In a recent post, Professor in Training expounds what is arguably a big dilemma that an academic scientist faces, trying to be at one time a successful, productive investigator and effective mentor for her advisees — graduate students, postdoctoral associates etc.
Fundamentally, the problem is one of different time scales. Junior colleagues need to learn how to do things on their own, in the process making mistakes, reinventing the wheel, taking a lot of time to do what an experienced researcher could have done in the blink of an eye. It’s inefficient, it is frustrating, but it is necessary. To my knowledge, no one has come up with anything better, when it comes to training.
But, why need it be a problem ? Because in the highly competitive environment in which a scientist operates, the pressing need to generate data and publish them quickly makes for a rocky marriage with giving a trainee the time needed to learn how to set up an experiment, prepare samples, learn how to use equipment properly, write a computer code from scratch, collect data, analyze them, write a first draft of a manuscript, and so on and so forth.
A Principal Investigator (PI) is often in the uncomfortable situation of making a decision between giving a trainee who has been assigned a scientific project all the time reasonably needed, possibly delaying completion and publication of time-sensitive research, or taking over and doing most/all the work herself, thereby depriving the trainee of the opportunity of acquiring the kind of hands-on experience that is sought.
Well, OK, so we all agree that this can be a problem but… an “unsolvable” one ? Is it really “not possible” to do what it takes “to advance [one’s] own career” while ensuring that “trainees get a first class education and training experience”, as PiT suggests ?
That seems a bit drastic. My observation is that, while some do it better than others, it is quite possible to achieve both objectives reasonably, that is to the satisfaction of everyone involved — scientists, trainees and administrators.
Goals — whose ?
First of all, a general remark: the reason why competition in science is so fierce — competition for funding, discoveries, prizes, everything — is that scientists themselves are by their own nature extremely, obsessively competitive. Simple as that.
Often times, the stress under which they operate is self-induced. This is typically the case with tenure-track faculty at second-tier institutions, or places with lesser emphasis on research than R1 universities. They will try to lay the blame on deans, directors and department chairs for (supposedly) setting unrealistic tenure benchmarks, when in fact those overly ambitious scholarly goals are really their own.
No sane dean or department chair at a non-R1 institution will ever expect, much less demand, that a newly hired faculty compete in research with peers at high-power institutions, enjoying lower teaching load and having access to facilities, funding, students and personnel that simply do not exist at a school whose stated mission does not include research as a main objective.
Everyone understands that, but don’t go telling it to the tenure-track faculty at a non-research institution, who still wants to operate as if that were Harvard, who is dead set on overcoming all odds and being the one who makes that jaw-dropping, Nobel-worthy discovery . Just take it with a grain of salt, when they tell you “I either get that grant, publish on that important journal, make that discovery, or I shan’t get tenure“.
(How do I know that ? Well, simple, that was me, talking, when I was at my previous institution).
Compromise is the name of the game
It is possible to achieve all objectives, namely produce good science timely and be an effective mentor. Just like in every complicated optimization problem, pragmatism, flexibility, and a few simple prescriptions, can go a long way toward meeting all of the conflicting needs.
This is what I would recommend a junior PI, based on my experience and observation (much of it seems really common sense):
If it is time-sensitive, don’t give it to a beginner.
This really seems like a no-brainer. It is one of those instances where academia seems to have a hard time figuring out what the rest of society has always known. If a project must be completed quickly, in order to keep up with the competition and/or have something concrete to beef up a grant proposal, thereby strengthening one’s chances to impress reviewers, no inexperienced researcher should be put in charge of it. Yup, it’s that simple.
Postdoctoral associates, possibly advanced graduate students, or in any case junior scientists eager to build an impressive record of accomplishments in a relatively short time, who [are supposed to] have the maturity and expertise to tackle challenging tasks, are perfectly suited to take on time-sensitive projects. It is, after all, their chance to show society “what they’ve got”. It is unfair, out of line and seriously ill-advised, to put undergraduate and beginning graduate students under pressure to deliver as if they were professional researchers, as if they already knew how to do what they are there to learn .
Students are not technicians
Keep students away from any kind of repetitive technical task that does not directly benefit them. Students should of course become technically versed, should learn how to use equipment, or become familiar with computer technology and operating systems, and yes, one of the benefits of college education is precisely the acquisition of technical skills through research. However, there is a line that should not be crossed, and that is when a student is tacitly “promoted” to the role of technician. That should be avoided.
I see this happening over and over again. I can understand where and how it originates, for not only are technicians expensive to hire and pay, whereas students are “cheap” — often times students actually know more than technicians. It goes like this, a student is told to learn how to use a specific piece of equipment, or assigned the task of managing a computer system, or what have you — possibly a valuable skill to acquire but nothing that can be regarded as scientific credentials. Next thing you know, that person becomes the “resident expert”, the guru to whom everyone (yes, the PI too) goes for help with that particular thing — including when it is not related to the student’s project. In essence, the student comes then to be regarded as an “asset”, but not in a good way. There are two serious drawbacks with that: a) the person will waste a lot of time helping others, time that will come at the expense of her own research, b) when the student eventually leaves (because they are supposed to, at some point…), a huge “expertise gap” will appear, possibly in some mission-critical aspect of the operation of the whole research group .
This type of situation should not come about, ever. A technical job should be assigned to a technician. If the department within which the PI operates does not provide one, and if the PI cannot hire one, then the PI should take on the task herself (yes, I was the system manager of my own computing cluster at my former institution). If one finds that demeaning or otherwise unacceptable, considerations along the lines of note  apply.
Well, what project shall I give a beginner, then ?
