SAS/STAT Software Features

Analysis of variance

Balanced and unbalanced designs.
Multivariate analysis of variance and repeated measurements.
Linear models.
More on analysis of variance.

Bayesian analysis

Built-in Bayesian modeling and inference for generalized linear models, accelerated failure time models, Cox regression models and finite mixture models.
Wide range of Bayesian models available via general-purpose MCMC simulation procedure.
Bayesian discrete choice modeling.
More on Bayesian analysis.

Categorical data analysis

Contingency tables and measures of association.
Bioassay analysis.
Generalized linear models.
More on categorical data analysis.

Causal analysis

Propensity score analysis.
Estimation of causal treatment effects.
More on causal analysis.

Cluster analysis

Hierarchical clustering of multivariate data or distance data.
Disjoint clustering of large data sets.
Nonparametric clustering with hypothesis tests for the number of clusters.
More on cluster analysis.

Descriptive statistics

Box-and-whisker plots.
Compute directly and indirectly standardized rates and risks for study populations.
More on descriptive statistics.

Discriminant analysis

Canonical discriminant analysis.
Stepwise discriminant analysis.
More on discriminant analysis.

Distribution analysis

Univariate and bivariate kernel density estimation.
More on distribution analysis.

Exact inference

Exact p-values and confidence intervals for many test statistics and measures based on one-way and n-way frequency tables.
Exact tests for the parameters of a logistic regression model.
Exact tests for the parameters of a Poisson regression model.
More on exact inference.

Finite mixture models

Modeling of component distributions and mixing probabilities.
Maximum likelihood and Bayesian methods.
More on finite mixture models.

Group sequential design & analysis

Design of interim analyses.
Perform interim analyses.
More on group sequential design and analysis.

High performance

14 SAS/STAT procedures are multithreaded.
12 SAS High-Performance Statistics procedures are available with SAS/STAT for single machine use.

Longitudinal data analysis

Marginal and mixed models.
Continuous and categorical outcomes.
More on longitudinal data analysis.

Market research

Simple and multiple correspondence analysis.
Two-way and three-way metric and nonmetric multidimensional scaling models.
Discrete choice models.
More on market research.

Missing data analysis

Multiple imputation.
Weighted generalized estimating equations.
Imputation for survey data.
More on missing data analysis.

Mixed models

Linear and nonlinear mixed models.
Generalized linear mixed models.
Nested models.
More on mixed models.

Model selection

Linear models.
Generalized linear models.
Quantile regression models.
More on model selection.

Multivariate analysis

Exploratory and confirmatory factor analysis.
Principal components analysis.
Canonical correlation and partial canonical correlation.
More on multivariate analysis.

Nonlinear regression

Automatic derivatives.
Bootstrapped confidence intervals.
More on nonlinear regression.

Nonparametric analysis

Kruskal-Wallis, Wilcoxon-Mann-Whitney and Friedman tests.
Other rank tests for balanced or unbalanced one-way or two-way designs.
Exact probabilities for many nonparametric statistics.
More on nonparametric analysis.

Nonparametric regression

Multivariate adaptive regression splines.
Generalized additive models.
Local regression.
Thin-plate smoothing splines.
More on nonparametric regression.

Postprocessing

Hypothesis tests.
Prediction plots.
Scoring.
More on postprocessing.

Power & sample size

Computations for linear models including MANOVA repeated measurements.
Computations for many hypothesis tests, equivalence tests and correlation analysis.
Computations for binary logistic regression and survival analysis.
More on power and sample size.

Predictive modeling

Classification and regression trees.
Partitioning of data into training, validation and testing roles.
Modern model selection methods such as elastic net and group LASSO.
More on predictive modeling.

Psychometric analysis

Multidimensional scaling.
Conjoint analysis with variable transformations.
Item response theory (IRT) models.
More on psychometric analysis.

Quantile regression

Simplex, interior point and smoothing algorithms.
Analysis of censored data.
Model selection for linear regression models.
More on quantile regression.

Regression

Least squares regression.
Principal components regression.
Quadratic response surface models.
Accurate estimation for ill-conditioned data.
More on regression.

Robust regression

M estimation and high-breakdown methods.
Outlier diagnostics.
More on robust regression.

Spatial analysis

Ordinary kriging in two dimensions.
Spatial point pattern analysis.
Variogram diagnostics.
More on spatial analysis.

Standardization

18 standardization methods.
More on standardization.

Statistical graphics

Hundreds of statistical graphs available with analyses.
Customization provided.
Base SAS “SG” procedures create user-specified statistical graphics.
More statistical graphics capabilities.

Structural equations

Structural equation models specified with popular modeling languages.
Parameter estimation and hypothesis testing for constrained and unconstrained problems.
More on structural equations.

Survey sampling & analysis

Sample selection.
Descriptive statistics.
Linear and logistic regression.
Proportional hazards regression.
Missing value imputation.
More on survey sampling and analysis.

Survival analysis

Nonparametric survival function estimates.
Competing-risk models.
Accelerated failure time models.
Proportional hazards models.
Interval-censored data analysis.
More on survival analysis.