Diminishing returns of information quality: Untangling the determinants of best answer selection

Online question and answer (Q&A) platforms are ideal live repositories for knowledge sharing and exchange. On Q&A platforms, a seeker may receive one answer to a question, none, or multiple answers. In the case of many answers, it is important to understand which answer will be selected as the best to inform how seekers retrieve knowledge-based solutions to inquiries. This study develops a best answer selection model that integrates the characteristics of the information exchanged as perceived by the knowledge community and those of the individuals exchanging such information (i.e., answers contributors and answers seekers). Using data from StackOverflow.com, we built conditional logit models to examine the determinants of the best answer selection. Results show that contributors' reputation and the quality of the information (measured as the net vote score received by an answer) are important determinants of best answer selection. This research theorizes on and uncovers a non-linear relationship between the net vote score received by an answer (a proxy for information quality) and its likelihood of being selected as the best answer by the seeker. We also find that the seekers' reputation and the contributors’ reputation influence how information quality affects the selection of the best answer.