(PS8-54) App Based Scrambled Sentence Task as a Measure of Interpretation Bias in Anxiety and OCD

Interpretation biases, the tendency to interpret ambiguous stimuli either positively or negatively, may play a role in the etiology and maintenance of anxiety. Among the tasks used to assess interpretation bias, the scrambled sentence task (SST) has proved to have good convergent validity and reliability across disorders. However, many studies rely on cross-sectional designs or assess the SST at only pre and post-time points. In the current study we examined a newly developed phone based SST that may show promise as a fast and readily available measure of interpretation bias in OCD and anxiety. An app-based measure is advantageous as it allows the examination of interpretation bias in real world settings on a day-to-day basis. Participants completed an app-based SST over three separate days, with each session comprising nine trials. During each trial of the task, they were presented with six scrambled words and were asked to create a grammatically correct five-word sentence. The resulting sentence conveyed a positive, negative or neutral valence (e.g., “The door was unlocked”, for negative). We calculated the proportions of ‘Positive,’ ‘Negative,’ and ‘Neutral’ responses for each participant. Additionally, we created a general bias measure by assigning numeric values to each response: 1 for Negative, 0 for Positive, and 0.5 for Neutral. We then computed a mean score for each participant, where higher scores indicate a more negative bias, and lower scores indicate a more positive bias. We used the Generalized Anxiety Disorder-7 to assess anxiety, the Patient Health Questionnaire-8 to assess depression. To assess internal consistency of the SST, we correlated odd and even trials and then applied the Spearman-Brown prophecy formula. The SST showed excellent reliability (r = 0.83). Moreover, the measure of bias obtained high correlation with standardized measures of anxiety. By leveraging technology, we introduce an app-based approach to measuring interpretation biases in real-world settings. Preliminary results suggest that the app-based SST may serve as an effective tool for assessing interpretation biases in real-time and across various disorders.