This article explores the precision and ease of different response formats for mental health assessments, comparing respondents’ preferences and clinicians’ beliefs. The findings highlight the potential of computational methods for free-text responses.