Shifting Episodic Prediction With Online Cognitive Bias Modification: A Randomized Controlled Trial

Abstract

Negative future thinking pervades emotional disorders. This hybrid efficacy-effectiveness trial tested a four-session, scalable online cognitive bias modification program for training more positive episodic prediction. 958 adults (73.3% female, 86.5% White, 83.4% from United States) were randomized to positive conditions with ambiguous future scenarios that ended positively, 50/50 conditions that ended positively or negatively, or a control condition with neutral scenarios. As hypothesized (preregistration: https://osf.io/jrst6), positive training participants improved in negative and positive expectancy bias, self-efficacy, and optimism more than control participants, ds and 97.5% CIs = -0.57 [-0.87, -0.27], 0.79 [0.42, 1.15], 0.28 [0.02, 0.53], 0.28 [0.04, 0.51], and, for expectancy bias, more than 50/50 participants, with gains maintained at 1-month follow-up. Unexpectedly, participants across conditions improved comparably in anxiety and depression symptoms and growth mindset. Targeting a transdiagnostic process with a scalable program may improve bias and outlook; however, further validation of outcome measures is required.

Behavioral Engagement and Psychosocial Outcomes in Web-Based Interpretation Bias Training for Anxiety

“…One approach to studying the relationship between engagement and outcomes in DMHIs has been to analyze associations between individual engagement metrics (or composites of metrics) and outcomes. [25][26][27][28][29][30] Another approach has been to group users by multiple engagement metrics (e.g., using data-driven cluster analysis) and analyze whether these groups show differential symptom reduction; this approach is promising because it can reveal potentially more complex engagement patterns or profiles in the data, which is especially useful in studying real-world engagement in implementation settings. [16,17,19,[30][31][32]…”

“…The present study analyzed whether distinct engagement subgroups, extracted in an exploratory cluster analysis from markers of task completion rate and time spent on training and assessment tasks, are associated with differential improvement in anxiety and interpretation bias in a trial of web-based CBM-I administered to anxious adults on our team's public research website MindTrails. We hypothesized that the group(s) with metrics suggesting higher engagement would have significantly better outcomes; in this case, after cluster analysis identified two groups differing on time spent across tasks, we expected the group that spent more time doing the program would improve more (based on the purported positive relation between extent of use and outcomes [23] and on the positive relation in a prior MindTrails study between mean time spent per CBM-I scenario and anxiety improvement) [28]. However, we also recognized plausible alternatives (e.g., taking longer can indicate distraction) [24] and ran post hoc comparisons of the groups on various baseline variables (e.g., demographics) to more fully characterize them.…”

“…Finally, we used those values to compute the mean time spent on an individual scenario across sessions for a given participant. [28]…”

“…Bias Some DMHI studies have found that groups characterized by higher engagement improve significantly more than other groups, [17] and some studies of individual metrics of time spent have found that greater time spent is significantly related to better outcomes. [26][27][28]32] We expected the group that spent more time (to the extent that this reflects higher engagement) to improve significantly more than the group that spent less time. However, prior findings on the associations between engagement groups (or individual time spent metrics) and outcomes are mixed, and the comparable improvements we found between the two groups in anxiety and positive interpretation bias in the present study are consistent with the findings of other studies in which all engagement groups improved during the intervention, with no group improving significantly more than the others.…”

Shifting Episodic Prediction With Online Cognitive Bias Modification: A Randomized Controlled Trial

“…

Study data, analysis code, and materials are available at https://osf.io/jp5ws (Eberle, Boukhechba, Sun, et al, 2021). Following the preregistration (https://osf.io/jrst6), the present study analyzed data collected from May 3, 2017, through September 9, 2018, for participants who enrolled on or before March 27, 2018.

…”