Residual

First taught in High school statistics

Definition

A residual is the difference between an actual data value and the value predicted by a model, such as a line of best fit.

02468100246810xyresidual = 6.5 − 5.8 = 0.7−0.6
Points above the line have positive residuals. Points below have negative residuals.
Advertisement

Example: A regression line predicts . At it predicts , but the actual value is :

  • A positive residual means the point is above the line (the model predicted too low).
  • A negative residual means the point is below the line.

Residual plots: graphing the residuals against checks whether a line is a good model. If the residuals look randomly scattered around 0, the line fits well. If they form a curve, a linear model is not the right choice.

The least-squares regression line makes the sum of the squared residuals as small as possible.

Advertisement