Manual for Research and Publication Ethics in Science and Engineering

Cho Eun Hee, Kim Young-Mog, Park Kibeom, Son Wha-Chul, Yoon Tae-Woong, Lim Jeong Mook, Hwang Eun Seong
ISBN-13: 978-89-5938-345-0-93190
Korean Federation of Science and Technology Societies
This is an open-access publication distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Publication Ethics

Publication Ethics

I. Writing and Submission Guidelines for Authors

1. Basic responsibilities of authors

Authors are accountable for the contents, research process, and results of their own research papers.

  • Authors should double-check their own research papers in detail for accuracy in calculations, data presentation methods, analysis of results, etc.

  • Authors must ensure that their research was carried out in an ethical manner, and has adhered to relevant regulations.

Authors must follow the regulations of academic journals in peer review, editing, and publication.

  • Authors must submit manuscripts that are original and must not submit the same paper to more than one academic journal.

  • Authors should respect the embargo regulations of an academic journal. One should not inform other media about the contents of the manuscript submitted to a journal until a designated time, and should also seek the agreement of the affiliated institution or the funding agency regarding this policy.

The author should cooperate fully with any requests made by the journal editor.

  • If any errors are discovered after the submission or the publication of the manuscript, even in case of a trivial error, the author must immediately inform the journal editor and discuss follow-up measures.

  • The author should submit data, regulations, procedures, software, lab notes, or other information promptly when requested by the editor.

  • Even after the publication of the manuscript, the author must cooperate fully with any requests made by the editor (or the readers); in addition, any errors or problems discovered regarding the manuscript must be resolved by cooperating with the editor.

2. Ensuring the objectivity, relevance, and transparency of the paper

The researcher must ensure objectivity, relevance, and reproducibility of the research described in the paper by conducting the experiment based on ample academic evidence and analyzing the results in a logical manner.

  • The research process and method of analysis suggested by the author in the manuscript must be clear so that other researchers can repeat and reproduce the same results.

  • Researchers must record the procedure and results faithfully and store the evidence in a systematic fashion. If at any point during the peer review, the reviewers or the journal editor requests the evidence, the researcher must be able to present it.

  • If there are any limitations present in the research, the researcher must indicate them clearly.

Any conflicts of interest, research procedure, and support received during the writing process of the manuscript that may have affected the analysis of the results must be clearly indicated.

  • Researchers must disclose all financial and non-financial conflicts of interest so that the journal editor, reviewers, and the readers are fully aware of such conflicts. Researchers can refer to the submission guidelines for conflicts of interest that must be revealed. (For information on the basics of conflicts of interest, please refer to ‘Chapter 5: Conflicts of Interest’.)

  • The research funding agencies must be revealed. If the funding institution provided support beyond monetary support, such as the design, performance, analysis, interpretation, translation, or report of the experiment, this must be indicated in the paper in detail. The source from which materials and equipment were provided must be specified along with any individual who supported the performance, statistical evaluations, or the conceptualization of the experiment.

3. First publication of original content

The author must ensure that the manuscript submitted is original and creative and has never been published in any other language or through a different medium.

  • The author must not submit the manuscript to more than one journal concurrently.

  • If the journal editor wishes to publish a manuscript previously published in another journal he or she should obtain permission from the copyright holder. The fact that the article has been reprinted should be disclosed with the source of the original publication.

☞ Even if the manuscript has never been published in a journal, if it has been presented or reported in other occasions, the individual must check submission regulations and ethics guidelines for each research report. In addition, if an author wants to publish a paper based on the results that were previously reported to a funding agency, he or she must check the agreements with the funding agency on the revelation of the research results.

If a single project generates multiple research papers, the related articles should be cross referenced so that the readers are aware of the situation.

  • If the results presented in a manuscript have been reported or if the same data is analyzed differently to create multiple manuscripts, the journal editor must be informed. A copy of any relevant manuscripts or manuscripts submitted to other journals must be submitted to the editor.

4. Ensuring research ethics

Results of research studies should be reported in an honest and accurate manner.

  • A significant level of data modification can be viewed as fabrication, and thus should be avoided.

  • In particular, when utilizing digital images (electron microscope images, X-ray images, electrophoresis photographs, etc.), researchers should be careful not to imply a biased conclusion. In certain cases, it may be unclear whether the manipulation was appropriate or within regulations. If such doubts are raised, the author must elaborate on his or her process and justify the decision in the manuscript so that the reviewers or the readers can make a proper judgment.

  • When writing a research paper, the results should not be deliberately excluded simply because they are difficult to explain, are not consistent with other data, or do not support the hypothesis or the conclusion of the paper

Any expressions used in the manuscript must all be written by the authors in their own words. The manuscript must also adhere to copyright and publication guidelines. If a source is protected under copyright, permission from the author of the source must be obtained prior to its usage.

  • When referring to others' work, the author must include a proper citation of the source. This also applies to sources written by the authors themselves.

  • When using data tables or diagrams from another manuscript or when using a substantial portion of another manuscript, the author should seek permission from the owner of its copyright.

Any researchers who are deemed to have played a significant role in the design and conduct of research and the writing of the manuscript must be offered authorship. (For authorship criteria, please refer to Chapter 2 Section 5 ‘Unjustified Authorship’.)

For research including experiments on humans or animal testing, the approval, permission, and enrollment conditions must be disclosed. Also, manuscripts should be written in a way that protects the research subjects as much as possible.

