Final Exam Practice–ANOVA and Logistic Regression
General Comments:
•
This file contains some practice problems on ANOVA and logistic regression. For practice on
contingency tables, regression, etc. see the practice files from our first two midterms. I posted
so many problems then that I don’t think it will particularly help to post more!
However
if there is a topic you feel short of practice on, let me know. You should also look over my
solutions to our midterms 1 and 2 as I have a tendency to ask again about things that the
class as a whole did poorly on
....
•
I don’t guarantee that these problems represent all possible questions I could ask on ANOVA
and logistic regression though I do think they are useful in giving you a sense of how I might
ask questions. As always, I feel that our homework assignments are the best representation
of what I feel is important for our class. In this case HW7 is the key. I have included the
warmup problems from that assignment here for reference.
ANOVA
(1) ANOVA basics (courtesy of Prof. Afifi):
A partly filled in ANOVA table is shown below:
Source
df
SS
MS
F
Between (Treatment)
3
16.42
Within (Error)
102.50
Total
31
(a)
Fill in the blanks in the table, and explain what each of the quantities SSB, SSW, MSB, MSW
and F tells you.
(b)
How many treatment groups are there? Briefly explain your reasoning.
(c)
A “completely randomized design” is an approach for carrying out an experiment where the
treatments are assigned to experimental units (subjects) completely at random. A “balanced de
sign” is one in which the number of subjects in each treatment group is the same for all treatments.
Is it possible that this table was derived from a balanced completely randomized design¿ Explain
your reasoning.
(d)
Obtain an approximate pvalue for a test of significance of the hypothesis that all of the treat
ment means are equal. In particular, would you conclude that result is significant at the
α
=
.
05
level? (Note: This can be done using an F table or using STATA.)
(2) Protein Intake (based on Rosner 12.112.5)
Researchers compared protein intake among
three groups of postmenopausal women: (1) women eating a standard American diet (SRD) (2)
women eating a lactoovovegetarian diet (LAC) and (3) women eacting a strict vegetarian diet
(VEG). The mean and standard deviation of protein intake as well as the group sizes are presented
1
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
in the table below. Use then to answer the following questions:
Group
Mean
SD
Number in group
STD
75
9
10
LAC
57
13
10
VEG
47
17
6
(a)
Perform an overall F test to determine whether there is a significant difference in mean protein
intake between the three groups. Be sure you an state the null and alternative hypotheses both
mathematically and in words, compute the test statistic by hand, get an approximate pvalue, and
explain your realworld conclusions.
This is the end of the preview.
Sign up
to
access the rest of the document.
 Winter '07
 Sugar
 Statistics, Statistical hypothesis testing

Click to edit the document details