Kolmogorov-Smirnov test or Shapiro-Wilk test which is more preferred for normality of data according to sample size.?
Compare to Kolmogorov-Smirnov test, Shapiro-Wilk test is more suitable for smaller sample size - some books said sample size < 100, pls refer SPSS version 18.0 for Windows - Analysis Without Anguish (Coakes & Ong, 2011). Other books said sample size < 50 including the following link:
Kolmogorov-Smirnov one-sample test is recommended for one-sample test and Kolmogorov-Smirnov two-sample test for independent samples in non-parametric statistics written by Siegel. When the sample is < 30 we use non-parametric tests in Finland. Shapiro-Wilk test is not mentioned in the book.