Implications of behavioral and cognitive reciprocity in social network data

Abstract

Cognitive maps of social behavior are of interest for theoretical as well as methodological reasons. The issue that has primarily been addressed, in comparing cognitive with behavioral social network data, is whether the participants' level of accuracy in reporting their social behavior is high enough to warrant the use of cognitive data as a substitute for behavioral data. The present paper focuses, instead, on what these imperfectly matched sets of data may tell us about low level processes of cultural codification, namely, those revealed by the transformation of behavioral interaction to cognitive report. Specifically, the analyses undertaken here indicate that the discrepancies between the two kinds of data are systematic, and that the cognitive data reflect an integration of the participants' direct relationships with each other into the set of relationships they observe as well as participate in. Participants tend to discount their high-frequency co-interactants for whom they, in turn, are relatively low-frequency co-interactants, in favor of those with whom they are reciprocally high. Thus, in reporting their direct relationships, they are in effect codifying them in terms of indirect relationships - namely, the relationships their co-interactants have with others in the set.