In one line: In an experiment the researcher changes one variable and measures its effect on another. No other method gives such strong evidence of cause.
Think about a cake that comes out flat.
Next time you bake it again and change only the oven temperature. Same recipe, same tin, same time.
This time it rises. Since the temperature was the single thing you changed, it is the only thing that can explain the difference.
The study on this page: A psychologist asks the same kind of question about behaviour: does having your phone on the desk lower your attention?
One group of students sits a 20-question attention test with their phone face down on the desk; another group has left the phone in another room.
Change one thing
the independent variable
- Set by the researcher
- Phone on the desk or in another room
Hold the rest steady
the controls
- Same test
- Same room and time limit
Measure the result
the dependent variable
- Score out of 20 on an attention test
The variable she sets herself is the independent variable (IV): phone on the desk, or phone in another room.
The variable she measures is the dependent variable (DV): each student's score on an attention test.
When the IV is the only thing that differs, a gap in the scores points to the IV as the cause. That link from cause to effect is causality.
Remember: One variable set. One variable measured. Everything else held steady.
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: Five parts turn a question into an experiment. Any experiment you read about can be taken apart into the same five.
Follow the phone study. Forty students sit a 20-question attention test. Some have their phone face down on the desk. The others have left it in another room.
The five parts
IV and DV, made measurable
Saying exactly how a variable is set or measured is called operationalising it.
Here: "phone nearby" becomes "own phone, silent, face down on the desk". "Attention" becomes "correct answers out of 20".
Hypothesis
A testable prediction about the effect of the IV on the DV.
Here: students whose phone is in another room will score higher than students whose phone is on the desk.
Experimental and control group
One group receives the treatment: the experimental group. The other does not: the control group, the baseline to compare against.
Here: phone on the desk is the experimental group. Phone in another room is the control group.
Random allocation
Each student is placed in a group by chance, for example by a coin toss.
Students who concentrate easily and students who do not are then mixed evenly across both groups.
Controls
An extraneous variable is anything else that might move the scores: the seat, the noise, the difficulty of the test.
Controls hold these steady. One test, one room, one set of instructions: a standardised procedure.
IV and DV · Prediction · Two groups · Chance · Controls
When a control fails: Imagine the phone-on-desk group sits by the windows and the other group sits by the wall.
Seating now differs along with the IV. It is a confounding variable: the view outside could explain the lower scores as easily as the phone.
Learn what examiners really want
See exactly what to write to score full marks. Our AI shows you model answers and the key phrases examiners look for.
Key idea: Before testing begins, the researcher decides who takes part in which version of the task. That decision is the experimental design.
The phone study has two versions of the task: phone on the desk and phone in another room. Each version is a condition.
The conditions can be shared out in three ways.
1. Everyone does both. Each student sits one test with the phone on the desk and another with it away.
2. One each. 20 students sit the test with the phone on the desk, 20 others with it away.
3. Pairs. Students are paired on last term's attention score, and one of each pair goes to each condition.
Repeated measures
- Each person takes part in every condition
- Strength: each person is their own comparison
- Strength: a smaller sample will do
- Limit: order effects; the aim is easier to guess
Independent measures
- Each person takes part in one condition only
- Also called independent samples
- Strength: no order effects; the aim stays hidden
- Limit: the groups may not be alike; more people are needed
Matched pairs
- Pairs are matched on a key variable, then split
- Strength: groups alike on what matters, no order effects
- Limit: slow to set up; no two people match exactly
Remember: Both = repeated measures. One each = independent measures. Pairs = matched pairs.
The first costs order effects. The second costs equal groups. The third costs time.
In repeated measures the danger is order effects.
Counterbalancing deals with it: half the students sit the test with the phone on the desk first, and half with it away first.
How to spot the design: Ask one question: did each student sit the test both with the phone on the desk and with it away?
Yes: it is repeated measures.
No: were students matched in pairs first? Matched: matched pairs. Not matched: independent measures.
Key idea: Sort any experiment with two checks. Who set the IV, and were people allocated by chance? Then was the setting controlled or real life?
The four types
True experiment (laboratory)
IV set by the researcher, groups decided by chance, in a controlled room.
Phone study: students sit the test in a quiet room set up for it (the laboratory). A coin toss decides who keeps their phone on the desk.
Real study: Loftus and Palmer (1974) showed students a film of a car crash. Each group was asked the same question with a different verb: how fast were the cars going when they "smashed", or "hit", each other? "Smashed" gave the highest speed estimates.
Good: only the phone differs between the groups. So a gap in scores can be put down to the phone: the study shows cause and effect.
