firstly.. not english user so please understand.. no intend to any offenses
i am having a problem selecting studies doing meta analysis
ex) if A was given 45mg 7day to certain disease patients and there was outcome in lets say FEV1 --study 1
A was given 45mg 14day to the patients with same diseasr and there was corresponding outcome in FEV1 -- study2
is study 1 and 2 capable of doing meta analysis?
why or why not..? please :)
It depends!
You can do meta-analysis if there is no a priori evidence that treatment duration makes a "difference". Otherwise, you would be meaninglessly combining "apples and oranges".
If, in your analysis, you find difference between the two duration in effect size (ie significat heterogeneity), you would use a random effects model (since there is a difference that is not explained and may be random). Otherwise, you would use a fixed effect model.
The meta-analysis would give the average effect of the treatment, regardless of duration.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis