Least Squares

Slide to the next stage, then move the blue points around the perimeter to find the line of best fit for the points shown. [br]Then keep advancing stages to see how well you do.[br]Click the button for a new set of points.
The object is to estimate the 'least squares' line. [br]Stage 1 shows you the points.[br]Stage 2 gives you the opportunity to estimate the line of best fit. [br]Stage 3 plots the residuals.[br]Stage 4 give you the 'square' of the residuals. [br]Stage 5 reveals the true least squares regression line.

What is R-Squared?

Drag the red point labeled "Drag Me." As this point moves, the sum of the squared residual errors changes, becoming minimum as the line [math]y = \bar{y}[/math] is rotated onto the Least Squares Regression Line. The percentage decrease in the sum of the squared residual errors between the line [math]y = \bar{y}[/math] and the Least Squares Regression Line is the value of [math]R^2[/math] (as a percentage).