White collar criminality: a prediction model
A prediction model for the purpose of maximally differentiating potential white collar offenders from non-offenders was developed and validated;The problems today in the selection of personnel for white collar positions are three-fold: the costs of white collar crime are on the increase, the applicant pool for white collar positions is on the decrease, and potential job incumbents often have little or no credit history or past work records from which assessments of risk are typically made;An essential personnel selection function, therefore, for the hiring of job applicants into sensitive white collar positions is the identification of individuals who may be prone to engage in financially irresponsible acts;Responses to five self-report instruments by 365 incarcerated white collar offenders and 344 white collar employees holding positions of authority addressed the relationships between three factors: behavioral tendencies of the individuals, their perceptions of personal and work-related situations, and behaviors in past situations;Forty-nine scales were reduced to 15 scales to form a discrimination function for purposes of classification. The function correctly classified 89.35% of the nonoffenders, and 90.41% of the offenders. Further analyses, based upon the results of the 15 scale discriminant analysis, identified a six factor discriminant model which correctly classified 87.96% of the non-offenders and 85.84% of the offenders. Cross-validation using a hold-out sample provided evidence for the stability of the weights in the above analyses as well as for an analysis using only male subjects. Base rate issues were addressed. For all of the models, the same two global constructs were identified under which were subsumed the 15 scales (or dimensions) and six scales (or dimensions) of the discriminant functions: extra-curricular activity and social conscientiousness.