1
Inclass Practice 4
(Bring this to class)
NAME:
_______________________________________________________
BUAD 310
USC ID#
:__________________________________________________________
Friday Lab time
:___________________________________________________
Dr.
ANSARI
PART 1  MULTIPLE CHOICE [5 points each]
1.
If DurbinWatson statistic is close to 0, and then we can conclude that,
a)
There is Heteroscedasticity
b)
There is autocorrelation between predictors
c)
Errors are not
normally distributed
d)
There is autocorrelation between errors
e)
None of the above
2.
In simple regression analysis as the scatter (spread) of points about the regression line becomes
greater, then
a)
Coefficient of determination will increase and error standard deviation will increase.
b)
Coefficient of determination will increase and error standard deviation will decrease.
c)
Coefficient of determination will decrease and error standard deviation will increase.
d)
Coefficient of determination will decrease and error standard deviation will decrease
e)
Coefficient of determination and error standard deviation will not be affected.
3.
In regression analysis, the
F
ts
, is given by,
a)
MSE/MSR
b)
(SSR/SSE)*(DFR/DFE)
c)
MSR/SSR
d)
(SSR/SSE)*(DFE/DFR)
e)
None of the above
4.
A consultant wanted to investigate the relationship between the price of a car (Y) and its age (X).
He performed simple regression on a sample of 50 cars. Which of the following would suggest a
significant relationship between X and Y ?
a)
Small pvalue for the estimated slope
b)
Small test statistic for the estimated slope
c)
Small pvalue for the estimated intercept
d)
Small test statistic for the estimated intercept
e)
None of the above
5.
In a simple regression model in which the data depends on time, a visual check of the assumption
that the values of the residuals are uncorrelated may be made by graphing:
a)
The residuals against the predicted values of y.
b)
The predicted values of
y against index.
c)
The residuals against the observed values of
y.
d)
The residuals against index.
e)
The normal plot of residuals.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document2
PART II
1.
A real estate broker who specializes in selling farmland in a large western state is interested in
developing a pricing formula of some type. He feels he could increase his business greatly if he could
accurately determine the value of a farmer’s land. The real estate broker hired a statistician to do
This is the end of the preview.
Sign up
to
access the rest of the document.
 Summer '07
 Lv
 Business, Normal Distribution, Regression Analysis, Errors and residuals in statistics, residuals, simple regression model, Regression Residual Error

Click to edit the document details