Abstract
The present study aimed to develop and validate a personality-based model of workplace deviance or counterproductive work behavior (CWB). The model consisted of salient dispositional factors considered to be predictive of CWB, either in a unidimensional, or multidimensional capacity, or both. Literature on concepts such as counterproductive work behavior, integrity testing, dark personality, the dark triad of personality and attribution styles informed the selection of traits for inclusion in the model. The shortcomings of selected personality based models of CWB were considered and a dual process account of CWB was postulated. In addition to considering normal personality and its relationship to CWB, this study also argued for the consideration of dark personality constructs as predictors of CWB. In particular, the dark triad of personality was suggested and it was argued that contemporary conceptualisations of dark personality constructs (i.e., psychopathy) enable their measurement in ‘normal’ populations. To examine the predicted relationships between the operationalised personality constructs of the model and the counterproductive criterion variables, the psychometric properties for the developed measures were examined in low stakes (n = 952) and high stakes (n = 401) conditions separately, and in combination. The combined sample included 1353 working adults ranging between 18 and 78 years of age (mean = 35.6, SD = 11.9). There were 619 (45.8%) men. The ethnic distribution included 534 (39.5%) Black/African, 381 (28.2%) White, 138 (10.2%) Coloured, 113 (8.4%) Indian, 11(0.8%) Asian and 176 (12%) unspecified. Construct validity results using confirmatory factor analysis and Rasch analysis found satisfactory fit for all items and scales in all three conditions. These results allowed for the examination of concurrent predictive validity. Bivariate correlation and linear regression analysis supported the concurrent validity for 11 of the 12 scales for the prediction of interpersonal and organisational CWB. In addition, quantile regression analysis provided a much more nuanced understanding of the relationships between the predictor and predicted variables in the study. Lastly, good evidence for convergent validity was found with the Hare Self-Report Psychopathy Scale. The results of this study supported the use of the personality variables of the model for the prediction of deviant behaviour in organisational settings.
D.Phil.