Lecture 2 (Thurs., Jan. 12): Sections 2.1, 2.2.
HW Assignment 1, Due Thurs., Jan. 19: 1.1, 1.2, 1.12, 1.14 (hint: write (n_1,...,n_c) as a sum of independent mult(n=1;pi_1,..pi_c)), 2.7
Lecture 3 (Tues., Jan. 17): Sections 2.2, 2.3.
Lecture 4 (Thurs., Jan. 19): Section 2.4 (mainly gamma), Section 3.1, confidence intervals for the odds ratio, relative risk, and risk difference via large sample methods (delta theorem).
HW Assignment 2, Due Thurs., Jan. 26: 2.2, 2.8, 2.12, 2.35, 3.1, 3.22, 3.24 (note typo: refer to problem 2.36)
Lecture 5 (Tues., Jan. 24: Delta Theorem continued.
Lecture 6 (Thurs., Jan. 26): Section 3.2,tests in I x J tables, We will skip most of 3.3, just read 3.3.1 on Pearson residuals, Section 3.4, CMH statistics.
HW Assignment 3, Due Thurs., Feb. 2: 3.4(a only), 3.11: a, skip b, for c, compute CMH statistics in SAS and interpret, add d give a c.i. for gamma, 3.30 (note that y1=n11 and y2=n21), 3.34, 3.13
Lecture 7 (Tues., Jan. 31): Section 3.5, Fisher's exact tests.
Lecture 8 (Thurs., Feb. 2): Section 3.5, exact unconditional tests.
HW Assignment 4, Due Thursday, Feb. 9:
Lecture 9 (Tues., Feb. 7): Section 3.5 and 3.6, exact tests for I x J tables, exact confidence intervals.
Lecture 10 (Thurs., Feb. 9): Begin logistic regression.
HW Assignment 5, Due Thursday, Feb. 16. 6: 4.1, 4.4, 5.2, 5.4 (plot by grouping age on age values or by using gam in r, use data), 5.33a,b
Lecture 11 (Tues., Feb. 14): Sections 5.1 and 5.2.
Lecture 12 (Thurs., Feb. 16): Section 5.3
HW Assignment 6, Due Thursday, Feb. 23: 5.7, 5.10, 5.17, 5.37a-c, 5.42
Lecture 13 (Tues., Feb. 21): Section 5.4: Multiple Logistic Regression.
Lecture 14 (Thurs., Feb. 23): Section 6.1. Model selection.
Mid-Term Exam - Tuesday, Feb. 28, 6-9 pm, 307 Mann Hall.
Lecture 15 (Thurs., Mar. 2): Model selection continued
Spring Break, March 6-10.
Lecture 16 (Tues., Mar. 14): Section 6.3. Pearson residuals, standardized Pearson residuals (6.2.1), dfbetas, generalized R^2 for logistic regression (6.2.5), ROC curves.
Lecture 17 (Thurs., Mar. 16: Inference for conditional association, K 2 by 2 tables, Mantel-Haenszel inference.
HW Assignment 7, Due Thursday, Mar. 23: 6.2 (details), 6.7a-d (details), 6.8.
Lecture 18 (Tues., Mar. 21):Section 6.7, Conditional Logistic Regresssion.
Lecture 19 (Thurs., Mar. 23):Section 6.4, 6.5. Power comparisons and sample size calculations.
HW Assignment 8, Due Thursday, Mar. 30:
6.7e(using cond. logistic reg.), 6.19c,
6.10, 6.26.
No Class (Instructor at ENAR), Tuesday, Mar. 28.
Lecture 20 (Thurs., Mar. 30): Sec. 6.6 and Introduction to bioassay (not in Agresti).
HW Assignment 9, Due Thursday, April 6: Problem 6.16 Note: you fit the log-log link by fitting y1=n-y (failures) to the complementary log-log link model. Also, what is the ED50 in terms of alpha and beta in the case of the log-log link (extreme value distribution for T)?
E4. Using the complementary log-log link (Gompertz cdf), find the ED50 of the beetle data (Table 6.14). Just leave it in the log dose scale. Also give a 95% confidence interval using the program ed50.pgm. You will need to put in .3665+intercept for the intercept and the covariance matrix from the genmod output using the covb option.
Lecture 21 (Tues., April 4): Return to Ch. 4 and review GLM approach. Then fit Poisson GLM models.
HW Assignment 9, Due Tues., April 18: 4.6,4.12,4.14 (add c. using GEE with the repeated statement in genmod, but see hints), 4.18 (give theta, b(theta), and derive E(Y) and Var(Y) from b(theta))
Lecture 22 (Thurs., April 6): Section 7.1: Models for nominal multinomial responses.
Lecture 23 (Tues., April 11): Sections 7.2 and 7.3: Models for ordinal multinomial responses.
Easter Break - No Class, Thurs., April 13.
HW Assignment 10, Due Thursday, April 20: 7.4 (data),7.8, 7.14 (use latent variable interpretation), 7.18.
Lecture 24 (Tues., April 18):Section 10.1: comparing dependent proportions, McNemar's test.
Lecture 25 (Thurs., April 20):10.2: Models for dependent pairs, 10.3: test of means for paired ordinal data (not in Agresti except for prob. 10.38 and testing for marginal homogeneity, 10.6: Bradley-Terry paired comparison logit models. measures and ordered categorical data.
HW Assignment 11, Not to be handed in, answers to be sent out April 28: 10.3a, 10.16
Lecture 26 (Tues., April 25): Section 11.3: GEE.
Lecture 27 (Thurs., April 27: Consulting problem with repeated measures and ordered categorical data.