About the Meta-Analysis Workshop » Learning Outcomes
After Day-1, you should understand:
- What a meta-analysis is, and how to perform one
- The key differences between meta-analysis and narrative reviews
- That the goal of a meta-analysis is to synthesize, and not simply combine, effect sizes
- The difference between fixed-effect and random-effects models
- How to use a forest-plot to understand and report a meta-analysis
- How to use CMA to compute effect sizes, perform a simple analysis, and create forest plots
- Common mistakes in meta-analysis, and how to avoid them
- Mistakes in choosing between fixed-effect and random-effects models
- Mistakes in understanding why a meta-analysis appears to conflict with a clinical trial
- Mistakes in using Vote-counting
- Mistakes in the goals of meta-analysis
- How to quantify and interpret heterogeneity
After Day-2, you should understand:
- How to compare the effect size in subgroups of studies
- How to use regression to assess the relationship between covariates and effect size
- Common mistakes in meta-analysis, and how to avoid them
- Mistakes in interpreting indices of heterogeneity
- Mistakes in choosing between fixed-effect and random-effects models for subgroups-analysis and meta-regression
After Day-3, you should understand:
- How to work with studies that report effects for two (or more) independent subgroups
- How to work with studies that report effects for two (or more) outcomes or time-points
- How to work with studies that compare two (or more) treatments to a common control group
- How to decide whether or not it makes sense to perform a meta-analysis
- How to assess the potential impact of publication bias
- How to perform a meta-analysis using studies that employed different designs (matched groups vs. independent groups), formats (some reported means, others reported t-tests) or outcomes (some worked with means, others with risks).
- How to use CMA to perform all of these analyses
- Common mistakes in meta-analysis, and how to avoid them
- Mistakes in working with multiple outcomes from the same sample
- Mistakes in interpreting publication bias
Testimonials
"Excellent introduction to the many features of meta-analysis."
Stephen J. Ganocy, PhD - Dept. of Psychiatry, Case Western Reserve University
"CMA is an incredible software package for the novice and experienced researcher alike. It offers a sophisticated array of analytical techniques, yet retains the flexibility to adapt to your specific application. The interface is intuitive and user‐friendly, and frees the user to focus on the analysis rather than the software. I use CMA in my graduate biostatistics course and recommend it to all of my faculty colleagues as well. The software is a joy to use!"
Dr. Michael W. Hubble - Associate Professor, Graduate and Distance Program, Director, Emergency Medical Care Program, Western Carolina University, Cullowhee, NC
"The meta‐analysis program is easy to use and provides me with information that would be hard to get by any other means. I analyze a lot of data and, believe it or not, use the meta‐analysis program every day. I am not a statistician, but the results I produce have routinely been verified by statistical consultants costing large amounts for their confirmatory analyses. Without verification, I have used the results to make important decisions regarding a potential billion dollar a year product. At night I wonder, “What would I have done without this great program.”"
Louis Fabre MD, PhD - Chairman, Fabre‐Kramer Pharmaceuticals Inc.