Abstract
Personality-based integrity tests are used in selection procedures to reduce the chance of hiring employees who are likely to engage in counterproductive work behaviour. The present study reports the internal psychometric properties of a new personality-based measure developed for this purpose. Data collected from 1353 working adults were used to investigate internal consistency reliability, and to examine construct validity with confirmatory factor analysis and Rasch analysis. Results showed that the reliability estimates for all the scales of the assessment were satisfactory. For the confirmatory factor analysis, inspection of the incremental (CFI and TLI) and absolute (RMSEA) goodness-of fit values found strong support for the construct validity of all the scales. Infit statistics from Rasch analysis provided further support for construct validity, with items from all the scales fitting the Rasch model. Combined the confirmatory and Rasch analysis demonstrated that unidimensional, coherent and meaningful latent constructs are being measured on the WRISc. Overall, results found excellent support for the internal psychometric properties of the instrument in a culturally diverse context.