AI evidence extraction

Oportunidades e desafios da utilização da inteligência artificial generativa no ensino universitário

A. L. Lopes, ChatGPT · Repositório do ISCTE-IUL · 2025

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

8/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

Methodology & findings

Study design

Reflexive analysis and conceptual discussion between the author and ChatGPT exploring pedagogical potential and challenges of generative AI in university education.

Primary method

None - this is a qualitative reflective analysis without statistical testing.

Main result

The study found that generative AI tools can support teachers in curriculum design, preparation of teaching materials, formative assessment and pedagogical monitoring, as well as provide personalized support to students through virtual tutors and immediate feedback. However, the authors note that "a IA generativa, quando utilizada de forma informada e orientada, pode constituir uma ferramenta pedagógica inovadora, potenciando a aprendizagem no ensino superior ao mesmo tempo que exige novas abordagens pedagógicas e institucionais" (generative AI, when used in an informed and guided manner, can constitute an innovative pedagogical tool, enhancing learning in higher education while requiring new pedagogical and institutional approaches).

Reports effect sizes.

Research paradigm

Interpretivist/Critical

Author conclusions

The authors conclude that "a IA generativa, quando utilizada de forma informada e orientada, pode constituir uma ferramenta pedagógica inovadora, potenciando a aprendizagem no ensino superior ao mesmo tempo que exige novas abordagens pedagógicas e institucionais" (generative AI, when used in an informed and guided manner, can be an innovative pedagogical tool enhancing higher education learning while requiring new pedagogical and institutional approaches). They also recommend institutional strategies including teacher and student training in generative AI literacy, ethical incorporation in teaching plans, and adaptation of assessment practices.

Risk of bias

No empirical data collection or validation; Reflective analysis without experimental controls; Potential confirmation bias through collaborative reflection with ChatGPT; No comparative analysis with control conditions; Context-specific (single institution) without broader generalizability claims tested

Limitations

The paper acknowledges risks of "alucinações" e imprecisões factuais produzidas por modelos generativos (hallucinations and factual inaccuracies produced by generative models), as well as "a possibilidade de dependência excessiva reduzindo o pensamento crítico dos/as estudantes" (the possibility of excessive dependency reducing students' critical thinking), and emerging ethical questions regarding academic integrity, plagiarism prevention, data privacy, and access equity.

Open questions raised

The chapter identifies the need for institutional strategies including teacher and student training in generative AI literacy, ethical incorporation into teaching plans, and adaptation of evaluation practices.
The paper identifies the need for critical integration of generative AI technologies in higher education through new pedagogical and institutional approaches, teacher and student training programs in generative AI literacy, and development of ethical frameworks for responsible use.
The paper identifies the need for institutional strategies including: teacher and student training in generative AI literacy, ethical incorporation into curriculum plans, and adaptation of assessment practices. Future research directions implicit in the work include studying the alignment between responsible AI use and active, collaborative, reflective learning pedagogies.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 73%

Explore related topics

AI Manuscript Writing Academic Writing Conceptual