HYPOTHESIS TESTING IN MULTIPLE REGRESSION

 

HYPOTHESIS TESTING IN MULTIPLE REGRESSION

Multiple linear regression involves two types of hypothesis testing. They are;

1.      Hypothesis testing about overall significance

2.      Hypothesis testing about individual parameter.

Hypothesis Testing about Overall Model Significance

Consider a multiple linear regression model with “p” predictors:

In ANOVA, the total variation in the dependent variable is partitioned into two parts. One part is called explain variation (the variation explained by regressors) and un explain variation (the variation which is due other sources accommodate in residual but not included in the model.)

ANOVA Table

Hypothesis testing about individual Slope

Consider a multiple linear regression model with “p” predictors:

Practice Question 

Given the data:

Solution: Apply OLS method on deviated form

The estimated multiple regression model in deviated form:

• Read More: General Linear Regression