In one line: An experiment creates the difference it wants to study, so nothing else can explain the result.
Does a short nap after learning improve memory?
A researcher sets the independent variable (IV) herself: nap, or no nap. She then measures the dependent variable (DV): how many word pairs each person recalls.
They vary together
two conditions are compared
- Nap group against awake group
The cause comes first
the researcher sets the IV
- The nap happens before the test
Nothing else explains it
the controls
- Same list
- Same delay
- Groups formed by chance
A correlation shows only the first of these. Because the experimenter decides who naps, and when, the order of events is fixed and other explanations can be closed off one by one.
Remember: Set the IV. Measure the DV. Control the rest.
Free preview
This is the free notes preview
You're reading the free notes. Aimnova Pro unlocks the full study experience — and you can try it with your first topic free to keep:
- FlashcardsLock in vocabulary and key terms with spaced repetition.
- Practice questionsAnswer exam-style questions and get instant AI marking.
- Mock exams & past-paper vaultSit full mocks and see exactly how examiners award marks.
- Personalised study planA daily plan built around your exam date and weak areas.
Key idea: A causal claim is as strong as the alternatives that the design has removed.
What makes an experiment show cause
A control condition
One group naps and one stays awake.
Without the comparison, nobody knows what recall would have been anyway.
Random allocation
Chance, not choice, decides who naps.
If people chose, those who nap easily might differ in other ways, such as how rested they are.
Controlled variables
Anything else that could affect recall (an extraneous variable) is held constant: the list, the delay, the instructions.
One that differs between the groups is a confounding variable.
Control condition · Random allocation · Controlled variables
Internal validity
- Was the change in the DV caused by the IV alone?
- Raised by control and random allocation
- High in a laboratory experiment
External validity
- Does the finding hold for other people and settings?
- Raised by lifelike tasks and varied samples
- Often lowered by tight control
Exam tip: The two kinds of validity pull against each other.
A study can be praised for one and criticised for the other in the same paragraph.
Stop wasting time on topics you know
Our AI identifies your weak areas and focuses your study time where it matters. No more overstudying easy topics.
Key idea: Expectation is an extraneous variable, and it needs a control of its own.
A trial compares a new antidepressant with no treatment. The treated group improves more. Some of that improvement comes from taking part, being looked after, and expecting to get better. This is the placebo effect.
Three controls for expectations
A placebo
The control group takes an inactive pill that looks identical.
Any difference between the groups is then the effect of the drug beyond the effect of expectation.
Double-blind
Neither the patients nor the people rating their symptoms know who has the drug.
This removes the experimenter's expectations as well as the participants'.
Wait-listing
For a talking therapy no placebo exists. The control group waits and is treated afterwards.
It answers the ethical problem of refusing treatment, but the waiting group knows it is waiting, so expectation is not fully controlled.
Placebo · Double-blind · Wait-list
Watch out: A wait-list controls for time passing, not for expectation.
People told to wait may even feel worse, which makes the treatment look better than it is.
Key idea: Chance is always one of the alternative explanations. A statistical test says how seriously to take it.
The nap group recalls 21 pairs on average and the awake group 18. Two groups of people would differ a little even if napping did nothing.
A statistical test gives the probability of a difference this large arising by chance alone. By convention, if that probability is below 5 in 100 (written p < 0.05), the result is called statistically significant.
What significance tells you
- Chance is an unlikely explanation
- The result is worth taking seriously
- More likely with larger samples
What it does not tell you
- That the effect is large
- That the IV was the cause
- That the finding matters in real life
Remember: Significance rules out chance. It does not rule out a confounding variable.
A badly controlled study can give a significant result for the wrong reason.
See how examiners mark answers
Answer exam-style questions with model answers. Learn exactly what earns marks and what doesn't.
How this is asked: The 15-mark question on an unseen study offers causality as one of four concepts.
An experiment with a flaw in its controls is a likely study to be given.
The pattern
- State the IV, the DV and the causal conclusion.
- Test the design: comparison, allocation, controlled variables, blinding.
- For each gap, name the confounding variable it lets in.
- Apply a second concept and show how it bears on the causal claim.
- Conclude on internal validity, and say what would repair the design.
The trap: "significant, so proven": A significant difference shows that something other than chance is at work.
It does not show that the something is the IV.
Discuss the following study with reference to two or more of the following concepts: bias, causality, measurement and/or responsibility.
A researcher tested whether a short nap after learning improves memory. Forty university students learned a list of 30 word pairs at 1 p.m. They were then asked whether they would prefer to nap or to stay awake: 22 chose to nap for 45 minutes in a quiet room, and 18 stayed awake and watched a film. At 3 p.m. all were tested on the word pairs by the researcher, who knew what each student had done. The nap group recalled more pairs, and the difference was statistically significant. The researcher concluded that napping improves memory.
Model answer plan
See the answer broken into clear points, with guidance on how each point earns marks.