I don't see much harm in it but i don't know how informative it will be either. I typically don't do meta-analyses (MAs) with that few studies. Yes, it is OK in the sense that it is technically possible. But, with that small number of studies I question if a systematic review approach, or perhaps incorporating both MA and SR, may be more informative. But that is just my opinion and from what I remember of your previous comments you work in a different field than me, so perhaps it is normal to do a MA with so few studies in that field.
@@LearnMetaAnalysis According to many colleagues I asked in my medical field no one does any publication bias if the Analysis includes less than 10 studies. I also found this paper which says similar things. www.ncbi.nlm.nih.gov/pmc/articles/PMC5482177/#:~:text=Second%2C%20investigators%20should%20use%20appropriate,methods%20are%20not%20very%20reliable) I don't understand fully how the statistics behind these tests, but I read that they are not as accurate with low study count. Yes it is common to see MA with even 4 Studies only in medicine. Being new to doing such papers, I think this weird change and honestly not so good quality of papers is due to many medical students trying to rush many papers without learning properly and these people themselves later on start giving courses which has so many errors or incomplete information and understanding as most analysis is done on RevMan software.
Amazing lecture. Your really simplified alot of ideas that seemed really complex for me to understand when I first started learning meta-analysis.
Thank you!
Should we always do publication bias? even if the study count is very low N
I don't see much harm in it but i don't know how informative it will be either. I typically don't do meta-analyses (MAs) with that few studies. Yes, it is OK in the sense that it is technically possible. But, with that small number of studies I question if a systematic review approach, or perhaps incorporating both MA and SR, may be more informative. But that is just my opinion and from what I remember of your previous comments you work in a different field than me, so perhaps it is normal to do a MA with so few studies in that field.
@@LearnMetaAnalysis According to many colleagues I asked in my medical field no one does any publication bias if the Analysis includes less than 10 studies.
I also found this paper which says similar things.
www.ncbi.nlm.nih.gov/pmc/articles/PMC5482177/#:~:text=Second%2C%20investigators%20should%20use%20appropriate,methods%20are%20not%20very%20reliable)
I don't understand fully how the statistics behind these tests, but I read that they are not as accurate with low study count.
Yes it is common to see MA with even 4 Studies only in medicine.
Being new to doing such papers, I think this weird change and honestly not so good quality of papers is due to many medical students trying to rush many papers without learning properly and these people themselves later on start giving courses which has so many errors or incomplete information and understanding as most analysis is done on RevMan software.