摘要：This experimental study aims to compare problem-solving abilities between students that are taught with guided discovery learning models and conventional learning models. This study involved all eighth graders as the research population. Furthermore, the sampling technique used is the simple random sampling technique in two classes in Junior High school SMPN 1 Kota Gorontalo by employing Superitem essay test as a data collection tool. This descriptive analysis is described in the frequency distribution and histogram tables, whereas inferential analysis was performed by using ANCOVA test. The results showed that guided discovery learning model is preferable to compare to the conventional learning model in improving mathematics problem-solving ability.
