Toward an Understanding of Recovery Resources in Diverse Populations: Factor Structure of the Assessment of Recovery Capital Instrument

Background and Purpose: Recovery capital is a theoretical construct encompassing the physical, social, human, and cultural resources that can support addiction recovery. The Assessment of Recovery Capital (ARC) is a 50-item dichotomous self-report developed and tested in the United Kingdom, using a primarily White, middle-class sample (Groshkova, Best, & White, 2012). Although the measure’s authors delineated 10 subscales (sobriety, psychological health, physical health, citizenship/community involvement, social support, meaningful activities, housing/safety, risk-taking, coping, and recovery experience), their analysis suggested a single factor. The objective of this study was to evaluate the ARC’s factor structure in a racially diverse, low-income US sample. As the theoretical literature suggests that recovery capital is not equally distributed across individuals or communities, we hypothesized that the factor structure of the ARC in a low-income sample would vary from the structure described in the original study.

Methods: Using a purposive sampling approach, low-income adults (under 200% of the federal poverty level) who self-identified as being in recovery from addiction to alcohol or drugs were recruited from community settings, such as soup kitchens and shelters, in Buffalo, NY. This ensured a sample of recovering individuals inclusive of those who did not access formal treatment. Participants completed an anonymous survey containing the ARC, demographics, and socioeconomic covariates and received a $5 gift card incentive. The sample (N=259) had a mean age of 46 years (SD=12.9) with an average monthly income of $440 (SD=$474). Almost a third of the sample were women (31%) and 61% identified as persons of color. Data were analyzed using MPlus to conduct a parallel analysis and exploratory factor analysis.

Results: Results from a parallel analysis and visual inspection of the scree plot suggested a five-factor solution. After deleting two items that had no variability, exploratory factor analysis, using a geomin and oblique rotation and robust weighted least squares estimator, also indicated five factors, explaining 29.4% of the variance. Model fit was adequate (X²=1023.89, p=.00; RMSEA=0.023; CFI=0.975; TFI=0.968; SRMR=0.084). Using a minimum cut point of 0.3, an additional item was excluded, retaining 47 items. Items were assigned to the factor with the highest rotated loading. Based on item assignment and the theoretical literature on recovery capital, we identified the five factors as sobriety, well-being, sense of purpose, security and belonging, and social support.

Conclusions and Implications. Using the totals from their 10 proposed subscales, the authors of the ARC suggested a single factor structure for recovery capital. However, when analyzed at the item level in our racially diverse, low-income sample, our findings indicate five factors, which overlapped only partially with the original subscales. This affirms our hypothesis and suggests that recovery capital may have differing dimensions in diverse populations. Future research should conduct a confirmatory factor analysis to test whether the data fit the five-factor measurement model yielded by our analysis. Developing accurate and sensitive measures of recovery capital is important to understanding the multiple dimensions that facilitate addiction recovery and ultimately addressing socioeconomic disparities in recovery outcomes.