Two rules cover this part of AP Statistics: you can conclude cause and effect only from a randomized experiment, and you can generalise to a population only when the sample was randomly selected from it. Without one of these kinds of randomness, the conclusion has to be narrower.
Observational study vs experiment
| Observational study | Experiment | |
|---|---|---|
| What the researcher does | Observes and records; no treatment imposed | Imposes treatments on experimental units |
| Cause and effect? | No, confounding is possible | Yes, with random assignment |
| Typical wording | “There is an association between…” | “Giving… causes a change in…” |
What your conclusion can claim
| Random assignment | No random assignment | |
|---|---|---|
| Random sampling | Cause and effect, generalised to the population | Association only, generalised to the population |
| No random sampling | Cause and effect for the units studied | Association only, for the units studied |
Example
Illustrative example. A survey finds that students who sleep more have higher grades. This is an observational study, so you can only state a positive association, not that more sleep causes better grades; a confounding variable such as workload or after-school activities may affect both.
Checklist before writing a conclusion
- Is it an observational study or an experiment?
- Was there random assignment?
- Was there random sampling from a population?
- Is the conclusion written in the context of the question?
Want a study plan that fits your child?
Tell us the curriculum, year group, exam date (if any) and goal. Our academic team will suggest a starting plan, 1-on-1 at our Bangkok centre or online.
Related articles
- AP Calculus FRQ: The Justification Examiners Actually Want
- What is AP (Advanced Placement)?
- Using the SAT for International Programmes at Thai Universities: What to Check
Learn 1-on-1: 1-on-1 AP tutoring
Frequently Asked Questions
No. Correlation measures the strength and direction of a linear relationship, not what causes what.
A variable related to both the explanatory and the response variable, so you cannot tell whether the observed effect comes from the variable being studied.
Yes. State which kind of randomness the study used, because it justifies the scope of your conclusion.
Max Academia offers 1-on-1 AP tutoring by subject. See also AP Calculus FRQ justification.

