To whet your appetite, here’s a little preview of key highlights from some of the analytical products. SAS/STAT® 12.1 provides new procedures for model selection for quantile regression, quantile regression for censored data, and multivariate adaptive spline regression. In addition, it includes many enhancements for the MCMC procedure, including handling missing data for both the response and covariates. SAS/ETS® provides Bayesian estimation methods for many univariate models in the QLIM procedure; many new tests in the PANEL and AUTOREG procedures for checking the statistical assumptions of the models concerning stationarity, cointegration, and structural change; and many new graphs, including dependency plots in the MODEL procedure.  The 12.1 release of SAS/OR® software takes advantage of multiple cores on the same computing platform for numerous operations, including problem creation in the OPTMODEL procedure, nonlinear multistart optimization, and a new decomposition-based algorithm for linear and mixed integer optimization. 

Check http://support.sas.com for 12.1 release announcements in a month or so.

If you aren’t attending JSM, you may still be interested in looking at some of the handouts we’ve prepared. These include highlights of the upcoming SAS/STAT 12.1 and SAS/ETS 12.1 releases,  information on ODS Graphics, and other analytics-related material.

Readers in the pharmaceutical industry might be interested in this macro that performs nonparametric randomization-based analysis of covariance for use in analyzing clinical trials data. The reduction of variance for the treatment effect provides more powerful tests and more accurate confidence intervals. This method was discussed in the paper "Up to Speed with Categorical Data Analysis" by Stokes and Koch (2011). Richard Zink, Research Statistician Developer in the JMP Life Sciences Group, has published an updated version of these macros in the Journal of Statistical Software, and they are available for downloading.

Readers may know that Paul Allison released the second edition of Logistic Regression Using SAS: Theory and Applications earlier this year. Well, your editor and co-authors have recently completed the third edition of Categorical Data Analysis Using SAS . And to complete the trifecta, Alan Agresti will be publishing the third edition of  Categorical Data Analysis this Fall.  You can take a look at the SAS books at the Publications booth at JSM next week.

Competing risks arise in studies when subjects are exposed to more than one cause of failure and when failure due to one cause excludes failure due to other causes. For example, the cause of failure in bone marrow transplantation can be relapse, death in remission, or other causes. The standard Kaplan-Meier method for survival analysis does not yield valid results for a particular risk if failures from other causes are treated as censored. A useful quantity is the cumulative incidence function, which is the cumulative probability of failure from a specific cause over time. The SAS macro %CIF implements appropriate nonparametric methods for estimating cumulative incidence functions. The macro also implements Gray's method (Gray 1988) for testing differences between these functions in multiple groups. This paper discusses these methods and illustrates the macro.

Finite mixture models provide a flexible framework for analyzing a variety of data. Suppose your objective is to describe the distribution of a response variable. If the corresponding data are multimodal, skewed, or heavy-tailed or exhibit kurtosis, they may not be representative of most known distributions. A finite mixture model provides a parametric alternative that describes the unknown distribution in terms of mixtures of known distributions. This paper provides an overview of the capabilities of the FMM procedure and illustrates them with applications drawn from a variety of fields.

This video describes the overall capabilities of the FMM procedure and applies it to a problem of overdispersion for credit data, including the use of the hurdle model. 

Reproducible research is an increasingly important paradigm, and tools that support it are essential. Documentation for many SAS analytical products has long been created from a single-source system that embeds SAS code in LaTeX files and generates statistical results from those files. Last newsletter, the paper that describes a SAS system for this process was included, and this month, we bring you an overview in video form. Check it out!

Simulations in your SAS/IML programs running too slowly? Then take a look at these tips compiled by Rick Wicklin in his blog, The DO Loop.  You may also be interested in Wicklin's post on defining a library of user-defined functions, which allow you to extend the capability of the SAS/IML language.

These and many other useful ideas come to you twice weekly in Wicklin's blog. Think about signing up!

The following CTW courses are being taught on Wednesday, August 1, at JSM.  You can sign up for them at the JSM registration desk.

SAS Procedures for Analyzing Survey Data (Pushpal K Mukhopadhyay)

Creating Statistical Graphics with ODS in SAS (Warren Kuhfeld)

Group Sequential Analysis Using SAS Software (Yang Yuan)

Structural Equation Modeling Using the CALIS Procedure In SAS/STAT Software (Yiu-Fai Yung and  Xinming An)

And now's the time to pick out a SAS regional user conference. All feature statistical tutorials and talks, plus the great opportunity for engaging with SAS users in your area.

Suppose that a model is fit to a set of independent groups by using the same predictors and you want to compare the parameters of these models across groups. Comparison of group parameters can be done the same way regardless of the model type (ordinary regression, logistic regression, Poisson regression, etc.) and involves specifying a single model that simultaneously estimates the intercepts and slopes for all groups. When a single model estimates all the parameters of interest, you can perform tests to compare them. This is also known as testing for the heterogeneity (or homogeneity) of slopes.