BLATANT BENEVOLENCE AND SOCIAL CAPITAL ATTAINMENT ON SOCIAL NETWORK SITES – A MULTI-METHOD APPROACH

Abstract

While information systems literature includes multiple studies on knowledge contribution or sharing, blatant benevolence in general has received less attention. To extend the knowledge of the blatant benevolence concept, this dissertation analyzes the blatant benevolence – social capital link on social network sites based on three different studies. The first study implements an interview method to explore antecedents, forms, and consequences of blatant benevolence. More importantly, it also explores the moderators that change the relationship between blatant benevolence and social capital attainment on social network sites. After conducting 126 interviews with social media users, we find that five most common online blatant benevolence forms are: donation behavior, volunteering behavior, participating at a soup kitchen, paying forward for a customer behind, and attending a mission trip. We identify several moderators, such as frequency of the posts, others vs. self-posting the prosocial behavior etc., which could impact the relationship between blatant benevolence and social capital attainment. The second study implements a 2x2x2 online experiment to empirically test the (i) relationship between blatant benevolence and social capital attainment on social network sites, (ii) moderating effect of frequency of the prosocial post, and (iii) others vs. self-posting the prosocial behavior. We find that blatant benevolence increases relational social capital and structural social capital. We also find that the frequency of the prosocial post increases the effect of blatant benevolence on relational social capital and cognitive social capital. Others posting the prosocial behavior, rather than self-post, increases the effect of blatant benevolence on relational and structural social capital. The third study relies on observational Twitter data to empirically test the main effect - relationship between blatant benevolence and social capital attainment - to validate the findings of the second study. We randomly identify 100,000 Twitter users from Twitter and download their personal profile within a certain period. From the 100,000 Twitter users, we identify users who posted prosocial contents within that period and labelled them as prosocial group. We then apply a propensity score matching to identify similar Twitter users as our control group. Applying the dynamic panel analysis, we compare the growth rate of the followers between the prosocial group and the control group by controlling the individual differences. Our results show that the prosocial group has gained more followers than the non-prosocial group - hence providing additional support to the significance of the main effect: blatant benevolence – social capital link.