Tuesday, August 6, 2013

Scientific Reproducibility, Hype, and the Glut of Ph.D.s

There is an interesting conjunction of articles in this week’s NATURE.  One opinion piece from an idealistic young graduate student deplores that fact that scientists must promote their work as being medically, economically or socially relevant, to an extent verging on ‘hype’ (1). In the same issue a news feature reports that the NIH is considering verification rules for some of the research it supports partially because of many comments from the pharmaceutical industry that much academic research cannot be reproduced (2). Finally, the issue contains an obituary of the Nobel prize winning physicist Kenneth Wilson (3).

Here is what relates these three articles- two are examples and one a counter example of the consequences of the overexpansion of contemporary science. Scientists oversell the pragmatic ramifications of their work largely because funding agencies require them to do so. For example, all biomedical investigators must deal with the NIH requirement for explaining the ‘significance’ of projected research as well as with the agency’s current emphasis on ‘translational’ research.  Other funding entities in the US have similar stipulations and the situation may be even worse in Canada and the UK where there is increasing emphasis on the commercial ramifications of research. However, as has been shown by many historians of science, the greatest impacts often flow from unfettered basic research rather than from work intended to address specific medical or technical problems. The career of Prof. Wilson is a good example of this. After being hired by Cornell in 1963 he did not publish a paper until 1969. Then in 1971 he published theoretical work that revolutionized areas of physics ranging from sub-atomic particles to fluid mechanics, thus earning himself the Nobel. Today it is hard to imagine anyone lasting for six years in an academic position without multiple publications (preferably ‘translational’ ones!).

Delving further, why are the funding agencies so insistent on research that claims to have immediate pragmatic ‘significance’. The reason is that the agencies must drum up political support for their enormous budgets, and it’s much easier to sell a Senator or Congressman on curing say prostate cancer than on elucidating some obscure molecular interaction. While society should generously support fundamental scientific research, both for its own sake and for its long term practical benefits, the level of support must be ‘right-sized’ to the state of the economy and the level of development of science itself. If maintaining public support for science means a constant process of overselling its short-term payoff, then perhaps the bloated science establishment needs some trimming.

One of the major problems, however, is that universities have used the relatively generous science budgets of the last few decades to train an enormous cohort of Ph.D.s who now must struggle for research funding. The current rather obscene degree of ultra-competition drives investigators to publish results prematurely leading to increasing concerns about reproducibility. As discussed previously on this blog (4) curtailing the production of Ph.D.s would be immensely helpful in obtaining the right balance between reasonable levels of publically supported science funding and the size of the scientific work force. Fewer, better quality Ph.D.s may actually produce a greater amount of high quality science than hordes of ill-trained Ph.D.s, many from 3rd rate institutions.

No comments:

Post a Comment