摘要：This study aims to determine whether the cooperative learning model type Numbered Heads Together (NHT) is more effective than the conventional learning model towards the mathematical problem-solving ability of VII grade of Salatiga State Junior High School 5 students the academic year 2019/ 2020.This research was Quasi-Experimental type with Non-Equivalent Control Group Design.The instrument of this study was a problem-solving ability test,which consisted of pre-test and post-test.Post-test results obtained an average of 73.24 in the experimental class and 58.62 in the control class.In other words,the average experimental class was higher than the control class.This result was confirmed by the Mann-Whitney U test that obtained a sig (2-tailed) value of 0.00 < 0.05,so H0 was rejected,while H1 was accepted;it meant that there was a difference in the average mathematical problem-solving ability of students between the experimental and the control class.Additionally,the average N-Gain of the experimental class was 0.54,and the control class was 0.29.The average increase in the experimental class was higher than the control class,and there were differences.As the Mann-Whitney U N-Gain test confirmed that obtained sig (2-tailed) value of 0.00 <0.05 so H0 was rejected,and H1 was accepted.Thus there was a difference in the average increase of students' mathematical problem-solving ability between experimental and control classes.
关键词：Effectiveness of Cooperative Learning Models;Numbered Heads Together;Mathematical Problem-solving Ability
