Practical Data Analysis with JMP, Second Edition, 2nd Edition

16 Residuals Analysis and Estimation

Overview

Conditions for Least Squares Estimation

Residuals Analysis

Linearity

Curvature

Influential Observations

Normality

Constant Variance

Independence

Estimation

Confidence Intervals for Parameters

Confidence Intervals for Υ|X

Prediction Intervals for Υ|X

Application

Overview

Regression analysis is a powerful and flexible technique that we can reliably apply under certain conditions. Given a data table with continuous X-Y pairs, we can always use the ordinary least squares (OLS) method to fit a line and generate results like those we saw in Chapter 15. However, if we want to draw general conclusions or use an estimated regression equation to estimate, predict, or project Υ values, it is critical that we ...