Premium
Population comparisons based on an extreme fraction of the sample: A caveat
Author(s) -
Gillett Raphael
Publication year - 1991
Publication title -
british journal of mathematical and statistical psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.157
H-Index - 51
eISSN - 2044-8317
pISSN - 0007-1102
DOI - 10.1111/j.2044-8317.1991.tb00962.x
Subject(s) - statistics , mathematics , sample size determination , sample (material) , fraction (chemistry) , sample mean and sample covariance , population , selection (genetic algorithm) , selection bias , econometrics , demography , computer science , physics , chemistry , organic chemistry , estimator , artificial intelligence , sociology , thermodynamics
Population comparisons based on the mean of a fixed number of the highest (lowest) sample values have been shown to be biased when the samples are of different sizes. The present study demonstrates that, surprisingly, the bias persists even when the mean is based on a fixed proportion, rather than a fixed number, of the highest (lowest) sample values. Although the magnitude of the effect is considerably diminished, it can still produce problems in statistical surveys and in applied settings, e.g. personnel selection. Methods of correcting the bias are outlined.