  • When requested by the editor, the author must provide a copy of the relevant documents such as the approval, enrollment form, or subjects' informed consent forms. In addition, the editor can request, when needed, the research proposal to verify whether the research was conducted according to the original research design. In such cases, the author should comply with the editor's requests.

  • The personal information of the subjects should be kept confidential. Any other information can be published only if agreed upon by the subjects. If some information may be offensive to the subjects or close parties of the subjects, additional consent is required even if the information is included in the scope of the original agreements.

  • Statistical analysis must adhere to the research design. Any addition or modifications to the method of analysis must be identified and distinguished from the original design.

II. Guidelines for Editors and Reviewers

1. Basic responsibilities of the editor

The editor must take responsibility for all matters related to the publication of the journal. The editor-in-chief must also establish regulations and procedures that ensure the quality and publication ethics of the journal and that allow continuous modifications and improvements.

  • The editor must take into account the opinions of the authors, readers, reviewers, and the journal editors. At the same time, he or she must inform these individuals of their responsibilities in the publication process. Furthermore, any other guidelines regarding policies, criteria, and the sharing of information before and after publication must be established and clearly delivered to the relevant parties.

The editor must respect freedom of expression. In addition, any secondary pressures that may affect the professionalism or morality of the manuscript must be restrained to ensure the independence of the editing.

  • Any activities regarding the management of the journal should be separated from the editing and evaluation of the manuscripts. Any sponsors or publishers should be prohibited from having an inappropriate influence on the publication of manuscripts or on manuscripts that have already been published. In addition, all the necessary information should be provided to the readers so that they could understand the roles and the characteristics of the sponsors or the publishers.

  • Regulations must be established so that the editor cannot be fired due to any criticism on the content of the journal by the publishing organization unless the editor-in-chief has made inappropriate edits or decisions that disrupt the objectives of the journal.

  • If advertisements are included in the journal, it must be ensured that the content and tone of the advertisement does not produce any conflicts with the publication guidelines. In addition, the advertisements should not be related to the content of the journal articles and be clearly separated from the articles.

The editor must establish and maintain a transparent, effective, and fair review process.

  • Timely publication of manuscripts should be ensured. A procedure to reconsider the editor's decisions must be in place in case of irregularities. The editor should also closely manage academic and ethical issues starting from the review process of the submitted manuscripts to the publication of the journal. Even after the publication of the journal article, if a problem arises regarding a manuscript, corrections, explanations, withdrawal, or an apology should be immediately issued, as appropriate, to ensure the integrity of the journal and the publication ethics. If ethical problems arise or any allegations of research misconduct are made, a policy must be implemented to deal with the issue whenever it arises and the situation must be monitored regularly.

The editor should clearly define the situations where conflicts of interest may occur during the editing and publication process of a journal and must also structure the procedure to deal with such situations appropriately.

  • At the most basic level, the authors, reviewers, editors, and the publishing organizations must each disclose any issues of conflicts of interest that are within their purview.

The editor should not leak information obtained during the editing or the peer review process to a third party. The editor should require reviewers to maintain confidentiality of such information.

  • In cases where the submission and peer review process are held online, any paths that may potentially allow other irrelevant individuals to gain access without permission should be blocked.

  • The editor should not utilize factitious means or abuse authority such as manipulating statistical indicators to improve the ranking of the journal.

  • The reference list of a manuscript should be selected purely based on academic criteria. No influence or pressure should be put on the author for any other reason.

☞ While the editor should keep information acquired during the editing process confidential, exceptions may be made to share information. For example, if an investigation is being conducted regarding misconduct, the editor can share information and cooperate with the investigation committee or the editor-in-chief of another journal to verify the academic value and the publication ethics of relevant manuscripts. The conditions that allow information sharing must be clearly stated in the submission or the peer review guidelines.

2. Academic value of the manuscript and ensuring publication ethics

The editor must establish appropriate measures to ensure that the publication of the journal promotes both publication ethics and the development of the academic field.

The editor must strengthen regulations and procedures so that he or she can evaluate whether or not the planning and performance of the experiment was carried out rigorously, the results were obtained through relevant methods, and the results and analysis were presented in a scientific manner.

  • For example, if a manuscript using statistical analysis is submitted, at least one reviewer should have statistical expertise to evaluate the statistical method, the sample size, the suitability of the analysis, and/or the potential biases.

The editor must clearly inform the authors that they must follow all domestic and international guidelines regarding the planning, performance, and publication of the research. Furthermore, the editor must only publish research results that have adhered to the aforementioned regulations.

  • Especially in cases involving human subjects or animal testing, the submitted manuscript should be processed by the editor and reviewers who are well-acquainted with the guidelines.

  • If there are national laws or regulations related to the research, the editor must ensure that the researchers have already consulted with the relevant ethics committee. It must be noted that even if the research had passed a pre-evaluation process, if ethical issues are raised during the peer review process, the article may be rejected.

  • If a research paper involving human subjects or animal testing is submitted and the relevant national regulations are unclear, the editor should consult with the author about any laws or regulations that may exist and should verify to what degree the research has adhered to the rules.

  • If the research involves human subjects, the editor must ensure that the subjects have volunteered for the research and have agreed to the publication of the research results. In addition, the editor should confirm whether the informed consent process and the content of the consent form were valid.

The editor must establish clear regulations regarding research misconduct such as violation of ethics regulations, fabrication, modification, plagiarism, or redundant publication. In addition, the editor-in-chief should reveal what types of follow-up actions he or she has planned to carry out during the editing process or after the publication in case of any research misconduct.

