Concurrent validation of a neurocognitive assessment protocol for clients with mental illness in job matching as shop sales in supported employment.

OBJECTIVES

Occupational therapists play a major role in the assessment and referral of clients with severe mental illness for supported employment. Nonetheless, there is scarce literature about the content and predictive validity of the process. In addition, the criteria of successful job matching have not been analysed and job supervisors have relied on experience rather than objective standards in recruitment. This study aimed to explore the profile of successful clients working in 'shop sales' in a supportive environment using a neurocognitive assessment protocol, and to validate the protocol against 'internal standards' of the job supervisors.

METHODS

This was a concurrent validation study of criterion-related scales for a single job type. The subjective ratings from the supervisors were concurrently validated against the results of neurocognitive assessment of intellectual function and work-related cognitive behaviour.

RESULTS

A regression model was established for clients who succeeded and failed in employment using supervisor's ratings and a cutoff value of 10.5 for the Performance Fitness Rating Scale (R(2) = 0.918, F[41] = 3.794, p = 0.003). Classification And Regression Tree was also plotted to identify the profile of cases, with an overall accuracy of 0.861 (relative error, 0.26).

CONCLUSION

Use of both inference statistics and data mining techniques enables the decision tree of neurocognitive assessments to be more readily applied by therapists in vocational rehabilitation, and thus directly improve the efficiency and efficacy of the process.