Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Language, Technology, and Social Media International Journal of Educational Practice and Policy
Saripudin
Universitas Pendidikan Indonesia
Humanities Education Languange, Linguistic, Communication & Media Social Sciences

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Automated Assessment Systems in Education: A Multi-Paradigmatic Analysis of Technological Capabilities, Pedagogical Implications, and Ethical Challenges Muh Yamin; Mumu Komaro; Saripudin
International Journal of Educational Practice and Policy Vol. 4 No. 2 (2026): June-July 2026
Publisher : PT. Global Research Collaboration

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.66314/ijepp.v4i2.546

Abstract

Artificial intelligence based automated assessment systems have developed rapidly over the past decade; however, studies that integrate technological, pedagogical, socio-cultural, and ethical dimensions simultaneously remain limited. This study presents a multi-paradigmatic analysis of 24 articles published in Scopus Q1 and Q2 indexed journals between 2020 and 2025. Three research questions are posed: the types and capacities of emerging automated assessment systems, their pedagogical and ethical implications, and the evaluative framework required. The analysis identifies six categories of automated assessment systems, with a dominant shift toward large-scale language models in recent studies. The findings indicate that technical superiority does not necessarily guarantee fairness or pedagogical validity. Three fundamental ethical issues are consistently identified: linguistic discrimination, lack of system explainability, and the indispensable need for human oversight. In response, this study introduces the TAPE-H Framework (Technology, Assessment Theory, Pedagogy, Ethics, Human Oversight) as an integrative evaluative model that assesses automated assessment systems holistically, moving beyond accuracy based metrics alone.
AI Ethics as Epistemological Governance: A Systematic Literature Review on Knowledge and Authority in the Age of Generative AI Cahyo Prianto; Mumu Komaro; Saripudin
Language, Technology, and Social Media Vol. 4 No. 2 (2026): June 2026 | Language, Technology, and Social Media
Publisher : WISE Pendidikan Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70211/ltsm.3026-7196.403

Abstract

This Systematic Literature Review (SLR) explores research trends on the ethics of generative AI use and how ethical issues are discussed within it, mapping geographical distribution and examining AI epistemic governance. This systematic literature review employs the PRISMA method, with a literature search conducted through the Scopus database, filtered based on keywords related to generative AI Ethics in Quartiles Q1 and Q2 and limited to the period 2020–2026, resulting in 39 articles for further analysis in this SLR. The research trend has continuously increased from year to year, and 2025 became the year with the highest number of studies addressing the ethics of generative AI use. This indicates a strengthening academic attention to ethical and epistemic issues in AI. The literature is dominated by themes of ethical concerns in the use of generative AI, such as bias, data privacy, transparency, accountability, misinformation, academic integrity, and cognitive dependence on generative AI. This study also finds that generative AI is most frequently positioned as a knowledge generator, while the combination of training data bias and cultural bias constitutes the most dominant epistemic issue. In the dimension of epistemic dependency, human dependence on AI is the most frequently discussed theme. This demonstrates growing concerns regarding the weakening of human autonomy, control, and cognitive capacity. From the perspective of authoritative actors, the scientific community occupies the strongest position, while multi-stakeholder governance emerges as the most widely supported governance model. These findings affirm that AI governance is understood as a complex issue that cannot be resolved by a single actor, but rather requires collaboration.
Co-Authors Cahyo Prianto Muh Yamin Mumu Komaro