Normalita dat

Abstract

The object od my work is description od graphic methods, also description of statistical tests which deal with normality. At first I desribe basic information about normal distribution of probability. Then I describe graphic mothods, which are used during tests of normality. The methods are Q-Q plot, P-P plot and N-P plot. After that I continue with statistical tests, where I describe Chi-squared test, Lilliefors Kolmogorov-Smirnov test, tests based on skewness and kurtosis, Shapiro-Wilk test and Anderson-Darling test. The most important part is simulation study made to compare these five tests according to errors I. and II. type. Selections from different types of probability distribution were generated by software RStudio. Tests and simulation studies for comparison of tests were made in the same software.