A comparison of the effects of programming and software application on mathematical problem-solving in secondary schools
Robert L. Mayes, Ray Kurtz · 1989
The purpose of this study was to research the effects of the use of software tools and computer programming on mathematical problem solving in the secondary mathematics classroom. A secondary goal was to determine if explicit and substantial instruction in a problem solving heuristic and specific problem solving strategies would improve mathematical problem solving skills. Two distinct treatment groups were formed. (1) The Non-computer Problem Solving Group was provided explicit instruction in the use of a general heuristic and specific problem solving strategies. (2) The Computer Problem Solving Group received the same explicit instruction as the non-computer group, but in addition received teacher led guided discovery lessons involving the application of software as a tool to explore mathematical concepts. The second group contained an embedded programming group which participated in programming exercises involving the exploration of mathematical algorithms. This group was dropped from the study due to the lack of students qualified in BASIC programming. The groups were formed from seven intact Algebra II classes in three different schools in the same city. A nonequivalent control group design was employed with the statistical treatment being a multi-factor repeated measures analysis. An initial and final measure of problem solving skill were administered at the beginning and end of the fall semester. Mathematical problem solving skill was measured using a product coding sequence which accounted for the level of the heuristic which was achieved on each problem. School attended and student mathematics quality score were used as blocking variables in the multi-factor repeated measures analysis. Results of differences between groups tested at the.05 level were: (1) between-subjects the school attended was significant; (2) within-subjects a significant gain in problem solving ability across schools and groups was found; and (3) within-subjects a significant three-way interaction among treatment group, mathematics quality score and difference contrast from initial to final measure was found. The results of a Newman-Keuls multiple comparison supported the use of the non-computer treatment for low level mathematics students, the computer treatment for medium level mathematics students, and either treatment for high level mathematics students.