The difference between a data observation and its corresponding fitted value obtained by regression analysis. The residual mean square is the sum of squared residuals divided by the appropriate degrees of freedom, and is an estimate of variance of random variation about the fitted model. Plots of residuals against data variables may suggest important modifications to the model. Plots of ranked residuals against percentage points of the normal distribution provide a check on the assumptions used in significance tests in regression analysis. Large residuals identify observations as outliers, whose exclusion from the analysis will make a large difference to the conclusions.