Correlation or Cause and Effect?
A misconception-busting classroom activity for Cambridge OCR A Level Sociology
One of the biggest challenges in teaching research methods is helping students move from everyday language (“X causes Y”) to more careful sociological thinking (“the data may show a correlation, but this does not prove causation”).
Students often use the phrase “cause and effect” too quickly, especially when they see a graph, a statistic, or a headline. This activity is designed to slow their thinking down and train them to ask better methodological questions:
- What does the data actually show?
- Is this just a correlation?
- Could there be other variables involved?
- What evidence would we need to make a stronger causal claim?
This makes it an excellent task for Cambridge OCR A Level Sociology when teaching:
- quantitative methods
- patterns and trends
- correlation
- cause and effect
- objectivity
- reliability
- links between theory and methods (especially positivist approaches)
Why this activity works
This task works so well because it starts with short, accessible examples that students can understand immediately, then pushes them to apply methodological language.
Instead of simply memorising:
- correlation = relationship
- causation = one variable causes another
students must actively classify findings and justify their decisions. That justification is where the real learning happens.
It also helps students become more careful readers of:
- media claims
- school data
- research summaries
- exam questions
In other words, this is not just a methods task but it also strengthens critical thinking and evidence literacy.
Activity overview: Correlation or Cause and Effect?
Purpose
To teach students to distinguish between:
- Correlation only
- Possible cause and effect
- Cannot tell from the data
Core idea
Students are given a set of short findings and must:
- classify each one
- explain their decision using methodological language
- suggest what further evidence would be needed to strengthen a causal claim
Step-by-step teacher instructions
1) Starter hook (5 minutes): “What’s wrong with this headline?”
Put a few exaggerated claims on the board, for example:
- “Phones cause anxiety in all teenagers”
- “Breakfast clubs guarantee better attendance”
- “More police causes more crime”
- “Revision causes high grades”
Ask students:
- What is the claim?
- Does the evidence prove this?
- What might be missing?
This gets them thinking about how easily causal language is used.
2) Teach the distinction clearly (5–10 minutes)
Give students a simple rule of thumb:
Correlation means two things are linked.
Causation means one thing directly causes the other.
Then add the key methodological reminder:
A correlation can suggest a possible relationship, but it does not prove cause and effect on its own.
You can model this with a quick example:
- Ice cream sales and sunburn both rise in summer
- Ice cream does not cause sunburn
- A third factor (hot weather) affects both
This is a great moment to introduce the idea of a third variable.
3) Main task: Classification challenge (15–20 minutes)
Give students a set of findings. They must sort each into one of three categories:
- Correlation only
- Possible cause and effect
- Cannot tell from the data
Then they must explain why using methodological language.
Example statements
- Students who revise more tend to score higher.
- Areas with more police officers report more crime.
- As ice cream sales increase, sunburn cases increase.
- A school introduces a breakfast club and attendance rises.
- Students who sleep more report better concentration in lessons.
- People who spend more time exercising report better mood.
- As the temperature rises, sales of cold drinks increase.
- Students who spend longer travelling to school are late more often.
- People who use fitness apps more often lose more weight.
- Students attending revision sessions tend to score higher in mock exams.
- Classes with stricter behaviour routines have fewer recorded disruptions.
- After a school introduces a phone ban, punctuality improves.
- Students who complete homework regularly achieve higher grades.
- Students in smaller classes make faster progress.
- Neighbourhoods with higher unemployment rates have higher recorded crime rates.
- Countries that spend more on education tend to have higher literacy rates.
- Areas with more CCTV cameras report lower street crime.
- Where youth services are cut, anti-social behaviour complaints increase.
- Children who attend nursery education have better educational outcomes later.
- Schools with more detentions have lower behaviour standards.
- People carrying lighters are more likely to get lung cancer.
- Children who own more books at home tend to achieve better in school.
This list can be downloaded below:
Answer Sheet
Key
- CO = Correlation only
- PCE = Possible cause and effect
- CT = Cannot tell from the data
1) Students who revise more tend to score higher.
Answer: CO
Why: Relationship shown, but no proof revision caused the higher score; prior attainment/motivation may affect both.
Extra evidence needed: Longitudinal tracking, controls for prior attainment, comparison groups, standardised revision measures.
2) Areas with more police officers report more crime.
Answer: CO
Why: Reverse causation likely (more police sent to high-crime areas); more police may also increase detection/reporting.
Extra evidence needed: Time-series data, comparison areas, controls for population density/reporting rates.
3) As ice cream sales increase, sunburn cases increase.
Answer: CO
Why: Third variable (hot weather/summer) affects both.
Extra evidence needed: Control for temperature/season; compare different weather conditions.
4) A school introduces a breakfast club and attendance rises.
Answer: PCE
Why: Plausible causal link, but other changes may have occurred at same time.
