Abstract:
The quantitative component of this study examined the effect of computerassisted
instruction (CAI) on science problem-solving performance, as well as the
significance of logical reasoning ability to this relationship. I had the dual role of
researcher and teacher, as I conducted the study with 84 grade seven students to whom I
simultaneously taught science on a rotary-basis. A two-treatment research design using
this sample of convenience allowed for a comparison between the problem-solving
performance of a CAI treatment group (n = 46) versus a laboratory-based control group
(n = 38). Science problem-solving performance was measured by a pretest and posttest
that I developed for this study. The validity of these tests was addressed through critical
discussions with faculty members, colleagues, as well as through feedback gained in a
pilot study. High reliability was revealed between the pretest and the posttest; in this way,
students who tended to score high on the pretest also tended to score high on the posttest.
Interrater reliability was found to be high for 30 randomly-selected test responses which
were scored independently by two raters (i.e., myself and my faculty advisor). Results
indicated that the form of computer-assisted instruction (CAI) used in this study did not
significantly improve students' problem-solving performance. Logical reasoning ability
was measured by an abbreviated version of the Group Assessment of Lx)gical Thinking
(GALT). Logical reasoning ability was found to be correlated to problem-solving
performance in that, students with high logical reasoning ability tended to do better on
the problem-solving tests and vice versa. However, no significant difference was
observed in problem-solving improvement, in the laboratory-based instruction group
versus the CAI group, for students varying in level of logical reasoning ability.Insignificant trends were noted in results obtained from students of high logical reasoning
ability, but require further study. It was acknowledged that conclusions drawn from the
quantitative component of this study were limited, as further modifications of the tests
were recommended, as well as the use of a larger sample size.
The purpose of the qualitative component of the study was to provide a detailed
description ofmy thesis research process as a Brock University Master of Education
student. My research journal notes served as the data base for open coding analysis. This
analysis revealed six main themes which best described my research experience: research
interests, practical considerations, research design, research analysis, development of the
problem-solving tests, and scoring scheme development. These important areas ofmy
thesis research experience were recounted in the form of a personal narrative. It was
noted that the research process was a form of problem solving in itself, as I made use of
several problem-solving strategies to achieve desired thesis outcomes.
