Residual
Also called: residuals, error term, regression residual
A residual is the gap between an observed value and what the model predicted. Residuals are where the diagnostics live: if they fan out with the fitted value you have heteroskedasticity, if today's predicts tomorrow's you have autocorrelation, and if they curve you have the wrong functional form. Plotting them is the single most informative thing you can do after fitting a regression.
Where this is taught
Definitions are the trailer. These free levels turn Residual into something you play — one bite-size lesson, with worked examples, a quiz and XP.