REDEFINING PROJECT BASED LEARNING IN ENGLISH CLASS

Abstract

This study used the experimental method with quasi experimental design, specifically, non-equivalent control group design. There were two groups, namely, experimental group in which it was given treatment by using PBL and control group in which it was given no treatment. For collecting the data, both groups were given speaking test and self-confidence questionnaire. For speaking, the data were collected using speaking test which was evaluated by two raters using analytic rubric consisting of five elements, namely accent, grammar, vocabulary, fluency and comprehension. Meanwhile, students’ self-confidence was evaluated using a ready-made-self-confidence questionnaire which measured (1) language use anxiety, (2) causal attribution, (3) perceived L2 competence, and (4) self-efficacy. Both data were analyzed by using t-test to see the difference in means (1) between pre- and post-test of students’ speaking achievements and of their self-confidence, and (2) of post-test of speaking achievements and of their self-confidence between the first and the second group. Regression analysis was also given to see the contribution of PBL to the variables in questions. The results showed that (1) there was a significant difference both in speaking achievement (t=8.828, p