Essays on Digital Platform IT Artifacts

Abstract

Digital platforms implement Information Technology (IT) artifacts to improve business performance. The transformation effects of such business innovations applied in various fields such as online entertainment platforms, social media, and online communities call for further investigation. My dissertation focuses on discovering the economic values of the IT artifacts implemented by digital platforms. First, I study an online reading platform that offers a by-content purchase method and a novel in-chapter online comment function to mitigate consumer uncertainty regarding the true quality of e-books. Specifically, I seek to uncover how such intervention affects consumer purchase decisions. Utilizing a Bayesian learning framework, I unveil the consumer learning process enabled by such a by-content purchase method. I find that consumers learn the quality of books within different genres from their direct experience at different paces. Moreover, I uncover the heterogeneous signaling effects of in-consumption comments within different topics on book quality. For instance, consumers facing more comments that comprise a neutral or emotional discussion over the plot or plead the author to post new chapters would revise upward their perception of the book quality. Finally, I find that consumers approaching the end of a book or exposed to a list of chapters with few informative names tend to skip more chapters instead of reading continuously. Second, I examine the effectiveness of a potential weapon that can combat the rapid spread of public health-related misinformation on social media platforms. Specifically, I focus on Twitter's intervention that aims to suppress misinformation by helping users find accurate information. In this study, I seek to provide a holistic view of the intervention's effectiveness by investigating its impact on true information diffusion. To this end, I employ a difference-in-differences model. Surprisingly, I find that the intervention suppresses not only the spread of false news but also true information. Further analysis reveals that true information is also suppressed because people have difficulty discerning the truthfulness of the information. Through a correlational analysis, I provide additional insights into the tweet characteristics that tend to mislead people's perceptions. Last but not least, I study individuals' engagement behaviors on an online freight driver-oriented community-based Question-and-Answer platform, with a specific focus on the difference in users' responding behaviors to urgent and non-urgent questions. To motivate users' voluntary contributions, the platform employs various IT artifacts. Through a multi-dimensional Hidden Markov Model, I provide empirical evidence on the heterogeneous effects of such motivating mechanisms and community characteristics on users' underlying motivation state transition as well as corresponding responding behaviors across urgent and non-urgent inquiries. Moreover, the analysis reveals the unintended effects of the formation of sub-groups inside the main community, alerting the online Q\&A forum. While a large main community increases users' contributions, a sub-group with more members and higher exposure to unsolved questions on the platform may discourage users from contributing more responses.