The Test–Retest Reliability and Predictive Validity of a Battery of Newly Developed Occupational Performance Assessments

Abstract

The purpose of the author in this study was to investigate the test–retest reliability and predictive validity of new occupational performance assessments. The researcher used a test–retest design with mixed quantitative and qualitative methods. The test–retest reliability of the perceived adequacy of engagement in occupations of priority to participants as measured on the assessments was r (15) = .54, p < .05. Perceived test adequacy predicted 27% [B = .517, t = 2.18, R² = .27, F (1, 13) = 4.75, p = .048] and retest adequacy predicted 67% [B = .820, t = 5.16, R² = .67, F (1, 13) = 26.61, p = .000] of variability in the retest frequency of engagement in occupations seen as a priority by research participants. Test satisfaction scores predicted 50% of variability in the test frequency of engagement in occupations [B = .707, t = 3.61, R² = .50, F (1, 13) = 13.01, p = .003]. It was concluded that adequacy and satisfaction scores could be used by occupational therapists in planning therapeutic interventions to facilitate future performance of occupations seen as important by clients.

KEYWORDS:

• instrumentalism
• assessment
• occupational performance
• reliability

ACKNOWLEDGMENTS

The researcher would like to thank Ms. Shaunna Berg and Ms. Carrie Larson, graduate occupational therapy students at The University of South Dakota, who assisted with data gathering and analysis for this study. Their participation was part of the requirements for a research course. Their contribution to the study was very much appreciated.

Notes

Note. *Pearson r is significant at p < .05. Numbers in bold type are the test–retest reliability coefficients.

Note. Interpretation of Cronbach's alphas is based on George and Mallery's (2003) guidelines.

Note. R² = percentage of variability in the dependent variable that can be explained by the independent variable; SS = Sum of Squares; df = degrees of freedom; F = ratio of sum of squared deviations between observed and predicted values to the sum of squares for the regression model; Beta (B) = slope of the regression line.