STAT 425 Week 03 - Multiple Linear Regression - Part 3

This video covers hypothesis testing for coefficients in the linear model, based on Gaussian error assumptions. We use coefficient t-tests for individual explanatory variables, and F-tests for groups of explanatory variables. The F test framework enables us to test between a reduced model and full model very generally, and includes the so-called F test for the regression, ANOVA F tests, and partial F-tests for subset models. We'll discuss the distribution theory behind the tests and their implementation in R.