I will attach my question.

Suppose that Y =X3+U. where Y' and U are nx1 vectors, X" is an nx X, vector, # is a K, x1 vector, U is statistically independent of X", E(U)=0, and E(UU') =J'Inxx; 0' >0 is a finite constant, and In is the nxn identity matrix. In many applications, K, is an increasing function of n but K, -+ 0 as n- co. Thus, there are more explanatory variables when n is large than when n is small. Let / be the ordinary least squares estimator of / in this model. Suppose that | X, [SM for all i=1 ..,; k=1,..., K, ; and some finite constant M. Suppose, also, that n" XX =Ixxx : where Ixxx, is the En x K, identity matrix.. Let z = (21:22....,Zx)' bea K, x1 vector with | |SM (k=1,..., K, ) all n. a. Under what additional conditions, if any, is z'S a uniformly consistent estimator of z/? Uniform consistency means that sup:; ISM [?'(8- 8) ( 0 as n + 00. b. Suppose z is a K, x1 vector of 1's. What is the rate at which z'(-8) converges in probability to 0? That is, find f(n) such that f(n)?'(8-#) has a non-degenerate limiting distribution