Homophily among VAA users: The case of Ecuador

Abstract

Users who recommended a Voting Advice Application online platform to others in their social networks were tied to any individuals who elected to follow the recommendation, in order to form a network dataset. This study examines the homophily hypothesis, the tendency of Inviters and Invitees to be disproportionately similar in demographics, political orientation and attitudes for VAA-generated data, employing random y-permutation statistical techniques borrowed from Social Network Analysis. The results confirm the homophily hypothesis, with Invitees being more similar with their respective Inviters, than would have been expected by chance or population make-up for all variables examined, demographic, political and attitudinal, with the exception of gender.