Confounding is best described as a phenomenon where a nuisance variable is associated with which of the following?

Confounding is best described as a phenomenon where a nuisance variable is associated with which of the following?

• A. A nuisance variable that is associated with both exposure and outcome and biases the observed effect unless controlled.
• B. A variable that modifies the effect of exposure on the outcome.
• C. A random measurement error.
• D. A characteristic of the sample that improves generalizability.

Answer: (A)

Confounding occurs when a third variable is linked to both the exposure and the outcome in a study, and this linkage distorts the apparent relationship between exposure and outcome. Because the confounder is related to both, it can create a spurious association or mask a real one, making the observed effect biased unless you measure and adjust for it. For example, if older age is associated with both higher exposure to a treatment and with worse health outcomes, failing to account for age could make the treatment look more or less effective than it truly is. Other possibilities describe different ideas: an effect modifier changes how strong the exposure–outcome relationship is in different groups, which is not the same as confounding. Random measurement error introduces imprecision or misclassification, not a biased association due to a third variable linked to both exposure and outcome. A sampling characteristic that improves generalizability relates to external validity, not bias in the exposure–outcome link.

Confounding occurs when a third variable is linked to both the exposure and the outcome in a study, and this linkage distorts the apparent relationship between exposure and outcome. Because the confounder is related to both, it can create a spurious association or mask a real one, making the observed effect biased unless you measure and adjust for it.

For example, if older age is associated with both higher exposure to a treatment and with worse health outcomes, failing to account for age could make the treatment look more or less effective than it truly is.

Other possibilities describe different ideas: an effect modifier changes how strong the exposure–outcome relationship is in different groups, which is not the same as confounding. Random measurement error introduces imprecision or misclassification, not a biased association due to a third variable linked to both exposure and outcome. A sampling characteristic that improves generalizability relates to external validity, not bias in the exposure–outcome link.