  • Standards regarding the management and evaluation of digital image files, images, and diagrams must be clearly stated in the submission guidelines. Especially in the fields where digital images play a critical role, the journal should be well equipped with enough manpower and an adequate procedure to ensure that the digital images have been managed appropriately.

  • It is recommended that editors use programs designed to search for textual similarities to ensure that no plagiarized manuscripts or manuscripts that have already been published in another journal are published. At the same time, the aforementioned editing policies should be clearly communicated to the authors.

  • The editor-in-chief must establish guidelines for peer review and editing policies that prevent attempts to publish a single research result in multiple manuscripts (salami publication) and publishing the same manuscript in multiple journals (duplicate publication).

Case study

Case study. Digital image management guideline 1[1]

In the Journal of Cell Biology (JCB) where photographic data is of the utmost importance, there exists a head photograph editor who reviews digital images of manuscripts that are expected to be published. In cases of black and white photos, it is said to be possible to detect manipulated photographs since simple changes in brightness and contrast can reduce the unity of the background. The JCB photograph review procedure is being used by many other journals such as Nature. JCB requires authors to submit the original digital image files along with the final manuscript and reviews all the submitted images. In many journals, if deemed necessary by the editor-in-chief, the author can be required to submit the original files. The image management regulations as stated in the JCB submission guidelines are as follows.

  • No specific feature within an image may be enhanced, obscured, moved, removed, or introduced.

  • The grouping of images from different parts of the same gel, or from different gels, fields, or exposures, must be made explicit by the arrangement of the figure (i.e., using dividing lines) and in the text of the figure legend.

  • Adjustments of brightness, contrast, or color balance are acceptable if they are applied to every pixel in the image and as long as they do not obscure, eliminate, or misrepresent any information present in the original, including the background. Non-linear adjustments (e.g., changes to gamma settings) must be disclosed in the figure legend.

3. Management of the peer review process

The greatest responsibility assigned to the editor-in-chief is to establish and manage a fair and effective peer review process. The editor-in-chief should provide a thorough and complete guide to the peer review process through the submission guidelines and identify the types of manuscripts that should be allowed to go through the peer review process and those that should not.

The final decision of whether or not a manuscript is published is up to the editor in charge. The editor must make an unbiased and fair decision on whether the topic of a manuscript agrees with the field of study of the journal, whether it contains enough creative ideas to make a significant contribution to the topic, and whether the research method and analysis are appropriate. The editor should establish a peer review process that is as fair and unbiased as possible, and disclose the process clearly so that both the authors and the reviewers would have a full understanding of the peer review process.

If the editor assesses the content or quality of the manuscript to be inadequate, he or she can reject the manuscript or recommend a different journal before the peer evaluation process.

The editor should send the reviewers' comments to the author in a constructive and helpful manner.

☞ In determining whether a manuscript will be published or not, the editor may utilize the review comments of experts in various ways. In some journals, the reviewers serve simply as advisors. In such journals, the editor may not directly ask the reviewer whether a manuscript should be accepted or not, and even if their opinions are referred to, they may not necessarily be followed. Other journals take a summative opinion of the reviewers, and the editor follows the majority. Regardless of the method, the journal should make the entire process clear for the readers and the reviewers in the guidelines.

Case study

Case study. Blind review process for fairness

The fair review of the editor and the reviewers is of utmost importance. To ensure fairness, two types of peer review processes have been devised in which the experts in the relevant field assess the value of a manuscript and publish only those deemed worthy. There are still many attempts being made to increase fairness and objectivity of the peer review system. However, it should be noted that the existence of a procedure does not ensure unbiasedness.

  • Single-blind review by anonymous reviewers

    In single-blind peer review, the identity of the reviewer is kept anonymous from the author and readers so that the experts of each field can express their opinions freely. The author only receives the comments of an anonymous reviewer. However, the identity of the reviewer can still be revealed if he or she so desires. Certain journals think disclosing the reviewer not only increases the quality of the evaluation but also makes it possible to acknowledge the reviewer's contribution in the publication of the manuscript. BMC Medicine, an online journal, has adopted an open peer review process and includes the editor's opinion, signed reviewers' reports, and the first and second draft of the manuscript in the publication. The EMBO Journal published by the European Molecular Biology Organization does not reveal the identity of the reviewer but publishes the review process file online as an appendix along with the final manuscript. The review process files include the comments of the editor and reviewers, the reply from the author, and the first and second draft of the manuscript.

  • Double-blind review, in which both reviewers and authors are anonymous

    In double-blind peer review, in order to enable the peer review process to be based solely on the academic value of the manuscript, and not be influenced by the reputation of the author or research institution, the peer review process remains completely anonymous to both the author and the reviewer. However, such attempts may be ineffective since, in many cases, the author can be inferred by the research topic and content of the manuscript. The author can also be found while searching relevant references. As a result, the peer review may actually not be an anonymous process. Because of this, some question the effectiveness of this method and argue that fairness is better achieved if the reviewers know the author.

4. When allegations of misconduct are made during the review process

When research misconduct or redundant publication is discovered during the review process, the editor should institute appropriate follow-up measures and make an announcement regarding this matter so that the authors, readers, and the reviewers can inform themselves of the situation.

