Err ... yes. I meant to have normalized this, with each of the $\beta_i$'s divided by $\sqrt{n}$. One might expect the same thing more generally true with each multiplied by $\alpha_i$, where $\sum \alpha_i^2 = 1$. Thanks for the quick responses!
Err ... yes. I meant to have normalized this, with each of the $\beta_i$'s divided by $\sqrt{n}$. One might expect the same thing more generally true with each multiplied by $\alpha_i$, where $\sum \alpha_i^2 = 1$. Thanks for the quick responses!
Err ... yes. I meant to have normalized this, with each of the $\beta_i$'s divided by $\sqrt{n}$. One might expect the same thing more generally true with each multiplied by $\alpha_i$, where $\sum \alpha_i^2 = 1$. Thanks!