Extra evidence needed: Before-and-after trend data, comparison school, control for seasonal/policy changes.
5) Students who sleep more report better concentration in lessons.
Answer: CO
Why: Self-report data; other variables (stress, health, screen time) may explain both.
Extra evidence needed: Longitudinal data, objective sleep/concentration measures, controls for stress/health.
6) People who spend more time exercising report better mood.
Answer: CO
Why: Exercise may affect mood, but mood may also affect likelihood of exercise (reverse causation).
Extra evidence needed: Longitudinal/experimental studies, controls for health and lifestyle.
7) As the temperature rises, sales of cold drinks increase.
Answer: PCE
Why: Causal link is plausible, but statement alone still does not prove it.
Extra evidence needed: Repeated data across times/places, controls for promotions/holidays.
8) Students who spend longer travelling to school are late more often.
Answer: PCE
Why: Longer travel may plausibly increase lateness risk, but transport reliability and family routines matter too.
Extra evidence needed: Controls for transport type, route reliability, departure time, weather.
9) People who use fitness apps more often lose more weight.
Answer: CO
Why: App use may reflect motivation; motivated people may be more likely to lose weight anyway.
Extra evidence needed: Random assignment to app use (if possible), controls for diet/exercise baseline, longitudinal data.
10) Students attending revision sessions tend to score higher in mock exams.
Answer: CO
Why: Selection effect — more motivated/higher-attaining students may choose to attend.
Extra evidence needed: Prior grade controls, matched groups, attendance intensity data.
11) Classes with stricter behaviour routines have fewer recorded disruptions.
Answer: PCE
Why: Plausible effect, but could also reflect teacher experience, class composition, or recording practices.
Extra evidence needed: Comparison classes, same teacher before/after, observation data, controls for cohort differences.
12) After a school introduces a phone ban, punctuality improves.
Answer: PCE
Why: The ban may have contributed, but other factors may have changed too.
Extra evidence needed: Trend data over time, comparison schools, evidence no other major policy changes occurred.
13) Students who complete homework regularly achieve higher grades.
Answer: CO
Why: Homework may help, but motivation, parental support, and prior attainment may drive both.
Extra evidence needed: Controls for prior attainment/home background, longitudinal data, homework quality measures.
14) Students in smaller classes make faster progress.
Answer: CO (can move toward PCE with stronger evidence)
Why: Smaller classes may be linked to resources, streaming, or pupil need.
Extra evidence needed: Matched comparison groups, controls for teacher quality and prior attainment, longitudinal tracking.
15) Neighbourhoods with higher unemployment rates have higher recorded crime rates.
Answer: CO
Why: Relationship may exist, but many third variables (deprivation, policing, demographics, reporting) may be involved.
Extra evidence needed: Multivariable analysis, longitudinal data, controls for deprivation/policing/reporting.
16) Countries that spend more on education tend to have higher literacy rates.
Answer: CO
Why: Spending may matter, but wealth, inequality, governance, and infrastructure also affect literacy.
Extra evidence needed: Longitudinal international comparisons, controls for GDP/inequality, policy implementation measures.
17) Areas with more CCTV cameras report lower street crime.
Answer: PCE
Why: CCTV may deter some crime, but other interventions may also be responsible.
Extra evidence needed: Before/after local crime data, comparison areas without CCTV, controls for policing changes.
18) Where youth services are cut, anti-social behaviour complaints increase.
Answer: PCE
Why: Cuts may contribute, but economic change, policing, or reporting behaviour could also explain rises.
Extra evidence needed: Time-series data, matched areas, controls for local context, complaint/reporting changes.
19) Children who attend nursery education have better educational outcomes later.
Answer: CO (can move toward PCE with stronger longitudinal evidence)
Why: Selection effects (family income, education, support) may explain both nursery attendance and later outcomes.
Extra evidence needed: Longitudinal cohort data, controls for family background, matched samples.
20) Schools with more detentions have lower behaviour standards.
Answer: CT
Why: More detentions may mean worse behaviour or stricter enforcement/recording; meaning is unclear.
Extra evidence needed: Behaviour observation data, detention policy comparisons, recording practice evidence.
21) People carrying lighters are more likely to get lung cancer.
Answer: CO
Why: Third variable (smoking) explains both lighter carrying and lung cancer.
Extra evidence needed: Controls for smoking behaviour; compare smokers/non-smokers.
22) Children who own more books at home tend to achieve better in school.
Answer: CO
Why: Book ownership may indicate parental education/income/cultural capital rather than directly causing achievement.
Extra evidence needed: Controls for parental education/income, longitudinal data, reading practice measures.
Quick teacher debrief questions (great for class discussion)
- Which examples were easiest to classify — and why?
- Which examples were most debatable?
- Why are sociologists careful about causal claims in real social life?
- How does this link to positivist preferences for quantitative data (patterns/trends) but also the limits of those data?
Leave a Reply