Phase 1: Prepare
Step 1 Warm-up recall
From M6.L3: what does an interruption cost besides its own time?
Private: nothing you type here is sent or stored.
Step 2 Learning goal
By the end of this lesson you can:
- ask five checking questions (Foundation);
- separate correlation from cause (Core);
- read a two-group comparison (Core).
Phase 2: Teach
Step 3 A familiar example
An advert says that students using a product score higher. Before you believe it, ask who was compared and who paid for the study.
Step 4 Clear explanation
Study advice comes from research, adverts and opinion. Five questions help you tell them apart.
| Question to ask of a study claim | Why it matters |
|---|---|
| Who was studied, and how many? | Small or unusual groups may not apply to you |
| Was there a comparison group? | Without one you cannot tell what caused the change |
| Were people assigned at random? | Otherwise groups may differ in other ways |
| What was measured, and when? | A quick test may not show lasting learning |
| Who benefits from the claim? | Sellers have a reason to overstate |
A common error is mistaking correlation for cause. In an invented survey, students who sleep 7 hours or more average 72 on a test, and those who sleep less average 64.
| Group (invented, 100 students each) | Average test score |
|---|---|
| Sleeps 7 hours or more | 72 |
| Sleeps under 7 hours | 64 |
| Difference | 8 points |
The difference is 72 − 64 = 8 points. That shows a link, not a cause. A hidden factor, such as a lighter workload or better health, could drive both sleep and scores.
An experiment that assigns people at random to sleep conditions can show cause; a survey cannot. Also look for size: 8 points on a 100-point test may or may not matter, and a small group can produce a large gap by chance. The review by Dunlosky and colleagues (2013) is a good example of weighing many studies together instead of trusting a single one. See also retrieval practice, one technique whose support comes from many experiments.
Step 5 Worked example (4 marks)
Group A averages 78 and group B averages 69. What is the difference, and can you say A’s method caused it?
- Difference: 78 − 69 = 9 points. (1 mark)
- Was assignment random? Not stated. (1 mark)
- Without random assignment, a hidden factor may explain it. (1 mark)
- Answer: 9 points; the cause is not shown. (1 mark)
Phase 3: Practise
Step 6 Learner practice (instant-marked)
1. Correlation means:
A link, not proof of cause.
2. 72 and 64 differ by:
72 − 64.
3. The best way to show cause is:
Random assignment balances other factors.
4. A hidden factor is one that:
It can create a link.
Phase 4: Close
Step 7 Recap
- Ask who, how many, compared with what, measured when and who benefits.
- A link is not proof of cause.
- Weigh many studies, not one.
Step 8 Exit ticket (ungraded)
A survey finds students with tutors score 6 points higher. Name one hidden factor.
Private: nothing you type here is sent or stored.
Step 9 Go further (Extension)
Take one study-tip claim from a website and answer the five questions.
Common mistakes
| Mistake | Why it fails |
|---|---|
| Treating a survey as proof of cause | Look for random assignment. |
| Trusting adverts | Ask who benefits. |
| Ignoring group size | Check how many. |
Limits. Even a good experiment applies only to people like those studied; check that you resemble them.