This study was conducted to find out whether there is a significant different on students’ writing performance of recount text between students who are grouped homogeneously and students who are grouped heterogeneously. The design of the study was a quasi-experimental research method with a quantitative approach. The population of this study will be the 10th grader at an SMA N in Semarang. The total
of the population is 431 students from the 12 classes. The writer used a purposive sampling by the English teacher recommendation and two classes were taken as sample, they were X-E.10 that was grouped homogeneously and X-E.12 that was grouped heterogeneously. The instrument of the study was a writing test. The writer conducted pre-test to both class before treatment and post-test after the treatment.
The result of pre-test and post-test were analyzed by using SPSS. The mean of the homogeneous groups pre-test score was 21,5 and the heterogeneous groups pre-test score was 22. The mean of post-test score of homogeneous groups was 24 and the
heterogeneous pre-test score was 24,5. The t-test result shows that there was a significant difference. The results show that the Sig. (2-tailed) is 0.024 < 0.05. it means that the Ho is rejected and the Ha is accepted. It can be concluded that there is a significant different on students’ writing performance of recount text between students who are grouped homogeneously and students who are grouped
heterogeneously.
Key Words : Homogeneous Grouping, Heterogeneous Grouping,
Collaborative learning, and Recount Text.