Stat/ME 424, Fall 2009
First Midterm Examination
(Total : 100 points)
•
This is a closed book exam, except for four sides of notes.
•
You are allowed to use calculators.
•
There are four problems in this exam.
•
Please put your answers in a blue book with your name on the cover.
Problem 1
(20 pts)
Questions (a)(c) involve the analysis of some data on alcohol consumption. The data were
collected in 1970. The variable
permale
is the percentage of males living in a state. Data were
collected from each of the 50 states of USA and Washington DC. The purpose of the analysis
is to see how alcohol consumption is related to the other variables. A simple linear regression
model was fitted trying to predict the alcohol consumption using the percentage of males and the
following partial output was obtained:
Coefficients:
Value
Std. Error
t
Value
p
value
(Intercept)
93.4260
207.8126
0.4496
0.6550
permale
4.2191
4.0777
1.0347
0.3059
Residual standard error: 32.05 on 49 degrees of freedom
Multiple RSquared: 0.02138
(a)
Suppose we replaced the percentage of males by the percentage of females in the above
regression model. Give the regression coefficients for this new model. Would the standard
error for the slope change?
(b)
In a regression problem, what (if any) is the unit of the
t
statistics?
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '10
 SCHNEIDER
 Linear Regression, Regression Analysis, Standard Deviation, Variance, Errors and residuals in statistics, cuckoo eggs

Click to edit the document details