Mutual exposure or close peer relationships do not seem to foster increased similarity in food, music or television program preferences.

We know very little about either the mechanisms through which preferences are created in humans, or the contributions of particular developmental forces: e.g. genetics, parental influence, peer influence, the media. Prior work has indicated surprisingly low correlations (averaging about 0.15) between the food or music preferences of young adult and the mid-point of the biological and rearing parents preferences for the same items. A likely candidate for a substantial influence is peers. In one study, we show that freshman college roommates, randomly assigned, do not become significantly more similar in their food or music preferences over the course of about seven months of mutual exposure. In a second study, we show that in three suburban third grade classrooms, the preferences for food or TV programs in children is not more similar to those of their best friends in the school, than it is to randomly selected other children of the same gender, in their classes. Although our findings are negative (lack of influence), in light of very strong expectations for positive findings, we take them to be notable. These findings add further mystery and puzzlement to the question of: where do preferences come from? The focus in this study is on individual differences among peers and forces that lead to preference similarity. The study does not speak to the influence of general peer norms.