Date of Award

5-2007

Degree Type

Report

Degree Name

Master of Science (MS)

Department

Mathematics and Statistics

First Advisor

John R. Stevens

Abstract

This project implemented a simple and useful SAS (Statistical Analysis Software) macro for meta-analysis. The macro was specifically designed for hierarchically dependent meta-analysis but may also be used for independent meta-analysis. Two data files are necessary for this macro. One data file contains the study names, control and treatment group sizes with corresponding means and standard errors, covariates to be tested, an indicator variable noting hierarchically dependent study groups, and a variable providing the degrees of freedom for interdependent studies. The second file contains a variance/covariance matrix for the studies. The macro returns useful information such as study estimates and standard errors, covariate estimates and standard errors, z-scores and p-values as well as a variance/covariance matrix for the covariate estimates. The macro also produces a ladder plot of study estimates including the final effect size estimate and, for hierarchically dependent meta-analysis, four plots showing the convergence of σ2 and ς, the parameters reflecting hierarchical dependence.

Included in

Mathematics Commons

Share

COinS