Journal article
Does the Assessment of Recovery Capital scale reflect a single or multiple domains?
Substance abuse and rehabilitation, Vol.8, pp.39-43
01/01/2017
DOI: 10.2147/SAR.S138148
PMID: 28790877
Abstract
Objective: The goal of this study was to determine whether the 50-item Assessment of Recovery Capital scale represents a single general measure or whether multiple domains might be psychometrically useful for research or clinical applications.
Methods: Data are from a cross-sectional de-identified existing program evaluation information data set with 1,138 clients entering substance use disorder treatment. Principal components and iterated factor analysis were used on the domain scores. Multiple group factor analysis provided a quasi-confirmatory factor analysis.
Results: The solution accounted for 75.24% of the total variance, suggesting that 10 factors provide a reasonably good fit. However, Tucker's congruence coefficients between the factor structure and defining weights (0.41-0.52) suggested a poor fit to the hypothesized 10-domain structure. Principal components of the 10-domain scores yielded one factor whose eigenvalue was greater than one (5.93), accounting for 75.8% of the common variance. A few domains had perceptible but small unique variance components suggesting that a few of the domains may warrant enrichment.
Conclusion: Our findings suggest that there is one general factor, with a caveat. Using the 10 measures inflates the chance for Type I errors. Using one general measure avoids this issue, is simple to interpret, and could reduce the number of items. However, those seeking to maximally predict later recovery success may need to use the full instrument and all 10 domains.
Details
- Title: Subtitle
- Does the Assessment of Recovery Capital scale reflect a single or multiple domains?
- Creators
- Stephan Arndt - Carver Bible CollegeEthan Sahker - University of IowaSuzy Hedden - University of Iowa
- Resource Type
- Journal article
- Publication Details
- Substance abuse and rehabilitation, Vol.8, pp.39-43
- DOI
- 10.2147/SAR.S138148
- PMID
- 28790877
- NLM abbreviation
- Subst Abuse Rehabil
- ISSN
- 1179-8467
- eISSN
- 1179-8467
- Publisher
- Dove Medical Press Ltd
- Number of pages
- 5
- Grant note
- Prairie Ridge Integrated Behavioral Healthcare, Mason City, IA, USA
- Language
- English
- Date published
- 01/01/2017
- Academic Unit
- Psychiatry; Biostatistics; Nursing; Injury Prevention Research Center
- Record Identifier
- 9985134048502771
Metrics
1 Record Views