Bad: an artificial setting, unlike real life.
Field experiment
IV set by the researcher, but in an everyday setting.
Phone study: in ordinary lessons, the teacher collects phones at the door on some days and not on others, and the end-of-lesson quiz scores are compared.
Real study: Levine and colleagues (2005) worked with Manchester United fans. As each fan walked between two university buildings, a runner fell in front of them and seemed hurt. Fans helped far more often when he wore a Manchester United shirt than a Liverpool or a plain one. The researchers changed one thing, the shirt, in a real place.
Good: students behave as they normally do.
Bad: less is controlled, and students have not agreed to take part.
Quasi-experiment
Groups made by something about the person.
The groups come from who the students are or what they already do, such as their age or a habit. That cannot be handed out, so chance cannot decide the groups.
Phone study: students who say they check their phone constantly are compared with students who rarely do.
Good: the only option when the IV cannot be handed out.
Bad: other differences stay uncontrolled. Students who rarely check their phone might also sleep more, so a better score may not be because of the phone.
Natural experiment
Groups made by something that happened to them.
An event in the real world changes things for people, without the researcher. It is a kind of quasi-experiment.
Not the same as quasi: quasi sorts people by what they are like; natural sorts them by what happened to them.
Phone study: the school bans phones from January for its own reasons. The researcher compares test results from before the ban with results after it.
Real study: Maguire and colleagues (2000) scanned the brains of London taxi drivers and of men who did not drive taxis. A part used for remembering places, the back of the hippocampus, was bigger in the drivers, and bigger the longer they had driven. Nobody made them taxi drivers: years of learning London's streets happened in real life.
Good: reaches questions nobody could set up, or ethically set up, on purpose.
Bad: the least control. Something else may have changed at the same time, such as a new timetable.
True · Field · Quasi · Natural
Natural experiments usually track change. In a longitudinal design the same students are measured before the ban and again afterwards. In a cross-sectional design different year groups are measured once, at the same time.
Can you say the phone caused it?: Only if chance decided the groups, like a coin toss. Then the groups start out alike, and the phone is the only difference.
Quasi and natural experiments: In a quasi or natural experiment, chance did not decide. Students who rarely check their phone might also sleep more or revise more.
So the study can show a difference, but it cannot rule out that something else about the groups produced it.
Never wonder what to study next
Get a personalized daily plan based on your exam date, progress, and weak areas. We'll tell you exactly what to review each day.
Control helps and hurts: Tight control is what lets an experiment name a cause. The same control can make the setting unlike real life.
It helps
- In the phone study, the same test, the same silent room and a coin toss mean only the phone differs.
- So the phone can be named as the cause.
It hurts
- A silent test room is not a normal lesson, where friends talk and phones buzz.
- So the result may not hold in a real classroom.
Strengths
- Shows cause: only the phone differed between the groups
- Room, test and timing held steady for all
- Standardised, so another school can replicate it
- Scores out of 20: objective and easy to compare
Limits
- Needs a coin toss: quasi and natural experiments lack it
- A silent test room is not a normal lesson
- Students may work out the aim and act to fit it
- Harm, such as real stress, cannot be given to people
Each point has a name.
Tight control gives high internal validity.
An artificial setting gives low ecological validity, and an artificial task gives low mundane realism. A silent test room with the phone face down is not a normal lesson.
Exam tip: Put the evidence from the study beside the judgement.
"Internal validity was high" is a label. "A coin toss set the groups and everyone sat the same test, so only the phone differed" is the point.
Your four class practicals
- Your class runs four small studies of its own, called class practicals: an experiment, an interview, an observation and a survey.
- Paper 2 asks about the ones your class ran.
- Keep a short note of each: aim, what you did, who took part, what you found, and the ethics.
How this is asked: Questions on research methods start from your own class practical. One of them asks you to design a study with a different method on the same topic, for 6 marks. Here that method is an experiment.
Six decisions, six short paragraphs
- The aim, and the type of experiment.
- The IV and the DV, both measurable.
- The design, the sample and how people are allocated.
- The procedure, step by step.
- The controls that keep it fair.
- The ethics: consent, the right to withdraw, a debrief.
The trap: an aim with no plan: "I would test whether phones affect attention" is an aim.
A design says how: who takes part, what each group does, and what number is written down at the end.
Your class practical was a questionnaire. Your class asked 25 students how many hours a day they spend on their phone and how easily they get distracted in lessons. Design an experiment to investigate the same topic you investigated in your class practical.
Model answer plan
See the answer broken into clear points, with guidance on how each point earns marks.