Year: 2024
Type: article
Abstract: The first goal of the article is to apply the modified Cramer-von Mises (CM) goodness-of- fit test for normality to a practical problem. The modification of the test involves varying the formula for calculating the empirical distribution function (EDF). The critical values are obtained using the Monte Carlo method for sample sizes 𝑛 = 10,20 and at a significance level of α=0.05. The second goal is to calculate the power of several tests for appropriately selected alternative distributions. The article shows that the values of constants 𝛼, 𝛽 in the EDF formula affect the power of the CM test. The effectiveness of the new proposal is illustrated by the analysis of real data sets.
Authors: Piotr Sulewski, Damian Stoltmann
Institution: Uniwersytet Pomorski w Słupsku
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4404858226