| Abstract: | Training evaluation is one of the most under-researched and neglected areas of industrial/organizational psychology. This article discusses the need for, and feasibility of, utilizing extended-control-group pretest designs in evaluation research; and reports a field application of such a design (the Solomon 4-Group) in an organizational training context. The effectiveness of a basic electricity training program for telephone installer-repairmen was measured, and the potential contaminating effects of pretesting were monitored. The results indicated that the training was potentially effective, but that pretest contaminations were present. To identify the pretest effects it was necessary to go beyond the Solomon model and consider complex interactions involving numerical aptitude level, pretesting and training. Pretest exposure depressed the posttest performance of trained subjects of medium and low numerical aptitude, while facilitating that of medium level untrained subjects. Pretest exposure had no effect on subjects scoring high in numerical aptitude in either the trained or untrained condition. Possible explanations for the moderating role of numerical aptitude are offered, and implications of the findings are discussed relative to applied organizational evaluation efforts and future research. |
