The effects of question cues on responses and responders in community question-answering: a goal-framing approach

• Main Author: Wu, Qian
• Other Authors: Goh Hoe Lian, Dion
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Social Sciences; Question cues; Community question-answering; Framing theory; Knowledge co-creation; User-generated content
• Online Access: https://hdl.handle.net/10356/174975
• Institution: Nanyang Technological University

On community question-answering (CQA) sites, individuals can post and respond to user-generated questions that bridge askers with responders. Current studies mostly regard questions as a tool to deliver askers’ goals. However, the extant studies neglect how questions shape responders’ goal interpretation. For responders, questions function as goal expressions framing the scope of their CQA participation (e.g., answering, commenting, following). By regarding questions as goal-frames for responders, this research employs goal-framing theory (GFT) to probe how user-generated questions are associated with responders’ CQA participation. Based on GFT, this research proposes that question cues are characteristics of question expressions emphasizing gain, hedonic, or normative goals on responders. Examples of question cues are complexity, specificity, rewards, emotional expressiveness, self-disclosure, collaborative climate, popularity, and politeness. Specifically, as CQA is becoming a critical approach to informal learning, user-generated questions from academic topics require further investigation. The disciplinary differences in academic problem-solving guide this research to further explore how STEM (i.e., science, technology, engineering, and mathematics) and non-STEM topics are associated with the relationship between question cues and responders’ CQA participation. There are three research objectives: (1) establish relationships between question cues and the three goal-frames (i.e., gain, hedonic, and normative goal-frames); (2) investigate the association between question cues and responses (i.e., answers and comments) across different topic types; and (3) examine the association between question cues and responders’ CQA participation across different topic types. To achieve these objectives, three interrelated studies were conducted and a theoretical framework was developed to identify question cues associated with different goal-frames. Study 1 aimed to fulfill the first and second objectives. A total of 14,599 postings were collected. Content analysis was conducted to identify different question cues. Next, response networks to different question cues were constructed and compared. Results revealed that question cues emphasizing gain goals (i.e., gain goal-frame cues) were positively associated with interactions with questions from the STEM topic, while hedonic goal-frame and normative goal-frame cues were positively associated with interactions among responses from the non-STEM topic. To address the first and the third research objectives, a cross-sectional survey was conducted in Study 2. A total of 211 active responders’ cue perceptions were collected, and structural equation modeling was conducted to validate the theoretical framework that associates question cues with the three goal-frames. K-means analysis identified three question cue combinations (QCCs) attracting active responders, and discriminant analysis illustrated how topic types were associated with these QCCs. The three QCCs were: (1) QCC1 emphasizing gain, hedonic, and normative goals, which was closely associated with non-STEM topics; (2) QCC2 emphasizing gain goals, which was related to STEM topics; and (3) QCC3 emphasizing hedonic goals, which was associated with both STEM and non-STEM topics. Based on the two studies, Study 3 focused on the third objective and explored the causal relationships between QCCs and responders’ CQA participation. An online experiment with a 2 (topic types: STEM, non-STEM) × 3 (QCCs) factorial design was conducted, and 109 participants were recruited. The results showed that high complexity and specificity attracted more reads, follows, and shares but fewer answers and comments than those without. Further, question cues emphasizing hedonic and normative goals attracted more answers and comments. Regarding disciplinary differences, non-STEM responders preferred the QCC emphasizing gain, hedonic, and normative goals, whereas STEM responders had divergent opinions regarding QCCs. This PhD research opens a new direction to investigate questions as goal expressions for responders. By adopting GFT, the findings reveal the disciplinary differences in the influences of question cues. The findings of this PhD research can guide CQA designers to facilitate askers to attract responders by employing QCCs. Responders’ active participation is critical to a sustained ecosystem of knowledge co-creation. Last but not least, the findings regarding human beings’ responses to distinct question cues can be used to train artificial intelligence algorithms and incorporate them in online knowledge co-creation.