If g_j, j=1..n are n variables generated through Beta distribution, and s_j(j)=log(g_j(j)), j=1..n, then how to determine the distribution of s_j.
Please do me some help to find it.
I appreciate your reply. Could you please elaborate a bit more that doing a transformation how could I know which distribution s_j belongs? Or your are assuming that s_j will still obey Beta distribution.