STAT22600 Winter 2017

Statistics 22600 = Health Studies 32600
Analysis of Categorical Data, Winter 2017
TuTh 1:30-2:50 PM, Eckhart 133

Textbook (Required)

Complete Syllabus: Click Here

Primary Course Website

Lecture Slides

Slides Topics Text Coverage
C01.pdf Likelihood and maximum likelihood estimation
Inference for a binomial proportion: Wald tests, score tests, likelihood ratio tests, exact binomial tests and the corresponding confidence intervals
Lab Activity: Comparing the coverage probability of Wald, score and Agresti-Coull confidence intervals
Chapter 1
C02A.pdf Diff of proportions, relative risk, odds ratio 2.1-2.3
C02B.pdf Pearson's chi-squared test of independence 2.4
C02C.pdf Fisher's exact test for small samples 2.6
C02D.pdf Association in three-way tables: partial tables and marginal tables, conditional and marginal associations, conditional and marginal odds ratios, conditional and marginal independence, homogeneous association, the Cochran-Mantel-Maenszel test for conditional independence Mantel-Haenszel estimate for the common odds ratio 2.7, 4.3.4
C03.pdf Overview of generalized linear models (GLM).
Data file: falls.dat, fallsUG.dat
3.1-3.2, 3.4
C04A.pdf Simple logistic regression: interpretation and inference.
Data file: horseshoecrabs.dat
4.1-4.2
C04B.pdf Multiple logistic regression 4.4
C04C.pdf
Multiway.pdf
Logistic regression with categorical predictors.
Logistic regression for multiway contigency tables
Data file: mousemuscle.dat
4.3
C05.pdf Model selection; Model checking (deviance, residuals); Watch out for sparse categorical data 5.1-5.3
C06.pdf Multicategory logit models:
Baseline-category logit models for nominal responses,
Cumulative logit models for ordinal responses
Lab Activity: Analysis of the wheat kernels data
6.1, 6.2
Poisson.pdf GLM for count data: Poisson regression, overdispersion and negative binomial regression
Data file: traincollisions.dat
3.3
C07.pdf Loglinear models for two-way and three-way tables; inference for loglinear models; the loglinear-logistic connection (Covered in 2016 but not in 2017) 7.1-7.3
C08.pdf Models for matched pairs: McNemar test, logistic regression for matched-pair data, population-avaraged and subject-specific models 8.1-8.2


Last Update: 6/28/2018