Although exposure and response prevention (ERP) is the most effective psychological treatment for OCD, only half of patients will reach symptom remission after completing it. Understanding the mechanism behind ERP is important to figure out how to improve this treatment. One theory of how ERP helps people with OCD is through expectancy violation, the process in which an individual’s prediction (or expectation) is disproven, allowing them to change their beliefs. Current research suggests that people with OCD may have difficulty learning and changing their beliefs — they may be unable to learn from expectancy violation during ERP, preventing them from changing their obsessive beliefs and ultimately preventing symptom improvement. There is a major gap in existing research, as no studies have tested factors that could impact whether people with OCD successfully learn from ERP.

This study will test two factors which may impact how well ERP works for people with OCD: pessimistic learning and cognitive immunization. Pessimistic learning is when a person changes beliefs more readily in response to negative, rather than positive, feedback or outcomes. Cognitive immunization is when one interprets inconsistent evidence in a way that fits their expectations without altering their beliefs. Participants with OCD will complete an online ERP treatment, and we will collect real-time data from them during exposure exercises to measure their expectancies and the outcomes of their exposures. This data will be used to see if those who are more prone to pessimistic learning and cognitive immunization will not benefit as much from ERP. Several technological advancements, including artificial-intelligence-based natural language processing (NLP) techniques, will be used to analyze data more robustly. Results from this study will ultimately guide the creation of stronger therapies for OCD, and signpost the path for future research with NLP to provide powerful, real-time insights to patients and clinicians during ERP.