VALIDITY OF AN EMPLOYMENT INTERVIEW FOR T HE PREDICTION OF TRAINING SUCCESS

A predictive validity study of a multimodal interview (in comparison w ith a cognitive abilities test) for the selection of apprentices in a financial organization is reported. The interview was developed to pre dict non-cognitive aspects of performance (''social competence''). Gra des at the end of the apprenticeship training can be predicted better by means of the cognitive abilities test than by interview (r = 0.55 v s. r = 0.21, all validity coefficients are corrected for attenuation). However, potential for doing qualified customer consulting can be bet ter predicted by the interview than by the cognitive abilities test (r = 0.27 vs. r = 0.12; z = 2.23; p < .05). In addition, the interview t ends to better predict potential for doing qualified customer consulti ng than to predict potential for doing qualified staff work (r = 0.27 vs. r = 0.18; z = 1.40; p < .10), whereas the reverse is true for the cognitive abilities test (r = 0.12 vs. r = 0.27; z = 2.08; p < .05). P rediction of training grades by means of the cognitive abilities test cannot be improved by the inclusion of the interview. However, predict ion of potential for qualified customer consulting cannot be improved by inclusion of the cognitive abilities test - beyond the interview.