Social discounting underlies individual altruistic decision-making, and it is frequently measured as the amount of hypothetical money one is willing to forgo for another person as a function of social distance. In the classic social discounting task, individual participants are asked to imagine their friends along a continuum of social distance, that is then used to estimate participant’s social discounting rate. While an ever-growing proportion of social interactions takes place over social media, no research has yet characterized social discounting in that context. Moreover, no research has estimated social discounting rate using real persons’ social distance, instead of the hypothetical continuum described above. Using existing social media indicators of social distance, it is now possible to estimate social discounting rate based on real people, which may lead to more accurate social discounting measurements and may expand the discounting model to real-life situations. Specifically, using computer algorithms to estimate the social distance from social media data makes it possible to assess the utility of numeric social distance indicators and the most appropriate ways to represent them. The proposed study examined the extent to which a hyperbolic model for social discounting fits social distance information retrieved from Facebook pages; and assessed whether there were differences in discounting rate when real or hypothetical social distance is used; also to further investigate whether discounting rates based on real persons are in fact based on perceived social distance by the participant, or on the imaginary social distance scale (i.e., an experimental artifact.)

It was found that the social discounting model can be applied in the social media context, even when real Facebook friends’ profiles were used as substitutes of numeric social distance indicators. Additionally, people showed similar altruistic tendencies in both the numeric and profile social discounting tests on the Facebook environment. These findings were qualified, however, by a high rate of nonsystematic data for the profile group; a rate much higher than traditional numeric paradigm. This discrepancy suggested that the allocation rates between numeric and profile approaches need further investigation to determine the factors affecting individuals’ generosity as a function of social distance indicators.