Homework SourseIf there is nondifferential misclassification of the reporti
If there is nondifferential misclassification of the reporting of past exposure in a cohort study such that the exposed and the non-exposed are misclassified to the same degree and the same direction, what will be the most likely effect of the misclassification?
Its effect will not bias the relative risk estimate
This misclassification is a form of selection bias
Its effect may weaken the association between the exposure and the outcome
Its effect may bias the relative risk estimate away from the null
One would not be able to determine what effect this would have on the estimate of the relative risk
C and D
| A. | Its effect will not bias the relative risk estimate | |
| B. | This misclassification is a form of selection bias | |
| C. | Its effect may weaken the association between the exposure and the outcome | |
| D. | Its effect may bias the relative risk estimate away from the null | |
| E. | One would not be able to determine what effect this would have on the estimate of the relative risk | |
| F. | C and D |
Solution
F. C & D both.
Differential misclassification can either overestimate or underestimate the true association, depending on the situation. It either pushes the study results away from truth toward the null hypothesis or away from truth in the opposite direction of the null hypothesis, depending on the situation.
