Dear Statalist,
I am trying to estimate leader FE with wild bootstrapped CI, as follows
Y1 is a binary variable.
The issue is estimates aren't centered in bootstrapped confidence bars for some leaders (This does not appear in regression without wildbootstrap). Is that due to small sample (20 leaders)? If that is the case, then I should obseve the same issue with another binary variable which is Y2. But for Y2, estimates are centered in bootstrapped confidence bars.
I am struggling to find the reason.
Any help would be appreciated.
Many thanks in advance.
I am trying to estimate leader FE with wild bootstrapped CI, as follows
Code:
forvalues i = 1/20 {
gen leader`i' = (leaderID==`i')
}
wildbootstrap reg Y1 leader* X1 X2 X3 X4 X5, noconstant cluster(leaderID) and
The issue is estimates aren't centered in bootstrapped confidence bars for some leaders (This does not appear in regression without wildbootstrap). Is that due to small sample (20 leaders)? If that is the case, then I should obseve the same issue with another binary variable which is Y2. But for Y2, estimates are centered in bootstrapped confidence bars.
I am struggling to find the reason.
Any help would be appreciated.
Many thanks in advance.
