ABSTRACT
How variables are related is central to most data analyses. The previous chapters regarding t-tests, ANOVA, and tests of proportions examine the relationship of the response variable to one or more grouping variables. More generally, this chapter begins the study of the relationships of continuous variables. Central themes are the scatterplot of two variables, possibly with a third grouping variable, and the corresponding Pearson correlation coefficient. Obtain these analyses with the lessR function Plot() and, just for the correlation, the function Correlation(). Bubble plots are also available for a small number of response categories to two Likert attitude survey items. For multiple variables, Correlation() returns the combined correlation matrix and scatterplot matrix of all the variables, plus an optional heat map. The nonparametric Spearman and Kendall coefficients are introduced as alternatives to the standard Pearson coefficient. These coefficients reflect the extent of the monotonicity of a relationship, a more general concept than linearity.
