am conducting a meta- analysis and our effect size metrics is odds ratio. Some literature have more than two categories (OR for each category) and others have only two categories. I would like to have a suggestion on how can I use these two different categories by dummy in to two categories for Meta – analysis. Second issue is; can I use correlation metrics together?
I would greatly appreciate your suggestions.
- What do you mean for categories? Do you mean situations in which different doses of pharmacological treatment is administered? In such situations I think you need to perform subgroup analyses with meaningful cut-offs.
- What do you mean for correlation? Odds ratios are appropriate for categorical outcomes while correlations are statistics used for continuous ones.
https://scholar.google.it/citations?user=pqD6_OsAAAAJ&hl=it
www.polygree.com
Dear Paolo Eusebi,
What I mean is two different literature use altitude is associated with tuberculosis incidence. But, they dichotomous indifferent ways(one articles in two and the other one in five).
I think it is really hard to perform meta-analysis in such situation.
https://scholar.google.it/citations?user=pqD6_OsAAAAJ&hl=it
www.polygree.com
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis