A 2-Day Seminar Taught by David Wilson, Ph.D.
How do you make sense of conflicting results across studies? It is common for a finding that is statistically significant in one study to be statistically nonsignificant in another. Does this reflect differences in the design of the studies? Do the effects actually agree in direction and magnitude even if they are different in statistical significance? What if we have 5 such studies, or 50, or 500? It would be surprising if all such studies were in perfect agreement with one another. Thus, we need a method to make sense of the variability in findings across studies.
Meta-analysis is a statistical solution to this problem and has become a widely-used method for synthesizing results across studies in the social and behavioral sciences and medicine, as well as several physical sciences, such as ecology. The key to meta-analysis is the effect size: it encodes the direction and magnitude of the finding on a common scale. Using specialized statistical methods, the average effect across studies can be estimated as well as an examination of the consistency of the effects. Analyses can also explore potential explanations for inconsistencies in findings, such as theoretically relevant design differences.
This course will provide hands-on instruction in conducting all aspects of a meta-analysis, including the systematic nature of the study search and retrieval process, systematic coding and data extraction, computation of various effect size types, and the estimation of both fixed and random-effects meta-analysis models. Students will also learn both categorical and regression based moderator analysis and methods for exploring publication selection bias. Finally, reporting and presentation of the results of a meta-analysis will be taught.
This seminar will use Stata for the worked exercises. However, no previous experience with Stata is assumed. Exercises can also be completed in SPSS and R (on request). To participate in the hands-on exercises, you are strongly encouraged to bring a laptop computer. A power outlet and wireless access will be available at each seat.
For Stata users, version 14 will be used for the examples, but the exercises can also be done with versions 12 or 13. Seminar participants who are not yet ready to purchase Stata could take advantage of StataCorp’s free 30-day evaluation offer or their 30-day software return policy.
Who should attend
This course will be helpful for researchers in any field who want to learn how to statistically synthesize results across independent studies using meta-analytic methods. Participants should have a basic working knowledge of introductory statistics through multiple regression analysis. No knowledge of matrix algebra is required or assumed but familiarity with standard statistical notation is useful.
LOCATION, Format, and MATERIALS
The class will meet from 9 am to 5 pm each day with a 1-hour lunch break from 12 pm to 1 pm at Temple University Center City, 1515 Market Street, Philadelphia, PA 19103.
Participants receive a bound manual containing detailed lecture notes (with equations and graphics), examples of computer printout, and many other useful features. This book frees participants from the distracting task of note taking.
Registration and lodging
The fee of $995 includes all course materials.
Lodging Reservation Instructions
A block of guest rooms has been reserved at the Club Quarters Hotel, 1628 Chestnut Street, Philadelphia, PA at a special rate of $149 per night. This location is about a 5 minute walk to the seminar location. In order to make reservations, call 203-905-2100 during business hours and identify yourself by using group code STA922 or click here. For guaranteed rate and availability, you must reserve your room no later than August 21, 2016.
If you make reservations after the cut-off date ask for the Statistical Horizon’s room rate (do not use the code) and they will try to accommodate your request.
- Logic of meta-analysis
- Scope and problem formulation
- Computing effect sizes
- Coding studies and database structure
- Fixed-effect meta-analysis
- Random-effects meta-analysis
- Categorical moderator analysis
- Regression based moderator analysis
- Creating forest plots
- Robust standard errors for dependent effect sizes
- Assessing publication bias
- Interpretation and reporting of results