Midterm 1 Solutions
Thursday 10/1/09
70 min. Closed book. Calculator allowed. Formula sheets and tables provided.
When asked to explain, comment, or discuss, use complete sentences and answer thoroughly
for full credit. Remember to answer the questions in
Some Review Problems to Work On
Well use the cars data set. These data were recorded in the 1920s and were used to examine the
relationship between how fast a car was going and how much distance it needed to stop.
We have 50 observations and two variables
Some Review Problems to Work On
Well use the cars data set. These data were recorded in the 1920s and were used to examine the
relationship between how fast a car was going and how much distance it needed to stop.
We have 50 observations and two variables
Data Analysis Exam 1
out Thursday 10/1/09
due Thursday 10/8/09 3pm
This exam is a week-long take-home data analysis exam. You are allowed to use your textbook
as well as other reference books you feel you might need. You should use the statistical softwar
Some Review Problems for Multivariate Linear Regression
When asked to explain, comment, or discuss, use complete sentences and answer thoroughly
for full credit. Remember to answer the questions in the framework of the problem (units,
etc); include what y
36-401 Formula Sheet Midterm 2
Some standard errors:
s2 = M SE (X X )1
s2 = M SE Xh (X X )1 Xh
Y
h
s2 h(new) = M SE 1 + Xh (X X )1 Xh
Y
Some model criterion:
Cp =
SSEp
M SE (f ull)
P RESSp =
n + 2p
(Yi Yi(i) )2 =
d2 where di =
i
ei
1hii
AICp = n ln (SSEp
Some Review Problems for Multivariate Linear Regression
When asked to explain, comment, or discuss, use complete sentences and answer thoroughly
for full credit. Remember to answer the questions in the framework of the problem (units,
etc); include what y
36-401 Midterm 2 Solutions
Thursday 11/12/09
75 min. Closed book. Calculator allowed. Formula sheets and tables provided.
When asked to explain, comment, or discuss, use complete sentences and answer thoroughly
for full credit. Remember to answer the ques
Data Analysis Exam 2
out Thursday 11/12/09
due Thursday 11/19/09 3pm
This exam is a week-long take-home data analysis exam. You are allowed to use your textbook
as well as other reference books you feel you might need. You should use the statistical softw
What Factors Help Predict Awarded Damages in Court Cases?
Introduction:
In order to better understand the civil trial system in the United States, particularly the amount of damages
awarded, the Civil Justice Survey of State Courts was undertaken in gener
Data Analysis Final Exam
out Thursday 12/03/09
due Friday 12/11/09 5pm - HARD DEADLINE, NO EXTENSIONS
This exam is a week-long take-home data analysis exam. You are allowed to use your textbook
as well as other reference books you feel you might need. You
36-401 HOMEWORK 3
Due: Thursday 9/17/09 at 3pm
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems, you do not need to han
1
36-401 - Homework 2 - Solutions
Due Thursday September 10, 2009
1. Textbook Problems
(a) Problem 1.6
i. Plot as in Figure 1.6 on page 11 of the book Look at Figure 1.6 on page 11 for an example.
Features I am looking for:
Three lines corresponding to l
36-401 HOMEWORK 4
Due: Thursday 9/24/09 at 2pm; my mailbox
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems; you do not
36-401 HOMEWORK 3 SOLUTIONS
September 15, 2009
(Remember that the R code shown in the solutions is for the benet of those who had problems with the
coding unless the problem specically asks for it, you should not be including your code in your write-ups.)
2.27(a) To test for a negative linear association between Age and Muscle Mass, we want to conduct the
test:
H0 : 1 0
Ha : 1 < 0.
(As before with this data set, 1 is from the regression of muscle mass on age.) Note that this is a one-sided
test because we
1
36-401 - Homework 4 - Solutions
Due Thursday September 24, 2009
1. Problem 2.1
(a) The condence interval for 1 does not include zero, so 1 is signicantly different from zero.
From this, we see that for every million people in each of the 50 marketing di
36-401 HOMEWORK 5
Due: Thursday 10/15/09 at 3pm
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems, you do not need to ha
36-401 HOMEWORK 5 SOLUTIONS
October 15, 2009
1. 2.24abcd
(a) We can get all the relevant information from summary(aov(copier.reg), but need to re-arrange
it to have a table in the format of Table 2.2.
Source of Variation
Regression
Error
Total
SS
76960
34
36-401 HOMEWORK 6
Due: Thursday 10/22/09 at 3pm
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems, you do not need to ha
36-401 HOMEWORK 7
Due: Thursday 10/29/09 at 3pm
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems; do not hand in code u
1
36-401 - Homework 6 - Solutions
Due Thursday October 22, 2009
1. 6.15
(a) EDA: Satisfaction is roughly symmetric with a mean of 61.57 and goes from 26 to 92. Age is
Histogram of Age
6
2
4
Frequency
6
4
0
0
2
Frequency
8
8
10
10
Histogram of Satisfaction
36-401 HOMEWORK 7 SOLUTIONS
November 2, 2009
1. Textbook 7.5
(a) SSR(X2 ), SSR(X1 |X2 ), SSR(X3 |X1 , X2 )
To compute this we re-run the regression analysis in the order needed and create the ANOVA
table.
> data.lm=lm(Satisfaction~Severity + Age + Anxiety
36-401 HOMEWORK 8
Due: Thursday 11/05/09 at 3pm
When asked to explain, comment, or discuss, use complete sentences and answer
thoroughly in the context of the problem for full credit. Use R when appropriate
for the homework problems, you do not need to ha
1
36-401 - Homework 8 - Solutions
Due Thursday November 5, 2009
1. Problem 6.5 bd
b Estimated Regression Function
First I perform the lm() command in R
> data=read.table("CH06PR05.txt")
> summary(lm(data$V1data$V2+data$V3)
Call:
lm(formula = data$V1 data$
Data Analysis 2 Sample Report
Introduction
Using a subset of the data acquired in 1975 and 1976 for the Study of the Ecacy of Nosocomial Infection
Control (SENIC), we study the relationship between properties of particular hospitals such as stang, locatio
Factors Associated with Hospital-Acquired Infections
Introduction:
The Study on the Efficacy of Nosocomial Infection Controls primary objective was to analyze
whether or not infection surveillance and control programs have reduced rates of hospital-acquir
36-401 HOMEWORK 9
Due: Tuesday 12/1/09
When asked to explain, comment, or discuss, use complete sentences and answer thoroughly
in the context of the problem for full credit. Use R when appropriate for the homework
problems; do not hand in code unless req