| Abstract: | Coding competencies from behavioral-event interviews according to the principles of a new approach to assessment (McClelland, 1973) produces assessments that are reliable and validly associated with success as an executive. These assessments are not influenced by length of protocol or by performance in the preceding year. Bias is not a problem if both the interviewer and the coder are blind to executive success. In contrast to a traditional psychometric approach based on regression analysis, an algorithm based on competency scores predicted managerial success and improved performance across a number of samples. This algorithm identified potential outstanding performers as individuals whose scores reached designated tipping points within clusters of substitutable competencies. Experts' judgments of competencies needed or shown by executives in various positions agreed only moderately with competencies shown to be important by the data from behavioral-event interviews. |
