| This dissertation, which comprises three independent essays, explores the dynamics of online user behavior and provides IS policy implications across three different applications. The first essay employs an econometric empirical analysis to examine the role of IT interventions on online users' gambling behavior, based on field data collected over 10 years from 3,526 online gamblers. The second essay develops a model of third-party review systems and uses incentive-aligned experimental approach in investigating the role of identity disclosure on the third-party review system's reliability. The third essay incorporates a natural language processing algorithms (i.e., text mining approach) to examine the textual characteristics of online reviews and study when consumers perceive online reviews helpful, based on 19,533 online reviews collected from Yelp.;Several policy implications for IT firms and society are proposed based on new discoveries. First, providing simple IT interventions (such as behavior monitoring and partial restrictions) to online gamblers are effective in breaking habitual patterns in the context of online gambling. Second, by disclosing reviewers' identity, the third-party review system loses objectivity because the reviewers become altruistic toward service providers. Findings can be easily extended into designing online review systems. Third, online reviews containing diverse topics and unpleasant emotions with high arousal (e.g., anger, anxiety, and disgust) are perceived more helpful. Online review providers can use these findings in identifying the most helpful reviews at an early stage and motivate consumers to provide more helpful information. |
