Statistical methods embraced by the terms regression analysis and analysis of variance are probably the most well-known and used in practical applications. They are based on the understanding that quantitative responses are often affected by one or a number of regressor variables. The assumed functional relationship between the regressor variables and the response is linear in unknown model parameters. This part deals with statistical tests on these model parameters, where it is either of interest if they are zero, and therefore the respective regressor variable is not relevant for the prediction of the response, or larger, smaller or equal to some pre-specified values. Chapter 16 treats the case of simple linear regression with one regressor variable as well as multiple linear regression with a set of regressor variables. Chapter 17 concentrates on analysis of variance where the effect of solely qualitative variables with finite numbers of possible levels on the response is of interest.