Subject: | |
From: | |
Reply To: | |
Date: | Tue, 5 Jul 2022 19:14:27 -0400 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
> How many papers were considered? By what criteria were they chosen or rejected? How much weight was given to each?
These issues are most often discussed in excruciating detail in a typical review paper.
> is the data examined and analyses or are the authors' conclusions all that was considered?
Again, the paper itself should explain, but data is examined to verify that the conclusions are supported by objective evidence.
I think everyone who does meta-analysis frets and lies awake worrying far more than they are credited for doing.
Men and women of good character are the norm in such pursuits, the advocates are easy to pick out as their advocacy tends to leave them isolated from those doing the actual work in their field.
It is easy to toss around accusations like "inherent bias" when one does not do the harder work of actually reading the included vs excluded papers (no, not just abstracts) to point to one or two that might have changed the results of the meta-analysis, and seem to stand up to scrutiny. Actually reading a paper and then reading some of what they cite to work out some details can consume several solid afternoons for ONE paper, a meta-analysis could easily take weeks to "review" by reading the papers included.
There should be a minimum ante of a tad more than just an accusation in a vacuum, lacking any quote or citation in support of the claim. I quite often start posts with "Hold on there, the paper says...", and most often I must go and find the paper, and refresh my memory to make sure it is correct to offer than gentle correction.
***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html
|
|
|