The editor-in-chief must establish a procedure to solve the issues raised by the reviewers. An editorial board or ethics committee meeting is usually held to review the problem. The committee identifies the severity of the issue, but does not make a final decision. Other appropriate measures, such as notifying the author's institution are followed.

☞ In Korea, there are many journals that have a separate ethics committee. In this manual, the role of the ethics committee is not specified separately. This is because the division of roles and responsibilities between the editorial board and the ethics committee can differ greatly depending on each journal's management system and situation. In any case, however, it is imperative that the management guidelines for the editorial board and the ethics committee be established and that the readers and the authors be clearly informed.

☞ If the journal is to organize an ethics committee to manage the ethical issues raised during the review or editing process, such a possibility should be clearly and separately stated in the submission guidelines. It is also possible to have the editorial board take care of the issues raised for submitted manuscripts, and have the ethics committee handle those for published papers.

If an error or some discrepancy in the data is discovered, the editor should ask the author to explain it. If necessary, the editor can request the data or research notes from the author.

  • If a minor correctable error is discovered prior to publication, the error can be revised and the paper can continue to go through the review process; the reviewers should be asked to reevaluate the revised manuscript when needed.

  • If a severe error that cannot be repaired without a major revision is discovered, the editor should inform the author and give him or her an opportunity to explain.

  • If the situation is not resolved despite the author's explanation, the journal should inform the author's institution of affiliation and request that they investigate the matter. The author should also be informed of the actions taken by the journal with respect to the institution. The affiliated institution should investigate whether the error is a simple mistake or intentional misconduct. The journal should ask the institution to provide the results of the investigation.

  • If the institution confirms that the discrepancy in the data is not due to any research misconduct, the manuscript should be resubmitted and reevaluated. However, if intentional misconduct is identified in the investigation, the journal should inform the author of the sanctions imposed by the journal.

☞ Discrepancies in data, errors, or irreproducibility of data can be due to research misconduct, but they can also be honest mistakes. Hence, if such problems occur, the editor should not hastily judge the situation with a particular explanation in mind. The editor's responsibility is to decide whether the submitted or edited manuscript can be published in its current form or to ask for an investigation from the author's affiliated institution when the problem appears to be severe.

If a submitted manuscript is under suspicion of plagiarism or redundant publication, the editor should check for plagiarism.

  • If the degree of plagiarism or redundant publication is not significant, the author can be asked to revise the manuscript to properly cite the source and continue through the review process.

  • If the plagiarism in the manuscript is so severe that it cannot be revised through partial modification, the corresponding author and all the co-authors should be informed. The information should include the journal's post-publication processes. The procedures may include refusal of further manuscripts from the relevant authors for a certain period, notification of the original author who committed the plagiarism, and notification of the research institution of affiliation.

  • Once plagiarism and duplication is confirmed, the editor should inform the copyright holder of the original manuscript (author and the publisher) and the author's affiliated institution of the situation. This can be understood as a part of the academia's self-regulation efforts. This prevents the problematic manuscript from being resubmitted, resulting in an increase in productivity of all journals and aids the effectiveness of the peer review system. This would lead the affiliated institutions to start investigation into the possible misconduct. It allows the identification of the one who is responsible and the severity of the errors the author has made; this is a way to protect authors with good intentions.

☞ Unlike fabrication or falsification of data, plagiarism and duplicate publication can be identified by the editor. Comparing a manuscript submitted to the journal with other manuscripts to check for plagiarism is also a responsibility of an editor. If it is discovered that the submitted manuscript has been plagiarized or contains redundant information, the editor should make a formal request for an investigation of the author's affiliated institution and ask for the results of the investigation.

Case study

Case study. Measures that the editor-in-chief should take when publication ethics are questioned

The Committee on Publication Ethics (COPE) has created flow charts to demonstrate the actions that the editor should take when ethical issues are raised about a published manuscript or a manuscript that is under review. It is extremely useful because it clearly shows the complete standard procedure that an editor should take. The following charts can be found on the COPE website (http://publicationethics.org/resources/flowcharts).

What to do if you suspect plagiarism

  • Suspected plagiarism in a submitted manuscript

  • Suspected plagiarism in a published manuscript

What to do if you suspect redundant (duplicate) publication

  • Suspected redundant publication in a submitted manuscript

  • Suspected redundant publication in a published manuscript

Changes in authorship

  • Corresponding author requests addition of extra author before publication

  • Corresponding author requests removal of author before publication

  • Request for addition of extra author after publication

  • Request for removal of author after publication

  • Suspected guest, ghost or gift authorship

  • How to spot authorship problems

Conflict of interest

  • What to do if a reviewer suspects undisclosed conflict of interest (CoI) in a submitted manuscript

  • What to do if a reader suspects undisclosed conflict of interest (CoI) in a published article

What to do if you suspect an ethical problem

  • What to do if you suspect an ethical problem with a submitted manuscript

What to do if you suspect fabricated data

  • Suspected fabricated data in a submitted manuscript

  • Suspected fabricated data in a published manuscript

What to do if you suspect a reviewer has appropriated an author's idea or data

  • What to do if you suspect a reviewer has appropriated an author's idea or data

5. Basic responsibilities of a reviewer

The reviewer should accept the request for review only if he or she has sufficient expertise on the subject for an adequate review. In the following situation, the reviewer should immediately return the manuscript describing the reasons for return.

  • One should avoid reviewing a manuscript if he or she lacks expertise or has participated in the requested research a similar research recently.

