Investigating an online social network using spatial interaction models

TL;DR

This study applies spatial interaction models to analyze factors influencing inter-institutional links on the German social network StudiVZ, highlighting the impact of geographic distance and institutional attributes on acquaintanceship rates.

Contribution

It introduces a spatial interaction modeling approach to social network analysis, emphasizing geographic and institutional factors affecting social ties.

Findings

Acquaintanceship decreases by 91% with each additional 100 minutes of travel time.

Geographic separation is the most significant factor influencing social links.

Institution type and regional history also significantly affect acquaintanceship rates.

Abstract

Using a spatial interaction modeling approach, we investigate the German collegiate social network site StudiVZ. We focus on identifying factors that foster strong inter-institutional linkages, testing whether the acquaintanceship rate between institutions of higher education is related to various geographic and institutional attributes. We find that acquaintanceship is most significantly related to geographic separation: measuring distance with automobile travel time, acquaintanceship drops by 91% for each additional 100 minutes. Institution type and the former East-West German divide are also related to this rate with statistical significance.