a set of graphical and numerical techniques routinely used by researchers to check for violations of assumptions in the application of regression analysis to particular data sets. For example, it is assumed that the relationship between the independent variables and the dependent variable is linear, that the variables have been measured accurately, and that any prediction errors resulting from the regression equation are independent and normally distributed with equal variance and a mean of zero. If the data do not possess such characteristics, the analysis may not be appropriate and thus its results may not be valid. See diagnostics; residual analysis.