Who counts as an author when reporting educational research?

The question posed in my title does not refer to citation counts or the prestige of having one's name on papers considered seminal in a field, but rather to the more basic question of how a person qualifies to be, or is disqualified from being, named as an author of a published research paper. Authorship is considered very important in the academic world, and decisions on whether to hire staff, offer them permanent established posts (‘tenure’), or promote them, are often made in part based on judgements about the quality of a researcher's list of publications. Therefore decisions about who should be named as an author on a paper that will be published in a research journal are very important to the researchers concerned.

Those outside of academia might suspect that even though this is important, it is unlikely to be problematic, as those involved in writing a paper will clearly know who they are. Yet, alas, things are not so straightforward, and occasionally journal editors become drawn into disputes about who should, or should not, be named as authors on a particular article. It has been known, for instance, for a journal referee to have been sent a submission to review, only to identify part of what they are reading as their own writing.

I wrote that!

Now that could obviously happen if there has been deliberate plagiarism, where one person knowingly copies some of the work of another and claims it as their own. That is a kind of theft of intellectual property (it would be infringement of copyright whether the original material is formally published or not) and fraudulent, and clearly unethical. Such deliberate cheating may occasionally occur, but as with other kinds of scientific fraud, such as fabricating data, is likely to be rare. At least, one seldom hears of such cases in educational research. It would be foolish to deliberately plagiarise in this way, as given the ubiquity of modern search engines it is very likely that any such act would be revealed after publication even if the problem was not spotted by reviewers or editors. More problematic is the issue of ownership of jointly shaped text.

On submitting work for publication authors are asked to confirm the originality of their work – that it has not already been published. Most journals will not knowingly republish work (and often another publisher will have already been granted a license to exclusively publish and distribute the original work). That seems a simple and clear principle, but actually it may not be so obvious exactly what it means in practice. This editorial is original…but I expect I have published earlier work using the word ‘original’, and perhaps even the phrase “it may not be so obvious exactly what it means in practice”. No one is likely to claim this makes this editorial previously published, but it does lead us to ask what degree of overlap should lead to a judgement that a piece of writing should not be considered original.

Often researchers work on a series of related studies within a programme of research, and even within a single project there are often likely to be a number of papers reporting different aspects of the work. When two papers report from the same project, they will be based on the same overall conceptualisation of the research topic, and derive from the same research design. Even if two such papers report very different aspects of results, they are inevitably going to share much in terms of the literature and methodology sections. Indeed it would look rather odd if two papers from the same project drew on very different literature or offered rather different justifications for research design. It seems quite reasonable that these sections of different papers are going to have much in common.

So this complicates the question of when two papers have too much overlap. Presumably they may share sentences in common. We might wonder if it is acceptable for them to have paragraphs, or even whole sections (‘research design’) in common. What seems reasonable is that both papers might start from the same draft, and then be developed so as to customise them in accordance with the specific foci and results reported in the different papers. The British Educational Research Association, (2000) recommends that a project should be reported in a detailed technical report and where this guidance is followed the technical report might be used as a common base text from which to develop different papers for submission to research journals. Ultimately, referees will use their judgement to decide whether a particular submission can be considered original, based largely on its overall contribution to new knowledge extending significantly beyond previously published work.

In terms of the scenario posed above, a problem arises when an author decides to reuse part of a previous publication but was not the sole author of the earlier piece. That need not be a problem if the same author team is involved in the new composition. However, when a reviewer is sent a submission and recognises his or her own prose in work which does not include him or her as an author, the issue will be raised with the editor. As so many of us undertake joint writing with a range of different colleagues, there is clearly scope for such issues to arise if we are not careful. When manuscripts move back and forth between computers for new iterations of writing, it may be very unclear by the end of the process who actually wrote what.

This suggests that it may be inappropriate to look to previous co-authored writing as a starting point for a new manuscript as a sole author or with a different team of authors. There is probably a strong case for saying that our thinking will move on over time, and often we might in any case be better advised to start new writing from scratch: but sometimes some of our most prized prose suffers from collaborative editing, and it is not unreasonable to look to put material intended for, but then edited out of, a collaborative article into other relevant writing. Given the ability to save copies of different versions of manuscripts, it is possible to keep a file of any original writing we may undertake that is to be contributed to collaborative writing, and so to return to our own original draft (before it was modified by, and spliced with the work of, our co-authors) as the basis for future writing. This certainly seems a sensible precaution to avoid the embarrassment of being accused of unprofessional behaviour by unintentionally misappropriating the work of our colleagues.

Writing and authorship

Although authorship issues can involve who wrote some sections of a manuscript, more problematic disputes may revolve around issues other than the identify of the writers. These disputes arise because authorship is not simply about who does the writing (APA, 2009). It is quite possible for major contributors to a research paper to actually have no part in drafting the manuscript, and only to see the work once it has been formed into a full draft. (It is worth noting that even in this situation it is important that all authors approve the version of a manuscript submitted for publication, and the corresponding author is expected to confirm this is the case when submitting.) So there can be a question over what counts as authorship, given that an author does not necessarily need to do the writing.

