The experimentation of the TTW and the NHT learning models on polyhedron viewed from the problem solving ability

Abstract

Abstract The study aims to obtain empirical data regarding the impact of models of learning models on mathematics outcomes when seen from the problem-solving ability. In this study, an experiment was conducted on the (Think talk write) TTW, (Numbered Heads Together) NHT and direct learning models. This study used a quasi-experimental study consisting of 2 experimental groups and 1 control group. In the experimental class, TTW and NHT were applied, while in the control class, direct learning was applied. The number of samples is 298 people, with details of 100 people from experimental class I, 99 people from experiment class II 99, and 99 people from the control class. The instruments in this study were tests and documentation and had met Validity and reliability. The research data were analyzed using two-way anava with different cells, and had met the prerequisite tests for normality and homogeneity. Research shows that (1) TTW are better than NHT and direct learning, and NHT are better than direct learning. (2) High student problem-solving ability provide better mathematics learning outcomes than medium or low student problem solving ability, moderate student problem solving ability are better than low student problem solving abilities, (3) there is an association between the learning model used and the student problem solving ability on student mathematics learning achievement.