Exposure misclassification bias in a cohort study can bias results in which way?

• A. It always biases toward the null.
• B. It depends on whether misclassification is nondifferential or differential.
• C. It never affects the estimate.
• D. It only biases toward the alternative hypothesis.

Answer: (B)

Exposure misclassification affects the estimated association, and its direction depends on the type of misclassification. When misclassification of exposure is nondifferential—that misclassification occurs at the same rate regardless of the disease outcome—the bias tends to pull the estimate toward the null, weakening the observed association. If misclassification is differential—varying by outcome status—the bias can shift the estimate in either direction, potentially exaggerating or hiding the true relationship. Because the direction isn’t fixed and hinges on nondifferential versus differential misclassification, this is the best answer.