Abstract
Sample means are not the whole story. Individual differences in aesthetic preference are a hallmark of aesthetic experience. We all find different things beautiful. To demonstrate the importance of individual differences in aesthetic preferences, we asked participants to rate, on a scale from 1 (not at all) to 7 (very much), the beauty of images that other participants thought would produce diverse ratings. We measured image typicality and the distribution of beauty ratings for each image. Then we determined the subset of images for which most participants rated one image as low in beauty and another as high in beauty. We present the Disputed Quartet — four images that are rated as typical and have highly variable beauty ratings with the same mean — and predict that most people will find in it an image they like and another they dislike. We tested 100 people and the mean difference between their maximum and minimum ratings was around 3 points on a 1-7 Likert beauty scale. The Disputed Quartet exemplifies how relying on means leaves out the variations in aesthetic preferences. So, statistically, let’s get past the first moment and consider the second moment.