Name generation in interpersonal political network data: Results from a series of experiments

We present results from three large scale survey experiments focused on the manipulation of political name generators. Using syntax that is widely employed outside of political science, we generate interpersonal political network data by varying the roles of alters, the time horizons of relationships, and the specific political nature of social exchanges. Across varying samples and electoral environments, we look for differences in these conditions on a wide range of common interpersonal network items, assess latency data on these treatments, and employ more detailed information on named discussants than most existing political ego-centric studies. We evaluate how well the now standard “compound” political name generator captures interpersonal political networks, finding that it does quite well save a few items of significant political importance. We discuss the implications of this research agenda for theories of social influence and the study of disagreement in democratic politics.

? We present three survey experiments focused on the manipulation of political name generators. ? Treatments vary the roles of alters, recall periods, and the political nature of social exchanges. ? Treatments yield distinct patterns of interaction, composition, and discussion initiation. ? The standard political name generator performs well; variations may add theoretical leverage.