Consider a model at the employee level, Vie = Bo + BiXie + Vie where the error term vie is specific to employee e at firm i. Suppose the individual data are unavailable. Instead the researcher has the firm data: yu, which is the average y for firm i, xi - the average x for firm i. On top of that the researcher has information on nu which is the size (number of employees) of firm i. Explain what would be the problem of regressing yu on Xi. (3 marks) How can the problem in (ii) can be solved? Explain in detail. (5 marks)