What you’ll learn
Distinguish correlation from causation.
Key idea
An observational difference describes association, not an assigned-treatment causal effect.
A group difference is sample association, not a causal effect; an alternative is plausible rather than established.
Worked example
Enter the exact value or ordered values in the format requested below.
Fictional OBSERVATIONAL summer survey, no assigned treatment: 14/20 ice-cream buyers and 2/20 nonbuyers report sunburn. Sunny weather is explicitly a plausible common factor that could influence both buying and sunburn, not established as the cause. Compute the numerical observed proportion difference, buyers minus nonbuyers. Do not call it causal effect.
- Sample shares: buyers 7/10, nonbuyers 1/10; buyer-minus-nonbuyer difference 3/5.
- This is an observed association, not a measured causal effect. The supplied plausible common factor and unassigned observational groups prevent this difference alone establishing causal direction.
Answer: 3/5
Quick Check
Interpret the result of 'Distinguish correlation from causation' in the data-generating context, including units and population limits.
Fictional OBSERVATIONAL summer survey, no assigned treatment: 18/30 ice-cream buyers and 9/30 nonbuyers report sunburn. Sunny weather is explicitly a plausible common factor that could influence both buying and sunburn, not established as the cause. Interpret the two-group frequency table and its causal limits.