Question: Construct a data set with two regressors for which both t-statistics are less than one even though each regressor is highly correlated with Y. Explain what this means. In particular, would the use of a stepwise regression procedure lead to the selection of neither regressor?