Toward Understanding Online Impression Management

How Twitter Users Control Textual Expressions Over Time

Abstract

Impression management on social networking sites is becoming ever more important as people live in an increasingly connected world where people frequently initialize, develop, and maintain relationships with others online. Previous studies have shown that people form impressions differently depending on their relationship with the audience. However, few studies have focused on its longitudinal aspect: how people manage their impressions by controlling expressions over time according to the audience. In this study, we investigated temporal changes in textual expressions (e.g., neurotic words), and then analyzed how such changes related to the audience size (i.e., followers), density (i.e., mutual connections), and feedback (e.g., Likes). An analysis of 5 million posts collected from 1.6 thousand Twitter users over a period of 2.5 years revealed that users who had developed more mutual connections with their audience tended to use more neurotic and conscientious expressions. Meanwhile, users who received more Likes from their audience were found to write less neurotic or conscientious languages. Our findings indicate that Twitter users gradually adjust their use of expressions through interactions with audiences, which may ultimately change the impression others have of them.