An exploratory study of Twitter metrics for measuring user influence

Influence plays a critical role in information communication, and the ubiquitous use of social media has made measuring influence on social media platforms a salient challenge. While previous studies have attempted to measure and investigate influence on Twitter, there is no consensus on its definition or relation to fundamental Twitter metrics. This study examined relationships between a composite influence measure of Twitter and fundamental social media metrics using a sample of tweets from a multi-year public campaign. Correlation analyses indicated that a user's number of followers had the strongest correlation with the composite measure. Principal components analysis was conducted for dimension reduction, and multiple regression analysis was performed using the resulting components. The findings revealed that a user's network was the most important predictor of the composite influence measure and that there was a negative relationship between campaign related activity and the composite measure. Implications of these findings are discussed. Overall, this study contributes to the understanding of and future efforts in the measurement of influence on social media.