low WHR as ideal, while 36.7% believed they would choose a moderate WHR as ideal. Only 8.2% believed that African-American men would choose a high WHR as ideal. This result reflects the cultural stereotype that African-American men prefer more curvaceous figures. Contrary to Caucasian participants’ expectations (and those of the primary investigator), most African-American men (58.0%) chose a moderate WHR as ideal for African-American figures. Another 36.0% of African-American men chose a low WHR as ideal for African-American figures and 6.0% chose a moderate WHR as ideal. For Caucasian figures, most African-Americans chose moderate and low WHRs as ideal (88.0%) in accordance with Caucasian participants’ expectations.
9.6. African-American men's expectations for Caucasian men
Examining what African-American men believed Caucasian men would choose, 73.5% believed that Caucasian men would choose an underweight figure as ideal in accordance with expectations. In accordance with expectations, Caucasian men chose underweight and normal figures as ideal for both African-American figures (92.0%) and for Caucasian figures (96.0%). For African-American figures, more Caucasian participants chose underweight figures as ideal as compared with the number who chose normal weight figures as ideal (56.0% vs.36.0%). However, for Caucasian figures, the numbers who chose underweight vs. normal weight figures as ideal was more evenly split (52.0% vs. 44.0%).For WHR, 22.4% of African-Americans believed Caucasian men would choose a low WHR as ideal. Another 49.0% believed they would choose a moderate WHR as ideal and 28.6% believed they would choose a high WHR as ideal. For each set of figures, Caucasian men preferred a moderate WHR (58.0% for the Caucasian figures and 68.0% for the African-American figures).
It seems that when considering each set of figures independently, there were few differences between groups on what constituted ideal weight but stronger differences on what constituted ideal WHR. However, African-American men (but not Caucasian men) seemed to hold different ideas of what constitutes an ideal weight and WHR, depending on the race of the figure. More specifically, African-American men (on the whole) preferred an African-American figure that is underweight with a moderate WHR but a Caucasian figure that was underweight or normal weight with a low or moderate WHR. Also notable is that more Caucasians believed African-Americans would choose a low WHR than actually did, while African-Americans accurately predicted that most Caucasians would prefer a moderate WHR.
9.7. Impact of participants' BMI, age, and SES on preferences?
Overall, based on self-report, participants were slightly overweight (see Table 1). The majority of the participants fell into the normal weight and overweight categories (44% and 37%, respectively). Another 10% qualified as obese, while 4% were very obese and 1% extremely obese. Interestingly, having a higher BMI did not necessarily translate to choosing a heavier figure as ideal. Similarly, comparing the weight and WHR preferences of those who were very high on SES with those who were very low on SES did not result in significant differences. Chi-square analyses of ideal and least favorite weight for both sets of figures and Mann–Whitney tests of ideal and least favorite WHR for both sets of figures did not yield significant differences between the youngest and oldest participants.10. Discussion
The results of this study are not consistent with previous research demonstrating that African-American men are more accepting of larger body sizes for women (Cohn & Adler, 1992; Fallon & Rozin, 1985; Greenberg & LaPorte, 1996; Thompson et al., 1996). Most striking, was the similarity between African-American and Caucasian men’s preferences for female weight. All men exhibited preferences for underweight and normal weight women, for both sets of figures. The results, however, support Singh’s (1994a) findings that African-Americans do not differ significantly from Caucasians in terms of preferred body weight for women. Because the figures in this study realistically depicted a full range of women’s weights and WHRs, it is unlikely that the results are an artifact of the stimuli. That is, men did not choose smaller sizes because this was all that was available. One possible interpretation of this finding may be that norms for African-Americans have changed. The findings differ from Freedman et al. (2004) findings based on data collected approximately 2–3 years earlier. It seems clear that in order to study ethnic–racial differences, care must be taken in understanding the samples used. It may also be that the results reflect that subjects were drawn from a major metropolitan area that is approximately 60% Caucasian, therefore may be more accepting of traditional Caucasian viewpoints. Partially supporting this position is that scores of the African-American sample on the acculturation measures were slightly lower than published norms. However, that the earlier study by Freedman and colleagues (2004)used a similar sample mitigates against this argument.The finding that African-Americans prefer underweight and normal weight women is also interesting given that the rates of obesity among African-American women continues to rise.
