Suppose you fit the linear model Y = β0 + β1X1 + … + βpXp. What would be the effect on bias and variance of your model if you include an additional predictor Xp+1, assuming that Y and Xp+1 are independent?