BioStat 571 Handouts

Tentative Outline of Lectures

Summary of Lecture -- 2007

• Week 1 : Examples from lecture
  • 03 Jan 2007 Notes: CorrData-Intro.pdf --- Intro / Overview
  • 05 Jan 2007 Notes: GLM.pdf --- Generalized Linear Model Review
  • seizure.dat --- Example data
  • seizure.q --- Example of GLM analysis S+ code
  • seizure.results --- Example output
  • seizure2.q --- Example GLM analysis with scale parameter for overdispersion

• Week 2 : Examples from lecture
  • 10 Jan 2007 Notes: MultModels.pdf --- multinomial regression models
  • wisc.data --- Example data
  • wisc.doc.txt --- Example documentation
  • wisc_newCDA.q --- Example of POM analysis S+ code
  • wisc_newCDA.out --- Example output
  • wisc_CRM.q --- Example of CRM analysis S+ code
  • wisc_CRM.out --- Example output

• Week 3 : Examples from lecture
  • 19 Jan 2007 Notes: OverdispQL.pdf --- Overdispersion and Quasilikelihood
  • rat_teratology.data --- Example data
  • rats.q --- Example of analysis that accounts for overdispersion S+ code
  • rats.out --- Example output
  • seizure3.q --- Example of analysis that accounts for overdispersion S+ code
  • seizure3.out --- Example output
  • negbin.q --- Negative Binomial Regression function from MASS library

• Week 4 : Examples from lecture
  • weil.data --- Example data - described in Liang and Hanfelt (1994) Biometrics. Also analyzed and discussed in Heagerty and Zeger (2000) Statisical Science
  • weil_eda_571.q --- Example of weighted analysis EDA S+ code
  • weil_eda_571.out --- Example output
  • weil_QL.q --- Example of analysis that accounts for overdispersion S+ code
  • weil_QL.out --- Example output
  • bod_regn.q --- S+ program for paired estimating equation analysis of overdispersed data
  • bod_regn_help.ps --- Postscript documentation for S+ program

• Week 5 : Examples from lecture
  • 29 Jan 2007 Notes: LM-WLS+EDA.pdf --- General Linear Model for Correlated Data -- WLS and EDA
  • variogram.q --- S+/R function to compute within-subject differences in order to plot variogram.
  • carot-baseline.dat --- Example data - described in Xu et al. (1992) JNCI
  • carot-baseline.txt --- Documentation
  • carot-longitudinal.dat --- Example data - described in Xu et al. (1992) JNCI
  • carot-longitudinal.txt --- Documentation
  • carot-combo.q --- S+ code used to create analysis data
  • cd4.data --- CD4 data
  • cd4.txt --- CD4 data
  • cd4_eda.q --- S+ code used for EDA

• Week 6 : Examples from lecture
  • 07Feb2007 Notes: LinearMixed571.pdf --- updated 15Feb2007
  • NewCFkids.data --- CF data
  • NewCFkids.txt --- documentation
  • cfkids-read.q --- Example data entry/structure program
  • cfkids-CDA-NewLMM.q --- S+ program for linear mixed model analysis
  • cfkids-CDA-NewLMM.out --- S+ output
  • 14 Feb 2007 Using SAS: --- There is SAS available on the UW UNIX machine "mead" -- see C&C UNIX Account Information --- If X-Win32 is installed on your computer/laptop then you can use SAS with its GUI. Just run X-Win32 first, then log in to mead and run SAS by typing `sas &' in the command line. The '&' means that you run the program in background. It enables you to type other commands in command line without terminating SAS session. Note: we will be able to use R/S+ for this class so SAS is not needed (optional).
  • NewCFkids-SAS.data --- data
  • cfkidsML.sas --- SAS program
  • cfkidsML.lst --- SAS output
  • cfkids-CDA-gee.q --- S+ program for weighted least squares analysis using gee(). Provides both model based standard errors and sandwich standard errors.
  • cfkids-CDA-gee.out --- S+ output
  • cows.data --- Example data - described in DLZ - Example 5.2, pages 100-103
  • cows.plot.q --- S+ program for EDA
  • cows.fit-gee.q --- S+ program for WLS estimation using GEE
  • cows.fit-gee.out0 --- results for WLS estimation with OLS beta and sandwich variance
  • cows.fit-gee.out1 --- results for WLS estimation with Exchangeable weighting and sandwich variance
  • cows.gee-test.q --- S+ program for testing of the quadratic terms
  • cows.gee-test.out --- testing results

• Week 7 : Examples from lecture
  • 16Feb2007 ----- StatisticalIssues.doc --- a brief report written for a general medical audience that explains the use of empirical Bayes methods for the performance evaluation of Hospitals.
  • Midterm Winter 2000

• Week 8 : Examples from lecture
  • 21Feb2007 ----- CatGEE.pdf --- notes regarding GEE and extensions.
  • hivnet-DataSetup.q --- Example data entry/structure program
  • HivnetWide.dat --- Informed Consent data
  • hivnet-CDA-gee.q --- Informed Consent data analysis using gee()
  • hivnet-CDA-gee.out --- Informed Consent data analysis results
  • hivnet-CDA-geese.q --- Informed Consent data analysis using geese()
  • hivnet-CDA-geese.out --- Informed Consent data analysis results
  • 23Feb2007 ----- ChagantyJoe-JRSS2004.pdf --- article that discusses GEE efficiency for binary data.

• Week 9 : Examples from lecture
  • 28Feb2007 ----- CatGLMM.pdf --- notes regarding generalized linear mixed models.
  • hivnet-CDA-CondLike.q --- Conditional likelihood analysis of IC data
  • hivnet-CDA-CondLike.out --- Informed Consent analysis using conditional logistic regression
  • hivnet.sas --- Informed Consent data analysis using GENMOD and NLMIXED in SAS
  • HivnetIC-SAS.data --- Informed Consent data for input into SAS
  • hivnet.lst --- Informed Consent data analysis results
  • hivnet-CDA-glmm.q --- Informed Consent data analysis using GLMM()
  • hivnet-CDA-glmm.out --- Informed Consent data analysis results

• Week 10 : Examples from lecture
  • 09Mar2007 ----- summary.pdf --- a summary of the second half of the course. THANKS!

  BioStat 571	Handouts	Homework	Data