Learning Assessment as a Strategic Tool for Improving Educational Quality
DOI:
https://doi.org/10.54801/k22g7k67Keywords:
learning evaluation, education quality improvement, CIPP model, HOTSAbstract
This article critically examines learning evaluation as a strategic instrument for enhancing the quality of education in the context of 21st-century challenges. Drawing on philosophical foundations (ontological, epistemological, and axiological perspectives), the study argues that evaluation is no longer merely a measurement tool but a dynamic mechanism for systemic improvement, curriculum development, and pedagogical transformation. Employing a Systematic Literature Review (SLR) based on the PRISMA protocol, this research synthesizes 68 core sources from reputable international and national journals (2015–2026), classical texts, and policy documents. The discussion is structured around four main themes: (1) ontological and epistemological concepts of learning evaluation; (2) contemporary typologies and models, including CIPP, formative-summative, and Higher Order Thinking Skills (HOTS)-based evaluation; (3) pedagogical implications for teaching quality, feedback mechanisms, and curriculum refinement; and (4) modern-era challenges such as algorithmic bias, digital divide, and accountability pressures, along with strategic reconstruction solutions. Findings affirm that a hybrid, reflective, and technology-integrated evaluation system can serve as a powerful lever for inclusive, equitable, and sustainable educational transformation. The article concludes with practical recommendations for policymakers and practitioners, particularly in implementing the Merdeka Curriculum in Indonesia, to position evaluation as the core driver of national education quality improvement
Downloads
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Abdul Malik Safrudin

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Data Lisence Terms
Choose how a default copyright date is selected or an article. This default can be overridden on a case-by-case basis. If you publish as you go, don't use the issue's publication date.


