Can students verify a source in a short teaching demo?
Laila has a fictional teaching demonstration with first-year students.
The question
The goal is for students to check a claim against a source. Laila plans 3 minutes of example, 5 of pair comparison and 4 for individual work and feedback.
- Model-generated summaryThe reported improvement persisted through a twelve-week follow-up.
- Original report excerptThe team collected follow-up observations for four weeks. The team collected no observations after that period.
“Teach students to identify what the report supports and point to the evidence.”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: What will students be able to do?
What would you do here? Open the move you’d make.
Ask students to identify a mismatch and mark the evidence
The strong move. You can check both understanding and use of the source.
Explain how different AI systems generate text
Tempting, but it misses something. That could provide background. It does not give students the source-checking practice required here.
Ask whether the summary sounds credible
Tempting, but it misses something. A plausible writing style cannot establish the reporting period.
See how a strong candidate says it
“Students will identify a mismatch between the generated claim and the report, then mark the sentence that tells us which claim is supported.”
Why it works: The objective names the decision and the evidence students must show.
Step 2: What will show that they learned it?
What would you do here? Open the move you’d make.
Collect the supported period and the cited line from each student
The strong move. That tests the observation and the reason for accepting it.
Ask each pair to read its improved summary aloud
Tempting, but it misses something. A good pair summary may hide which student found the source and how they checked it.
Have students ask a chatbot for a corrected paragraph
Tempting, but it misses something. The generated paragraph does not show that each student checked the source.
See how a strong candidate says it
“I would first demonstrate how to match the claim to the report. In the individual check, every student would state the supported period and mark the line that establishes it.”
Why it works: An individual answer gives evidence that a student can do both parts of the task.
Step 3: How would you teach without devices?
The room has no network access. The host asks candidates to run the teaching demonstration without student devices.
What would you do here? Open the move you’d make.
Keep the comparison and individual check on paper
The strong move. All students can practice the skill with the source in front of them.
Use the time to lecture on model errors instead
Tempting, but it misses something. That avoids the device issue by removing the activity that demonstrates learning.
Ask for a written explanation of AI ethics from memory
Tempting, but it misses something. That may test a useful concept. It does not test the evidence check in this demo.
See how a strong candidate says it
“I would use printed copies of the same claim and report. Students can compare them in pairs, then mark the period and its supporting line on paper.”
Why it works: The delivery changes while the objective and individual evidence stay the same.
Step 4: What would you reteach with two minutes?
Six of ten individual responses identify the four-week period. Four of ten cite the report text. The responses were tallied separately, and two minutes remain.
What would you do here? Open the move you’d make.
Demonstrate the evidence check and ask each student to try it again
The strong move. This uses the remaining time to test the specific skill that fewer students showed.
Say six students have mastered the whole objective
Tempting, but it misses something. Six chose the correct period. The data do not establish which students also cited the report.
Spend the rest of the demo explaining how to write the answer elegantly
Tempting, but it misses something. Better prose leaves the gap in evidence selection untested.
See how a strong candidate says it
“Six of ten students identified the four-week period. Four of ten marked a source line. I would point out the source sentence once, then ask each student to mark the supporting line in a short new example.”
Why it works: The two counts show that identifying the value and giving evidence are separate outcomes. A brief check targets the missing evidence.
The answer, in short
- Decision
- Keep a source-based exercise and check the cited evidence.
- Because
- A student can show the learning objective without a device.
- What would change it
- A new learning objective would require a new activity and a different check.
Common traps
- Check the learning evidence from each student.
- Keep separate counts separate unless you know who is in both.
Check your answer
- I named an observable objective.
- I defined what a response must show.
- I adapted the activity to the room.
- I used the responses to decide what to teach again.
Last reviewed Nov 11, 2026