Consider the following computer results, which describe regression analyses involving two independent variables X1 and X2 and a dependent variable Y. Assume that your goal is to assess the relationship of X1 with Y, controlling for the possible confounding effects of X2.
a. Using an appropriate regression coefficient as your measure of association, determine whether confounding exists. Explain.
b. Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients. What conclusion would you draw? Explain.
c. What is the moral of this example?