Calculating the f-statistics
The f-statistics are another test to see whether the slope is equal to zero. The null hypothesis is that the slope is zero. The f-statistics test whether we can reject this. To calculate it, we have to define the mean squared error first.
The regression for the mean squared error is the explained variation regression (SSE) divided by the regression degrees of freedom minus 1. In this example, we have the regression for two variables. The degree of freedom is 1:
Regression Mean Squared Error = SSR / Regression Degrees of Freedom Regression Mean Squared Error = 610.277 / 1 = 610.277
The residual mean squared error is the unexplained variation residual sum of squares (SSE) divided by the degrees of freedom of the residuals. The residual degrees of freedom are the number of records of the data source minus 2. In this case, we have 23 records. The degrees of freedom are 21:
Residual Mean Squared Error = Unexplained Variation Residual Sum of Squares...