There is a general principle here which is quite clear. An author of an academic work is someone who made a substantial intellectual contribution to the work. Academic ethics, and the guidelines used by most journals, therefore require both that:

(a) all those who made substantial intellectual inputs to a work should be named as authors;

(b) only those who made substantial intellectual inputs to a work should be named as authors.

The principle seems very straightforward. Unfortunately, application of the principle requires interpretation and evaluation of the level and nature of the input from different contributors to the research. Probably matters are not helped by a common human tendency to recognise our own involvement and contribution in a project as significant, whilst being less aware of, or not valuing as much, the contributions of others.

Doing research and authorship

Some research in education is carried out by a single researcher, who is responsible for all aspects of a project, and then there should be no question over the authorship of project outputs. Conversely, large projects may have several investigators who collectively contribute to the overall conceptualisation of the project, whilst individual investigators manage particular researchers who execute discrete parts of the design. Even projects with a single lead investigator may involve a team of researchers, perhaps with their own specific roles in the project – such as being involved in particular aspects of data collection and analysis.

Judgements need to be made about which activities amount to substantive intellectual inputs, or are better seen as being necessary but purely technical contributions. Such judgements are likely in part to depend upon the kind of research being undertaken. For example, if research has a ‘confirmatory’ nature – such as survey work or quasi-experimental work – then research assistants who collect data through questionnaires or completing observation schedules (designed by others) may be judged to be technicians undertaking supervised tasks. They are carrying out skilled work – but could readily be replaced by another trained research assistant with no loss to the project. Their role is essential, and they may be worthy of a named acknowledgment for undertaking the role, but they are not likely to be considered as authors of research outcomes. The same applies to someone hired to enter data into a computer so that it can be analysed by an existing software programme according to protocols designed by researchers, or someone acting as an administrative assistant for a complex project.

In projects of a more interpretive nature (where data is not firmly framed by categories established at the outset, but rather needs to be analysed through an emergent analytical frame), those carrying out interviews, or making observations, may need to have greater engagement with the purposes and conceptualisation of the project. Data collection may require considerable interpretations drawing upon an understanding of the rationale of the project and the theoretical perspective being adopted – for example in making detailed observation notes relating to project themes, or being able to ask appropriate follow-up questions in response to participant comments in interviews. Data analysis is likely to require much more than following a protocol provided by more senior members of the project, and there may need to be a stage of reaching inter-rater reliability involving detailed discussion between researchers.

In these cases it may be that data collection and analysis can only be effectively undertaken by members of the project team who are able to make significant and substantial intellectual contributions to the project, and so are entitled to be considered authors of outcomes. A simplistic but useful question to ask may be whether the team members concerned could have been substituted during the project by replacements who are formally qualified but had no special knowledge of the particular research, without diminishing or disrupting the project. If team members could not be replaced by alternative qualified candidates provided with minimal briefings and instructions, then it seems likely these team members should be considered as authors of the project outcomes.

Thinking about research and authorship

As the main issue in determining authorship is intellectual input, an author does not always need to have been involved in the collecting and analysis of data, either directly or even in a supervisory role. Key intellectual work on a project includes the conceptualisation of the research area, the development of methodology (including instrumentation and analytical methods) and a research design, and the interpretation of results. A person who takes a major role in these stages of a project is making an important input and is likely to deserve to be considered an author of project reports.

As always, there are judgements to be made about the degree of input which justifies more than an acknowledgement (which can be considered somewhat equivalent to a citation) and entitles a contributor to authorship. Offering comments on a draft research plan or paper will most often justify only acknowledgement unless those comments highlight major issues, and are extensive enough to contribute to substantive redirection of the work. On the other hand, prolonged engagement in discussions about the project that influence major research decisions may justify authorship even without hands-on work with data or supervisory responsibility for research assistants.

Judgements about such matters may be nuanced, and it is not surprising that occasionally there are sharp differences of opinion among colleagues. However, where those involved are keen to make the right decisions it may be useful to keep in mind that there is an ethical imperative for researchers to ensure both that deserving authors are not excluded from authorship, and equally that authorship is only offered on the basis of a genuine substantive contribution that merits that level of credit.

Supervising and enabling the research of others

Research supervisors and research group heads may be keen to support the development of research students or other junior researchers by offering them authorship on papers deriving from work within the research group. This is a perfectly correct aspiration, but must be based on involving the junior researchers in the work of the project at a suitable level to justify credit as authors. That ensures they obtain valuable research experience, and acquire or practice new skills, and so develop as researchers (as well as deserving authorship). Simply including students in projects to undertake routine tasks may be justified where it is an expected part of their duties (more often the case in some countries than others), or if it provides desired opportunities for supplementing income, but does little to develop the researcher or allow them to merit author status.

