On selection of models for continuos meta analysis data with incomplete variability measures
The choice between the fixed and random effects models for providing an overall meta analysis estimates may affect the accuracy of those estimates. When the study-level standard deviations (SDs) are not completely reported or are “missing” selection of a meta analysis model should be done with more...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Pushpa Publishing House
2011
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/7217/ http://irep.iium.edu.my/7217/ http://irep.iium.edu.my/7217/1/FJMS-_Final_Version.pdf |
| Summary: | The choice between the fixed and random effects models for providing an overall meta analysis estimates may affect the accuracy of those estimates. When the study-level standard deviations (SDs) are not completely reported or are “missing” selection of a meta analysis model should be done with more caution. In this article, we examine through a simulation study, the effects of the choice of meta analysis model and the techniques of imputation of the missing SDs on the overall meta analysis estimates. The results suggest that imputation should be adopted to estimate the overall effect size, irrespective of the model used. However, the accuracy of the estimates of the corresponding standard error (SE) are influenced by the imputation techniques. For estimates based on the fixed effect model, mean imputation provides better estimates than multiple imputation, while those based on the random effects model are the more robust of the techniques imputation used. |
|---|