The validity of composite photographs for assessing masculinity preferences

Among non-human animals, exaggerated male traits may signal immunocompetence and are often attractive to females. This finding has not been reliably replicated in human populations, however, where preferences for feminised male faces are often observed in experiments. Some authors have suggested that certain experimental approaches, and in particular the use of 'composite', or averaged, photographs, may lack ecological validity and elicit misleading responses. This may account for reports of preferences for femininity.

To date, however, experiments have varied on numerous, often confounded, methodological dimensions, making the contribution of different factors difficult to assess. To address this, we conducted an experiment to examine the influence of stimulus type on participant responses. Masculinity preferences were measured using both composite photographs and the photographs from which they were composed, while all other methodological variables were held constant. In contrast to predictions by previous authors, composites elicited significantly stronger overall preferences for masculinity, especially when textural information was reduced during the averaging process. The existence of preferences for feminised male faces is unlikely to be explained by the use of composite photographs alone.