Effectiveness of problem-based learning to optimize student learning outcomes in regression analysis course

This study aims to determine the effectiveness of the implementation of ProblemBased Learning (PBL) model in optimizing student learning outcomes in the regression analysis course at Mathematics Department, Udayana University. The type of research used is action research, with a classroom action research design model. Sources of data in this study are students of the fourth semester who took a course of Regression Analysis. Data collection techniques are tests, interviews, and observations. Statistical analysis technique with analysis of variance is used to test student scores on the three cycles performed. The action procedure is designed in three cycles, with stages of planning, implementation, observation and reflection. The results showed that descriptively the average of student learning outcomes in cycle I = 54.15; Cycle II = 75.25; and cycles III = 90.15 which means an increase in students' understanding and reduced misconceptions that occur in using regression analysis. Further analysis using analysis of variance obtained Fcount= 34,76 with Pvalue = 0,000 indicates the null hypothesis is rejected, it's means that the three cycles have different effects on student learning outcomes, with a tendency to increase students' ability at every stage of the cycle.