This chapter discusses the ideas of sequences of models for describing, interpreting, and assessing data. The power of the model-based approach is substantially due to the computer facilities available for fitting more complex models to data. The chapter describes the use of computer packages for all calculations and focuses on conceptual ideas of the models. It introduces various factor terms into a linear regression model to represent variation of the intercept and slope of the linear relationship among various data sets. The chapter examines different ways of using models with a mixture of factors and variables for interpreting data involving structured relationships and for improving the interpretation of experimental data. It considers the addition of a regression term to an experimental design factor model aimed at improving the precision of treatment comparisons. A common application of covariance analysis is to correct for accidents that have occurred to some experimental units.