Abstract

This study provides a predictive measurement tool to examine perceivedanxiety from a longitudinal perspective, using a non-intrusive machine learningapproach to scale human rating of anxiety in microblogs. Results suggest thatour chosen machine learning approach depicts perceived user state-anxietyfluctuations over time, as well as mean trait anxiety. We further find areverse relationship between perceived anxiety and outcomes such as socialengagement and popularity. Implications on the individual, organizational, andsocietal levels are discussed.

 

Quick Read (beta)