Factor validation and Rasch analysis of the individual recovery outcomes counter

Abstract

Objective: The individual recovery outcomes counter is a 12-item personal recovery self-assessment tool for adults with mental health problems. Although widely used across Scotland, limited research into its psychometric properties has been conducted. We tested its’ measurement properties to ascertain the suitability of the tool for continued use in its’ present form.

Materials and methods: Anonymised data from the assessments of 1743 adults using mental health services in Scotland were subject to tests based on principles of Rasch measurement theory, principal components analysis and confirmatory factor analysis.

Results: Rasch analysis revealed that the six-point response structure of the individual recovery outcomes counter (I.ROC) was problematic. Re-scoring on a four-point scale revealed well-ordered items that measure a single, recovery-related construct, and has acceptable fit statistics. Confirmatory factor analysis supported this. Scale items covered around 75% of the recovery continuum; those individuals least far along the continuum were least well addressed.

Conclusions: A modified tool worked well for many, but not all, service users. The study suggests specific developments are required if the I.ROC is to maximise its’ utility for service users and provide meaningful data for service providers.

Implications for Rehabilitation

• Agencies and services working with people with mental health problems aim to help them with their recovery.

• The individual recovery outcomes counter has been developed and is used widely in Scotland to help service users track their progress to recovery.

• Using a large sample of routinely collected data we have demonstrated that a number of modifications are needed if the tool is to adequately measure recovery.

• This will involve consideration of the scoring system, item content and inclusion, and theoretical basis of the tool.

Keywords:

• Mental health
• recovery
• factor analysis
• Rasch measurement theory
• validity

Acknowledgements

The authors would like to thank Nigel Henderson and Jane Cumming (Penumbra).

Disclosure statement

No potential conflict of interest was reported by the authors.