Software for IBC2004 Satellite Workshop
Sydney, Australia, July 2004

Logistic Regression Analysis of Two Phase Data

S-Plus functions, datasets and documentation for the analysis of two-phase data with binary outcome as described by N.E. Breslow and N. Chatterjee, ``Design and analysis of two-phase studies with binary outcome applied to Wilms tumour prognosis," Applied Statistics 48:457-68, 1999

• S-Plus documentation
• S-Plus functions
• Three grouped datasets from the National Wilms Tumor Study
  1. NWTS.dat Complete data
  2. NWTSc.dat Case-control sample
  3. NWTSb.dat Balanced sample

WARNING: Apparently due to changes in the way version 6.1 of S-Plus handles prior weights with binomial outcome data, the program WL for Horwitz-Thompson ``weighted likelihood" estimation included in the above set of S-Plus functions gives incorrect answers. You are strongly advised to compare your output on the test datasets with that shown in the Applied Statistics article. An alternate WL function is available that does work correctly with version 6.1.

Analysis of Case-Cohort Data

• S-Plus documentation
• S-Plus functions
• nwtco.asc Continuous data from National Wilms Tumor Study (see documentation )

Analysis of Exposure Stratified Case-Cohort Data

• S-Plus documentation
• S-Plus functions
• nwtco.asc Continuous data from National Wilms Tumor Study (see documentation )