Finite-n Corrections to Normal Approximation |
| |
Authors: | J Vrbik |
| |
Institution: | 1. Department of Mathematics , Brock University , St. Catharines, Canada jvrbik@brocku.ca |
| |
Abstract: | ABSTRACT We show that the sampling distribution of a statistic which follows the Central Limit Theorem can be approximated by a Normal distribution, modified according to the statistic's first few moments. One can thus achieve an impressive accuracy even for relatively small samples. |
| |
Keywords: | Central Limit Theorem Cumulants Sampling distribution |
|
|