  • If a reviewer finds it difficult in to evaluate the entire content of the submitted manuscript or is able to evaluate the manuscript regarding only certain aspects, then it must be noted in the reviewer's report. The reviewer's field of expertise and focus of review should be described clearly.

If the reviewer agreed to review the manuscript, then the reviewer's comments should be submitted in a timely manner within the given peer review period.

The peer review should be done without consideration of any non-academic factors such as the author's nationality, religion, political beliefs, or gender.

  • If the peer review must be avoided due to a conflict of interest, one must adhere to the journal's regulations. If a reviewer is in any of the following situations, and the journal does not provide a guideline, the reviewer should notify the editor.

  • The reviewer should declare all personal, financial, academic, occupational, political, and religious conflicts of interest such as currently working or soon to be working in the same institution as the author(s), or if one of the authors has applied to the reviewer's institution to work. If any uncertainties are present, the reviewer should seek advice from the editor (Refer to Chapter 5: Conflicts of Interest).

The reviewer should respect the confidentiality of the content of the peer reviewed manuscript and the peer review process. Any information obtained during the peer review process must be kept secure, and any information obtained during the review process should not be used to the reviewer's own advantage.

  • Once peer reviews are complete, the manuscript should either be returned or disposed of. After online review processes, the reviewer must delete the copy after returning the review.

  • Individual meetings with the author should be avoided, and even if the identity of the reviewer is revealed, any questions or issues that require consultations with the author during the peer review should be reported through the editor.

The responsibility of the peer review lies with the individual invited by the editor.

  • If the reviewer seeks the advice of an expert in the specific field, even if the expert is a colleague in the same institution, the reviewer must still seek approval from the editor. Some journals put responsibility on each reviewer to seek assistance from another colleague as needed. If another researcher has contributed to the reviewer's comment, the editor should be informed of the names of those who provided assistance so that proper credit can be given. Even if the journal has allowed seeking assistance, if assistance was received, it is still wise to indicate this in the reviewer's comments.

If any major mistakes, misconduct, or misbehaviors in the data, or conflicting conclusions, errors, plagiarism, or duplicate publication are found during the review, the editor should immediately be informed.

6. Reviewers' comments

The manuscript should be evaluated to see if its characteristics suit the journal, if it contains new and scientific significance, if it is scientific and logical in the presentation of results and analysis, and if the content is coherent and understandable.

The reviewer's report should be organized by categories. They should also address the strengths and weaknesses of the manuscript appropriately and thoroughly so that the editor will be able to make an unbiased judgment.

In a critique, weaknesses must be clearly pointed out and suggestions for addressing such weaknesses should be provided.

  • The reviewer should check for illogical or biased assertions and make suggestions to clarify the interpretations of the results. He or she can suggest methods to further verify the argument.

  • When criticizing the manuscript's argument, the reviewer should provide the basis of such criticism, for example, by citing relevant references. Even if the reviewer decides that the manuscript is not up to publication standards, he or she should provide the reasoning behind the decision and advice for possible improvements in detail.

Unless stated otherwise by the editor, the reviewer should evaluate the entire manuscript.

  • If the reviewer does not have the expertise to evaluate one section of the manuscript but thinks this section should still be reviewed, he or she should notify the editor. In such cases, the reviewer can seek an expert's opinion or tell the editor that the review was conducted while assuming the section in question was sound so that the editor can make the proper judgement.

The reviewer's report should be written objectively and fairly. In order to obtain objectivity for the review, one must not have any conflicts of interest with the content of the research or the author. If there is a potential conflict of interest, the reviewer should decline the invitation with an explanation or must seek advice from the editor. If it is believed not to be a significant conflict and hence does not require declining the invitation, the reviewer can accept the invitation, but the situation must be declared in the reviewer's report.

☞ The reviewer must be able to advocate for both the author and the journal. 2 All effort must be made to provide both unbiased and constructive criticism on the side of the author while refraining from personal attacks. Review should be completed in a timely manner so that the manuscript can be processed without delay. Furthermore, the reviewer must make an effort to only allow manuscripts of the highest quality to be published in the journal by accurately evaluating the quality of the manuscript: more specifically by (i) reviewing any errors in the research process or conclusion, (ii) verifying that the research results support the conclusion, (iii) confirming that other research findings were properly attributed, and (iv) checking that the research has produced an original and meaningful conclusion.

III. Follow-up Measures for Published Manuscripts

1. Corrections

When journal editors encounter the following situations, they should consider issuing a corrigendum or erratum to rectify the error: (1) when errors are made during editing, (2) when minor errors in calculations or in descriptions are found after publication, or (3) when there is a need to edit the list of authors such as removing an author who does not meet authorship criteria or adding a deserving author.

2. Expression of concern

If a question raised is so severe that it questions the validity of the manuscript but has not been definitively verified, an expression of concern can be used to explain the situation to the readers and eliminate any potentially harmful effects.

  • The expression of concern can be published in the following circumstances: (1) when research misconduct is suspected, but an investigative committee has not yet been formed or the conclusion of the investigation is uncertain, (2) when the validity of the investigation into the misconduct is in question, (3) if the authors or the experts of the field have conflicting opinions on the research, (4) if the result of the research has the potential to harm the readers, (5) if a manuscript that did not receive authorization has been confirmed to have been submitted as a duplicate (in this case, a separate expression of concern for redundancy can be published).

