The Control of Type I Error and Power in Statistics for Spearman's Rho and Kendall's Tau Correlation Coefficients by Monte Carlo Method

• Activity Number: 227
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Social Statistics Section
• Author(s): Chittanun Sitthisan*+
• Companies: University of Northern Colorado
• Keywords: Spearman's Rank Correlation Coefficient ; Kendall Tau Correlation Coefficient ; Type I error

Abstract

The purposes of this research were to study the ability to control Type I error in statistics for testing the significance of Spearman's Rank and Kendall Tau Correlation Coefficients. This research also studies the power of the test in statistics for Spearman's Rank and Kendall Tau Correlation Coefficients when the correlation coefficients of population are equal to 0.00, 0.50, and 0.90. The population used attributed as the Normal, Positive skewness, and Negative skewness distribution having the correlation coefficient of 0.00, 0.50 and 0.90. The sample sizes of the study were 10, 20, 30, and 50. This research is an experimental research by Monte Carlo Method using the R statistics package to simulate the population. The findings showed that Spearman's Rank Correlation Coefficients could control Type I error following as the level of significance specified both at .05 and .01 in every sample sizes and every distributions. The findings also indicated that controlling in Type I error for Spearman's Rank Correlation Coefficients statistics was better than Kendall Tau Correlation Coefficients statistics. The research findings also suggest that the results in the Power of test in sta