Premium
Characterization using normal or log‐normal data with multiple censoring points
Author(s) -
Hawkins Douglas M.,
Oehlert Gary W.
Publication year - 2000
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/(sici)1099-095x(200003/04)11:2<167::aid-env395>3.0.co;2-q
Subject(s) - censoring (clinical trials) , statistics , mathematics , normality , estimator , parametric statistics , sample size determination , normal distribution , standard deviation
Probability plotting methods based on the Kaplan–Meier estimator of the cumulative distribution function provide an effective way of checking the normality or log‐normality of samples with multiple left censoring points. The plots also provide graphical estimates of the mean and standard deviation of the underlying data. The paper shows that the distributional and performance properties of the methods depend largely on the number of complete readings rather than on the original sample size or the percentage of censoring that occurred. On the view that parametric methods are generally preferable to non‐parametric where both can be applied, this argues for choosing analysis methods on the basis of the absolute number of uncensored readings rather than their proportion in the original sample. Copyright © 2000 John Wiley & Sons, Ltd.