☞ In the past, the expression of concerns were usually issued in medical journals to forewarn readers of problems that have been raised in the articles and to prevent any harm that may be caused to the reader and the public. For example, when a a new drug is being used throughout the world based upon the conclusion drawn in an article that the drug does not have any side effects, the editor can publish an expression of concern informing the readers of the problem while waiting for a conclusive decision so that the risks of the drugs can be reduced.

☞ If allegations of other research misconduct are raised, and if more time is required to conclusively evaluate the validity of the manuscript regarding the research methodology, the editor can publish an expression of concern to inform the readers. Instead of hastily retracting the manuscript without the suspicions being fully confirmed, the editor can explain the situation through an expression of concern. This reduces the dangers caused by extreme measures such as retraction and also allows the readers to obtain accurate information regarding the published article. After the expression of concern is issued, if the suspicions of research misconduct are cleared, the situation can be explained again to the readers. If the author of the paper is concluded to have committed misconduct, it should be retracted promptly to provide the readers with accurate information.

The editor should minimize the impact of the article in question by publishing the ‘expression of concern’ as promptly as possible. The announcement should be posted in the online database as soon as the decision is made, and should also be immediately published in the printed version of the journal.

  • The fact that the ‘expression of concern’ regarding the relevant manuscript has been issued should clearly be detectable while searching for the published article in the database. Whenever the article is searched or is downloaded, the expression of concern should be linked to the article. The journal editors and the manager of the journal database must make the greatest effort to ensure that any expression of concern or the revised articles are linked to the relevant article so that the readers can obtain the most accurate information.

When writing an expression of concern, the editor of a journal should clearly identify the article and reveal the specific circumstances that caused the problem.

  • The editor should ensure that the article in question can easily be verified in the expression of concern by clearly identifying the author and title at the beginning of the expression of concern; the sections and contents in question should be elaborated in as much detail as possible. In particular, whether the reason is due to research misconduct, or due to an honest mistake, or due to differences in the analysis of results must be clearly stated in the expression of concern. Under no circumstances should the expression of concern include hostile remarks or libelous comments, and should reveal only the contents in question in detail.

3. Retraction

Retraction is a process which announces to readers that the publication cannot be relied upon, and thus should be ignored. The editor should retract the published article within the following circumstances.

  • Publications are retracted when the results presented cannot be accepted due to research misconduct, or due to honest errors such as miscalculation or experimental errors. Retractions are also used when the research is confirmed to have been plagiarized or conducted in an unethical manner.

  • When a duplicate publication is confirmed without clear prior disclosure or proper permission, the reason should be clearly stated and the second duplicated manuscript should be retracted. Among the duplicated publications, the first published manuscript can remain published as long as the validity of the study results is not questioned. However, the fact that some of the content has been duplicated should be revealed in an ‘expression of concern.’

☞ In the following cases, (1) if only a portion of the manuscript has an error due to an honest mistake and (2) the validity of the manuscript can be sustained after correction, or (3) if only the author has been changed and not the content, then issuing a corrigendum should be considered first.

☞ If plagiarism has been confirmed in only a very small portion such as one or two sentences of the article, the editor should take the readers and the author who has been plagiarized into account and should make a decision on either issuing a corrigendum or retraction of the manuscript. During this process, the opinion of the original author should be fully taken into consideration.

☞ If only a small portion of the article has been duplicated, the editor should decide whether it is better to retract the whole publication or to issue an expression of concern that identifies the duplicate content and allows quoting from a previous publication. The decision can be based on the amount and the importance of the duplicated content. The editors should make an effort to make the best decision from the reader's perspective and must remember that the decision should be based not on punishing the author who committed misconduct, but on ensuring the journal publication ethics.

The notice of retraction should include specific information on who initiated the retraction and why the article is being retracted.

  • At the beginning of the retraction note, the title and the author of the manuscript should be clearly identified so that the manuscript can be easily linked to the retraction.

  • The reasons behind the retraction should be as detailed as possible. In particular, it should be very clear whether the reason for retraction is due to research misconduct or due to an honest mistake. If the error occurred due to miscalculation or during the experiments, (i) how the error took place in what steps, (ii) problems that resulted from the error, and (iii) which part of the study were not affected by the error, should all be clearly stated. If the error occurred due to research misconduct or duplicate publication, this fact should be accurately summarized and explained to the readers.

☞ Articles get retracted when they have been published without securing validity. This is to ensure the publication ethics, and it should not be considered a punishment for the author. Publications are retracted due to research misconduct such as falsification, fabrication, and plagiarism. But retractions are also applied in cases of honest mistakes such as miscalculation or experimental errors. Retractions should be considered a purification process to improve the standards of academia. In no cases should hostile or libelous comments be made regarding the author, and only the problematic aspects of the article should be discussed in detail.

☞ The editor should make efforts to create a retraction note that provides accurate information to the readers and is acceptable to all parties involved. However, when the retraction was specifically due to research misconduct, the authors quite often request that instead of clearly stating the fact, the retraction note be written in a roundabout manner. In such cases, if possible, it is recommended that exact phrases are quoted from the report of the investigative committee of the research institution. The editor, while fulfilling his or her duties to truthfully inform the readers, could cooperate with the author regarding the expressions and phrases but still should issue the retraction note in a timely manner.

Retraction of a manuscript should be issued in the following ways.

  • Publish the retraction note on the online database promptly after the decision has been made and publish the retraction note in the first printed version right after the decision has been made.

