Measuring Community Integration: Development and Psychometrics of the Community Connections and Engagement Scale

The objective of this community-based participatory research project was to develop a clinically useful, psychometrically-sound scale to measure community integration for adults with severe mental illness. Two researchers and an administrator of a behavioral health agency (BHA) recruited a group of providers, half with lived-experience of severe mental illness. Through a series of five focus groups, provider participants guided identification of four major domains of community integration and the development of 95 scale items; items and domains were reviewed by three external researchers with subject matter expertise. Initial pilot: BHA providers administered the scale to clients (n = 51) with 19 completing it twice to investigate internal consistency, test-retest reliability, and restricted variance and collinearity of items. Further piloting: providers at two BHAs administered the scale to clients (n = 178) to conduct exploratory factor analysis and analyze internal consistency. After initial pilot, 50 items remained post item reduction for restricted variance and collinearity, with Cronbach's alpha of .95 and test-retest reliability of .90. After a larger pilot, a four-factor solution emerged, aligning conceptually with the four domains as anticipated; 33 items loaded (factor loadings ≥ .4), with RMSEA of .069 and overall Cronbach's alpha of .89 (subdomains ranging .78-.86). The scale has good preliminary psychometric properties and appears to be feasible for use in BHAs for the purposes of research and evaluation, with clinical utility for assessment and treatment planning.