In one line: Change is a difference between measurements, and measurements differ for many reasons besides real change.
A clinic offers twelve sessions of therapy for anxiety. At the end, patients' scores on an anxiety questionnaire are lower. Has the therapy changed them?
Perhaps. But scores also vary from day to day, people answer differently once they know a questionnaire, and most anxious people feel somewhat better after a few months whatever is done. A claim of change has to separate the signal from this noise.
Remember: A difference is what you measure. A change is what you have to argue for.
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: Every design for measuring change makes the same move: the same variable is measured at more than one time.
Designs that capture change
Before and after
One measurement before an event and one after it.
The clinic gives the anxiety questionnaire at session 1 and at session 12.
Repeated measures
The same participants provide every measurement, so each is compared with themselves.
Stable differences between people are removed from the comparison.
Longitudinal
The same people are measured repeatedly over months or years.
The clinic tests its patients again after six months and after two years.
Before and after · Repeated measures · Longitudinal
Longitudinal
- The same people over time
- Shows change within a person
- Slow and costly; people drop out
Cross-sectional
- Different groups at one time
- Shows a difference between groups
- Quick; the groups may differ in other ways
Watch out: People who drop out are rarely a random selection.
If the patients who felt no better stop coming, those who remain make the therapy look better than it is.
Practice with exam-style questions
Answer exam-style questions and get AI feedback that shows you exactly what examiners want to see in a full-marks response.
Key idea: Four things produce a difference between two measurements when nothing has really changed.
Four false changes
An unreliable measure
A measure that gives different results on different days shows "change" every time it is used.
A practice effect
People score differently the second time because the test is familiar.
Regression to the mean
People selected for extreme scores tend to score nearer the average next time.
Patients usually seek help when they are at their worst, so some improvement is expected by chance.
Time alone
Many problems ease with the passing of time, and outside events affect everyone.
Unreliable measure · Practice · Regression · Time
A control group, measured at the same two times but not treated, is exposed to all four. Whatever change it shows is subtracted from the change in the treated group. What is left can be credited to the treatment.
Exam tip: Match the threat to the detail in the study.
"Selected for high scores" points to regression. "The same test twice" points to practice. "No comparison group" points to time.
Key idea: Effectiveness is more than a lower score at the end. It depends on the comparison, on how long the change lasts, and on whether the change matters.
Four questions about effectiveness
Compared with what?
Improvement has to exceed that of an untreated or differently treated group.
Does it last?
A follow-up months later shows whether the change has been kept.
Is it meaningful?
A change can be statistically significant and still too small to notice.
A drop of two points on a sixty-point scale may leave a patient's life exactly as it was.
How many people?
For populations, the measure is prevalence: the proportion who have the problem at a given time.
A prevention programme is effective if prevalence falls.
Compared · Lasting · Meaningful · How many
These questions apply at each level at which psychologists try to produce change: the individual patient, a local community, and global strategies.
Remember: Significant is a statement about chance. Effective is a statement about people's lives.
Feeling unprepared for exams?
Get a clear study plan, practice with real questions, and know exactly where you stand before exam day. No more guessing.
How this is asked: An unseen study that reports an improvement and calls a treatment effective is a natural choice for the 15-mark question on Paper 2.
Measurement and causality carry most of the analysis; bias is often close behind.
The pattern
- State what was measured, when, and in whom.
- Ask who is missing from the second measurement.
- Go through the false changes that fit the design.
- Separate a significant change from a meaningful, lasting one.
- Conclude on the claim of effectiveness and what would test it.
The trap: only the people who finished: Results for those who completed a programme describe the people it suited.
Ask what happened to the rest.
Discuss the following study with reference to two or more of the following concepts: bias, causality, measurement and/or responsibility.
A company wanted to show that its eight-week sleep app treats insomnia. Of the 1,100 adults who downloaded the app and agreed to take part, 300 completed all eight weeks. These 300 kept a sleep diary in the week before starting and in the final week. Their average time to fall asleep dropped from 48 minutes to 31 minutes, a statistically significant change. The company concluded that the app is an effective treatment for insomnia.
Model answer plan
See the answer broken into clear points, with guidance on how each point earns marks.