(PS10-B26) Evaluating the Quality of Cognitive Restructuring in Participant Responses to an Online Behavioral Intervention

         Cognitive restructuring (CR) is a common component of self-guided behavioral intervention technologies (BITs) for mental health (Thew et al., 2022). In BITs that include CR, participants are prompted to reframe their own negative thoughts, sometimes without any additional support from trained mental health professionals or coaches (e.g., Wasil et al., 2021). While BITs that include CR have been effective in treating various mental health concerns, little work has been done on investigating the quality of the CR that users engage in. As such, it is difficult to determine if BITs that include CR are teaching participants to use the skill. This study aims to fill in this gap by examining the quality of participants’ responses to a CR exercise in a trial of an online BIT.

         Participants were taken from a randomized control trial of the “Overcoming Thoughts” platform, a novel platform for treating depression and anxiety symptoms developed in collaboration with Mental Health America and hosted on their Screening-to-Supports website. On the platform, participants were prompted to enter a thought “they were struggling with right now” and allowed to pursue a cognitive or behavioral intervention path. Participants who selected the cognitive path were guided through a CR activity, where they were asked to rework the initial thought by reframing it into a “healthier” version. Throughout the trial, 67 participants used the cognitive path and contributed 356 responses to the restructuring prompt. Of these participants, 53.7% identified as female, 43.3% identified as male, and 3.0% identified as non-binary/other. For race, 65.7% identified as White, 11.9% identified as Asian, 10.4% identified as more than one race, 9.0% identified as Black or African American, and 3.0% identified as “Unknown/Not Reported”. Hispanic/Latino participants were 22.4% of the sample. Participants’ average age was 34 years (M= 34.5, SD=12.8).

To evaluate the quality of participants’ CR, the initial and restructured thoughts will be qualitatively coded. Initial thoughts will be coded by type (e.g., interpersonal, work-related, etc.). Restructured thoughts will be coded by quality and the use of common CR questions (i.e., alternatives, evidence, implications). Research assistants will be trained on the codebook and all responses will be double coded to allow assessment of interrater reliability.  Regular checks of interrater reliability and recalibration will occur during the coding to ensure consistency across raters. After coding, analyses will be conducted to determine the average ratings on all core features of CR across the sample. Additional analyses will also be conducted to determine if the average ratings on all core features of CR are moderated by the type of initial thought entered.  

To our knowledge, this is the first project to evaluate the quality of CR conducted on digital platforms. Given the increased interest in BIT as treatment options, it is important to understand the extent to which users can learn and apply CR without the presence of a trained individual to guide them. This study can also inform the use of CR more generally by identifying common types of thoughts people struggle with and how challenging it is for people to apply CR to these different types of thoughts.