Pearson's and Spearman's Correlation

A correlation analysis provides a quantifiable value and direction for the relationship between the two variables, but the output generated cannot determine cause and effect. The two commonly used correlation analyses are Pearson's correlation (parametric) and Spearman's rank-order correlation (nonparametric). The Pearson and Spearman analyses provide the researcher with a p-value (i.e., significance level) and an r-or p-value (i.e., strength of the relationship). This chapter discusses the assumptions of the correlation analysis in more depth. The following assumptions must be satisfied in order to run Pearson's and Spearman's correlation: data type; distribution of data; and random sampling. The chapter further compares Pearson's and Spearman's tests. Statistical programs are used to run a correlation analysis to determine a significant relationship (p-value) and the strength of the relationship (r, ρ).