How do correlation and causation differ, and what role do confounding variables play in PT research?

How do correlation and causation differ, and what role do confounding variables play in PT research?

• A. Correlation is association; causation implies one causes the other; confounders may include age, comorbidities, pretest differences, placebo effects
• B. They are the same
• C. There are no confounding variables in PT research
• D. Only sample size matters

Answer: (A)

In research, correlation means two variables move together, but it does not prove that one causes the other. Causation means changes in one variable produce changes in another. In physical therapy research, confounding variables are extra factors that are related to both the intervention and the outcome and can create a misleading impression of cause-and-effect or obscure a real effect. For example, age, comorbidities, baseline functional status (pretest differences), and placebo effects can influence both who receives a therapy and how much they improve. If these are not controlled, you might attribute improvement to the therapy when part of the improvement comes from these other factors. That’s why this answer is the best: it clearly distinguishes correlation from causation and acknowledges that confounders can influence results and need to be managed through study design and analysis to make valid causal inferences. The other statements fail because they claim correlation and causation are the same, deny that confounding variables exist in PT research, or imply that only sample size matters, which misses the importance of study design and confounder control.

In research, correlation means two variables move together, but it does not prove that one causes the other. Causation means changes in one variable produce changes in another. In physical therapy research, confounding variables are extra factors that are related to both the intervention and the outcome and can create a misleading impression of cause-and-effect or obscure a real effect. For example, age, comorbidities, baseline functional status (pretest differences), and placebo effects can influence both who receives a therapy and how much they improve. If these are not controlled, you might attribute improvement to the therapy when part of the improvement comes from these other factors.

That’s why this answer is the best: it clearly distinguishes correlation from causation and acknowledges that confounders can influence results and need to be managed through study design and analysis to make valid causal inferences. The other statements fail because they claim correlation and causation are the same, deny that confounding variables exist in PT research, or imply that only sample size matters, which misses the importance of study design and confounder control.