Abstract
The Iterative Truncation Method, a generalization of Hoffmann's method, has been widely applied to determine the normal range, but it has some defects; (1) danger of eliminating normal data,(2) smaller estimation of standard deviation,(3) little consideration on the types of distribution.
Supposing theoretical normal distribution, it is made clear mathematically on the estimation of standard deviation that the bias is so large and k (truncation limit factor) smaller than √3 leads to zero estimation.
