<?xml version="1.0" encoding="UTF-8" ?>
<xml>
<records>
<record>
<title>Analysis of variance: Comfortless questions</title>
<authors>
<author>L.V. Nedorezov</author>
</authors>
<affiliations>
<affiliation>
Research Center for Interdisciplinary Environmental Cooperation RAS, Saint-Petersburg, Russia
</affiliation>
</affiliations>
<journal>Computational Ecology and Software</journal>
<issn>ISSN 2220-721X</issn>
<homepage>http://www.iaees.org/publications/journals/ces/online-version.asp</homepage>
<year>2017</year>
<volume>7</volume>
<issue>3</issue>
<startpage>91</startpage>
<endpage>99</endpage>
<publisher>International Academy of Ecology and Environmental Sciences</publisher>
<location>Hong Kong</location>
<date>
<received>13 April 2017</received>
<accepted>20 May 2017</accepted>
<published>1 September 2017</published>
</date>
<keywords>
<keyword>analysis of variance</keyword>
<keyword>tests for normality</keyword>
<keyword>goodness measure</keyword>
</keywords>
<abstract>
In this paper the simplest variant of analysis of variance is under consideration. Three examples from textbooks by Lakin (1990) and Rokitsky (1973) were re-considered. It was obtained that traditional one-way ANOVA and Kruskal - Wallis criterion can lead to unreal results about factor's influence on value of characteristics. Alternative way to solution of the same problem is under consideration too.
</abstract>
<doi>DOI 10.0000/issn-2220-721x-compuecol-2017-v7-0008</doi>
<url>http://www.iaees.org/publications/journals/ces/articles/2017-7(3)/analysis-of-variance-Comfortless-questions.pdf</url>
</record>
</records>
</xml>