  • The article should be retracted as promptly as possible. This is to minimize the probable damage to other researchers, who may waste their time and effort or draw incorrect conclusions based on the invalid publication. This is also to minimize cases where it may lead to a meta-analysis that reproduces the same misleading results.

  • Only articles that have already been published can be retracted. If retraction was issued after being published online but before being printed, the retraction note should be linked to the online article using the same methods stated above. However, the printed journal does not need to publish the retraction note.

The retracted article should be open to the public but also clearly state that it has been retracted. When the article is found through a database search, it should be clear that the manuscript has been retracted.

  • The retraction note should easily be visible on the page that contains the article. The retracted article should be available to everyone regardless of the journal or online database disclosure policy.

  • It should be clearly indicated in the retracted article file that the paper has been retracted. It should also be attached so that the retraction note is automatically downloaded together whenever the manuscript is downloaded.

Retractions should be requested by the author or the publisher and authorized by the editor. The editor can also decide to retract a manuscript under his or her own authority.

Among the authors, an author or all of the authors of the manuscript can request a retraction from the editor. In such cases, the retraction must be justified, and if only a few authors are requesting the retraction, the circumstances must also be clearly explained.

IV. Utilization of Similarity Test Software

1. Two pillars of similarity tests: search engine and database

The search engine and the database are the two pillars of similarity tests. This is because the subjects of comparisons are compiled in a database and the search engine is used to check how similar the data in the database are to the selected text.

Similarity can be tested using an index of internet sources using Google or other search engines, or by utilizing a separately compiled database and search engine.

  • CrossCheck, which will be explained in detail below, utilizes a database created by CrossRef. The search engine is one created for commercial use by iParadigms. Other search methods, such as the free web-based search service, eTBLAST (http://etest.vbi.vt.edu./etblast3) developed by Viginia Bioinformatics Institute; MEDLINE, a database for biomedical academics; and arXiv, a database developed by NASA for mathematics, physics, and computer science, can all be used.

  • PlagScan (https://www.plagscan.com) is a program that compares the selected text with online data and also allows users to add their own text to the PlagScan database. The Korean service CopyKiller (https://www.copykiller.co.kr) also allows the selected text to be added to its own database. Hence, there are cases where the similarity rises if the same text is searched again. In addition, there are websites such as Copyscape (http://copyscape.com) that use the Google search engine to check how similar a selected webpage is compared to other existing webpages.

  • When comparing Korean sentences, the similarity test software developed in English-speaking countries may not be well-suited to handling particles or changes in word order effectively. Professor Hwan-Gue Cho, a computer scientist at Pusan National University is developing DeVAS and DeVAC (http://devac.cs.pusan.ac.kr), programs specifically tailored to detect plagiarism in Korean documents.

Manuscripts that have yielded copyright to publishers are not easily found on the internet. As a result, there is a limit to testing similarity for new manuscripts to be submitted using only online data. Hence, there is a need for journals to utilize academic database search tools such as eTBLAST and CrossCheck While it is important to use a good search engine to thoroughly screen for similarities, it is more critical to establish a broader database. Here, cases using the most widely used academic database, CrossCheck, will be illustrated as examples, but first, CrossRef, the organization that provides the CrossCheck service, will be introduced.

2. CrossRef: The organization providing the CrossCheck service

CrossRef, which provides the CrossCheck service, is an organization that registers and manages digital object identifiers (DOI).

  • A digital object identifier (DOI) refers to unique identification number assigned to digital data. CrossRef links a DOI to the digital data and connects the DOI and the site providing the data.

  • Even if the URL of the saved data changes, users who use the DOI numbers to access the data will not be affected. This is because CrossRef is responsible for linking the DOI to the newly changed site address. The DOI is a permanent link.

  • The format of DOI is similar to “http://dx.doi.org/prefix/suffix”. For example, a manuscript published in 2007 from the journal Automatica from Elsevier has the DOI “http://dx.doi.org/10.1016/j.automatica.2007.xx.xxx”.

CrossRef was launched in early 2000 as an association of many scholarly publishers. As of November 2013, there were 24 employees and 18 members of the Board of Directors. Furthermore, many working groups and committees have been established. According to 2012/13 statistics, 3 around 4,500 publishers and academic societies participate in CrossRef and over 63 million journal articles have been given DOI names. Among the content items, 7 million books and around 50 million journal articles have a DOI, comprising 11% and 82% of documents with a DOI, respectively.

3. CrossCheck: similarity test service provided by CrossRef

CrossCheck refers to a service provided to member publishers and academic journals by CrossRef to screen and prevent plagiarism, and it consists of the iThenticate search tool and CrossCheck database.

Crosscheck service = CrossCheck Database + Search Tool iThenticate

CrossCheck is the name of the service and at the same time the name of a database.

The search tool iThenticate is a software program that compares and analyzes texts.

  • The search can be conducted on the web (http://www.ithenticate.com), and has also been set up to allow searching from the homepages of journals that subscribe to the CrossCheck service.

  • The search tool iThenticate is a program developed by iParadigms, who also developed Turnitin. This is why CrossCheck is a paid service. Academic journals, once they become members of CrossRef, are provided with the CrossCheck service and are required to pay for an annual license or to purchase credits depending on the number of searches.

The CrossCheck database that has compiled comparable texts is based on the data provided from CrossRef member journals.

  • According to 2012/13 statistics, the CrossCheck database includes 38 million items.

  • In addition to CrossCheck, third party databases such as PubMed and arXiv.org are available, and internet data can also be included in the search indexes.

