squenshl Messages 468 Reaction score 4 Thread starter May 8, 2011 #1 Give an example to show that if not assuming independence of X1, X2, ..., Xn it is possible to show that Var(1/n * sum from k = 1 to n of Xk) >> [tex]\sigma^2/n[/tex]
Give an example to show that if not assuming independence of X1, X2, ..., Xn it is possible to show that Var(1/n * sum from k = 1 to n of Xk) >> [tex]\sigma^2/n[/tex]
g_edgar Messages 606 Reaction score 0 May 9, 2011 #2 What do you get if [itex]X_1 = X_2 = ...[/itex], which is sort of the EXTREME example of non-independence?
What do you get if [itex]X_1 = X_2 = ...[/itex], which is sort of the EXTREME example of non-independence?
squenshl Messages 468 Reaction score 4 May 9, 2011 #3 Since X1 = X2 = ... = Xn. This implies that Var((1/n)*nX1) = n2[tex]\sigma^2[/tex]/n2 = [tex]\sigma^2[/tex] >> [tex]\sigma^2/n[/tex] Last edited: May 9, 2011
Since X1 = X2 = ... = Xn. This implies that Var((1/n)*nX1) = n2[tex]\sigma^2[/tex]/n2 = [tex]\sigma^2[/tex] >> [tex]\sigma^2/n[/tex]