Collaborative Study and Paired Test Taking in Collegiate Level Linear Programming Instruction

The purpose of this investigation was to examine the effects of collaborative learning strategies on formulating solutions to linear programming word problems that were designed to incorporate problem-solving skills. Forty-six students majoring in business at a small southwest college in Virginia participated in the study. After an instruction session, a study period, and a question and answer discussion, participants completed the test instrument based upon random assignment to three treatment groups. These included individual study with individual test taking (control), paired study with individual test taking, and paired study with paired test taking. All participants returned in 17 days to complete a posttest individually having received no further instruction in linear programming theory.

The following null hypothesis was examined: No differences in treatment means measuring problem-solving abilities would be found based on students' test and posttest scores using two treatment groups of collaborative study, with collaborative or individual test taking, and a control group of individual study with individual test taking. After satisfying the assumptions of no difference in ability in the treatment groups, establishing significant influence of ability on test score and posttest score variables, and establishing homogeneity of regression, an analysis of covariance (ANCOVA) was used to test the null hypothesis. The null hypothesis was rejected. Treatment had a significant effect on the variance for the test score variable, F = 3.92, p < .05, and for the posttest score variable, F = 4.44, p < .05. Newman-Keuls post hoc test showed significant differences in the adjusted means of the test score variable between the individual study with individual test taking group (72.22) and the paired study with paired test taking group (87.86). For the posttest score variable, the Newman-Keuls post hoc test revealed significant differences between the adjusted means of the individual study with individual test taking group (36.25) and the paired study with individual test taking group (59.20), and between the adjusted means of the individual study with individual test taking group (36.25) and the paired study with paired test taking group (55.77). Implications of findings and recommendations for further research were discussed.