Consider a hypothetical program at SFU which gives incoming freshmen free laptops if their family income is below 20,000 dollars. a) (10 points} Suppose I had data on freshman at SFU and estimated the following regression GPA = g +lcomputer+u, where computer is a dummy variable equal to 1 if the individual owns a computer and 0 otherwise. Is 51 going to biased? Why or why not? b) (10 points} Now suppose I use family income as an instrument for computer. Is this going to be a valid instrument? Why or why not? 0) (10 points} Explain how you might use a regression discontinuity framework to estimate the effect of computer use on GPA. You do not necessarily need to include a regression equation, but please describe your setup in detail.| d) (10 points} Suppose that we plot the data and it looks like the below. "What would you say is the eect of the program on GPA? ' 20,000 dollls . 'i '9 .1. can" i a?\" N ' ~33" IQ- I 0 50000 100000 150000 200000 Family Income e) (10 points) Suppose a university ofcial asked you whether you thought that the program was e'eetive in raising GPAs through computer use. In light of the previous graph, what would you say