Effect of exposure measurement error in a case-control study of child pedestrian injuries.

We examined the effect of the duration of the exposure measurement period on the odds ratio in a case-control study of traffic flow and the risk of child pedestrian injury. Varying the length of the measurement period resulted in differential exposure misclassification and bias in the odds ratio. Differential misclassification may have arisen from measurement error in traffic flow estimation over a short period or from inclusion of irrelevant exposure information with a longer measurement period.