This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.
|CE_16C||Mon, 8/2/2010, 1:00 PM - 5:00 PM||CC-8 (East)|
|Meta-Analysis: Statistical Issues for Combining the Results of Multiple Studies — Continuing Education Course|
|Instructor(s): Ingram Olkin, Stanford University|
|Meta-analysis enables researchers to synthesize the results of multiple studies designed to determine the effect of a treatment. The movement for evidence-based medicine, evidence-based education, etc. has led to an increased reliance on the methodology of meta-analysis. Another component for the increased use of meta-analysis in almost every field is the information explosion, which means that individuals can no longer read all the articles that appear in any one field. Meta-analysis consists of several parts prior to statistical analysis: formulation of the problem, a literature search, coding and evaluating the literature, inclusion-exclusion criteria, after which statistical issues can be addressed: methods for combining studies, sensitivity analyses, models for adjusting bias, missing data, how to deal with unpublished papers. The statistical methods will include nonparametric, parametric, and regression models. Examples of published meta-analyses in medicine and the social sciences will be used to illustrate the various methods. The first half of the workshop will be introductory, and the second half will focus on statistical methods.|