A correction of the local pooled error (LPE) method to replace
the asymptotic variance adjustment with an unbiased adjustment
based on sample size.
Description
Two options are added to the LPE algorithm. The original
LPE method sets all variances below the max variance in the
ordered distribution of variances to the maximum variance. in
LPEadj this option is turned off by default. The second option
is to use a variance adjustment based on sample size rather
than pi/2. By default the LPEadj uses the sample size based
variance adjustment.