So you want to test if the mean of a numeric outcome differs across two groups. Question: Are your measurements paired or independent?
You have paired data if every subject in your sample has a measurement under two different conditions. Each pair of measurements will be related because the were obtained from the same subject.
- You measure people’s blood pressure before and after taking a drug.
- To see if a new brand of sunscreen prevents sunburn, you put sunscreen on one arm of each study participant and don’t put anything on the other arm.
You have independent measurements when two different groups of sample subjects are present, and each subject has a signal measurement under one condition only.
- You want to see if students living in dorms spend the same amount of money on food as those who live off campus.
- In a drug trial, you give one group of patients the drug and another group a placebo.