Conclusion
From the results of this study, it can be concluded that most of the students had good metacognitive skills from the average score for 1-19 questions. However, negative statements consist of easy to give up behavior in solving problem. The improvement in their metacognitive skills was not statistically significant. This showed that metacognitive skills cannot be improved significantly with one treatment, but it does not mean there was no improvement at all. Using questionnaire will generate inaccurate data, because students