A related issue is whether research students should include their supervisors as authors of papers they write about their thesis projects. The decision here needs to be based upon the same considerations as we have seen in other cases – that is whether a substantial intellectual contribution has been made by the supervisor. The occasional brilliant research student may feel they have little need of substantive intellectual input into their work by their supervisor or other advisors: but in general if a research supervisor is not making substantive intellectual input into the development of the student project we might wonder if they are fulfilling their role as a supervisor. Most supervisors suggest ideas and directions for a project, offer advice at all major stages of the work, and regularly critique the students' ideas and writing. That would normally amount to a substantive intellectual input.

In practice there are cultural differences across educational and institutional contexts. For example, sometimes a student is assigned a project to work on (as is common in the natural sciences) which is already outlined by the supervisor and has been designed to fit with a particular conceptualisation of the field; whereas in other national contexts it is more usual for the student to identify a research issue or problem and offer an outline research proposal as part of the process of application and admission to the university. In some countries a thesis is normally based around a set of published/presented papers acknowledged as authored by a team that includes the student; whereas in other national contexts the thesis is a stand-alone monograph that is presented as ‘all the student's own work’. In the latter case, it is well-understood that the student is supported by a supervisor or advisory team, but that the sole authorship of the thesis reflects how the final research decisions, the actual collection and analysis of data and the drafting and honing of the thesis is undertaken by the student. It is important to recognise in such cases that it is taken for granted within the University that the named supervisor(s) made substantive intellectual contributions to support the student's development and research project, and the naming of a single author for the student thesis is anomalous with respect to how authorship will be understood when publishing the work in academic journals.

Research supervisors are well aware that some students seem to manage their projects with modest input that is largely at the checking-in-and-confirming-all-is-well level, whilst other students seem to need frequent and very specific guidance (almost amounting to being told what they should do next) at most stages of the research process. Given the variation in experience here, it seems sensible to suggest that a default assumption is that a research supervisor will normally have provided input into the student project at the level that amounts to authorship of any papers substantially deriving from the thesis project – but that the student and supervisor should discuss and agree whether this is so in relation to particular submissions. It is certainly quite conceivable that sometimes publishable work from a thesis project may be best seen as sole-authored. A research review based on a literature chapter, for example, may in some cases only have had supervisory input at a general level worthy of acknowledgement, rather than authorship. In general, though, it is likely that a research student's work will reflect significant and substantial intellectual input from the advice and feedback offered in regular supervision. The student usually takes major responsibility for research decisions in their own thesis project, and so would normally be the first named author, although if the project is part of an ongoing programme of work in the lab or group, the authors may still agree it is appropriate for the supervisor to be named as the corresponding author.

If research supervisors generally deserve to be considered as authors on papers reporting their students' work, we might wonder about others who are involved in facilitating the work of researchers. Probably the issue will be most pertinent in the case of a head of institution/lab or research group. The key question here is likely to be whether the input is both substantial and relates to the intellectual content of the work. So a head of group who establishes a research programme and strategy for the group; and who holds regular meetings to discuss the progress of ongoing projects within the group; and who offers detailed guidance and critique on projects, might be considered to make a substantial intellectual input into the work of the different projects. However, a head whose role in relation to a particular project is largely administrative (perhaps in terms of ensuring the funding, infrastructure and research environment are in place to allow researchers to go about their work) may be vital for the work to proceed, without making the kinds of contribution that would be considered to amount to authorship.

Authorship decisions must be consensual

Occasionally journal editors find a dispute develops over a submitted paper when another potential but unacknowledged author comes forward to claim authorship. This is not an ideal situation for the academics involved, especially if a paper has to be withdrawn from consideration for publication whilst the parties look to resolve their difficulties. As the RSCPublishing, (2012) ethical guidelines for authors acknowledge “there is no universally agreed definition of authorship”, yet as I have explored here there is a clear principle that needs to be sensibly, and sensitively, applied in each particular case. Ultimately it is the responsibility of the corresponding author submitting a manuscript to a journal both to ensure that named co-authors have approved the version submitted, and to be confident that all those entitled to authorship of the work are properly credited in an agreed order. Problems should be avoided if the issue is considered at an early stage of a project, and if those involved in the work are in clear agreement about the authorship of any particular output from the point at which the writing is planned. Discussing the issue within the team can help to clarify any different perspectives on the extent and significance of particular contributions and allow potential problems to be considered – and hopefully resolved – before there is a manuscript for the agreed authors to put their names to.

References

1. American Psychological Association, (2009), Publication Manual of the American Psychological Association (6th ed.). Washington DC: American Psychological Association.
2. British Educational Research Association, (2000), Good Practice in Educational Research Writing. Southwell, Notts.: British Educational Research Association.
3. RSCPublishing, (2012), Ethical Guidelines and Conflict of Interest. From http://www.rsc.org/Publishing/Journals/guidelines/EthicalGuidelines/EthicalGuidelinesandConflictofInterest/sect3.asp.

---