<![CDATA[Mathematical learning models with direct learning used so far can be one of the causes of poor mathematics learning outcomes. One way to overcome mathematics's poor learning outcomes is by applying learning models that involve more student activities. Several learning models involve more student activities so that learning takes place more meaningfully for students, cooperative learning. This study aims to determine which is more effective between the Questions student’s Have (QSH) type of cooperative learning model and Numbered Heads Together (NHT) cooperative learning model on students' mathematics learning outcomes. This study's population were eighth-grade students consisting of 7 classes, samples taken by purposive sampling to class, obtained by class VIII-G as experimental class 1 with QSH learning, and class VIII-F as experimental class II with NHT learning. The type used in this study is experimental research with a quantitative research approach. The research method is a post-only control design. The instruments used are post-test questions in the form of multiple-choice tests. Instrument testing uses the Validity Test, different power, reliability test, and level of difficulty. Data analysis techniques consisted of the initial ability normality test, initial ability homogeneity test, and comparing the average scores of students' mathematics learning outcomes. The study results obtained the average test scores of class G students' mathematics learning outcomes as experimental class I with QSH learning at 70.61. The average test scores of class F mathematics learning outcomes as experimental class II with NHT learning amounted to 59.70. There are differences in students' mathematics learning outcomes using the QSH type cooperative learning model with the NHT type cooperative learning model. The difference in the average value of the test results of the experimental class I with QSH learning and experimental class II with NHT learning was 10.91 with the average value of students' experimental learning outcomes of the experimental class I with QSH learning was higher than the experimental class II with NHT learning. This means that the Questions Students type cooperative learning model has more effective than the Numbered Heads Together type learning model of students' mathematics learning outcomes.]]>
