ABSTRACT

Many variables observed in real life are related. The type of their relation can often be expressed in a mathematical form called regression. Regression model relate a response variable to one or several predictors. This chapter aims to evaluate the goodness of fit of the chosen regression model to the observed data, estimates the response variance, tests significance of regression parameters, and constructs confidence and prediction intervals. Analysis of variance (ANOVA) explores variation among the observed responses. A more universal, and therefore, more popular method of testing significance of a model is the ANOVA F-test. It compares the portion of variation explained by regression with the portion that remains unexplained. Careful model selection is one of the most important steps in practical statistics. In regression, only a wisely chosen subset of predictors delivers accurate estimates and good prediction. Coefficient of determination shows the portion of the total variation that the included predictors can explain.