StaTips Part VI: Bivariate correlation

Sa？etak A very common situation in medical research, including orthodontics, is when a researcher has to verify the association between 2variables, best referred to as bivariate correlation. Bivariate correlation is an analysis that measures the strength of relationship between twovariables through the calculation of different correlation coefficients. The most common correlation coefficients are: Pearson (r), Kendall(rho), Spearman (rho) and the point-biserial (rpb). The choice of the correct coefficient is based on the type of data to be analysed and, forsome of them, the existence of assumptions for using parametrical tests. Indications on how to choose the correct coefficient and abouttheir interpretation are provided