Which term is used when the association between two variables is due to the influence of a third variable?

Which term is used when the association between two variables is due to the influence of a third variable?

• A. Confounding
• B. Bias
• C. Causation
• D. Interaction

Answer: (A)

Confounding happens when the association you observe between two variables is actually driven by a third variable that influences both of them. In practice, you see A linked to B, but the link weakens or disappears once you account for C, which affects both A and B. That’s what makes the observed relationship spurious rather than a direct A→B effect. Think of a simple example: you notice that people who carry lighters have more lung disease. It isn’t the lighter causing disease; smoking is the third factor that’s associated with both carrying a lighter and with lung disease. If you adjust for smoking, the apparent link between lighters and disease diminishes. This concept is distinct from bias, which is a systematic error in how a study is designed or conducted; from causation, which would mean A directly causes B; and from interaction (effect modification), where the effect of A on B changes depending on the level of a different variable.

Confounding happens when the association you observe between two variables is actually driven by a third variable that influences both of them. In practice, you see A linked to B, but the link weakens or disappears once you account for C, which affects both A and B. That’s what makes the observed relationship spurious rather than a direct A→B effect.

Think of a simple example: you notice that people who carry lighters have more lung disease. It isn’t the lighter causing disease; smoking is the third factor that’s associated with both carrying a lighter and with lung disease. If you adjust for smoking, the apparent link between lighters and disease diminishes.

This concept is distinct from bias, which is a systematic error in how a study is designed or conducted; from causation, which would mean A directly causes B; and from interaction (effect modification), where the effect of A on B changes depending on the level of a different variable.