Mathematical problem-solving ability is an essential competency that students at all educational levels must master through structured thinking based on Polya stages. The Guided Discovery Learning (GDL) model is considered an innovative alternative capable of developing this ability through active and constructive guided discovery processes. This study aims to systematically examine the implementation of GDL and its relationship with students' mathematical problem-solving ability across various educational levels based on verified articles from 2017–2026. The novelty of this study lies in its cross-level synthesis that explicitly identifies research gaps, namely the absence of specific experimental empirical evidence for junior high school students and the unsynthesized simultaneous interaction of multiple moderating variables. The method used is Systematic Literature Review (SLR) following Zawacki-Richter et al. (2020) procedures, sourced from Google Scholar, SINTA, and Scopus, resulting in 19 included articles. The results show that GDL consistently improves students' mathematical problem-solving ability at junior high and senior high school levels based on Polya indicators, with key moderating factors including instructional tool quality, digital media integration, and students' affective aspects. The conclusion shows that GDL is a valid, effective model recommended for broad implementation in mathematics learning, particularly at the junior high school level which requires further experimental investigation.
Copyrights © 2026