JMIR Publications

ABSTRACT

Background:

Health information consumers find question and answer (Q&A) communities as an alternative way of meeting their information needs. However, the quality of their questions can directly impact what answers they receive from the community and whether they receive an answer at all.

Objective:

This study aims to improve our understanding of quality of questions from health information consumers in online Q&A communities.

Methods:

To this end, we develop a research construct of question quality and its operationalization for the first time, then empirically identify the content and language determinants of question quality. Moreover, we propose a novel approach to measuring the overall quality of questions by adapting the k-means clustering algorithm and automated measures of indicators of question quality. To validate our proposed constructs and determinants of question quality, we collected and analyzed questions for kidney disease from both expert-curated and community-based Q&A communities. The research construct was validated by manual assessment of health experts, and the determinants were tested using regression analysis.

Results:

High-quality questions are more likely to include demographic and medical information than lower-level quality questions. At the same time, asking questions at the various stages of disease development was more likely to reflect high quality than lower-level quality questions. Low quality questions were generally shorter with lengthier sentences than high quality questions.

Conclusions:

The findings of this research can guide health consumers in better formulating their health information questions and support online Q&A communities in managing health questions effectively.