Is it ethical for departments to have publication requirements for students

Question

This question and this question suggest that it is not uncommon/unheard of for departments to require publication by students (generally masters and doctoral) either to get credit for a course or graduate.

This seems to me like a way of off loading the assessment of students to peer reviewer, and therefore it seems ethically questionable. Is there a pedagogical reason to require externally peer-reviewed publication?

Answer 1

For an individual course (assuming a time-limited scenario, e.g. a semester, quarter, or even year), it's certainly ridiculous to have a publication requirement. The timing is far too variable, and even the best researchers can have papers rejected or discover before submission that they've been scooped. If it were possible to reliably publish a worthwhile paper within the timespan and amount of effort required for a typical course, then there would be a lot more papers published.

A requirement of submission is not as absurd, but still unreasonable. I see no advantages over a mock submission sent to the professor teaching the course, but there are severe disadvantages: it can put pressure on students to make inappropriate submissions, and it can waste the time of editors and reviewers on submissions that may be withdrawn once the course requirements are complete. My cynical interpretation is that submission requirements deliberately use the threat of professional embarrassment to encourage students to work harder, and that seems unethical.

The one scenario where publication requirements can make sense is for a degree program. For example, some PhD programs require that a dissertation must be based on peer-reviewed publications. Of course that's reasonable only in fields with a sufficiently rapid publication cycle, and even there I'm not fond of this idea personally (but I wouldn't call it unethical).

Answer 2

First things first: a doctoral student should be expected to publish something other than a thesis. Even if she has no desire to enter into a research profession after the PhD, publication of scholarly articles should be an important milestone in the process.

Beyond that, however, expecting that a student—whether a master's student or a PhD student, or even worse, an undergraduate—publish a paper as part of a single "course"—is absurd for a multitude of reasons. First, in the context of a single educational course, the time spent will almost certainly be unsuitable for the preparation of a manuscript; moreover, given the lag times in between submission of an article to a journal or conference and its acceptance, it is unlikely that it can be completed within either a trimester or even a semester, which means that incomplete grades will likely be par for the course.

Thus, since it serves no real valid educational purpose for the students—since the work isn't being evaluated by the educational staff whose job it is to provide instruction—such behavior is extremely questionable, and very likely unethical. This is doubly so if the only criterion for grading is the acceptance of the paper in an external journal, particularly since there are so many "pay-to-publish" journals out there that will publish anything, given the page charges.

Now, I do require that students prepare something like a research article for one of the courses I teach. However, I do that as an exercise in preparing them for writing research articles. I have no expectation that they would bother to submit these papers to actual journals—the material just isn't sufficient for that. However, in terms of learning how to write a paper—mentioning relevant literature, explaining their methodology, clearly demonstrating and illustrating their results, there is nothing comparable. You learn to be a researcher by doing research—and that includes writing about research!

Answer 3

The answer to this question may depend on the country and possibly on the field. The following is based on a European CS perspective.

As other answers have noted, asking for a publication within one year or within one semester is - at least for comparably inexperienced (pre-Master) students - a rather unrealistic or at least highly uncontrollable goal (because successful publication is dependent on a number of external factors, not the least of which is what else is submitted to the same venue).

Asking for publications as a prerequisite for graduation, on the other hand, is partially a different topic, which depends highly on the degree aimed for:

• A Master's degree is commonly thought of a certification for specific verifiable skills. Everyone who invests the necessary effort and fulfils the pre-defined requirements is supposed to be able to achieve that degree. As such, I agree it is ethically questionable to outsource the evaluation of a part of the preconditions for that degree to an external body rather than evaluating the work oneself in a somewhat stable manner (i.e. in a way that two students who did more or less the same work get more or less the same mark). Peer reviews wouldn't judge the work by its fulfilment of the rules imposed for the Master's degree, but based on its contribution, suitability to the venue, and its relative value to all other current contributions.

