(Data file: Transact) The data in this example consists of a sample of branches of a...

Transcribed Image Text:

(Data file: Transact) The data in this example consists of a sample of branches of a large Australian bank (Cunningham and Heathcote, 1989). Each branch makes transactions of two types, and for each of the branches we have recorded the number t1 of type 1 transactions and the number t2 of type 2 transactions. The response is time, the total minutes of labor used by the branch. Define a = (t1 + t2)/2 to be the average transaction time, and d=t1- t2, and fit the following four mean functions M1: E(timelt 1, t2)= Bot+But1+B₂t2 M2: E(timelt 1, t2)=B02 + B32a + Bad M3: E(timelt 1, t2)=B3+Bast2+ B₁3d M4: E(timelt 1, t2)= Bos + B₁at1+ Bat2+Ba+Bud 4.2.1 In the fit of M4, some of the coefficients estimates are labeled as "aliased" or else they are simply omitted. Explain what this means and why this happens. 4.2.2 What aspects of the fitted regressions are the same? What aspects are different? 4.2.3 Why is the estimate for t2 different in M1 and M3? (Data file: Transact) The data in this example consists of a sample of branches of a large Australian bank (Cunningham and Heathcote, 1989). Each branch makes transactions of two types, and for each of the branches we have recorded the number t1 of type 1 transactions and the number t2 of type 2 transactions. The response is time, the total minutes of labor used by the branch. Define a = (t1 + t2)/2 to be the average transaction time, and d=t1- t2, and fit the following four mean functions M1: E(timelt 1, t2)= Bot+But1+B₂t2 M2: E(timelt 1, t2)=B02 + B32a + Bad M3: E(timelt 1, t2)=B3+Bast2+ B₁3d M4: E(timelt 1, t2)= Bos + B₁at1+ Bat2+Ba+Bud 4.2.1 In the fit of M4, some of the coefficients estimates are labeled as "aliased" or else they are simply omitted. Explain what this means and why this happens. 4.2.2 What aspects of the fitted regressions are the same? What aspects are different? 4.2.3 Why is the estimate for t2 different in M1 and M3?

Answer rating: 100% (QA)

421 Some of the coefficient estimates in the M4 fit are tagged as aliased while others are simply om... View the full answer