4. Using CrossCheck

The procedure for testing similarities is as follows:

  • Login to http://ithenticate.com and create a folder to submit the selected manuscript. The search parameter can be designated while setting up the folder. Among the available search indexes, one or all among the databases <CrossCheck>, <Internet>, and <Publications> can be selected. If <CrossCheck> and <Internet> are selected, both the CrossCheck database and internet database collections are included in the data searched. In order to include a third party database as described above, <Publications> can be selected.

  • An exclusion list can also be determined. Quoted material, fixed phrases, and bibliographic material can be excluded from the test, and sections such as ‘Abstract’ or ‘Methods and Materials’ can also be excluded. The reason ‘Methods and Materials’ is allowed to be excluded from the test is because there are topics in which the author has no choice but to use wording similar to that of other published documents.

Fig. 1. shows a screenshot of a folder-setting page that excludes quoted material, bibliographic material, and fixed phrases and that includes CrossCheck, Internet, and third party databases in the search.

  • Once the settings of the folder have been verified, the similarity test begins. The program supports a variety of file formats such as PDF, Microsoft Word, and HWP (Hangul word processing file).

Fig. 1

Folder settings display

ch3f1

Fig. 2. presents a similarity report using two uploaded documents. By clicking on the similarity percentage, one can see which documents the selected manuscript is similar to and by how much. The selected manuscript and the similar manuscripts can be compared in many different ways.

Fig. 2

A screenshot with two files uploaded

Similarity is shown as a percentage.

ch3f2

Figs. 3. and 4. show screenshots when using the text mode and the document viewer, respectively.

Fig. 3

A case in which the document viewer is being used

ch3f3
Fig. 4

A case in which the text mode is being used

ch3f4

5. The significance of a similarity test using a search tool such as CrossCheck

The goal of using a search tool is plagiarism prevention. A program such as CrossCheck cannot be a mechanical tool for checking plagiarism.

  • The similarity report shows the degree of the similarity of the documents and not the degree of plagiarism. Just because the similarity is high, it does not mean that it has been plagiarized, and vice versa. However, the higher the similarity percentage is, the more closely the potential for plagiarism should be examined further.

  • It is important to examine in detail in what ways the submitted document is similar to other existing manuscripts as shown in Figs. 3 and 4, instead of simply relying on the numerical similarity value. What is important is the content, and the degree of similarity between the content of the submitted manuscript and the existing manuscripts can only be evaluated by the editors and reviewers with expertise.

  • Usage of CrossCheck can help prevent a researcher who may have committed plagiarism from submitting the manuscript to CrossCheck member journals; in addition, it can prevent a researcher from plagiarizing from the CrossCheck member journals.

  • Each journal can use the similarity test in its own way as a reference in the manuscript evaluation process. For example, if the similarity percentage is higher than a certain value, the author can be invited to rewrite and resubmit the manuscript without entering the peer review process.

Similarity detection software such as CrossCheck should be used to increase the integrity of researchers. Its goal is not to mechanically check for plagiarism.

References

1 

DJ BenosKL KirkJE HallHow to review a paperAdv Physiol Educ2003274752

2 

Committee on Publication Ethics (COPE)Codes of conduct and best practice guidelines[Internet]COPEcited 2015 Dec 17Available from: http://www.publicationethics.org/ resources/ code-conduct

3 

Committee on Publication Ethics (COPE)Flowcharts[Internet]COPEcited 2015 Dec 17Available from: http://publicationethics.org/resources/flowcharts

4 

CrossRef2012/13 CrossRef annual report[Internet]OxfordCrossRefcited 2015 Dec 16Available from: http://goo.gl/C0wZkA

5 

CH HamOH KwonSY KimGood publication practice guidelines for medical journals2nd ed.SeoulKorean Association of Medical Journal Editors2013

6 

International Committee of Medical Journal Editors (ICMJE)Recommendations for the conduct, reporting, editing and publication of scholarly work in medical journals[Internet]ICMJE2014cited 2015 Dec 17Available from http://www.icmje.org/recommendations/

7 

Ministry of Education, Science TechnologyNational Research Foundation of KoreaGood research practices' guidelinesSeoulNational Research Foundation of Korea2011

8 

Singapore Statement on Research Integrity[Internet]cited 2015 Dec 17Available from: http://www.singaporestatement.org

9 

E WagerCoping with scientific misconductBMJ2011343d6586

10 

E WagerV BarbourS YentisS KleinertRetractions: guidance from the Committee on Publication Ethics (COPE)Croat Med J2009505325

11 

E WagerS KleinertResponsible research publication: international standards for authorsA position statement developed at the 2nd World Conference on Research IntegritySingaporeJuly 22-24, 2010T MayerN SteneckPromoting research integrity in a global environmentSingaporeImperial College Press/World Scientific Publishing201130916

Notes

[2] DJ Benos KL Kirk JE Hall How to review a paper Adv Physiol Educ 2003 27 47 52.

[3] CrossRef 2012/13 CrossRef annual report [Internet] Oxford CrossRef cited 2015 Dec 16 Available from: http://goo.gl/C0wZkA

    About
    Publication
    Events
    Certification
    Membership
    News

    Copyright by Korean Council of Science Editors
    The Korea Science & Technology Center 2nd floor 22 Buk-7-gil, Teheran-ro, Gangnam-gu, Seoul, 06130, Korea
    Tel: +82-2-3420-1390 Fax: +82-2-563-4931
    Privacy Policy