The following regression equation was calculated for class of 24 CPA II students. Required: a) Calculate the t ratio and the 95% confidence interval for each regression...

      

The following regression equation was calculated for class of 24 CPA II students.
cpaII112019956.png
Standard error (0.0190) (0.034) (0.018)
Where y=students score on a theory examination
x1 = Students rank (from the bottom) in high
school x2 = Students verbal aptitude score
x3 =A measure of students character
Required:
a) Calculate the t ratio and the 95% confidence interval for each regression coefficient.
b) What assumptions did you make in (a) above? How reasonable are they?
c) Which regressor gives the strongest evidence of being statistically discernible?
d) In writing up a final regression equation, should one keep the first regressor in the equation, or drop if?
Why?

  

Answers


Raphael
cpaII112019956i.png
b) The assumptions include:
- Error or residuals are independent and normally distributed for a given value of x.
- Expected value of error is equal to zero.
- Variance of errors is the same for all x's.
These assumptions are set up to enable one to come up with a projection of the population from the
sample. So they are reasonable.
c) X1 gives the strongest evidence of being statistically discernible because the t statistic calculated is within
the required range.
d) The decision to keep or drop the first regressor will be based not only on t-test, but also looking at the r2 and standard error of the regression in general. The main objective is to include the regressor that
reduces standard error of regression and r2 value is large. Other than just having the t test alone. In this case since t calculated is within the required range and standard error of regression is low, then it will be appropriate to include the first regressor x1 in the final regression
raphael answered the question on January 1, 2019 at 19:02


Next: Advantages that hard disks have over diskettes
Previous: Does finding a no linear relationship between two variables mean no relationship?

View More CPA Quantitative Analysis Questions and Answers | Return to Questions Index


Exams With Marking Schemes

Related Questions