click below
click below
Normal Size Small Size show me how
Regression
Significance
Question | Answer |
---|---|
What are the three f tests? | 1) Simple (overall fit of single regression) 2) Wald (irrelevant variable test) 3) Chow (tests for structural differences – compares beta hat's of regressors) |
What does the f-value measure? | Measures the significance of the overall fit of the regression (R2) |
What three items are needed to calculate critical f-value? | 1) Degrees of freedom for numerator = number of independent variables? 2) Degrees of freedom for denominator = ? 3) Significance level (generally 5% or 1%)? |
What does the Simple F-test tell you? | 1) Used to test the significance of the overall fit of a single regression model. 2) Ratio of what you’re explaining (ESS) to what you’re not (RSS), controlled for degrees of freedom. |
What does the Wald F-test used for? | Used to determine the impact of adding an additional variable (helps eliminate variables that do not meet a certain level of significance). GOOD FOR TESTING FOR IRRELEVANT VARIABLES |
What is the Chow F-test used for? | 1) Used to test for structural changes in a regression model – ie: beta hats of regressors differ between time periods or across locations. 2) Null always assumes no structural changes. |