With that being said, here are several ways to reduce and possibly avoid confirmation
bias as a reviewer and as a reader of reviews.
Not exact matches
The
reviewer wrote, «[i] t would probably... be beneficial to find one or two male biologists to work with (or at least obtain internal peer review from, but better yet
as active co-authors)» to prevent the manuscript from «drifting too far away from empirical evidence into ideologically
biased assumptions.»
Although a scan of the published record for potentially applicable studies turned up 1,553 possible candidates, independent
reviewers dismissed most of them
as containing either potential
biases or major deficiencies in their design or data reports.
A study published in the 27 September edition of JAMA, the Journal of the American Medical Association, examined how
bias impacts the peer review process of medical journals by comparing what happens when study authors are identified for reviews, known
as a single - blind review, and when the identity of study authors is kept from
reviewers, known
as double - blind reviews.
Double Blind Book Reviews will use a double blind
as the name implies, author and
reviewer will not know the other's identity to remove
as much
reviewer bias as possible.
This makes a problem when big name
reviewers become
bias,
as most are, and review games poorly by comparing them to their favorite games.
Secondly, even being less generous to Jones, — and it is undoubtedly bad luck to draw the chief of the institute whose work you are criticising
as one of your
reviewers or stupidity for submitting it to a journal where they are on the review board, take your pick — Science, and the peer review process, is bigger than one
biased reviewer (or even a nest of
biased reviewers).
As I read it, three
reviewers were said to have been «skeptically
biased» by Wagner with no stated reasons other than it is the only explanation for how the paper was accepted even though curiously he also states:
There is another
bias,
as the «peer - review» process itself forms a
bias, because peer -
reviewers mostly agree with the majority that makes the consensus.
Did Rado purposely leave out any mention Connolley's Wikipedia work
as just one of two
reviewers of the entire Ofcom complaint because it might lead people to discover that Connolley has the appearance of being heavily
biased against the very people the complaint was complaining about?
«We didn't like»... right, no
bias there, when the Galaxy Tab S was widely considered by most
reviewers as the best tablet in it's price range and many considered it the outright best Android tablet period and it DEFINITIVELY had by far the best tablet screen, bar none, including Apple, and you didn't even give it it's own paragraph.
Two
reviewers will independently screen studies to assess eligibility,
as well
as extract data from, and assess risk of
bias of included studies.