Suppose you are given a data set {(yi, mu) |z' : 1, . . . , n}, and you obtain the OLS estimates from the tted regression model :1} = B0 + ,31271. However, the underlying truth is that the response variable also depends on another unobserved independent variable :32. That is, the true model is 9:30 +51% +52%: +6- In other words, your t the data using a mis-specied model. Then is the OLS estimator of the slope ,8} in the fitted simple linear regression model unbiased? Justify your answer by deriving E(,81). And if B 1 is biased, nd its bias