Using a Latin Square Design to Determine the Most Effective Mathematics Teaching Method

Charles K. Assuah

This study determined the most effective teaching method among four (4) teaching methods (i.e., direct instruction, inquiry-based, cooperative, and guided discovery). The study adopted a 4 x 4 Latin-Square design, using four (4) classes, four (4) teachers and four (4) teaching methods to record students’ scores. Participants consisted of forty (40) public high school students and four (4) teachers in a school district in the western region. The results indicated that main effect for teacher was significant, F (3, 15) = 37.50, p < 0.05,  = 0.95, main effect for class was significant, F (3, 15) =755.83, p < 0.05,  = 0.99, main effect for teaching method was significant, F (3, 15) = 37.50, p < 0.05,  = 0.98.  (i.e., there were significant differences in students’ scores at all levels with respect to teacher, class, and teaching method). Tukey’s HSD post hoc multiple comparison tests indicated that students’ scores for direct instruction was significantly lower than their scores for cooperative, t (6) = -9.50, p < 0.05,C.I = [-11.95, -7.05], students’ scores for direct instruction were significantly lower than their scores for guided-discovery, t (6 ) = -15.50, p <  0.05, C. I = [-17.95, -13.05], students’ scores for inquiry-based instruction were significantly lower than their scores for cooperative, t (6) =- 9.50, p < 0.05, C. I = [-11.95, -7.05], students’ scores for inquiry-based instruction were significantly lower than their scores for guided-discovery, t (6) = -15.50, p < 0.05, C.I = [-17.95, -13.05]. Nonetheless, students’ scores for direct instruction were not significantly different from their scores for inquiry-based instruction, t (6) = .00, p > 0.05, and students’ scores for inquiry-based instruction were not significantly different from their scores for guided-discovery, t (6) = -15.50, p > 0.05.This study has demonstrated that a Latin-Square design has a greater ability and robustness in detecting treatment differences as compared to a one-way analysis of variance. The study finally concluded that mathematics teachers should be encouraged to use more guided-discovery and cooperative learning methods, in their instructions.

https://doi.org/10.31871/IJNTR.5.7.14