Previous Chapter Chapter 5: Regression Analysis: Assumptions and Diagnostics Next Chapter
Chapter 5: Regression Analysis: Assumptions and Diagnostics
As shown in the previous chapter, ordinary least squares (OLS) regression links the values of dependent variable Yi(i = 1, 2,…, n) to the values of a set of independent variables Xik by means of a linear function and an error term εi:
where k ranges from 0 to p-1. This model thus contains p regression parameters (namely effects of p −1 predictors and one intercept: X' equals 1 for all cases). The linear function is called the linear predictor or the structural part of the model, while the error term is the random or stochastic component of the model. In general, regression analysis can be used for two purposes: (1) to describe the ...