Reevaluating the Psychometric Properties of Symptom Measures Within Longitudinal Contexts: The Yale-Brown Obsessive Compulsive Scale and Dimensional Obsessive-Compulsive Scale

As research and clinical settings increasingly emphasize questions of change, it is crucial that our mechanistic and outcome variables are established as reliable and valid measures of such change. However, there is often a mismatch between the purposes for which symptom measures were developed and validated versus their application. Traditional psychometric theory has focused largely on between-person change, whereas increasingly research and clinical questions concern within-person change. We examined the psychometric properties of two commonly used measures of obsessive-compulsive symptoms (Yale-Brown Obsessive Compulsive Scale, YBOCS; Dimensional Obsessive-Compulsive Scale, DOCS) within a longitudinal treatment context (N = 570). Regarding reliability, we applied traditional (i.e., internal consistency at each week) and novel methods that allow for examination of the reliability of both within- and between-person change (i.e., variance partitioning based on generalizability theory). We examined longitudinal concurrent validity by correlating per-person slopes of obsessive-compulsive and depression symptom measures obtained via mixed-effects models. Within-person change reliability was acceptable or good for the YBOCS and DOCS total scores (.77,.83), suggesting that these measures are capable of capturing meaningful changes that exist within persons over time, and between-person change reliability was excellent (.99-1.0). Per-person slopes analyses supported the longitudinal concurrent validity of both measures. Our data support the continued use of the YBOCS and DOCS as measures of obsessive-compulsive symptoms for the purpose of many longitudinal research questions. The current study provides a template for reestablishing the psychometric properties of other commonly used measures in the context of longitudinal investigations.

Automated summary

This study investigated the psychometric properties of two commonly used measures of obsessive-compulsive symptoms and found that they are reliable and valid in capturing within-person changes as well as between-person changes. The analysis of per-person slopes using mixed-effects models confirmed the longitudinal concurrent validity of these measures. This research supports the continued use of these measures for longitudinal research purposes and provides a template for evaluating the psychometric properties of other commonly used measures in longitudinal investigations.