Getting to the Heart of Groups – Analyzing Social Support and Sentiment in Online Peer Groups

The success of online peer groups in addressing social problems has attracted researchers’ attention investigating their inner workings. While social support, for instance informational or emotional support, is considered the theoretical foundation of peer groups and sentiment a key indicator for groups’ efficacy, the linkage between both concepts is still unexplored. To address this gap, we analyze a unique dataset from a field experiment at the German Federal Employment Agency providing online peer groups in the context of unemployment. Building on sentiment classification, content analysis and methodically inspired by event study methodology, we quantitatively analyze the impact of messages providing different types of social support on the online peer groups’ sentiment. We not only investigate whether a message has impact, but also how much and how long. Results show that, while informational support is exchanged most frequently, only emotional support has substantial and sustainable impact on peer groups’ sentiment.