| Popis: |
The purpose of this research was to develop reliable tests of nonprocedural M1 tank driver skills that could serve as quantitative measures for tank-driver performance in simulators. Eleven driving tasks were derived from an Army Research Institute criticality survey. Analysis of these tasks resulted in decisions to test nine of the tasks, but only those aspects that related to the driver and were feasible for testing. The tests were tried out on 77 soldiers in two M1 OSUT classes. The data were used to assess scorer agreement and internal consistency, to estimate validity and utility based on reliability and variability and to direct revisions and recommendations for future testing. For each of the nine tests, the data indicated that driver performance could be measured reliably. Tests were designed so that useable quantitative data could be obtained. Although refinements and broader applications of the tests are required, the tests should serve the purpose for which they were designed. |
