Question

Assume your crude measure of association is 9 and you stratify by a variable that has two levels. Your first stratified measure of association is 4. Which of the following would your second stratified measure of association have to be to see confounding?

Answer 1

To observe confounding with a crude measure of association of 9 and one stratified measure of 4, the second stratified measure must differ substantially from both, indicating that the stratifying variable is confounding the association.

Step-by-step explanation:

To see confounding when your crude measure of association is 9 and the first stratified measure is 4, the second stratified measure would have to be different from both of these measures, but still not equal to 1 (no association). Confounding implies that the separation of data by strata reveals different measures of association, suggesting that the initial measure of association was confounded by the stratifying variable.

For example, if the second stratified measure of association was 3, this would indicate confounding because the stratified measures (4 and 3) are both substantially different from the crude measure of 9 and from each other, implying that the variable by which we stratified has a confounding effect on the association between the main variables of interest.