Suppose the true regression model is given by (4-8). The result in (4-10) shows that if either
Question:
Suppose the true regression model is given by (4-8). The result in (4-10) shows that if either P1.2 is nonzero or β2 is nonzero, then regression of y on X1 alone produces a biased and inconsistent estimator of β1. Suppose the objective is to forecast y, not to estimate the parameters. Consider regression of y on X1 alone to estimate β1 with b1 (which is biased). Is the forecast of y computed using X1b1 also biased? Assume that E[X2 | X1] is a linear function of X1. Discuss your findings generally. What are the implications for prediction when variables are omitted from a regression?
Fantastic news! We've Found the answer you've been seeking!
Step by Step Answer:
Related Book For
Question Posted: