Lisa Oshita - Survival Analysis of iFixit's Online Question and Answer Forum

Abstract

Community-driven online question and answer forums (CQA) are becoming increasingly valuable sources of information. These platforms house an expansive amount of crowd- sourced knowledge in the form of thousands of questions and answers posted everyday. There are forums that cover a broad range of topics, like Yahoo! Answers, and forums focused on specific topics, like computer programming-focused Stack Overflow. An example of the latter is iFixit’s Answers forum. As the main goal of iFixit is to enable people to fix their own devices by providing them with the knowledge and tools necessary, the Answers forum, which features questions related specifically to device repair, is a valuable source of information for this community. Thus, it is important that questions receive timely answers. This paper presents a survival analysis on the time until a question receives its first answer. We developed a Cox proportional hazards model to predict the failure probability of a question, or the probability that a question receives an answer before a certain time. Though we identified signifcant predictors, the predictive accuracy was low (R2 = 0.15). Our findings indicate that the most important predictors were the device category of the question (questions pertaining to Apple products received answers faster than others (HR = 2.56, 95% CI = (2.31, 2.82))) and factors related to the question’s title (e.g., whether or not it was phrased as a question (HR = 1.31, 95% CI = (1.22, 1.41))). Future studies could investigate if the factors identified as signifcant in our study can be generalized to other CQAs.