Abstract
The standard error of an equipercentile equating is derived for four different situations. Some numerical results are checked by Monte Carlo methods. Numerical standard errors are computed for two sets of real data. (Author)