Attractive, approving, and trustworthy faces are perceived as more similar to one another than their negative counterparts. The same goes for liked compared to disliked acquaintances, celebrities, groups and, in fact, all sorts of positive compared to negative words and pictures. That is, “good is more alike than bad” is a robust property of people’s information ecology.
Good is more alike than bad may be due to the moderateness of positivity. On many, if not most, evaluatively relevant dimensions (e.g., temperature), one range of moderate, positive amounts is flanked by two ranges of extreme, negative amounts (e.g., too cold and too warm). This inverted u-shaped relation between amount and evaluation results in good is more alike than bad for stimuli evaluated based on one dimension alone. Multidimensional evaluation further constrains and diversifies positivity and negativity, respectively, and thus intensifies good is more alike than bad. My research tests this and other possible explanations.
Why bother about good is more alike than bad? The literature is full of valence asymmetries in information processing: Positive compared to negative information is less distracting, is recognized faster, is categorized more broadly, is generalized more readily, is remembered less accurately, is less decisive, is less likely a difference between people and things, is harder on the mind when judging others, etc. Further, similarity influences information processing from attention to memory to evaluation. Thus, good is more alike than bad may explain and predict a variety of behavior-relevant valence asymmetries in information processing (independently of the alternative explanation “bad is stronger than good”), which I examine in my research.