Peer review can never be perfect because people are never perfect. However, there is no better method to assess the quality of a publication. On the other hand, one should not overestimate this method either. One could also imagine a review of the peer review process for a given journal, but this would not be perfect either. Here are some discussions that are also about improving peer review:

https://www.researchgate.net/post/How_can_we_obtain_a_better_peer_reviewing

https://www.researchgate.net/post/How_prepared_are_article_reviewers_to_make_more_gentle_reviews

https://www.researchgate.net/post/How_would_you_deal_with_expert_referees_reviewers_that_send_unprofessional_reviews

https://www.researchgate.net/post/Open_peer-review_is_full_transparency_possible

https://www.researchgate.net/post/Why-the-article-reviewing-peer-review-in-Journal-and-Conference-is-a-voluntary-work

https://www.researchgate.net/post/Is_peer-review_censorship_or_necessary_evil

https://www.researchgate.net/post/Is_blind_peer-review_honest_Does_it_help_or_disfavor_achieving_an_objective_review_of_journal_papers

https://www.researchgate.net/post/Which-i-more-suitable-for-scientfic-journals-blind-peer-review-or-open-peer-review-or-combination-between-the-two

https://www.researchgate.net/post/How_many_researchers_think_publishers_journals_should_pay_for_their_content_and_also_for_peer_review

https://www.researchgate.net/post/Reviewers_participation_should_be_considered_too

https://www.researchgate.net/post/If_there_is_ever_going_to_be_a_time_peer_reviewers_will_be_paid_for_their_services_the_time_is_now

https://www.researchgate.net/post/Is_there_any_source_where_I_can_read_peer-reviews

https://www.researchgate.net/post/Do_peer-reviews_made_by_unknown_reviewers_need_permission_to_be_shared

https://www.researchgate.net/post/Why_the_publishing_policy_for_most_of_journals_are_not_clear

https://www.researchgate.net/post/On_the_reviewing_of_journals

https://www.researchgate.net/post/Do_you_really_think_that_the_referees_of_scientific_journals_are_really_qualified

https://www.researchgate.net/post/Do_you_belive_in_luck_in_publication_did_you_came_a_cross_a_situation_where_you_felt_that_Peer_reviwing_was_a_matter_of_luck_in_respect_to_reviwers

https://www.researchgate.net/post/Unreasonable_review_times

In my opinion peer review is the most scientific approach at this time and like most scientific exercises it comes with many errors and limitations. Hypothetically if we want to replace human intervention in peer review, we have two possible approaches (1) no peer review and (2) AI (artificial intelligence) driven peer review. The former can potentially do more harm than good for the progress of science. The later is too futuristic and at the moment cannot comprehend all the nuances of scientific work. Therefore atleast at the given point of time, peer review by 2-3 experts is the most scientific approach to publish and approve research.

Peer review is expected to be scientific. Its unfortunate however that regional, specialty and personal reviewer's biases lead them to unguided comments whose answers may not improve the article eventually.

To my mind, there should be scientific specifications for manuscript review.

Reviewers who find some studies difficult should refuse to review it rather than making comments which reveal their incompetence.

The science is hidden in the reality that no one knows it all and allowing other researchers to read and correct is expected to uplift the article outlook.

Thank you for your response, Tolulope Omolola Ajala, Biswajoy Ghosh , and Wolfgang R. Dick

Among the responses, the only suggested way that would make peer-review scientific is to make it public, as this would make reviewer's comments observable, falsifiable/verifiable.

Especially, for rejected articles, reviewer's comment and editor's decision need to be public, so as to verify/falsify the quality of journal, reviewer's expertise level, and editor's bias.

This way would also make reviewers and editors re-think of their ethical-responsibility for rejecting/accepting an article.

Dear Seffetullah, as already indicated by Wolfgang R. Dick this is certainly among the most often discussed questions on this platform. As with many other issues, a possible answer to this question strongly depends on the discipline of research. Thus I can answer only in view of our field of research, chemistry. I published some 400+ papers in leading chemical journals over the past 45 years. In my personal experience peer-review at least in our discipline is generally an absolutely scientific process. All high-ranked chemical journals carefully select their potential reviewers. When you register on a journal homepage as an author and / or reviewer, you are normally asked to specify your areas of expertise. This forms the basis for the selection of potential reviewers by the editors. Thus when you mention the "reviewers' expertise levels or even their levels of understanding an article" I can say that this level is generally very high in chemistry. I never experienced an occasion where a reviewer did not understand our manuscript, and I claim for myself that I have a rather good overview af various fields of chemistry. This is why more than 40 chemical journals regularly ask me to review manuscripts for them. The "request to revise/re-write a manuscript" is absolutely normal as everyone (including me) makes mistakes. I would say that in 90+ percent of all cases the reviewers' comments greatly helped improving the quality of our manuscripts. This is perhaps not the answer you wanted to hear, but it's my personal experience.

