In the paper a test of the hypothesis $\mu+c \sigma \leq M$, $\mu - c \sigma \geq m$ on parameters of the normal distribution is presented, and explicit formulas for critical regions are derived for finite sample sizes. Asymptotic null distribution of the test statistic is investigated under the assumption, that the true distribution possesses the fourth moment.
@article{104428,
author = {Franti\v sek Rubl\'\i k},
title = {Testing a tolerance hypothesis by means of an information distance},
journal = {Applications of Mathematics},
volume = {35},
year = {1990},
pages = {458-470},
zbl = {0727.62027},
mrnumber = {1089926},
language = {en},
url = {http://dml.mathdoc.fr/item/104428}
}
Rublík, František. Testing a tolerance hypothesis by means of an information distance. Applications of Mathematics, Tome 35 (1990) pp. 458-470. http://gdmltest.u-ga.fr/item/104428/
Matematická statistika, Praha, SNTL 1978. (1978)
Mathematical Methods of Statistics, Princeton University Press 1946. (1946) | MR 0016588
Linear Statistical Inference and Its Applications, (Czech translation). Praha, Academia 1978. (1978)
On testing hypotheses approximable by cones, Math. Slovaca 39 (1989), 199-213. (1989) | MR 1018261
On the two-sided quality control, Apl. Mat. 27 (1982), 87-95. (1982) | MR 0651047
Correction to the paper "On the two-sided quality control", Apl. Mat. 34 (1989), 425-428. (1989) | MR 1026506