Case Study 1 by Sandi MA

2025-10-02 07:21:52 UTC

Yes, but we have to consider or adapt the questions. We should make sure that the questions made suit with the need analysis.

2025-10-02 07:23:17 UTC

it is still reliable to some point but still needs human refinement before we actually use it. If we find that the questions are valid and useful we can use it immediately

2025-10-02 07:23:27 UTC

According to the presentation and the article you have explained to us, AI cannot be relied 100%, it is just a complement so my answer is 50/50 there are cases that we can rely on it for assessment purposes but still, English teachers need to put their expertise

2025-10-02 07:23:32 UTC

It can be but AI can't replace our judgment and validation.

2025-10-02 07:23:39 UTC

I don’t think ai generated question is valid and reliable for assessment purposes because it can produce bias and no human involvement in creating the assessment cannot be relied on

2025-10-02 07:23:41 UTC

In my opinion, AI on its own cannot guarantee the validity and reliability of an assessment, as it still heavily relies on teachers' judgment (learners' level, cultural context, and so on).

2025-10-02 07:24:17 UTC

AI-generated questions are not automatically valid and reliable, but they can become so depend on evidence collected

dhiyaathagz — 2025-10-02 07:24:55 UTC

I think AI can support item creation, but still it lacks context, curriculum awareness, and cultural sensitivity. Therefore, human judgment is still essential to review and test the questions.

2025-10-02 07:26:21 UTC

Yep they can quickly generate questions and even refine them, but that doesn’t automatically make them valid or reliable. True validity and reliability still require human expertise, careful alignment with objectives, and proper testing.

2025-10-02 07:26:25 UTC

I think the reliability of AI-generated questions depends on the prompt, and teacher review and validation are needed.

2025-10-02 07:27:37 UTC

Honestly, I'm not sure, but AI cannot replace us as a human, sometimes AI makes mistakes, so we still have to check and make sure all what we've asked AI to do.

2025-10-02 07:33:09 UTC

AI created instruments can save time and inspired ideas, but they are not automatically valid. They become valid only after human review, alignment with theory and ideally some testing.