What happened
In a Google for Education account, Universiti Putra, Malaysia professor Nurfadhlina Mohd Sharef describes replacing traditional coding tests with verbal interviews in her data mining teaching. She reports that students use Gemini in Google Colab, then defend modelling choices, troubleshoot problems and identify potentially incorrect AI recommendations. She also reports using Gemini Canvas and Gemini Notebook to support evaluation at scale.
Why it matters
The assessment focus is students’ judgment over AI output, rather than code production alone. For educators and course designers, the practical distinction is between a working model and a student’s ability to explain why its choices are appropriate. The account offers an example of assessing that distinction through conversation. However, this vendor-published report does not provide measured learning outcomes or evidence that the approach transfers to other subjects or training settings.
What to do next
Use the account as an assessment-design example, not proof of effectiveness. Consider whether learners should explain which AI suggestions they accepted, which they questioned and how they justified final decisions. Before adapting verbal interviews, decide what evidence of judgment you need and how you would evaluate it consistently. The professor’s claim of evaluation at scale lacks operational detail here; the evidence also does not establish product availability or privacy arrangements for your institution.
About this briefing
Reviewed by TutorFlow Editorial. We link the primary source, preserve its publication date, and distinguish reported claims from TutorFlow analysis. Our commentary focuses on practical decisions for educators and training teams.
Discussion (0)
Open to everyone