hypothetical question

[student_786]

Hi. I have just commenced study of econometrics and have a few questions.

We were presented with an ungraded homework exercise- in which I am struggling to find an answer to, and I was hoping someone could provide some assistance. We dont have to answer it, but out of genuine interest, i would love to know how to approach such an issue.


The question is as follows:

Data obtained on test scores and student-teacher ratios in region A and region B. Region B, on average, has lower student-teacher ratios than region A.
The following regression is run:

Yi=B0 + B1X1i +B 2X2i +B3X31+ui
where Xi is the class size in region A, X2 is the difference in class size between region A and B, and X3 is the class size in region B.

The regression package shows a message indicating that it cannot estimate the above equation. What is the problem here and how can it be fixed?



Thanks in advance to anyone kind enough to shed some light on this question.
regards

[startz]

Hint: This is a question about multicollinearity.

[student_786]

Hint: This is a question about multicollinearity.

Thanks startz. Since i have only just commenced studies in this subject area, i'm still struggling to get my head around the application of concepts.

Would you mind explaining briefly how the problem could be fixed?

My initial thought would be to exclude a particular variable from the model since one of the variables may exacerbate the multicolliniarity problem.
But if that is the case, how do you best determine which variable to exclude?

Thanks again in advance.
regards

[startz]

Hint: This is a question about multicollinearity.

Thanks startz. Since i have only just commenced studies in this subject area, i'm still struggling to get my head around the application of concepts.

Would you mind explaining briefly how the problem could be fixed?

My initial thought would be to exclude a particular variable from the model since one of the variables may exacerbate the multicolliniarity problem.
But if that is the case, how do you best determine which variable to exclude?

Thanks again in advance.
regards

You're actually being asked a "trick" question, although a very good one. There are only two variables to be manipulated (the two regional variables), but there are three coefficients. Any one of the coefficients is redundant. For a good exercise, write out the equation three times eliminating one variable in each, and ask yourself how to predict the effect of a change in the first the regional variables in each case.