Final Assessment

AI Tools for Content Creation — Final Assessment

Thirty questions across the orientation and ten modules — including scenario, prompt-critique, output-evaluation, assessment-design, visual/accessibility and responsible-use items. Every question explains why answers are correct or incorrect. Pass mark 80%. Attempt again after focused review if needed.

1. What is the course's core promise about using AI for content?

Create, evaluate, improve and responsibly deploy content, teacher in control.Let AI make teaching decisions.Generate content unchecked.Replace the teacher.

2. Educational content is best understood as what?

A designed learning experience serving an outcome.Whatever looks attractive.As much text as possible.Only images.

3. Why must a teacher review every AI output?

AI errs; the teacher is accountable for what students receive.AI is always correct.Review wastes time.Only for exams.

4. What must you never enter into a public AI tool?

Identifiable student data or photos.A lesson topic.A grade level.A subject name.

5. Which parts make up a high-quality educational prompt?

Role, task, audience, context, constraints, format, quality.Just the topic.A single word.The student's name.

6. What is the skilled response to a disappointing AI output?

Diagnose the gap and iterate.Accept whatever it made.Always start over.Abandon the task.

7. Why ask AI to critique its own output?

To surface gaps to check — not as proof of correctness.Because its self-review is final.To skip teacher review.It never helps.

8. A worksheet should be designed starting from what?

The learning outcome.A blank template.The number of questions.The clip-art.

9. Why verify an AI-generated answer key?

AI keys can be wrong; verify every answer.Keys are always right.Only if neat.Never needed.

10. What is the rule for slide text?

One idea per slide; detail in speaker notes.As much text as fits.Full paragraphs.No structure.

11. What is the risk of a vivid AI story for teaching?

Invented facts, bias, or entertaining without teaching.Stories are always safe.They can't mislead.They need no review.

12. What is the danger of a misleading analogy?

Where it breaks down it plants a misconception.Analogies never mislead.It saves time.It is always accurate.

13. When should an image be used in a resource?

When it genuinely supports understanding.On every slide always.Never.Only for decoration.

14. Why must AI-generated images be reviewed for text and labels?

They often garble text and mislabel parts.Their text is always perfect.Labels never matter.Only colour matters.

15. What must every meaningful educational image have?

A text alternative (alt text).A watermark.A caption in Latin.Nothing.

16. When is a concept map the right visual?

To show relationships among concepts.To show a step-by-step process.To decorate a page.Never useful.

17. Why validate an AI-generated concept map?

It can show plausible but incorrect links.Its links are always right.Maps can't be wrong.No review needed.

18. Why does AI often produce a quiz that doesn't test the outcome?

It defaults to recall unless the level is specified.It always aligns.Outcomes are irrelevant.Recall tests everything.

19. What makes an MCQ distractor diagnostic?

It reflects a real misconception.It is obviously silly.It is the longest.It repeats the answer.

20. What must you verify about an AI model answer?

That it is correct and complete.That it is long.Nothing.Only its format.

21. When is an analytic rubric the better choice?

When you want per-criterion feedback.When you want the fastest marking.Never.Only for exams.

22. What makes a rubric criterion observable?

It describes evidence you can see.It uses the word 'good'.It is short.It has no levels.

23. How does biased rubric language disadvantage students?

It rewards fluency and background over learning.It speeds marking.It has no effect.It only affects spelling.

24. What do the six VERIFY steps stand for?

Validate, Evaluate, Review, Inspect, Fit, Yield.Verify, Edit, Repeat, Ignore, Finish, Yes.View, Erase, Read, Import, File, Yield.No fixed meaning.

25. Which parts of AI content should you be most suspicious of?

Specific dates, statistics, quotes and citations.Only the title.Only the font.Nothing.

26. Why must meaning never rest on colour alone?

Some learners can't perceive colour.Colour is always clearest.It looks nicer.It never matters.

27. How does localisation differ from translation?

Localisation adapts examples, currency and context too.They are identical.Localisation only changes fonts.Translation adapts context.

28. What does an audit log provide?

Evidence a human reviewed and corrected the resource.A final grade only.The word count.Nothing useful.

29. What does the capstone content pack require of every included resource?

Fact-checked, reviewed, accessible, no student data.Only that it looks polished.Only that it is long.Nothing specific.

30. Across the whole course, who is accountable for AI-assisted content?

The teacher, always.The AI tool.The student.No one.