Paper Authorship and Importance of Publication

[swagatopablo]

I am doing PhD in telecommunication, networking and want to pursue an academic career. I understand that to get a good post doc position, I need good publications. Unfortunately, publications became the de facto scale to measure one's performance. But authorship is a tricky issue and must be managed diplomatically to get maximum mileage from your research. In this regard, I have a few questions for experienced researchers. I know that this forum is primarily for graduate school applicants but still I believe there are some senior graduate students, research fellows or may be even faculty members. A lot of current or prospective graduate students may find the answers relevant.

 

1) Signing a research integrity declaration was compulsory at the start of my program and it specifically mentioned that only the people who contributed to the idea, analysis or writing should be included as authors and not necessarily the PI of a project or the professor who brings the funding. But as I see around me and for myself, the supervisors are frequently detached from ground level details of a research topic and yet, it's somewhat of custom to include the supervisor as a coauthor. In fact, I am yet to come across a single paper from anywhere where a professor isn't included. I don't know whether it's the same for all schools, but my question is whether a publication without my supervisor counts as part of my work towards finishing my PhD or it's my duty as a graduate student to give papers to my supervisor?

 

2) Suppose one person (whose name is X) forms the core idea, does the theoretical analyses and contributes to the actual development. But apart from these tasks, there are things like writing long codes, running the simulations and then writing a paper-tasks which require less intellectual capability, less knowledge but sometimes may be more time consuming, demanding a bit more familiarity with a specific software package etc. Very often, these tasks are performed by someone else (suppose Y) and when a paper is written, the author names are X, Y, <<Supervisor>>. Whether this practice is desired or not, seeing this around I got the impression that in general only the first authors are the true contributors deserving credits. I want to know from more senior researchers (who got more experience in different research institutes, universities and different countries), whether this is a common practice or exception? In any case, when I apply for a research position afterwords and my prospective recruiter glances upon my publications, am I supposed to have most of my papers as the sole author or first authorship will suffice?

 

3) My supervisor encourages me to do good research on interesting problems without giving much thought to publications. I may appear cynical but seeing some blog posts and hiring committee opinions led me to believe that research without publication is useless. It's like saying working as a consultant not for money but for love of the job. Off course all the PhD graduates (even from the same school) aren't worth the same in academic/research job market. So what's going to be my measure of performance after I graduate from a reasonably reputed university? Is a prospective recruiter going to read my thesis, understanding my ideas for himself/herself and then judging their worth? Potentially, each PhD thesis can open a horizon of new research and it's not a trivial task to go through a thesis in a short span of time, passing a quick judgement even for a professor. Again for PhD students off course you don't have GPA (as a crude numeric measure) like masters or bachelor students. So how exactly does it work?

Edited June 16, 2013 by swagatopablo

[TakeruK]

I'm in my third year of grad school but this is what has been my experience so far, in regards to your questions:

 

1. Signing something like that is common and sometimes even individual collaborations (especially those spanning multiple institutions) have their own agreements and contracts. Sometimes these agreements are university-wide and written in very general terms to accomodate all the differences in field. For example, I know that in the humanities, it is much more common for supervisors to not be included on the papers. I know some people in the humanities who only meet their supervisor something like once or twice a year, and the supervisor just provides very general guidance in the PhD program while the student is solely responsible for coming up with their own research topic and writing their own papers.

 

In my field, in the sciences, the supervisor and student work together a lot more. Often, the student talks with several professors about their research interests and maybe the two of them come up with an idea together, or the student works on a project that was originally the idea of the professor. In this case, it is clear from the agreement/contract that the supervisor should be a coauthor on the paper, since they have at least contributed to the "idea". Often, in the sciences, we would meet with our advisors regularly and brainstorm ideas together -- either one on one, or in a group meeting. The advisor might tell us to perform the analysis a certain way, or provide code from previous work etc. This obviously makes it even more clear that the advisor should be on the authorship list. Finally, there is a lot of edits and rewriting during the paper writing stage, and especially for new researchers, the advisor might actually rewrite a large chunk of the paper.

 

Overall, what I'm saying is that it's perfectly fine for supervisors to appear on all of their students papers and still keep academic integrity / keep that agreement valid. Supervisors provide more than just funding! They might not be in the lab performing the experiments or the analysis, but chances are, the research problem was their idea, or the technique was theirs, or they trained a post-doc / senior student who is now supervising you more directly. Basically, if the project would not have been possible without the ideas or work of the supervisor, then the supervisor should appear on the authorship list. Thus I wouldn't say it's your "duty" to "give" papers to your supervisor. Instead, I would say it's your job (i.e. you were hired) to work in your supervisor's lab, on your supervisor's projects. If you wanted, you could argue that graduate students are simply a tool to carry out the supervisor's plans. This is true, and some supervisors will treat their students as simply machinery to get work done. That's why it's important to find a good match in supervisor/program so that you get something you want out of your work.

 

To answer the last part -- usually the requirement for counting the paper towards your PhD (i.e. to include it as part of the thesis) is that the student be the first author (or primary contributor), for the work to be done at the school as part of the PhD program, and for the work to be published in a peer-reviewed journal.

 

2. First authorship is usually seen as the most important position. Coauthorship is not bad, it shows that you have done work, but at the PhD level and beyond, the real "currency" is first authorship. This could depend on the field/journal, but I think pretty much no one reads a paper with 4 authors and expect that all 4 authors contributed equally. The difference between first/second/third/etc author's contribution varies a lot and it's not always clear. Some journals now require authors to write a paragraph at the end stating who actually did what. In one collaboration I was part of, the contract was that the main PI for that analysis (usually a post-doc) would be first author and then the rest of the collaboration would be listed in alphabetical order. So, in this case, only the first position has meaning. When you read the academic webpages of people looking for jobs, they usually write things like "X papers, Y first authored". So, I would say that being first author is valued a lot more than another authorship position, but it's NOT the case that you should solely do things that get your first authorship. 

