Scaffolding in a computer-based constructivist environment for teaching statistics to college learners
Abstract
To prevent the user from over-relying on technology support, the researcher proposed a 3-D contingent scaffolding model to systematically vary the technology support in response to the learner's performance in a learning task consisting of a sequence of sub-tasks. The research reports on the implementation of this model in a computer-based instructional program on a topic in statistics and a study of its effectiveness with college students. In this study, the 3-D contingent scaffolding model was implemented in a computer-based instructional program, "Hypothesis Testing--the Z-Test" in order to establish baseline data for integrated-media-based instruction or a hypermedia-based learning environment. This scaffolded instruction was evaluated in terms of knowledge maintenance and knowledge transfer by comparing it to full-support instruction and least-support instruction. A secondary interest was to determine whether the predictive utility of static (what the learner already knew) and dynamic (the learner's ability to learn measures of learners' ability varied in different instructional conditions. The dynamic measure was indicated by the percentage of successful trials in the learning process. The static measures involved seven baseline variables: courses taken, self-rating of skills, and self-rating of preference in mathematics/statistics and pretest. The findings showed evidence that the scaffolded computer-based instruction promoted knowledge maintenance and improved independent knowledge application. The scaffolded instruction promoted learning consistently across individuals. However, further investigation and evidence is needed to confirm its superiority on knowledge transfer. The results also revealed that the dynamic measure of the learner's ability was a better predictor of the learning outcome for subjects using this scaffolded instruction than the static measures. In media-assisted instruction, the 3-D contingent scaffolding model provided a systematic way to link the concept of scaffolding to the integrated media design features using both support building and fading techniques. In statistical education, this model was implemented successfully on the topic of hypothesis testing. The results of this study also showed that this topic could be taught to college students with low mathematical background and statistics beginners.
Degree
Ph.D.
Advisors
Lehman, Purdue University.
Subject Area
Curricula|Teaching|Educational software|Statistics|Higher education
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.