| Abstract: | This study was motivated by the need to design and develop computer-based learning materials based on cognitive theories of learning, and to develop methods of data collection and evaluation for the purpose of testing the effectiveness of instructional design components used in a computer-based problem solving environment. The objective of this study was to discover relationships that exist among knowledge structures and students' requests for visual and textual information and time spent attempting to solve a problem in a computer-based learning environment. The method used was to: (1) design and develop a computer-based problem solving environment for eighth grade earth science students, (2) apply instructional treatment in a classroom setting, and (3) evaluate the knowledge structures and students' interactions with the problem solving environment. Simple correlations were obtained from among all the variables and a principal component analysis was conducted on the pre-test subtests and interaction variables. The highest correlations among all possible combinations were among pre-test application, analysis and synthesis; and between requests for textual information and requests for visual information. Requests for visual information was negatively correlated with pre-test application, analysis and synthesis. The principal component analysis yielded two cannon factors related to procedural knowledge and need for information. These two factors accounted for eighty-five (85) percent of the variation in the data. An analysis of time spent in the pre-test and post-test indicated that those variables may have been an intervening factor in the correlational analysis. An ordered regression analysis of the five levels of learning measured in the pre- and post-test was conducted. Pre-test subtest scores, number of requests for textual and visual information, and time spent in the three major components of the system were used as predictor variables. Post-test subtest score was the dependent variable. Different combinations of independent variables contributed to prediction of post-test scores at the various levels. Recommendations included changes and additions in the instructional design process and in the research design. Problem solution should be included on the computer rather than on paper so that efficacy of the problem solution and immediate feedback could occur. It was also recommended that a qualitative measure of motivation and/or attitude and a test measuring use of imagery be added to the analysis of knowledge structure variables. Students who aren't familiar with the learning format should be allowed to ask procedural questions. The continuation of research with a computer-based problem solving environment has merit and should continue to be explored on other populations in different instructional settings. |
|---|