The peer review process does not successfully keep out bad science.
Some of the flaws of the peer review process include:
Bias: Reviewers may be biased against certain research or researchers (particularly when they have been nominated for review by the authors), which can affect their evaluations. For example, reviewers may be more likely to reject research that challenges their own views or that comes from researchers they do not know.There is also a bias in valuing positive results more than negative results.
Lack of diversity: Since identities of peer reviewers are often not known, it is difficult to assess their backgrounds, which can result in a lack of representation and potentially biased evaluations.
Limited expertise: Reviewers may not have the necessary expertise to evaluate certain types of research, which can lead to incomplete or inaccurate evaluations.
Slow turnaround time: The peer review process can be slow, with some studies taking months or even years to be reviewed and published. This can delay the dissemination of important research findings. (See Preprints)
Lack of transparency: The peer review process is often not transparent, as peer reviewer identities are unknown and reviews are not always public. (see Open Peer Review)
Agreement among reviewers is weak (Kravitz et al. 2010)
Rejections/acceptation decisions are inconsistent
Major errors and flaws are inconsistently discovered (Baxt et al. 1998, see Andersen 2013 for a case study)
Examples of flawed peer reviewed research
Efforts to improve peer review
Opening up the peer review reports would increase transparency and accountability of peer review. See The Turing Way for more information about Open Peer Review.
PREreview on how to peer review constructively.
Sustainable peer review via incentive aligned markets
Sharp criticism of controversial ancient-human claims tests eLife’s revamped peer-review model