- Analysis of Variance
- Bayesian Analysis
- Categorical Data Analysis
- Cluster Analysis
- Descriptive Statistics
- Discriminant Analysis
- Distribution Analysis
- Exact Inference
- Finite Mixture Models
- Group Sequential Design and Analysis
- Longitudinal Data Analysis
- Market Research
- Missing Data Analysis
- Mixed Models
- Model Selection
- Multivariate Analysis
- Nonlinear Regression
- Nonparametric Analysis
- Nonparametric Regression
- Post Processing
- Power and Sample Size
- Predictive Modeling
- Psychometric Analysis
- Quantile Regression
- Regression
- Robust Regression
- Spatial Analysis
- Standardization
- Structural Equations Models
- Survey Sampling and Analysis
- Survival Analysis
- SAS/STAT Procedures A-Z

The CAUSALMED procedure estimates causal mediation effects from observational data. You use the CAUSALMED procedure, rather than the CAUSALTRT procedure, when you need to estimate direct and indirect effects of binary or continuous treatment variables on an outcome variable. The indirect effect refers to the treatment effect that has influence on a mediator variable, which then has a direct effect on the outcome variable. The CAUSALMED procedure generalizes linear models that have binary, negative binomial, Poisson, or normal distributions for the outcome and have binary or normal distributions for the mediator. You can include interaction effects between the treatment and mediator variables and among the covariates.

PROC CAUSALMED allows you to:

- Specify outcome (Y), treatment (T), and Mediator (M) variables.
- Use covariates to remove confounding between Y, T, and M.
- Compute standard errors and confidence intervals for effects estimates using asymptotic formulas or by bootstrapping.
- Request percentages for estimated effects relative to total effect, for example percentage mediated and percentage due to interaction.
- Specify various two-way, three-way, and four-way decompositions of total effects.
- Specify covariate levels for evaluating conditional mediation effects.
- Do BY group processing.

The following effects are available:

- Total effect
- Controlled direct effect
- Natural direct effect
- Natural indirect effect

For further details see the SAS/STAT User's Guide:
The CAUSALMED Procedure

( PDF | HTML )