This study analyzes the characteristics of questions asked on social question-and-answer (Q&A) sites and explores how the characteristics of those questions affect deliberation quality. We collected COVID-19 virus discussions from an English (Quora) and a Chinese (Zhihu) social Q&A site. Content analysis found that certain question characteristics (i.e., length, type, emotion) significantly influenced user engagement and content quality. Longer questions were more likely to trigger answers with higher user engagement yet lower content quality. Conversational questions tended to develop answers with a higher engagement level, while informational questions were more likely to lead to higher content quality and more diverse content. Neutral questions might trigger higher user engagement level and higher rationality. Cross-platform and cross-culture aspects are also discussed.