If you want to use the meta-analysis as your research method to investigate the relationship between two variables (ie x as independent and y as dependent variable), what do you have to do with a research in which the dependent variable is broken into two variables ( y1 and y2).
Shakil Ahmed
You can sum up those (y1 and y2) those variables and count it as one and see whether you can calculate the OR/RR or not? If yes, then you can include the study if not, unfortunately, you have to exclude it.
0 votes
0 thanks
Hytham K. S. Hamid
You can do subgroup analysis well supported by RevMan software, it will also allow you to compare y1 and y2
0 votes
0 thanks
Wojciech Feleszko
or , depending on you protocoll, you may take only the last data (eg. incidence of asthma, that may increase with time).
0 votes
0 thanks
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis
More Saman Firouzi's questions See All
Similar questions and discussions