• A doctoral degree, in contrast, is not a degree you can reproducibly achieve for doing certain defined tasks. In a way, upon starting to work for a doctoral degree, one leaves the world of studying and doing things in order to fulfil pre-defined requirements, and one enters the world of being an active researcher who does things in order to find new knowledge - which of course means that the final result is partially outside of one's realm of influence. A doctoral degree is a degree that is granted as a result of contributing to the body of knowledge. Hence, any particular work effort that served for someone to get their doctoral degree will not do for anyone else (because that particular contribution is not new any more), and in the hypothetical situation that all knowledge in the universe has been discovered, no-one can get a doctoral degree any more. This has rather nicely been illustrated in a webcomic. In this case, requiring peer-reviewed publications makes sense in several respects:

  • Peer reviewers evaluate the work exactly by the desired criteria, i.e. whether the work is a contribution that is currently notable enough to warrant publication.
  • Rather than being just an idea that someone uttered, the new finding is validated by the reviews and thus has reliably extended the overall amount of knowledge.
  • Due to the aforementioned fact that the amount of knowledge that mankind is familiar with is constantly changing, what was a disadvantage in the case of the Master's degree is exactly the desired outcome here: By trying to make it a peer-reviewed publication, the work is evaluated in comparison to the current state of the art and all other work currently being done.

tl;dr: Binding a Master's degree to successful peer-reviewed publications is questionable as the degree should be granted whenever you prove your skills by completing a pre-defined set of conditions such as exams. Binding a doctoral degree to successful peer-reviewed publications, on the other hand, is the entire point, because a doctoral degree says This person has successfully expanded the knowledge available to mankind. These accomplishments are not measured by exams, but the same way as the quality of all research is measured: by peer reviews.

Answer 4

The problem certainly seems fishy and whether it is unethical or not probably depends on what the (local) rules and regulations governing courses will be as well as the possible ramafications of the process relative to those regulations.

The main issue for me is what will be the purpose of having publications as a requirement for students (not at graduate level). There is of course nothing wrong if undergraduate papers can be published, and experiencing the publication process can be valuable. But as one of the posts referred to in the question states, it seemed as if the requirement was to submit, not to publish. I do not see any particular value in that experience that could not be replicated within the department itself.

Since any publication process requires quite a chunk of time I can see a major problem in the timing. In my system, a course should be possible to complete within the stipulated time (corresponding to the number of credits). Imposing a system where publication is part could (really would) clearly violate such limitations. So, a follow-up question would be "are there systems where courses can be open ended?" or "can courses be required to last until an un-controllable goal is achieved?". If the system allows such cases, then the requirement would be ok from a legal point of view. I would still think it is in a grey zone.

Your point of off-loading the assessment is part of the shady picture. I would have thought that each university regulatory system would have some, at least, recommendations on how quickly assessments must be done and, again, assessing through journal peer review would require at least as much time as the length of a typical course itself. So as a whole, I think that using paper submission and/or publication as part of course requirements can be violations of local assessment regulations and very poor behaviour. I would call it unethical if it means unnecessarily prolonging the students graduation or impacting on their ability to get student's loans (equivalent) due to not getting credits in time (impacting their ability to take responsibility for their timing).

Answer 5

I would like to add some argument to the existing answers, which I first give in distilled form for clarity and on whose details I will elaborate afterwards:

Main Argument

As scientific publications are reports of research, requiring students to publish essentially requires them to successfully perform research. The essence of research is to explore the unknown and thus its results (and success) are not predictable. Relatedly, successful research requires amongst others luck. And luck is something that successfully finishing a course, degree, etc. should not depend on too much.

Some comments

• I think publications which do not require original research such as reviews are not what we are talking about here. (It may be a way out though: I know somebody, who at some point strongly considered fulfilling the publication requirement of a PHD by publishing a review after many unpublishable research approaches.)
• While failed research idealistically is also worth reporting (as it has produced the knowledge that some approach does not work), it certainly is much more difficult to publish (leading to the publication bias). This holds even more so, if the reason for failure is not that you have run into some long cul-de-sac but there are many short culs-de-sac.
• There is some research which is very straightforward (and thus has little luck involved) as it basically consists of applying some established method to a new object where no big surprises are to be expected. For example: Determine the distributions of volumes of organ A of rats fed with diet B. Such research is however arguably better published in one big study going through a series of objects (e.g., organs and diets) instead of thousands of small publications, which all are essentially the same.
• When the impact of luck is sufficiently low that assessing students by publications can be considered fair depends on several factors such as the publication threshold of the field or the typical duration of experiments. Arguably this only happens at PHD level in most fields, if at all.
• Another related problem is that people being pressured to produce successful research are more likely to fake data, plagiarise or commit other types of fraud. Now, while this is a general and not totally unavoidable problem with students, requiring this to be published may also harm others who try to build upon the respective publication.