This study critically examines test and non-test evaluation methods in learning Islamic Religious Education (PAI) at the high school level. This research uses the literature review method with primary data in the form of 30 national and international scientific articles published between 2013-2023 from Google Scholar, ScienceDirect, DOAJ, and Moraref databases, as well as policy documents such as Merdeka Curriculum and KMA No. 183/2019. Data were collected through systematic selection using inclusion-exclusion criteria based on relevance to the topic of PAI evaluation in Indonesia. Data analysis was conducted using thematic synthesis technique to identify patterns, challenges, and recommendations in the implementation of PAI evaluation. The results showed that PAI evaluation is still dominated by conventional test methods based on memorization, while non-test methods such as observation, portfolio, and self-assessment that are more effective in assessing noble character and psychomotor skills have not been optimally implemented due to limited technical guidance and teacher training. In fact, evaluation in line with the principle of learner-centered learning demands instruments that are able to measure deep understanding, critical thinking skills, and character strengthening. However, only 30% of PAI teachers understand the preparation of Higher Order Thinking Skills (HOTS) questions, and 65% have difficulty in designing valid instruments for affective aspects. Other challenges identified include unpreparedness of school infrastructure and teachers' administrative burden. Therefore, this study recommends a hybrid evaluation model that combines HOTS-based tests with non-test methods, teacher training in instrument design, and collaboration with religious institutions to strengthen contextual assessment and support more meaningful learning for students.
Musanna et al. (2025) studied this question.