For a meta-analysis, is this possible that i transform some data for ease of calculations from the selected studies if the author has not mentioned properly but we find few points to assume that what could be the possible grades as per description given, it is purely for the ease of analysis only...
can we assign grades based on desscription given in studies?
Please guide who do have extensive experience in Meta-analysis, thanks
It is possible, no one can stop you from doing so, but there will always be the question of whether your decision to transform, and how you do the transformation, is scientifically/clinically sound. I'm not sure if "facilitating ease of analysis" as purpose for data transformation is reasonable in general, because you are already adding value/subjective judgment from your end that may not be shared by the authors of studies in question.
There are at least two suggestions I can immediately think of that you should strongly consider before resorting to data transformation for the sake of "ease of analysis":
1. Contact the (corresponding) authors for the missing details in their paper you think are helpful for your systematic review.
2. Do not include the included studies in question to your quantitative synthesis or meta-analysis, but only in the qualitative synthesis part. Systematic reviews are not all about the forest plots anyway.
I agree with with Aedrian Abrilla on her constructive suggestions: Points 1&2
Some data transformations can be performed easily using RevMan Calculator or according to the Cochrane Handbook. However, many transformations cannot be easily computed. Thus, I would agree with Aedrian's suggestions.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis