Effect of Violations of Normality Edgell and Noon, 1984 On the Correlation Coefficient t-Test
Is the t-test for correlation coefficients robust to violations of its assumptions?
Overview Review t-test of the Correlation Coefficient Violations Bivariate Normal Assumption Independence Assumption
Review Violation of Normality Violation of Independence Bivariate normal assumption Both variables come from normal distributions Both variables come from normal distributions OROR One variable is from a normal distribution and the variables are independent One variable is from a normal distribution and the variables are independent Independence assumption Value of one variable is not influenced by the other Value of one variable is not influenced by the other t 2= r 2 / ((1-r 2 )/df)
Review Violation of Normality Violation of Independence Run 10,000 samples Very Non-normal distributions Range of sample sizes Determine the proportion of samples that were significant at the.05 and.01 level Method
Review Violation of Normality Violation of Independence Distributions Exponential Distribution
Review Violation of Normality Violation of Independence Distributions Uniform Distribution
Review Violation of Normality Violation of Independence Distributions Cauchy Distribution
Review Violation of Normality Violation of Independence Results
Review Violation of Normality Violation of Independence Results
Review Violation of Normality Violation of Independence Method Run 10,000 samples Range of sample sizes Zero correlations with dependency Determine the proportion of samples that were significant at the.05 and.01 level
Review Violation of Normality Violation of Independence Method Zero-Correlations with dependency 1) Second variable is the square of the First Variable 2) Mixed Bivariate Normal Distributions - Population is aggregate of smaller subpopulations
Review Violation of Normality Violation of Independence P=.5ρ1=.3P=.5 ρ2= -.3 ρ =0 Mixed Bivariate Normal Distributions
Review Violation of Normality Violation of Independence Results
Violations of Normality Robust at.05 Robust at.05 At.01, only sensitive to extreme departures from normality At.01, only sensitive to extreme departures from normality Conclusion Is the t-test for correlation coefficients robust to violations of normality?
Conclusion Is the t-test for correlation coefficients robust to violations of independence? Not Robust But Non independent variables are not likely to have a correlation of zero Non independent variables are not likely to have a correlation of zero t-Test could be considered a test of the hypothesis of independence t-Test could be considered a test of the hypothesis of independence