1.
{20 pts] Circle T {True} or F {False} for each of the following statements.
{1}.
(2}.
{a}.
(4}-
(5}.
[5}-
[7}-
{a}.
{9}.
{m}.
{2 pts} [ T ] Patient condition (good. fair. serious. critical] is an ordinal categorical variable .
{2 pts} [ T ] The Fisherâ€˜
STA 617 Chp8 Loglinear Models
1
STA617
Advanced Categorical Data Analysis
Instructor:
Changxing Ma
Department of Biostatistics
716 Kimball, University at Buffalo
Phone: (716) 829-2758
Email: [email protected]
Days, Time:
M W, 9:00 AM - 10:20 AM
Dates:
08
STA 617 Chp9 Loglinear/Logit Models
Impact Model
http:/mch.peds.ufl.edu/calculator/impactmodel.pdf
http:/mch.peds.ufl.edu/research/techreports/2003_200
4/risk_factors_associated_with_outcomes_for_medicaid.
pdf
1
STA 617 Chp9 Loglinear/Logit Models
Impac
STA 617 Chp9 Loglinear/Logit Models
9.7 Poisson regressions for rates
In Section 4.3 we introduced Poisson regression for
modeling counts. When outcomes occur over time,
space, or some other index of size, it is more relevant
to model their rate of occur
Homework 1
Name_
STA 617: Adv Categorical Data
Analysis
(25 Points in total)
8.6 (8 points)
a. From G2=31.6695 with DF=48, the p-value must be > 0.05 (actually p=0.9). The model is not lack of
fit. But we might be able to further simplify this model (back
Homework 2
Name_
STA 617: Adv Categorical Data
Analysis
(25 Points in total)
9.2 (8 points)
a. Explain why a loglinear model should include the
AS term.
Answer: Since A and S are explanatory variables, a model should contain the AS term, which
forces the
STA 517 Chp4 Introduction to Generalized Linear Models
4.3 GENERALIZED LINEAR MODELS
FOR COUNTS
count data - assume a Poisson distribution
counts in contingency tables with categorical response
variables.
modeling count or rate data for a single discre
STA 617 Chp10 Models for matched pairs
10.4 Symmetry, Quasi-symmetry and
Quasi-independence
1
STA 617 Chp10 Models for matched pairs
2
STA 617 Chp10 Models for matched pairs
3
STA 617 Chp10 Models for matched pairs
4
STA 617 Chp10 Models for matched pairs
STA 617 Chp10 Models for matched pairs
10.5 Measurement Agreement
between Observations
1
STA 617 Chp10 Models for matched pairs
2
STA 617 Chp10 Models for matched pairs
3
STA 617 Chp10 Models for matched pairs
4
SAS code
STA 617 Chp10 Models for matched p
Name:_
STA617:AdvancedCategoricalData
Due Wednesday, Oct 21st (50 Points)
This is an observational cohort study of infants born in the state of Florida in a given year
(depends on your dataset). It is an exploratory epidemiological evaluation of a seconda
STA 617 Chp11 Models for repeated data
1
11.3 MARGINAL MODELING:
Generalized Estimating Equation (GEE)
At each combination of predictor values, ML fitting
assumes a multinomial distribution for the IT cell
probabilities for the T observations on an I-cat
STA 617 Chp11 Models for repeated data
Analyzing Repeated Categorical
Response Data
Repeated categorical responses may come from
repeated measurements over time on each
individual
or from a set of measurements that are related
because they belong to th
STA 617 Chp9 Loglinear/Logit Models
9.4 Modeling ordinal associations
The loglinear models presented so far have a serious
limitation - they treat all classifications as nominal.
If the order of a variables categories changes in any
way, the fit is the
STA 617 Chp10 Models for matched pairs
10.3 Marginal Models for Squared
Contingency Tables
1
STA 617 Chp10 Models for matched pairs
2
STA 617 Chp10 Models for matched pairs
Model has 2(I-1) marginal probabilities by I parameters,
so df=I-2 for testing fi
STA 617 Chp10 Models for matched pairs
Summary
Describing categorical random variable chapter 1
Poisson for count data
Binomial for binary data
Multinomial for I>2 outcome categories
Others
Limitation: one parameter only, can be adjusted by
scale pa