Chair: TESSMER, MARTIN
DAI-A 60/05, p. 1442, Nov 1999

Mental models are a type of structural knowledge. Research has indicated that the use of mental models can enhance student learning, improve troubleshooting skills, enable the application of learned concepts and improve transfer of problem solving skills. However, there is little consensus in the instructional design research community concerning an operational definition or appropriate measures of the construct. There has not been a construct validity study with a sufficient sample population and diversity of learning outcomes to offer ample evidence concerning the nature of the construct. The purpose of this study is to test the hypothesis that mental models are a discrete learning outcome, having convergent validity with theoretically similar high-level outcomes and divergent validity from theoretically dissimilar outcomes. The study tested 127 subjects of aviation electronics training in a military school in the Southeast. Eight learning outcomes were measured, including two designed to assess structural knowledge. Structural equation modeling was used to construct and evaluate competing models in a multivariate environment. The resultant three-factor model was a good fit for the data set, with excellent marks on the absolute and incremental fit indices. The results of the study lend support to the hypothesis that mental models are a unique learning outcome. Because learning outcomes are essential to the instructional design process, the measurement tools reference list may prove to be useful resources. Also, the discussion concerning the testing of concepts, problem solving and knowledge structures have direct application for the design and development of instructional materials.