An Introduction to Categorical Data Analysis || Appendix A: Software for Categorical Data Analysis

✍ Scribed by Agresti, Alan

• Publisher: John Wiley & Sons, Inc.
• Year: 2007
• Tongue: English
• Weight: 156 KB
• Edition: 2
• Category: Article
• ISBN: 0471226181
• DOI: 10.1002/9780470114759.app1

✦ Synopsis

All major statistical software has procedures for categorical data analyses. This appendix has emphasis on SAS. For information about other packages (such as S-plus, R, SPSS, and Stata) as well as updated information about SAS, see the web site www.stat.ufl.edu/∼aa/cda/software.html.

For certain analyses, specialized software is better than the major packages. A good example is StatXact (Cytel Software, Cambridge, MA, USA), which provides exact analysis for categorical data methods and some nonparametric methods. Among its procedures are small-sample confidence intervals for differences and ratios of proportions and for odds ratios, and Fisher's exact test and its generalizations for I × J and three-way tables. Its companion program LogXact performs exact conditional logistic regression.

SAS FOR CATEGORICAL DATA ANALYSIS

In SAS, the main procedures (PROCs) for categorical data analyses are FREQ, GEN-MOD, LOGISTIC, and NLMIXED. PROC FREQ computes chi-squared tests of independence, measures of association and their estimated standard errors. It also performs generalized CMH tests of conditional independence, and exact tests of independence in I × J tables. PROC GENMOD fits generalized linear models, cumulative logit models for ordinal responses, and it can perform GEE analyses for marginal models. PROC LOGISTIC provides ML fitting of binary response models, cumulative logit models for ordinal responses, and baseline-category logit models for nominal responses. It incorporates model selection procedures, regression diagnostic An Introduction to Categorical Data Analysis, Second Edition.