The influence of physical appearance on a fair share

There are many factors that influence people’s concept of a fair share. Personal preferences, additional information, as well as previous experience make humans choose different strategies when interacting with different opponents under otherwise similar circumstances. To account for such factors computationally, we previously introduced a utility-based model called priority awareness. In this paper, we present a study with human participants that provides additional evidence that humans use priorities when they need to think about a fair share. We perform this study by using a method called human computation, which can be used to explicitize implicit information by asking many people to perform only a few tasks. More specifically, we look at the impact of physical appearance on what humans consider a fair share. In an online survey, we present people with a few fictive opponents represented by a photo, and ask them to play an Ultimatum Game with these opponents. We aggregate the human data into a ranking and find that appearance has a strong influence on what people consider a fair share, i.e., people associate a different priority with different opponents, purely based on immediate appearance. The same method may be used to elicit other implicit influences on human priorities. Such influences may then be taken into account by developers of, for instance, multi-agent systems.