In conducting a multiple regression analysis data analysts frequently find themselves with "an embarrassment of riches" in terms of having many possible independent variables to use. What potential problems can result from parameterizing a regression model with, say, 100 independent variables?