The cause-and-effect trap
A free TEAS® scientific reasoning explainer with an original practice question
A correlation means two things tend to change together. It does not show that one causes the other. Something else could drive both, or the link could run the other way.
Records, surveys and other observations can show a correlation. To show cause, a study has to change one thing on purpose, keep everything else the same, and compare the results with a control group.
When a question asks what the evidence supports, pick the conclusion that claims no more than the data shows. Answers that add a reason the data never measured are a common trap.
Try a sample question
A hospital reviews its records. On average, patients who had more visitors went home sooner than patients who had fewer visitors. Which conclusion do these records best support?
- A.More visitors make patients recover faster
- B.The two are linked, but the records don't show cause
- C.Any patient with visitors will go home sooner
- D.Patients with fewer visitors stay longer because they are sicker
Show the answer
B. The two are linked, but the records don't show cause
The two are linked, but the records don't show cause. When two things rise and fall together, that is a correlation. To show cause, a study has to change one thing on purpose and keep everything else the same. These records only observe what happened.
Why the other answers tempt:
- A. The records show that visitors and shorter stays go together. They don't show that visitors cause the shorter stays. Something else could explain both.
- C. The pattern is an average across many patients. It doesn't predict what happens to every single patient.
- D. This might be true, but it is a guess about the cause. Nothing in the records measures how sick the patients were.
Seen together is not the same as caused by. Only an experiment shows cause.
Checked against OpenStax Introductory Statistics 2e, 1.4 Experimental Design and Ethics. Original question, not from a real test.
Find the traps you fall for
This trap is one of many. The free Science Check shows which ones you fall for and why each wrong answer was tempting. No signup needed.