Good luck with your work and best wishes, Frank Edelmann

Dear Frank, Frank T. Edelmann , thank you for sharing your individual experience and professional views. In fact, my inquiry is aimed at hearing every possible answer.

I am not sure if field/discipline matters, given that peer-review itself is in general anonymous (not observable, not falsifiable/variable, especially for rejected papers).

As to whether or not high-ranked journals are careful in selecting reviewers, how can we know that, especially for rejected papers? Some journals publish a list of reviewers, but most journals do not do that.

For example, how can I know who were the reviewer and associate editor (of a very high-ranked journal), handled a submission for 11 months (under first round review) and rejected, as one of major reasons: "author(s) did not provide examples of items" despite the fact that the submission included descriptions and tables of all the items analyzed statistically.

Rejection for such reason is not just an individual experience or uncommon.

Also, to complete the quotation: "revise/re-write a manuscript according to reviewer's aim rather than author's aim." An example for the sake of clarity, a submission about parent-child communication is aimed at understanding how parents encourage/discourage their children/students at home. But reviewer asked to replace the aim with how teachers encourage/discourage students in the classroom.

Otherwise, of course, request for a major/minor revision is of the nature of peer-review.

It depends on what you call "scientific".

If you write an ingenious paper, a scholastic soul will probably reject it. But you should first define what you mean by scientific.

A review is essentially an opinion. An opinion may be based on relevant facts, and valid inference; but it may also be based on wrong assumptions (non-facts) and invalid reasoning. It may also be partial, malicious, etc.

Dear Mr Mario Radovan,

Here, my comments to your remarks

• "A review is essentially an opinion": Okay!
• “An opinion may be based on relevant facts, and valid inference”: Normal! and this is the goal of the peer reviewing.
• "but it may also be based on wrong assumptions (non-facts) and invalid reasoning": There, it is not good but recoverable by an exchange between the author and the reviewer.
• " It may also be partial, malicious, etc.": there it becomes serious because there is a priori no reason for that.

Everything you say is very true. It remains to develop the reasons which lead (in some cases) to the "attitude of the reviewer" in your last remark (fourth point here).

That being said, the authors also have their flaws (fortunately not all).

Unfortunately, in this kind of situation it is the author who depends on the reviewer not the other way round.

"Unfortunately, ... it is the author who depends on the reviewer not the other way round." - Correct; but it is difficult to avoid this

In sum, reviewing is needed, but reviewers are generally not perfect (impeccable) as well as authors are not. In my view, reviewers are not a big problem for (good) authors if they do their job in a couple of weeks. If the author is not happy with the review, s/he can send the paper to another journal. The worst thing is to wait six months without any answer.

Dear Mario Radovan,

The worst thing is not to wait six months but to wait fourteen (14) months (real case).

The editors also are not perfect, I invite all the speakers here to read the steps of a "singular editorial procedure". see the following url:

https://www.researchgate.net/post/Rights_of_the_author_Versus_rights_of_the_editor_14_months_lost_for_nothing

Hi. It can be scientific if the peer-reviewer more expertise in the field of the author of the article; otherwise , it is not scientific at all. This is my opinion. Thank you.

An interesting article: https://financialpost.com/opinion/the-peer-review-crisis

(Thanks to Khalid M. Saqr, see https://www.researchgate.net/post/What_is_that_one_thing_that_you_really_hate_about_publishing)

Is the peer-review itself scientific?

In any case, peer review, i.e. by experts in the field, is infinitely more reliable than a quick recommendation by a Researchgate member.

The Need to Review Peer Review

"(1) the current system fails to utilize digital technologies of open review which have advanced multiple other scholarly disciplines; (2) the author nomination system facilitates the publication of friendly reviews; (3) anonymity in the review system permits reviewers to promote unchallenged, biased, or personally motivated publication decisions; and (4) impact factors might inspire editors to publish articles that might not otherwise be accepted." (Anderson, 2013, p.337).

Anderson, E. (2013). The need to review peer review: The Regnerus scandal as a call to action. Journal of Gay & Lesbian Mental Health, 17(3), 337–351. https://doi.org/10.1080/19359705.2013.789459