STA 4504/5503
Categorical Data Analysis
Sample Exam 1
Spring 2011
1. Indicate whether each of the following is true (T) or false (F).
(a)
T
In 22 tables, statistical independence is equivalent to a population odds ratio
value of = 1.0.
(b)
F
A British stu
STA 4504/5503
Sample questions for exam 2
Courtesy of Alan Agresti
1. True-False questions.
(a)
For General Social Survey data on Y = political ideology (categories liberal, moderate, conservative), X1 = gender (1 = female, 0 =
male), and X2 = political p
STA 4504/5503
Categorical Data Analysis
Sample Exam 1
Spring 2011
1. Indicate whether each of the following is true (T) or false (F).
(a)
In 22 tables, statistical independence is equivalent to a population odds ratio
value of = 1.0.
(b)
A British study r
ASSIGNMENT # 1 UNGRADED PORTION
Note: For each test of hypotheses that you do ( unless otherwise indicated ), you must include H0, Ha, calculated
test statistic, calculated rejection region, calculated P-value, and a conclusion written completely in terms
COMPARING PROPORTIONS IN 2 X 2 TABLES
Example 1: Cohort Study Western Electric employees longitudinal study over 10 years.
Does working under tension increase the risk of coronary heart disease (CHD) ?
X = Do you work
under tension ?
Yes
No
Y = Major Coro
CHI-SQUARE GOODNESS-OF-FIT TEST
The Chi-Square Goodness-of-Fit Test is based on the multinomial distribution and is used to test whether or not
the population being sampled is of a particular type. It works best for testing for discrete distributions. The
TABLE STRUCTURE FOR TWO DIMENSIONS
Case 1: Two categorical variables, X and Y, which are not in a response-explanatory type relationship, i.e.,
we are interested in seeing if the two variables are related in some way but there is no real reason to think o
Chapter 1: Introduction
1.1 Categorical Response Data
Methods for response variable (a.k.a. outcome variable, dependent
variable) Y whose measurement scale is a set of categories.
Explanatory variables (a.k.a. predictors, covariates, independent
variables
COMPLETELY RANDOMIZED DESIGN EXAMPLES
Higher wholesale beef prices over the past few years have resulted in the sale of ground beef with a higher
fat content in an attempt to keep retail prices down. Four different supermarket chains were examined, and
fo
ASSIGNMENT # 2 GRADED PORTION
Due by class on Thursday, January 26,
Be sure to read the questions carefully and answer each part that you are asked.
1.
Problem 2.6 on page 56 in the text. Dont forget to interpret in terms of the problem.
2.
Problem 2.10,
ASSIGNMENT # 2 UNGRADED PORTION
Be sure to read the questions carefully and answer each part that you are asked.
1.
Problem 2.4 on page 56 in the text. Dont forget to comment.
2.
Problem 2.8 on pages 56 57 in the text.
3.
Consider the data below taken fro
MEASURES OF ASSOCIATION AT LEAST ONE VARIABLE NOMINAL
Measure of Association: A statistic that quantifies the strength of the dependence ( or relationship ) between
two categorical variables.
Measure of Association: This can be used for any type of categ
ASSIGNMENT # 1 GRADED PORTION
Due by class on Tuesday, January 17.
Note: For each test of hypotheses that you do ( unless otherwise indicated ), you must include H0, Ha, calculated
test statistic, calculated rejection region, calculated P-value, and a con
INFERENCE FOR THE BINOMIAL PARAMETER
Likelihood Function: the probability of the observed data, expressed as a function of the parameter
Example:
Consider a binomial variable with n = 5 and unknown
Suppose we observe Y = 5 then P( 5 ) =
5!
5 ( 1 - ) 5-5
SYLLABUS STA 4504 / STA 5503
CATEGORICAL DATA ANALYSIS ( 4504 ) SECTION 7516
CATEGORICAL DATA METHODS ( 5503 ) SECTION 7519
Term:
Spring 2017
Professor:
Dr. David Groggel
Graduate Assistant:
Reza Sadeghi
Office Hours: Dr. Groggel
Griffin-Floyd Hall
Room 1
Chapter 1: Introduction
1.1 Categorical Response Data
Two Types of Categorical Variables
Nominal: unordered categories
Ordinal: ordered categories
Methods for response variable (a.k.a. outcome variable, dependent
variable) Y whose measurement scale is a s
Chapter 1: Introduction
1.1 Categorical Response Data
Methods for response variable (a.k.a. outcome variable, dependent
variable) Y whose measurement scale is a set of categories.
Explanatory variables (a.k.a. predictors, covariates, independent
variables
Chapter 1: Introduction
1.1 Categorical Response Data
Methods for response variable (a.k.a. outcome variable, dependent
variable) Y whose measurement scale is a set of categories.
Explanatory variables (a.k.a. predictors, covariates, independent
variables
STA 4504/5503
Categorical Data Analysis
Spring 2012
Course Information
Time and Location:
Tues 8:30 10:25 a.m. (per. 23)
Thur 9:35 10:25 a.m. (per. 3)
100 Grin-Floyd Hall (FLO)
Instructor: Dr. Brett Presnell
Oce: 225 FLO
E-mail: [email protected]
Oce
Formulas for exam 1:
Binomial P (y ) =
n!
y (1 )ny , y = 0, 1, 2, ., n,
y !(n y )!
n1+
n11
Hypergeometric P (n11 ) =
n2+
n+1 n11
n
n+1
Poisson E (Y ) = ,
odds = /(1 ),
V ar (Y ) =
(nij ij )2
,
ij
n11 n22
=
n12 n21
ij = (ni+ n+j )/n,
nij
,
ij
G2 = 2nij
STA 4504/5503
Sample test questions for Exam 1
Name:
1. For the following statements, answer true(T) or false(F).
a.
In 22 tables, statistical independence is equivalent to a population
odds ratio value of = 1.0.
b.
A British study reported in the New Yor
STA 4504/5503
Sample questions for exam 2
1. True-False questions.
(a)
For General Social Survey data on Y = political ideology (categories liberal, moderate, conservative), X1 = gender (1 = female, 0 =
male), and X2 = political party (1 = Democrat, 0 = R
TESTING INDEPENDENCE
H0: X and Y are independent variables
Ha: X and Y are dependent variables
We test these hypotheses using the goodness-of-fit
approach, i.e., find expected values assuming H0
true and compare them to the observed values.
Full Multinomi