مقارنة اختبار Z المعدل-للتباين مع اختبار ليفين المعدل بوجود القيم الشاذة باستعمال المحاكاة

Abstract EN

The problem of non-homogeneity variance one of the basic assumptions of the analysis of variance and analysis of linear regression and other statistical methods for it is one of the important topics in statistics because of their different applications in the humanities and other areas, and most of the researchers is a big mistake, a real application data without testing and thus get a shaded results, from this perspective has become imperative for the researcher test and check data and information to determine the nature of the data and to explain the relationship between different factors and analysis of results in a more accurate and comprehensive.

Thus, the importance of research in the use of statistical methods that create a substantive exact description of his experience on the foundations appear far from the interpretations and personal factors that have a direct impact may affect the search results, Finally, It has been accessed the most important the homogeneity of variance tests modified in the event of outlier values for the diagnosis of the problem of non homogeneity of variance and has the comparison between the Modified Z-variance and test Modified Levene test using simulation and based on standard power of the test.