Relationship development on Twitter : a mixed-method approach to trust building and social capital accrual

Abstract

Humans have a fundamental need for social connection and belonging. Twitter, a popular social network site (SNS), provides a new mode of interaction for people to develop interpersonal relationships with others. Given Twitter’s unique features, it is an ideal context to study how the mode of interaction can shape the development and impact of interpersonal relationships. In this dissertation, I investigate this issue by comparing relationships on Twitter to previous findings in two key theoretical areas of relationship development: trust and social capital. As a secondary aim, I make multiple methodological contributions in the area of SNS research. Data for this dissertation come from two separate studies: a pilot survey study of 99 Twitter users and a longitudinal, mixed-methods study of 494 Twitter users that incorporates user log data from Twitter. Chapters 2 and 3 are dedicated to improving the measurement of SNS engagement, an important and widely used variable in SNS research. Chapter 2 is a systematic review of the psychometric properties of all the scales that have been developed to measure SNS engagement. I found that common practices are not in sync with validity evidence in this area, and that the most popular measure of SNS engagement, the SNS Intensity (SNSI) scale, has serious concerns for its validity. In Chapter 3, I attempted to provide a fully valid version of the SNSI scale by testing an improved version of it, the Kuru & Pasek SNSI (KP-SNSI) scale. Unfortunately, the KP-SNSI scale did not have sufficient structural or discriminant validity. However, I did identify a suitable measure of SNS usage, which I then used in place of SNS engagement in my remaining analyses. Chapters 4 and 5 are substantive investigations focused on two essential components of interpersonal relationships: social capital and trust. In Chapter 4, I investigated the development and impact of social capital on Twitter. I found that Twitter users do develop both bonding and bridging capital (the two main subtypes of social capital). I found that bonding capital lead to greater well-being, but bridging capital lead to greater loneliness. While the association between bonding capital and well-being is well-known in offline contexts, bridging capital’s effect on loneliness is surprising, and likely due to unique aspects of Twitter as a mode of interaction. In Chapter 5, I formulated and tested two competing models of trust development on Twitter. I found that the model based on social influence is a better explanation of my data. In addition, I found that self-disclosure and trust may be mutually reinforcing on Twitter, and that social network structure apparently does not affect the development of trust on Twitter. In Chapter 6, I reviewed my findings and contributions. I identified a number of methodological contributions that may benefit other SNS researchers in the future. In reviewing how Twitter shapes interpersonal relationships, I find that the picture is nuanced and complex. I conclude that it is important remember that Twitter is simply a tool, whose effect is determined largely by the humans who use it, for better or for worse.