Abstract

Indonesia has implemented student learning methods; it can affect student learning outcomes. This study aims to determine the effectiveness of project-based learning models and think pair share to improve the learning outcome of English learning at SMP IT AL-Madani Semarang. This study using Quasi-Experiment design, pre-test and post-test design. Sampling technique used was using cluster random sampling. The research data was analyzed by using independent sample t-test at α = 0.05. The results showed that: (1) learning models using project-based learning on average 81.77 (2) learning models by using think pair share with an average of 74.56. (3) There are differences in cognitive learning outcomes between the project-based learning group and think pair share with a significance value of 0.002 < 0.05. Meanwhile, the gain test concluded that there are differences between project-based learning and think pair share models in the moderate category. The conclusions showed that project-based learning model is more effective than think pair share model in improving students' cognitive learning outcomes. Meanwhile, think pair share model can be used as an alternative on English learning in the subject of “we love what we doâ€. The benefits of this study is as a source of information in answering problems that occur in the learning process, particularly to improve the learning outcomes and are useful in the world of education, especially in the field of curriculum development science.