The effect of drill question sequencing on learning and user satisfaction in computer-assisted instruction in molecular geometry
Abstract
In the literature on computer-assisted instruction (CAI), three patterns for the presentation of drill problems have been described as optimal: (1) a fixed series of random problems, (2) a sequence entirely user-selected, and (3) a sequence programmed for mastery learning. An experimental investigation of the effects of these three problem sequencing strategies on learning was performed using a randomized post-test-only design. Subjects were general chemistry students practicing labeling and pattern recognition tasks in a computer-based tutorial on molecular structure. The experimental treatment was programmed into the practice problems which followed a lesson on hybridization. Experiments were conducted in trailer sections of general chemistry courses at both Vincennes University, a comprehensive, resident junior college (n = 35), and Purdue University, a major midwestern land-grant university (n = 210). A 2 $\times$ 3 unbalanced ANOVA of mean scores on a ten-item post-test revealed a significant main effect for the school. No significant main effect for treatment group was observed, nor were interactions between school and treatment significant at the.05 level. Planned contrasts revealed inferior performance by the user-control group at the junior college. A similar set of contrasts failed to confirm comparable differences between groups at the land-grant university. The mastery group did not perform significantly better than the group with a fixed arrangement of problems at either site. This may be due to the use of insufficiently rigorous mastery criteria. When the three sequencing strategies were described to them, students from all three groups expressed a preference for the mastery-based sequence. Additional feedback from the students offered ideas for improving the program and for implementing CAI into the curriculum. Response to the program was uniformly positive. Most students preferred it to text-based homework, and wished for more CAI in their study of chemistry.
Degree
Ph.D.
Advisors
Bodner, Purdue University.
Subject Area
Science education|Curricula|Teaching|Educational software
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.