In recent years, active learning and knowledge-sharing platforms (henceforth ACKs) have gained recognition as powerful educational tools enabling users to learn and practice myriad topics, such as programming and foreign language from any place and at any time. However, the high rate of user dropouts and low user engagement are impeding users’ endeavors to learn, collaborate, and share their knowledge on these platforms. Although a vast body of extant studies has examined user engagement and incentives in several ACKs such as Connectivist MOOCs, significant research gaps remain unaddressed: (1) The research on some salient ACKs, including inquiry-based ACKs, remains sparse. The most prominent examples of inquiry-based ACKs include Question-based Learning platforms (QLs) such as Math Pla...[ Read more ]

In recent years, active learning and knowledge-sharing platforms (henceforth ACKs) have gained recognition as powerful educational tools enabling users to learn and practice myriad topics, such as programming and foreign language from any place and at any time. However, the high rate of user dropouts and low user engagement are impeding users’ endeavors to learn, collaborate, and share their knowledge on these platforms. Although a vast body of extant studies has examined user engagement and incentives in several ACKs such as Connectivist MOOCs, significant research gaps remain unaddressed: (1) The research on some salient ACKs, including inquiry-based ACKs, remains sparse. The most prominent examples of inquiry-based ACKs include Question-based Learning platforms (QLs) such as Math Playground and Duolingo, as well as Question Answering websites (QAs) such as Stack Overflow and Ask Ubuntu; (2) there is a paucity of explicable dropout forecast models for inquiry-based ACKs that can determine the underlying reasons for users dropping out of these platforms; and (3) a lack of awareness about the reasons behind gamification failure in inquiry-based ACKs.

This thesis aims to address these research gaps by adopting a mixed quantitative and qualitative research approach. The thesis comprises two key segments investigating user engagement and incentives in QL and QA platforms, respectively. Each platform entails its unique design attributes and subtle nuances that, in turn, require a thorough investigation.

When examining QLs, we first characterize the engagement patterns (moods) of users over time in a large-scale QL typically used to impart training in computational and programming to undergraduate students (users). Subsequently, we present a novel hybrid dropout prediction model benefitting from the utilization of students’ engagement moods in order to enhance the accuracy of dropout predictions in QLs. According to our findings, users working on QLs exhibit collective preferences to answer questions premised on the engagement mood category with which they are associated. Any deviation from these collective preferences significantly increases the probability of user dropouts. Gamification denotes a popular strategy to avoid or mitigate user dropouts in similar scenarios. Nevertheless, in the capacity of an external incentive, it is often fraught with its own share of problems. Within this thesis, our subsequent study adopts a qualitative research approach to explore one of the most pressing concerns in gamification, i.e., an adverse phenomenon alluded to as gamification misuse, in a large-scale gamified QL. We undertake careful thematic analysis to identify the most common factors underpinning gamification misuse, before classifying them into two groups: active and passive. To mitigate or prevent the occurrence of gamification misuse in their future designs of gamified learning platforms, we also provide gamified QLs with a practical set of suggestions.

In the process of studying QAs, we first investigate user dropouts in QAs through the lens of flow theory, a well-known psychological theory. The theory posits that users tend to be highly engaged in their experience when the tasks (assignments) encountered by them are congruent with their skill levels. Accordingly, we present a method of new task assignment that may help decrease user dropouts in QAs. We then explore promotional gamification schemes in QAs in a subsequent study. Promotional gamification refers to a temporary gamification scheme added atop an already gamified QA to increase user engagement for a short time-span (e.g., during the holiday season). This thesis demonstrates that the addition of more gamification schemes to a pre-existing gamified platform does not always increase user willingness or engagement to contribute more to QAs. On the contrary, it risks increased user dropouts and overjustification after the removal of additional gamification schemes.

Overall, this thesis provides unique insights to inform researchers’ and practitioners’ understanding of user engagement and incentives in inquiry-based ACKs, potentially enabling them to reduce users’ dropout rates, improve their learning experiences, and obviate unnecessary mishaps such as gamification misuse and overjustification.