 

3. Doing things purely for science/research and not for authorship is, in my opinion, a little idealistic and naive. It's also more common for people who are in established positions (e.g. tenured professors) to have more idealistic views on their field. However, your supervisor makes a good point because the opposite is also not a good idea. You don't want to be so publication driven that you publish for the sake of publishing/producing and generate bad papers. Publications follow you around forever, so you don't want your name attached to work you aren't proud of.

 

Your supervisor is right in the sense that you should strive to do good science and good work. Papers will come naturally out of the process. You should do your due diligence by selecting projects/science to work on that you think people will be interested in. You should plan for the future by picking a PhD topic that you think would get you hired in the position you want in the future. I think it's a mistake to simply follow your passions and try to find a PhD thesis that fulfills all of your research desires. Instead, you should pick one that other people (i.e. those who will hire you) are interested in. You don't have to be in love with your PhD topic and after working for years on a topic, most topics, no matter how interesting to start with, will probably become very tedious!

 

Your measure of your worth after graduation is going to be a combination of your papers, your advisors' letters of recommendation, and any impact/influence you've made (e.g. at conferences and seminars -- many senior students will try to attend as many conferences as possible and use their connections to give seminar talks at schools they're interested in so that they market themselves and their skills).

 

I don't think a prospective employer will be reading your PhD thesis. Instead, the hiring process at most North American schools involves first making a shortlist out of all the applicants (never seen this part but not sure how it's done), and then the short-listed candidates are invited to be interviewed by a large number of professors and students (usually the students "interview" is a group discussion) at the school. They will also give some talks about their research as well. Through this process, the department can hire the person that is the best fit (in terms of research interest as well as interpersonal relationships).

[fuzzylogician]

That a PI or advisor is listed as a last author doesn't affect how your contribution as first author is viewed. Most fields have clear customs about what first means, what the middle authorship means and what last means. If you are first and your PI is last, these contributions do not contradict each other and each is associated with a completely different role. This really depends on the conventions of individual fields but for PIs (especially untenured ones, but others too) the number of papers on which they last author is essentially the most direct measure of the productivity of their lab. This helps them obtain funds both within the university and without and in a sense if you are working in that lab and enjoy the funding, it is in your best interest to have your PI be a co-author on your work, because productive labs do better at obtaining funding than unproductive ones.

 

Consider what happens if the advisor is not a co-author on your papers; I work on an interface between two subfields, one theoretical and one experimental. In the experimental subfield, the professors who I work with are always co-authors on all my papers and in the theoretical subfield they never are. Mind you, they may be some of the same people, and their contribution to my work is more or less the same, but depending on the type of work and the resulting paper they may or may not end up being co-authors. Now, these professors are measured against others from fields where it is customary to have professors as co-authors on all student papers. These fields may be very closely related and the distinctions may be obscure. Deans and other Higher-Ups use the number of publications as a direct measure of productivity. It's not easy to explain why Prof X publishes 1-2 papers a year while Prof Y has 6-7, never mind that many are actually students' papers and in terms of direct productivity the two professors produce quite similar amounts of work, and they also do a similar amount of advising. The advising that goes into our sole-authored papers doesn't count, or counts for less. There is no good way to prove that a certain professor contributed some amount to a student's paper, especially since we don't have just one advisor and instead work with several people on each project. Anyway, the bottom line is that their last-authorship is just not in competition with your first. 

 

As for second- and third- authorship, this obviously counts for less than first, unless it's a field where authors are ordered alphabetically, but I wouldn't be so quick to discount the work done by others that goes into your paper. Writing, in particular, is very hard and in my field it's a chore that is left to the first author, not relegated to someone else. Doing time-intensive analyses and coding is also not something to dismiss. At the end of the day, unless there is a paragraph stating who did what on a paper, there will be some amount of guessing as to who did what (therefore, if the contributions are very uneven I recommend insisting on writing such a paragraph). Normally people will guess that the first author did most of the work and others contributed to certain aspects. But what really matters for jobs is how your advisor describes your contribution to each paper in their letter of recommendation. You will also describe your contribution in your research statement, so potential employers will have a better understanding of what you did. I don't think there is any reason to prefer sole papers to first authored ones, but this depends on the field. In some it's highly unlikely that one person can complete a project alone from beginning to end, but in others this can be done. If so, it's probably good to have sole-authored work to show that you can do that, but also co-authored work. Ask your professors about your field and also check the CVs of recent graduates who got jobs like the ones you'd like to have when you graduate--what is the ratio of sole:co-authored works on their CV? Aim for a similar mix. 

 

Finally, I think your advisor has a point about doing good work that is interesting to the community, because that's the kind of researcher that will get hired later. You want to work on problems that are attractive and you want to develop skills that are normally sought after in job searches. Ideally you want to find a topic that you are interested in, but you don't want to get sucked up into doing something that only you are interested in. If you work on something others like, you will get attention and hopefully also publications. You'll also have an easier time creating a research program and describing a future for such a project that others will be excited about, once you're looking for jobs. A dissertation is not the end of a project, it's merely the beginning. As TakeruK says, you don't have to be in love with the project, just find it reasonably interesting and with a reasonable future. Those kinds of projects also tend to produce publications, if successful, so that's another plus. But from where (I think) you are in the process, being too publication-driven is not the best idea - publications should not be forgotten but before then you actually need to sit down and do good work for a while, and preferably on something on that interface between what the community values and what you find personally exciting.