Item

Abstract

The intrinsic variance of a beauty rating is the essential baseline for evaluating how beauty ratings vary within and across individuals. Given independent repeated measures, it’s easy to estimate the variance of the distribution underlying one judgment. However, in the context of beauty, effects of assimilation, contrast, and recall-memory might produce sequential dependence, spoiling the independence. In seven experiments, we measure how much of the variability across beauty ratings can be attributed to recall memory and sequential dependence. We find that recall memory contributes little to the variability of repeated judgment (less than 0.7 standard deviation on a 7-point Likert scale) for both subjective beauty ratings and an ellipticity judgment with ground truth. We also find that order effects, including assimilation and contrast, become measurable only when the stimuli are very similar, regardless of task. Lastly, we find that, after discounting the small recall-memory contribution, the intrinsic variance of beauty judgment increases slightly with stimulus set size. Our results indicate that the response to a given stimulus in a diverse set is affected by the number and not the value of other stimuli, which parallels Miller’s (1956) 7±2 limit on one-dimensional perceptual categorization. Overall, we show that, provided the stimuli are not very similar, recall memory and sequential dependence hardly affect the intrinsic variability of beauty judgment. Hence, the variance of repeated ratings does estimate the intrinsic variance of beauty judgment of one stimulus in a diverse set.