10/06/2026
Muy interesante, food for thought para nuestros cursos para docentes en el marco de AI-UpskillEd 👉
When it comes to AI use in research, not all AI use is equal!
That is the main message I took from a recent paper by Zhu and Yang (2026) on generative AI use and graduate students’ research competence.
The authors separate AI use into three categories.
First, there is technical support AI use: data processing, coding, chart creation, formatting, and other practical tasks.
Then there is text development AI use: outlines, language polishing, paragraph logic, and writing structure.
Finally, there is transformation AI use: using AI to expand thinking, reflect on limitations, connect ideas across disciplines, and build multidimensional perspectives.
This distinction is important.
Using AI to format a table is not the same as using AI to question a research design. Using AI to polish a paragraph is not the same as using AI to reflect on the limits of a study.
But what caught my eye in this study is the inclusion of critical thinking.
The study found that critical thinking helps explain the relationship between AI use and research competence. In simple terms, AI seems more beneficial when students critically evaluate, revise, question, and integrate AI outputs instead of accepting them passively.
Critical thinking turns AI from a shortcut into a research support tool.
But the authors also add an important caution. The mediating effect of critical thinking is real, but modest.
Research competence is shaped by many factors: supervision, research training, motivation, institutional support, discipline, experience, and access to good mentoring.
AI and critical thinking matter, but they are not the whole story.
This is the kind of nuance we need in higher education.
Link in the first comment!
References
Zhu, H., & Yang, S. (2026). The impact of generative AI use on graduate students’ research competence: The mediating role of critical thinking and the moderating role of research self-efficacy. Behavioral Sciences, 16(2), 304.
