7 7th Turorial
7.1 Recap
Previously we have been discussing multivariate linear regression that have the formula Y=Xβ+ϵ. Now assume we have two linear model M1 and M2 where M2 is a simplification of M1 such as (M1:yi=β0+β1 x1i+β2 x2i+β3 x3i and M2:yi=β0+β1 x1i) A question arises here is which model is preferred and that is equivalent to test
H0=β2=β3=0 vs H1=β2≠0,β3≠0
The decision rule is to reject H0 if F=(D2−D1)/qD1/(n−p)>Fq,n−p,α
where n is the number of observations, p is the number of parameters in M1 and q is the number of parameters fixed to reduce M1 to M2. For the example above, p=4 and q=2, and D1 and D2 are the SSR of M1 and M2 respectively. Equivalently, we could use the notation SSE(...). For instance SSE(X1,X2,X3) denotes the sum of squared error for a multiple linear regression that includes X1, X2 and X3 to draw the model.
7.2 Exercises
Exercise 7.1 Download the csv file for the dataset here
- Estimate the liner model for the given data and interpret its coefficients.
- Discuss the efficiency of the model by two different approaches.
- Write the ANOVA table that factorize the sum square regression X1 and X2 given X1.
- Use partial F to test whether you can remove X2 from model.
- Calculate R2 , r2Y,2.1 , rY,1.2 and r2Y,2
- Estimate the corresponding standard model and discuss its coefficient.