Shapiro-Wilk Normality Test

R is one of the most powerful programming languages for statistical testing; owing to its ability to use packages directly without resorting to programming tools or  languages in easy manner. Recently this language became popular for statisticians and data scientists and the packages are more spread among developers

Due to the importance of data, it is ineluctable to test the data before using them and perform any parametric test, it is important to test whether the data are following or not normal distribution. Thus, Shapiro-wilk and Kolomgorov-smirnov are widely used for this purpose. In this article, Shpiro-Wilkwill be presented with solved example

The Shapiro-wilk test, in statistics, is used in order to determine whether the data fits to normal distribution or not. The Shapiro-Wilk is more suitable for sample < 50. In this test, the null hypothesis (Ho) can be determined as the data are normally distributed; however, the alternative hypothesis (H1) can be determined as the data are not following the normal distribution

For example, for the following data (Age of employees in company): 42 54 47 47 44 42 30 52 48 58 46 36 49 55 41 52 57 61 50 47 54 47 50 38 31 47 56 38 55 47 35 35 39 41 44 53 47 42 44 45 56 38 58 52 54 49 53 43 35 56 38 42 35 40 34 35 45 54 47 45 47 46 36 45 46 47 34 45 44 50 38

The test can be done Using R language by writing the following code: “Shapiro.test(write here the uploaded data name)"

For the previous example, the solution using R is shown in the below figure. From the results, it can be seen that P-value (0.2445) ≥ 0.05;thus, Ho is accepted. In other words, the age of employees follows normal distribution