In: Statistics and Probability
Coefficient of Determination
Coefficient of determination, denoted R² or r² and pronounced "R squared", is the proportion of the variance in the dependent variable that is predictable from the independent variable.
A measure of goodness of fit for a linear regression model is represented by the coefficient of determination, or R^2R2, and it is broadly used to assess the quality of a linear regression model
How do you compute the Coefficient of Determination?
Most often, the coefficient of determination is computed using some type of statistical software package. But using the actual Math definition is useful to arrive to an important interpretation for R-Squared.
Mathematically, the coefficient of determination is computed as
R^2=SSR/SST
where SSR stand for the regression sum of squares and SST stands for the total sum of squares.
SST=SSR+SSE
For example, if the coefficient of determination is R^2 = 0.473, what does that tell you? It indicates that 47.3% of the variation in the dependent variable is explained by the corresponding linear regression model.