Many people have an intuitive feel for residuals in least square models and know that the sum of squared residuals is a goodness-of-fit measure. Generalized linear regression models use a different but related idea, called deviance residuals. What are deviance residuals, and how can you compute them? Deviance residuals (and