This study investigated the manager's use of information in performance appraisal in making decisions about employees. The main research question addressed by the study was whether or not managers weighted the information in the performance appraisal differently when faced with different types of decisions. Associated research questions were: (1) the identification of the personal variables (experience, seniority, education, etc.) with which this differential weighting was associated, if any; and (2) whether or not an analytical tool could be devised for: (a) identifying and clarifying the actual policies for ranking, judging, and making decisions about personnel; (b) developing or expanding a manager's awareness of their decision making processes; and (c) facilitating training of managers in making decisions using the multiattribute information of the performance appraisal. / The sample subjects were 80 managers of both sexes from a large state university. Their organizational levels ranged from first line supervisors to vice presidents, deans and directors of institutes; their educational levels ranged from high school completion to Ph.D.s; and their backgrounds included education, behavioral sciences, engineering, business and humanities. All of them had experience in evaluating personnel. / The model employed was linear multiple regression. The experimental design was fractional factorial (after Addleman). Managers were provided with two sets of sixteen performance evaluations each representing respectively sixteen candidates for a promotion, and sixteen candidates for a merit money allocation. The managers were asked to rank the candidates according to their preference for the above purposes. Each performance evaluation had five factors: initiative, dependability, job knowledge, quality of work, and quantity of work. Each factor had three levels: satisfactory, above satisfactory and outstanding. Both series of sixteen candidates included exactly the same cells, differing only in the names of the "candidates." / The results of the study indicated that the managers of this sample weighted quality and quantity of work more heavily when ranking for merit money distribution, and weighted job knowledge more heavily when ranking for promotion purposes. The differential weighting of the factors of dependability and initiative was not statistically significant ((alpha) = 0.05), though the lack of significant results for initiative may have been due to insufficient test power. / The experiment failed to show a significant association between preference for weighting differentially and the personal variables considered: sex, educational background, educational level, experience in management, and experience in evaluation. This analysis used oneway ANOVA, with t contrasts between means of weight coefficients (betas). The results were sparse and did not present an interpretable pattern, probably due to insufficient cell sizes (five to seven subjects per cell in some cases). Nevertheless, some results indicated that for both promotion and merit money, males emphasize more quantity of work than females (p = .008) and the reverse was true for quality of work, where females emphasized more quality than males for both promotion and for merit money distribution purposes (p = .05). Another unexpected result was the low weight attributed to quantity of work for both decisions ((mu)(,promotion) = -.022 and (mu)(,merit money) = .095) which was interpreted as due to the minimum level of "satisfactory" used for each of the factors in the performance evaluation. / Recommendations focused on the importance of having a better model of judgment when using performance evaluation information for decisions, on the necessity of having stronger construct validity on the side of the criterion variables of the performance evaluation, and, finally, on the convenience of having separate performance evaluations for different purposes. / Source: Dissertation Abstracts International, Volume: 41-10, Section: A, page: 4480. / Thesis (D.B.A.)--The Florida State University, 1980.
| Identifer | oai:union.ndltd.org:fsu.edu/oai:fsu.digital.flvc.org:fsu_74323 |
| Contributors | TELIAS, MOISES GUILLERMO., Florida State University |
| Source Sets | Florida State University |
| Detected Language | English |
| Type | Text |
| Format | 201 p. |
| Rights | On campus use only. |
| Relation | Dissertation Abstracts International |
Page generated in 0.0025 seconds