Code to accompany the paper "When does reputation lie? Dynamic feedbacks between costly signals, social capital, and social prominence" by Marion Dumas, Jessica L. Barker, and Eleanor A. Power.
Models were run in Python, with figures generated in R.
The only data used in this paper draws on earlier work by Power (see "Discerning devotion: Testing the signaling theory of religion" in Evolution and Human Behavior, open access version here). The data represent the change in reputational standing for those residents who fulfilled vows in the annual festival for the goddess Māriyammaṉ, in the village of "Teṉpaṭṭi" (a pseudonym) in 2013.
The reputational nominations (see details in the paper referenced above) from before the festival were gathered from virtually all adult village residents (98%) in April 2013. The festival took place in August 2013. In the days immediately following the festival, reputational nominations were again gathered from a stratified random sample (stratified by caste/religion) of 50 adult residents. To make these comparable, we look at the change in the percent of possible reputational nominations (of any time) that a person received from before to after the festival.
For our measure of social capital, we draw on social support name generator questions also asked in April 2013. To allow for the easiest comparison to the analytical and agent-based models considering here, here we use in-degree (i.e., the total number of people who named a person as someone they would turn to for help). This is then normalised, as we do with Si in the models, so that this is a close approximation of our measure of social prominence/capital.
For more on the relationship between religious practice and social support here, see "Social support networks and religiosity in rural South India" in Nature Human Behaviour and "Collective ritual and social support networks in rural South India" in Proceedings of the Royal Society B, open access version here.