Post-doc dilemma

When a PhD student is nearing completion of his/her PhD, the worries, concerns and dilemma pertaining to a post-doc start to loom.  This situation is strongly dependent on the broad area of PhD and also on the specific area within a particular discipline. For instance, to my knowledge and observation, PhD graduates in basic sciences such as physics or mathematics mostly or usually do one or more post-docs, while PhD graduates from say Electrical or Computer Science Engineering usually bypass the post-doc stage - not all such engineering PhDs, but mostly. For instance, if we randomly pick up 100 fresh PhD graduates in Physics or Mathematics, and 100 fresh PhD graduates in Electrical/Computer Engineering, what how would the statistics look like with regard to the percentage of graduates going for post-doc in physics/math vs. those in electrical/computer engineering? I do not have the statistics with me to claim anything, but based on my observation and interaction in the last 4.5 years, I would estimate more than 90% of physics/math graduates go for post-doc, while I won’t be surprised if the corresponding post-doc fraction in electrical/computer engineering discipline is less than 20% or even 10% for that matter. [All discussions here refer to obtaining PhD in USA, in a top 50 school.]

The reason is obvious: electrical/computer science PhD graduates have a reasonably higher number of industrial positions available to which they can apply and get absorbed (in USA again) while the number of such industrial positions open to math/physics PhD graduates is certainly fewer than those open for engineers. Thus, to my knowledge, engineering PhD graduates usually join industry after their PhD while their counterparts in basic sciences eventually join academia or national labs etc. as scientists after doing one or more post-docs. Now, the engineering PhD graduates who aim to join academia usually tend to do a post-doc after PhD which is mostly to ramp up their publication profile, increase citations, get more experience and also more recognition among the people in his/her community. Also, to my observation, the duration of post-doc in engineering is less than that in basic sciences. For instance, it is not uncommon to find people doing 3-year long post-docs in basic sciences, or even more than two or even three post-docs for that matter, while I will be surprised to find anyone doing a more than 3-year post-doc (in a university) in Electrical/Computer Science Engineering, not to speak of more than one post-doc! It’s just the nature of the discipline I believe.

Alright, now I shall exclude the basic sciences, for I am not in basic sciences. Electrical engineering, that’s where I am! What should I look forward to in a post-doc? I am not keen on more than 2-year post doc due to multiple reasons which I shall not elaborate here. Also, as I wrote, it is really uncommon to find anyone in my area doing a 2+ year post-doc. Now, what should I look forward to in a 2-year post-doc? Usual things to look for are – a superior brand name of the school, a more well-known academic pedigree, more exposure/experience, more publications and citations, more recognition within the community, etc.

To my limited knowledge, while considering faculty applications, the committee mostly looks into two things more carefully – the brand name of the school of the applicant (for instance, the names of MIT, Stanford or Berkeley can weave magic), and the number of publications/citations of the applicant. Quick note here: in most of the branches of engineering, publications in Nature/Science or such other high impact factor journals is very uncommon because many of the progress in engineering is incremental in nature such as increasing efficiency/power/output/etc in some applications etc. and so are not breakthroughs in fundamental understanding to be reported in Nature/Science. So, it is rare to find engineering papers in such journals. Ironically, the inventions in technology which have literally revolutionized our human lives everyday such as semiconductor lasers, LEDs, transistors, ICs etc. were all published in IEEE or such other journals, not in Nature/Science J.

Anyways, so brand name of the school and number of publications in journals most widely read in the specific field (IEEE for Electrical Engineers, for instance) are the two most critical criteria to evaluate an applicant’s profile. Now which of the two factors weighs more will most likely depend on the committee which evaluates. But then, here is another very important thing to consider while applying for post-doc: should I stick to the same area X where I did my PhD, or should I shift gears and jump to another area Y? That brings me a very important question I ask myself: in a 2-year post-doc, what is the realistic probability that I shall gather a reasonable recognition for myself within the community if I jump to area Y? I will be totally new face in area Y, and in two years, shall I be able to make my mark in the community so that the experts at least know my name or my existence? That will depend on how prolific I am in publishing good journal papers and giving conference talks in that 2-year window. Being in an experimental area, jumping in to a new group with a new research topic will require a significant ramp-up time to get acclimatized to the labs, equipments, people, staff members, etc. and by the time I feel comfortable with the group members and technical staff and start doing experiments smoothly on my own, it will be at least six months over. And since I would like to start applying for faculty positions in the final year of post-doc (since starting from application to getting an offer letter window can be one year in many cases), it implies I have less than one year to do core research in my post-doc. Irrespective of the brand name of the school or the post-doc advisor, jumping to a new area Y thus seems a very bad idea (for me, for my 2-year post-doc plan) for I can imagine how difficult it will for me to publish a lot in that small period. Ruled out.  

Thus, a post doc in my area X seems a more optimistic and wise idea. And on a brighter side, I am kind of known in my community to some extent. Several professors who are leading top-level research in my area in several US universities and some government officials who fund those projects actually know me in person now. Since I have presented at a good number of conferences, some people even in Europe have interacted with me and know me in person as well. So, my existence is kind of known by some experts in my area X. So, a strong 2-year post-doc in this area will most likely establish my name within the community even more strongly. So, that means a post-doc in my area ! Great.

Apart from the brand name of the school where to pursue post-doc from, another very critical thing to consider is – will the post-doc advisor allow me to work in multiple projects in parallel? Examples abound where a post-doc is put strictly in to one funded project; if the project works, it’s well and good that he/she will get a couple of papers, or else if the project doesn’t work, then that’s it – a totally unproductive post-doc in terms or publications. There are actually such unfortunately not-working projects where post-docs come out dry when they are made to work only in one project – it’s all advisor dependent. So, can I work in multiple projects?

So, the ramp-up time should be very less so that I slide in the system smoothly and fast and start churning out papers in good number. The group should be ready to welcome me, and ready to involve me in their projects as well irrespective of what the advisor wants. Then here’s another thing which though obvious is very important: the number of publications in post-doc should be at least proportional to that during PhD. Ideally and usually, the idea is that in post-doc, one can pump out more and better papers in a faster pace than in PhD, for, the requisite training to work more efficiently, think more clearly, design experiments faster and to write papers faster has been already achieved in PhD. So if I have ‘n’ publications in my 5-year PhD, then it will be reasonable to aspire for at least ‘n’ publications in a 2-year post-doc in the worst case. It will be extremely disappointing, frustrating and negative-impression-creating to have very few publications coming out of my post-doc – the brand name of school/advisor doesn’t matter in that case (to me). Personally, I shall be super pissed with myself in such a situation especially because I shall be graduating my PhD with a reasonably decent number (~ 17) of publications.

Thus, conclusion of this long and introspective essay is that, since I’m going to graduate with a PhD from a top-20 school in Electrical Engineering, either I should pursue a post-doc in a top-10 [or another top-20 school higher up in ranking than mine], or, join a prolific and kickass research group even in a below-20 ranked school which publishes high quality papers very frequently (in which case I trade-off brand name of post-doc school!). Tough part: the combination of everything together (in my area) is almost impossible to find!