Table of Contents#### Download Safari Books Online apps: Apple iOS | Android | BlackBerry

### 16.4 Evaluating Model Assumptions and Goodness of Fit

Entire Site

Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

An important step in multiple regression analysis is to evaluate model assumptions and how well the model fits the data. As with simple regression analysis, residuals play an important role in this step. Residuals are used in determining such quantities as RMSE and R, but they can also be informative in their own right, especially when plotted against other variables. Large residuals are especially noteworthy. Residual patterns can reveal problems with the data or inadequacies in the model.

In this section, we use the model with the three X-variables Total Sales per year, Total Sqft, and FTE Managers, even though the two-variable model (excluding FTE Managers) appears superior. The reason for choosing this model is to demonstrate that some tools actually show that a variable such as FTE Managers is not needed in the model. These tools help with finding a good equation, a topic discussed in Chapter 18.