Objective Research Funding? An Approach to quantify the Value of Experiments

The distribution of research funds is clearly
not based on purely objective criteria. Most countries have different ways of
how to deal with this issue – all face different, but serious problems.
Bruce Knuteson (MIT)
has developed a formula of
which he claims is able to estimate the scientific merit that a proposed
experiment will give back per monetary unit before
we actually perform it.
Knuteson’s formula
estimates the gain to be obtained by a proposed experiment in terms of the reduction in information entropy the experiment is
expected to provide. This is a seductive concept: Large scientific projects,
think of ITER at Cadarache or the Large Hadron Collider at CERN – cost the public a lot of money. The ‘right’
distribution of research money is thus not only of interest to promote the future success of
scientific research, but also of larger societal interest.
The Swiss newspaper Neue
Zürcher Zeitung (NZZ)
reports on Knuteson’s formula. Can it really help to
provide objective and rational criteria for funding the right type of research?

Kunteson’s formula falls back to a well-known
concept of information theory – the information entropy, which quantifies the
degree of uncertainty associated to a state of knowledge. This is very tempting.
Distribution of research funds according to their capability to give surprises
may counterbalance the trend to push predominantly mainstream research, while
equally good and important research that is located far or not even far from
the mainstream faces serious funding deficits.


Real-life – even the one in the lab – is,
however, far too complicated for a single cardinal figure to represent all
relevant features. The formula requires
an a priori guess of the likelihood of the occurrence of the outcomes of the
experiments; completely unprecedented events are thus discounted. In another
framework, Nassim Taleb has termed these non-predictable outcomes most lucidly
‘Black Swans’. Moreover it is exactly the reproduction and validation of
already known experimental results that distinguishes science from pseudo-science.


We have to deal with a lack of purely objective
criteria for which type of experiments are to be funded – and this depends
crucially on our perception of scientific research and what is worth investigating.
Currently we just can hope that the shortcomings of different national research funding
methods may level out as science is an international enterprise.

  • Facebook
  • Twitter
  • Reddit

Comments are closed.


Subscribe Via Email