A “safe” project, one that will not lead to any ground-breaking discovery, but one with a well-defined publishable result , which will be regarded as useful by the community and will allow the person to
- become familiar with methodology
- get started with literature review
- develop an appreciation for the investigative process
- start learning how to write an article
It is often possible (and a good idea) to have a beginning graduate or an undergraduate student contribute to the research effort of a senior graduate student or a postdoc, in the process learning the “trade” directly from her more experienced peer . For an undergraduate student, that may well be the only research project in which she is engaged before starting graduate school, assuming that that is her choice. Once a beginning graduate student has contributed to the completion of a first, significant project, then she can be assigned something of greater breadth and more fundamental scope, on which she will be working continually as the leading investigator for, say, two to three years. That project, which will ideally yield a few publications (at least one on a high profile journal), will constitute her scientific training ground.
In a way, as long as one is working on a cutting edge project, one could argue that there is always some pressure to obtain results before competitors do, and I think it is appropriate for graduate students to appreciate that. However, I really do not think that it is ever a good idea to engage a trainee at any level on a project that has a well-defined deadline, whether that be motivated by the need of incorporating new results in a grant proposal or other reasons.
About those papers…
Teaching a student how to write a manuscript can be one of the most extenuating, frustrating, difficult tasks of the job of an academic. The temptation of telling a student “You know what ? Forget it, get started on your next project, I’ll write this myself” is sometimes very difficult to resist, especially if a student, for one reason or another, does not seem to learn. I am guilty of doing that myself. It should not be done. Students must be able to communicate effectively in writing, or their entire career will go nowhere, regardless of how good they are, in research and possibly outside as well (yes, we need to tell them this very thing in these very terms).
No matter how sick and tired we are with that work, no matter how appalling it is to read a third version that looks even worse than the first one, no matter how much they hate our guts for that, we need to send them back and have them rewrite it, again, and again, and again until they get it right. If, at some point, we come to the conclusion that they really cannot do it on their own, we need to have them sit with us in our office and write the paper together, line by line. This is what my own PhD advisor did with me, and I don’t care how much I hated it back then, he did me a favor.
Same goes for talks — their chances of landing a job depend on how well they can present their work to a general audience. They need to practice, a lot; their first talk may have to be rehearsed for hours and hours. We need to teach them how to make slides, how to stay within the allotted time, how to convey the cogent information and leave out irrelevant details, how not to make everyone in the audience car sick by constantly moving the laser pointer in circles, all over the screen (not that I am thinking of anyone in particular — I know you are reading), and so on. We cannot expect them to pick up all of that by themselves (albeit some will). It would be irresponsible.
Sure, all of this will cut into our research (and leisure) time, and at the end of the day it will often times be “just another article”. We may not get nearly as much out of it as we put in, in terms of time, but that is not the point (and any academic who says something along these lines is in the wrong line of work) — it’s an integral part of the profession.
How can chairs and deans make our lives easier ?
It is simple: by acknowledging that mentoring students and postdocs is a lot of work. By that I do not mean lip service — we do not know what to do with it. I am talking officially recognizing supervising activity at evaluation time, including for promotion and tenure purposes. Everything else being the same, the faculty who supervises a greater number of students and postdocs makes a greater contribution to the mission of the institution . Individual student mentoring should at the very least be applied toward the fulfillment of a faculty’s teaching duties (justifying a classroom teaching load reduction). I also think that an institution should look favourably at the record of a tenure-track faculty who has proven capable of engaging a large number of students in her research program — even if that may not translate in a proportionally greater publication output (it usually does not, for the reasons mentioned above).
 Just to be clear: Being so driven and competitive is not a bad thing, it is a good thing. And it is also good if some of that determination, stubbornness and enthusiasm rub off on junior colleagues and students. However, one’s narcissism and personal obsession ought not become a problem for others. It is a delicate balancing act.
 If postdocs are not there, if the money to hire them is not available, if due to the very nature of the institution postdocs are not common where you are, and/or it is difficult or impossible to hire them, you are at a competitive disadvantage. You can either accept and embrace your institution’s different character, and attempt to set up a meaningful, possibly highly rewarding but scaled down research program, compared to that which you would be running at a R1 (more here), or you can try to do the best you can with limited means while looking for a more suitable position — obviously, in order to keep up with competition at R1 schools, you will have to work much, much harder, precisely because of the meager resources at your disposal. But, to expect your undergraduate students to turn into postdoc-level scientists is silly.
 I have seen computational research groups literally in disarray, after the student who had been in charge of installing or maintaining a crucial piece of software left.
 I can only speak (somewhat) knowledgeably about my own field, but my sense is that such projects always exist. In my case, calculations of equations of state, or of physical quantities measured or measurable in experiments for which theoretical predictions are missing or incomplete (and therefore a direct comparison is possible), or comparative studies of different computational methodologies, are my usual choices.
 This is a recurring theme in my posts (see here for instance). One of the most powerful, compelling arguments for maintaining a decent size (not too large) research group, is precisely that of promoting the interaction between junior scientists of different seniority. I am convinced that one will learn at least as much from fellow students as from her own supervisor. It is also the most natural venue in which one can develop mentoring skills.
 Like everything else, abuse on the part of an individual faculty is possible, and should be prevented. One ought not furnish any kind of incentive to faculty to keep advisees around for longer than needed, just for the purpose of listing a large number thereof on their annual activity report, thereby collecting extra brownie points or being granted reduced teaching. While obviously each case is different, in general a department ought be concerned with a pattern of students supervised by the same faculty taking consistently longer than average to complete their degrees.