Avalilação PREreview Estruturada de Artificial Intelligence in Higher Education: A Systematic Review of Contributions to SDG 4 (Quality Education) and SDG 10 (Reduced Inequality)

Does the introduction explain the objective of the research presented in the preprint?

Yes

The introduction section of the preprint explicitly states the research's objective. It first identifies a gap in the existing literature, noting that few studies have systematically reviewed how AI interventions in higher education contribute to sustainability goals.

Are the methods well-suited for this research?

Highly appropriate

The methods used in this study are highly appropriate for answering the research questions. The authors employ a Systematic Literature Review (SLR), which is the gold standard for synthesizing existing evidence on a specific topic. The rigor of their approach is demonstrated through several key practices: Adherence to a Framework: The study is explicitly guided by the PRISMA 2020 framework, a widely accepted best-practice guideline for systematic reviews. A PRISMA flow diagram (Figure 1) is included to transparently illustrate the study selection process. Comprehensive Search: A detailed and transparent search strategy was used across five relevant academic databases. Bias Reduction: To ensure consistency and reduce bias, two reviewers independently screened all 75 full-text articles. High Interrater Reliability: The authors calculated and reported a Cohen's Kappa of 0.82, indicating "almost perfect" agreement between the two reviewers, which is a strong indicator of methodological reliability. Systematic Data Handling: The study used a structured template for data extraction and a systematic, multi-reviewer process for thematic data synthesis.

Are the conclusions supported by the data?

Somewhat supported

The conclusions are consistently thorough and provide a realistic interpretation of the data without overreaching or drawing conclusions not reflected in the results. The conclusions presented in the preprint are highly supported by the findings from the 51 synthesized studies. The authors provide a balanced summary that accurately reflects the key results of their systematic review. Despite the synthesis, I believe that the conclusion requires a significant justification in terms of international standards, such as UNESCO recommendations, and close reflection from statistical data.

Are the data presentations, including visualizations, well-suited to represent the data?

Somewhat appropriate and clear

Strengths: PRISMA Flow Diagram (Figure 1): This is the only visualization in the document, and it's well-executed. It clearly shows the study selection process with appropriate numerical detail at each stage (1673 initial records → 1360 screened → 75 full-text reviewed → 51 included). This is the standard and expected visualization for systematic reviews. Structured Tables: Table 1 presents the search string clearly with logical organization by topic categories (AI terms, SDG terms, education level terms). Clear Textual Organization: The findings are organized thematically with bullet points and subsections that make the data patterns accessible. Limitations Quantitative Data Underutilized: The review mentions specific numbers (51 studies, various coding agreements, etc.) that could be effectively visualized to show patterns, but relies primarily on narrative description. Thematic Synthesis Presentation: While the three research questions structure the findings well, visual aids (like concept maps or frameworks showing relationships between themes) would enhance comprehension of the complex interrelationships.

How clearly do the authors discuss, explain, and interpret their findings and potential next steps for the research?

Somewhat clearly

Strengths in Discussion and Interpretation: Strong Connection to Framework: The authors effectively link findings back to sustainability education theory, interpreting how AI supports transformative learning, systems thinking, and equity. Organized Structure: The discussion systematically addresses each research question, making it easy to follow their interpretation of findings. Nuanced Treatment of Contradictions: They acknowledge competing evidence (e.g., AI both enhancing AND potentially restricting critical thinking), showing depth rather than oversimplifying. Context and Literature Integration: Findings are interpreted within broader research trends (e.g., comparing their results to Barrera Castro et al., 2024; Gu, 2024), demonstrating scholarly engagement. Weaknesses: Limited Depth on Mechanisms: While they note what AI does (improves outcomes, enhances accessibility), they provide less insight into how or why these effects occur. The discussion often restates findings rather than deeply interrogating them. Underdeveloped Practical Implications: The "Practical Implications" section lists recommendations (invest in infrastructure, promote AI literacy) but lacks concrete guidance on implementation strategies or prioritization. Next Steps Clarity: The "Recommendations for Future Research" section identifies gaps (need for longitudinal studies, mixed-methods, direct sustainability outcomes) but doesn't propose specific research designs or priorities. It reads more as a list than a strategic research agenda. Missing Critical Analysis: Given the emphasis on equity (SDG 10), there's surprisingly limited discussion of power dynamics, who benefits from current AI implementations, or how structural inequalities might be perpetuated despite good intentions. Overall: The authors provide clear, well-organized interpretation of their synthesis, but the discussion could be more insightful. They explain findings competently and identify logical next steps, but lack the depth and specificity that would elevate this to "very clearly." The interpretation is sound but somewhat surface-level.

Is the preprint likely to advance academic knowledge?

Highly likely

Significant Contributions: First Systematic Synthesis: This is explicitly the first systematic review examining AI in higher education specifically through the lens of SDG 4 and SDG 10, filling a genuine gap in the literature. Methodological Rigor: The PRISMA-guided approach with clear inclusion/exclusion criteria, interrater reliability (κ = 0.82), and transparent screening process provides a replicable model for future work. Practical Infrastructure Solutions: The identification of specific partnership models (cloud grants, AI-as-a-service, public-private funding) that reduce entry barriers offers actionable knowledge for resource-constrained institutions. Balanced Perspective: The review acknowledges contradictory evidence on critical thinking and doesn't oversell AI's benefits, which adds credibility and nuance to the field.

Would it benefit from language editing?

No

The preprint demonstrates strong academic writing with only minor language issues that don't impact comprehension.

Would you recommend this preprint to others?

Yes, it’s of high quality

Why Recommend the preprint: Fills a Real Gap: This is the first systematic review specifically examining AI in higher education through SDG 4 and SDG 10 lenses—a timely and needed contribution. Methodologically Sound: The PRISMA approach, clear screening process, and acceptable interrater reliability (κ = 0.82) provide a trustworthy foundation. Comprehensive Scope: Synthesizing 51 empirical studies across multiple databases offers valuable breadth for researchers entering this field. Practical Value: The infrastructure partnership examples and thematic organization make this useful for practitioners and policymakers, not just academics.

Is it ready for attention from an editor, publisher or broader audience?

Yes, after minor changes

The review currently describes what's happening but doesn't engage with the authoritative frameworks (UNESCO's 194-member-state commitments) that guide global education policy. These changes would transform it from an academic literature map into a policy-relevant resource that education ministries and institutional leaders would actively use. Key Additions Required: 1. UNESCO Policy Framework Integration Connect findings to UNESCO Recommendation on Ethics of AI (2021) Reference Beijing Consensus on AI and Education (2019) Link to Education 2030 Framework (Incheon Declaration) Ground recommendations in established international commitments 2. Deepen SDG 10 (Equity) Analysis Address digital divides, marginalized populations, and algorithmic bias Examine who benefits vs. who is excluded from AI implementations Apply UNESCO's equity principles to interpret findings Move beyond surface-level accessibility discussions 3. Add Policy-Oriented Sections New subsection: "Policy Context: AI and Global Education Agendas" Restructured: "Policy and Practice Implications" organized around UNESCO pillars New subsection: "Policy Recommendations" for governments, institutions, and international bodies

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.