Saturday, 26 September 2015

The PhD quality

I am told that IIT Roorkee is going to award about 235 Ph.D. degrees in the forthcoming convocation on October 03, 2015. This marks an increase of almost 100% on YoY basis. This upward trend of producing PhDs is going to continue in foreseeable future, if the thinking of the apex advisory council is of any guidance. The 100% rise is indeed eye popping even if we take into account the fact that PhD is not exactly a time bound programme. This increase in the PhDs can not be attributed to the increased faculty strength in recent times as they have not been around long enough, as yet. Since the good candidates are not exactly queueing up to join our PhD programme, this extra-ordinary increase in the PhDs produced calls for discussions and to re-examine our system of checks and balances to maintain quality of our output and to assure that it is not being compromised.

How should one define the quality of students (doctorates, in particular, since there is no grading) that we graduate? The process of the making of a doctorate is akin to the manufacturing process wherein raw material is processed to produce a finished product. So, the definition of quality from manufacturing industry should be a good guidance:

''In manufacturing, a measure of excellence or a state of being free from defects, deficiencies and significant variations. It is brought about by strict and consistent commitment to certain standards that achieve uniformity of a product in order to satisfy specific customer or user requirements. ISO 8402-1986 standard defines quality as "the totality of features and characteristics of a product or service that bears its ability to satisfy stated or implied needs." If an automobile company finds a defect in one of their cars and makes a product recall, customer reliability and therefore production will decrease because trust will be lost in the car's quality.'' Read more: 

Every milestone in the formal educational set-up is associated with the development of well-defined set of skills. The undergraduate programmes are aimed at "How to" skills whereas the post-graduate programmes should aim at "Why" to encourage and develop critical thinking. In an ideal situation, this would be construed to have been achieved if the candidate gets a couple of peer reviewed publications under his/her belt by the time of graduation. But it is not a fool proof system. The scholarly journals are only interested to see the novelty in the submission and have no mechanism to check how the work was actually carried out and reported. It is quite possible that the candidate worked like a robot carrying on the detailed instructions of the supervisor to produce the results and the supervisor wrote the majority of the manuscript with very little input coming from the student. Obviously, the candidate gets the credits in authorship but has learned little to progress on the path to developing an attitude of critical enquiry. Ideally, the candidate becomes eligible to be conferred a doctorate degree when he/she develops the faculty of critical enquiry to the extent that he/she can take up an independent career in research and to supervise/train other candidates in this process. There is, unfortunately, no way to gauge/ascertain this capability other than an honest assessment of the supervisor. If the supervisor, for whatever reasons, fails to perform this basic screening then it is a rather remote possibility that the candidate may not get the license to philosophize and supervise independent research and produce questionable PhDs in the process.  It is not without reason that PhD is such a valued and respected academic accomplishment and it is our solemn duty to maintain its high esteem. I quote a paragraph in this context:

"PhD is hard. It is meant to be hard not because inflicting pain is necessarily fun, nor because some scientists are 'dementors', and not because your PhD is expected to solve the mysteries of the universe. It’s hard because it is an apprenticeship in science: a frustrating, triumphant, exhausting, and ultimately Darwinian career that will require everything you can muster. A PhD is essentially a test. Don’t fool yourself into thinking that you pass this test by passing your PhD. Wrong. The fact is that passing a PhD is like getting a certificate of participation. Why? Because almost everyone who starts a PhD and sticks around long enough ends up getting one. No, the real test is what happens after your PhD. That’s when you’ll know whether you’ve really passed. Do well and it will open the door to a career of unparalleled intellectual freedom." [from Tough love: An insensitive guide to thriving in your PhD]

Recently, I had a chance to go through a PhD thesis on analytical modelling of vibration of plates. I was quite amused to note that the main argument of the thesis was in finding virtue in an approximate numerical solution scheme that affects the computed frequency parameter in fourth or fifth place of decimal in comparison with the benchmark case computed using a different numerical approximation. As a practitioner of the vibration theory, I can swear on oath that I would be more than happy to get a reliable estimate of the natural frequency to unit's place or at most one decimal place. Moreover, I wonder if it ever occurred to the candidate that for all his singing praises for the adopted numerical scheme, the difference could probably have been caused by the standardization of the floating point operation since late 1980s (because the benchmark result dates back to the time before the advent of floating-point standard and hence the coding and platform of computing made a lot of difference in the accuracy of numerical computations). Probably, the candidate has never heard the golden advice, "The purpose of computing is insight, not numbers!" This lack of appreciation of insight into the basic mechanics of the problem is the biggest failing of our training process that is called PhD and is worrisome. This fresh PhD now has the license to supervise a PhD and will have a cascading effect on this academic lineage of PhDs.

While we are discussing the PhD students and their work, another issue that is worth debating is the authorship of the publications. Should the supervisor be a co-author? If he is a co-author of the research papers published prior to submission of thesis, then is it not a case of conflict of interest when the supervisor examines the PhD thesis of the candidate? In this scenario, will it be a better proposition to insist on at least one single author journal publication by the candidate before the thesis is submitted? This sole authorship in a journal article will also go a long way in helping the academic career of the student while ensuring that the candidate owns 100% responsibility for at least part of the work being examined.

I am sure that I have done enough to stir the hornets' nest and welcome a healthy discussion/debate on this issue. Having said that I am also sure that we all have some interesting anecdotes to share about our days as graduate students that it'll make for an interesting discussion. Let me share something that I had conjured up on the basis of an empirical analysis of available data:
The Supervisor's Law of Inertia:
"A supervisor tends to produce graduate students who are his/her academic replicas, unless influenced by some external factors."

A few links to recommend to your PhD students about the PhD process: 

Some Important Things Most Students Never Ask About Graduate School 
Post a Comment