It is possible that African-American males are responding to more recent media images depicting smaller, thinner African-American women. This finding may have also been a reflection of the sample being somewhat more acculturated to Caucasian culture as evidenced by lower scores on the AAAS. This position is further supported by this sample reporting that they had flexible dating practices and were willing to date outside of their ethnic group.Despite the overall similarity, some differences in preferences for WHR between groups, did emerge, however. Among those who chose a low WHR as ideal, more were African-American. Similarly, for those who chose a moderate WHR as ideal, more were Caucasian. For same-race figures, there was more group consensus among Caucasians that a low WHR was not attractive, while African-Americans were split in choosing a low and moderate WHR as unattractive. This may reflect the presence of two subgroups within the African-American sample; those who disliked low WHRs and those who disliked high WHRs. Those who disliked low WHRs appear to have preferences more akin to those of Caucasian men and find a moderate WHR most attractive, in accordance with what Singh posits as “universally appealing”. The other subgroup consists of African-American men who dislike a high WHR. Curiously, in this study these findings were not found to result from differences in acculturation level. It is possible that acculturation not relating to choice of figure resulted from a restricted range of scores on the acculturation measure. It is also possible that the measure of acculturation does not capture aspects of African-American culture that would influence preferences for female size and shape.
As predicted in the assessment of cultural stereotypes, most African-Americans believed Caucasians would choose an underweight figure as ideal. Most Caucasians believed African-American men would choose underweight and normal weight figures, which they did. With respect to weight, African-Americans clearly hold the stereotype that Caucasians like thinner people.
For WHR, approximately half of the African-American subsample believed that Caucasian men would choose a moderate WHR as ideal, while the other half was split between a low and a high WHR. Caucasian men, did in fact, choose a moderate WHR as ideal for both Caucasian figures and African-American figures. Similarly, more than half of the Caucasian men believed that African-American men would choose a low WHR as ideal, consistent with cultural stereotypes. Most African-American men in this study chose a moderate WHR as ideal while some chose a low WHR as ideal. Consistent with cultural stereotypes, it appears that Caucasians believed that a low WHR would be more popular amongst African-Americans than it was in actuality. This may suggest that while African-American men’s preferences for women’s body shape are changing in a direction more consistent with the preferences of Caucasian men, that the cultural stereotype has not yet “caught up” to this change and is consequently not yet recognized by Caucasians.
Consistent with the primary findings, approximately 93% of men who participated in this study, regardless of race, endorsed having flexible dating practices as evidenced by either currently dating someone outside one’s own racial group, or by expressing a willingness to date someone outside one’s own racial group. Because the participants responded to this question1 so uniformly, any differential standards applied to Caucasian vs. African-American women based upon dating practices could not be assessed according to group preferences. These more liberal than expected dating practices may reflect the changing culture and is certainly consistent with the finding that there is a blending in perception of ideal body shape. This may also have influenced (or be reflective of) the lower level of acculturation level reported by African-American participants. It is also possible that this result reflects the participants’ desire to appear “politically correct” more so than their actual dating practices. However, many participants indicated that they have in the past, dated women from a variety of ethnic backgrounds, rather than a mere willingness to date someone from their own. It should be noted that demand characteristics may also have led some participants to attempt to choose a figure as ideal (for the Caucasian set of figures) that they believed approximated the figure of the experimenter, who was also Caucasian. However, this does not seem likely given that the task did not involve a high degree of interaction with the experimenter despite her presence in the room. Future studies might include a Caucasian and an African-American experimenter to assess the impact of experimenter race on the expressed preferences of men.
The current study addressed several limitations of previous research. The figures depicted in the current study more accurately reflected the bodies of real women and through the use of two sets of figures (one Caucasian and one African-American), allowed for participants with different preferences for each group to express such preferences in a differential manner. Overall, it appears that over time, men are increasingly showing preferences for women who are very thin. This trend is likely fueled by media depictions of women. Because the media is controlled, for the most part, by Caucasians, it makes sense that the images of women projected to the public mirrors the thin ideal that has become increasingly prevalent in American culture.
The results of this study suggest that African-American men do not prefer heavier women. If women are not protected from restricting types of eating disorders by men’s preferences, then one might expect the rate of restrictive eating pathology to rise for African-American women. Some research has suggested that the rates of restrictive types of eating disorders in African-American women are increasing to the prevalence rates of Caucasian women (O’Neill, 2003) while other research suggests that the differential rates persist (Striegel-Moore et al., 2003).
Although the findings from this study extend the existing research on men’s preferences for female body size and shape, it should be noted that weight and WHR may not be the most important features to men when considering the attractiveness of a potential mate. Facial characteristics, personality, and overall appearance may be more important. Given that African-American women have been shown to have more multi-dimensional body images, African-American men may also be less inclined to focus on actual body size and shape, and may pay more attention to a woman’s overall attractiveness, including her style of dress, hair, and personality. Therefore, asking African-American men, in an experimental situation, to then focus on body shape and size, may have left some of them at a loss as to how to choose figures on such narrow criteria. As well, using two-dimensional line drawings to assess preferences for WHR may be problematic in that such drawings do not capture the three-dimensional nature of women’s bodies. Future studies could make use of new software available, such as Poser, in order to allow participants to view figures from all angles before selecting an ideal figure. Despite these limitations, however, it is intriguing to speculate that there may be an important shift in the perception of beauty. The impact of the shift, however, awaits further study.
Aww poor sheboons
somebody save the sheboons
