This study reconsiders the “double standard of aging” hypothesis, which states that older women are judged as being less attractive than older men. We considered the subjects sexual orientation (i.e., heterosexual, gay, lesbian, bisexual) and showed that older women were not judged more negatively than older men per se. Male subjects exhibited a stronger youth bias than females, but only if the target picture was a potential sexual partner. Females showed an equal amount of preference for youthful faces, independent of the targets sex and regardless of whether the target was a potential sexual partner.
Studies analyzing attitudes towards aging and the elderly have often found that older women are judged more negatively than older men (e.g., Berman et al., 1981, Deutsch et al., 1986 and Laurence, 1964). This phenomenon was first referred to as the “double standard of aging” by Susan Sontag (1972), who suggested that modern urbanized societies allow two standards of male beauty: the boy and the man, but only one standard of female beauty: the girl. An explanation for this is provided by evolutionary theories: they assert that youthfulness is a more salient cue for men than for women when faced with the task of mate selection, due to the fact that women’s fertile years are typically more limited than men’s, and more precipitously terminated by the onset of menopause (cf. Bailey et al., 1994 and Kenrick and Keefe, 1992).
Previous studies have suggested that men are attracted to relatively younger women (Kenrick & Keefe, 1992), that men place greater emphasis on physical attractiveness because of its value as an age cue (Buss, 1989, Feingold, 1990 and Thornhill and Gangestad, 1999) and health indicator, that women’s femininity is perceived to decrease with age, which is not the case for men’s masculinity (Deutsch et al., 1986), and that age is a more salient cue for men than for women (Kogan, 1974 and Kogan, 1979). Thus, most studies have been supportive of the double standard of aging hypothesis, although there have been exceptions for attractiveness ratings (Zebrowitz, Olson, & Hoffman, 1993), for attitudes towards the elderly (Kite et al., 1991 and Öberg and Tornstam, 2003), and for the double standard in self-perceptions (Wilcox, 1997). In general, a more substantial and consistent effect of a double standard of aging was found when photographic stimuli were used as opposed to verbal descriptions (for an overview of the earlier studies see Kogan & Mills, 1992).
In most of these previous studies involving picture ratings, the sexual orientation of the subjects was not recorded and we can thus assume that the subjects were mostly heterosexual. As a consequence, there has mostly been a confound between the sex of the subject and the sex of the target as a potential mate. For male subjects, the targets displaying potential mates were female, and for female subjects, the targets displaying potential mates were male. This confound makes it impossible to tell if the reason for the documented sex of subject by sex of target interaction lies in the target or in the subject. An evolutionary explanation, as proposed for example by Kenrick and Keefe (1992), would suggest that it is due to the subject as it is more important for men than for women that their sexual partner is youthful. But it is also conceivable that sex of subject by sex of target interactions result because women’s appearance is indeed more susceptible to the effects of aging, due for instance to greater physiological or hormonal changes, or that beauty standards are different for men and women through socio-cultural norms.
A study by Silverthorne and Quinsey (2000) included homosexual men and women in their samples and found indeed that the variance could be explained by the subject better than by the target. That is, homosexual and heterosexual men would prefer younger partners of their preferred sex than would homosexual and heterosexual women. However, since Silverthorne and Quinsey asked their participants specifically for ratings of the targets’ sexual attractiveness, only the ratings for the preferred sex (opposite sex for heterosexual participants, same sex for homosexual participants) can be interpreted, since the ratings for the targets belonging to the other sex were all very close to the minimum, across all age groups.
The goal of the present study was to elucidate the effect of age on attractiveness ratings for both the preferred and the unpreferred sex. Specifically, we were interested in whether men showed a stronger preference for youthful faces than women in general, even when the face was not a potential mate, or whether the stronger preference for youthfulness was limited to potential mates. For this, we used a similar paradigm as Silverthorne and Quinsey, including homosexual as well as heterosexual subjects in order to avoid the usual confound between the sex of the subject and the sex of the target. Additionally, we also included bisexual subjects. However, instead of asking the subjects how sexually attractive they found a particular target face, we asked them only how attractive they found the face, in order to get interpretable results for all combinations of target sex, subject sex, and subject sexual orientation.
Furthermore, there are some methodological concerns about Silverthorne and Quinsey (2000) study that we wanted to address with the present study. One surprising finding by Silverthorne and Quinsey was that the maximum ratings were not consistently given to the youngest targets, but partly to those closer to the participants’ own age. We suspect that this pattern may not be generalizable, but may rather be explained by characteristics of the sample, specifically the age range of the participants, since the participants’ age was not included as a covariate in the analysis. For our own study, we expected the youngest pictures to show the highest ratings if the age of the participants is treated as a covariate in the statistical analysis. Furthermore, Silverthorne and Quinsey raise the concern that their small picture sets (3 pictures per age group) may have led to artifacts in the results. We avoided this problem with the following measures: (1) by conducting a preliminary study where the age and attractiveness of a larger number of pictures were rated, before selecting the final pictures sets, (2) by using larger sets of pictures in the final stimulus material, and (3) by having two complete picture sets to which the participants were randomly assigned, and whose results could be compared in order to get a measure of internal reliability.
