For technical-knowledge workers seeking information about how to complete software tasks, online social question and answer (SQA) sites represent a valuable resource as an emerging form of software documentation. However, because answerers on these sites respond to questions on a volunteer basis, not all questions receive answers. Current research shows that askers provide contextual information in varying amounts, yet researchers have not yet reliably described contextual information types, disagree on whether more or less information associates with answerability, and have not yet compared the coherence of answered and unanswered questions. To assist technical-knowledge workers posting questions on SQA sites, this study explores the relationship between contextual information and answerability and between logical coherence and answerability.

This study analyzes 3,529 contextual-information t-units and 690 comment t-units from social how-to questions about Microsoft Word that askers posted on the popular SQA site Super User. Content analysis enabled a close examination of not only the amounts of contextual information that askers provided, but also the types of information, relationships among types, and relationships between types and answerability. Establishing and using three reliable codebooks related to social how-to questions, to contextual information, and to answerers’ follow-up comments, the study presents descriptive statistics and examples of contextual-information types and comment types. Further analyzing contextual-information types, the study presents and explores statistical differences in the distinctness, magnitude, variation, efficiency, word count, and logical coherence of contextual information in answered and unanswered questions.