Comparison of Spurious Correlation Methods Using Probability Distributions and Proportion of Rejecting a True Null Hypothesis

Subscribe to access this work and thousands more

ABSTRACT

The problem of spurious correlation analysis, e.g. Pearson moment-product correlation test is that the data need to be normally distributed. This research work compares spurious correlation methods using some non-normal probability distributions in order to obtain the method with the best degree of association among them. The methods were compared using proportions of rejecting true null hypothesis obtained from t and z test statistics for testing correlation coefficients. Data from Normal, log-normal, exponential and contaminated normal distributions were generated using a simulation method with different sample sizes. The results indicate that, when the data are normal, exponential and contaminated normal random distributions, Pearson's and Spearman's rank have the best proportion of rejecting the true null hypothesis. But, when the data are log-normal distribution, only Spearman's rank correlation coefficient has the best proportion of rejecting the true null hypothesis. Thus, Pearson's and Spearman's rank have the best degree of association under normal, exponential and contaminated normal distributions. While for log-normal distribution only Spearman's rank has the best degree of association.

Subscribe to access this work and thousands more