Lane, William2021-08-162021-08-162021-05https://hdl.handle.net/11299/223084University of Minnesota M.S. thesis. May 2021. Major: Computer Science. Advisor: Lana Yarosh. 1 computer file (PDF); ix, 48 pages.Mental Health Technologies offer great promise in disseminating mental health content and training, to people with limited access to clinical settings. For example, ”cognitive reappraisal” is an emotion regulation technique where someone changes their perspective on a negative thought to change its emotional impact. However; cognitive reappraisal is a skill that takes practice to learn and must be applied outside of clinical settings. In this work we present Flip*Doubt a web application that employs crowd-workers to help users reformulating their negative thoughts. Flip*Doubt was deployed in two, one month field deployments. From these deployments we contribute a qualitative understanding of how a crowd-powered cognitive reappraisal system might be used in the wild as well as two codebooks capturing important context in both negative thoughts and their reformulations. Additionally, we contribute a quantitative analysis using machine learning techniques to explore the predictability of a quality measure in cognitive reappraisal data. The results introduce contextual features of effective cognitive reformulations that may be explored in further research as well as guidelines for incorporating crowd- powered cognitive reappraisal features in mental health technologies. In the discussion section we cover some challenges in predicting reformulation quality from a cognitive reappraisal dataset and offer strategies for future investigation.enThesis or Dissertation