When the covariance is divided by two standard devia- tions, the range of the covariance is rescaled to the inter- val between −1 and +1, thus the interpretation of correla- tion follows as in the case of Equation (11).
= Sample standard deviation of
y
Correlation is sometimes criticized as having no clini- cal interpretation or meaning [16]. This criticism is mitigated by taking the square of the correlation coefficient which is often called COEFFICIENT OF DETERMINATION. [17] expressed coefficient of determination 2r proportion of common variation in the two variables (that is the “strength” or “magnitude” of the relationship). He emphasized that it is important to know this magnitude or strength in order to evaluate the correlation between variables. The square index is interpreted as proportion of variation in one variable accounted for by differences in the other variable. According to [16],