An Attempt to Identify Indicators of Competence on Mechanical Maintenance Tasks.
Abstract
Army service schools are often faced with the problem of evaluating student performance on job tasks that are extremely time consuming or resource constrained. Solutions may lead to methods of testing in which students are tested on parts of tasks only. Methods of part-task testing are useful only if the sample of task elements comprising the test are predictive of whole-task proficiency. The work reported describes attempts to develop procedures that enable test developers to identify elements that predict task performance. One approach follows closely the task analysis review procedure used in skill qualification test development. Another requires subject matter experts to select task elements to test, given specific selection criteria. In a third approach, performance data from an Army training study were examined to determine empirically the most likely testing point(s) within each of the tasks tested. Results of the research indicate that the identification of task-element samples predictive of whole-task performance is a problem not easily solved. Subject matter experts did not agree well in their selection of most-predictive elements, and an empirical study of existing performance data indicated that most-predictive elements have little in common even among similar tasks. Until particular types of elements can be verified as predictive of performance for particular types of tasks, procedures for part-task test development cannot be set forth. Future approaches to identifying and classifying predictive task elements should be based on experimentally generated performance data supported by comprehensive diagnostic scoring. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1979
- Accession Number
- ADA095722
Entities
People
- Charlotte H. Campbell
- James H. Harris
- William C. Osborn
Organizations
- Human Resources Research Organization