Saturday, April 23, 2011

Public funding of science, and access to information

On multiple blogs over the last few months, I've read comments from lay-persons (that is, nonscientists) that say, in essence, "As a citizen, I paid for this research, and therefore I should have access to all the data and all the software necessary to analyze that data."  The implications are (1) research funded by the public should be publicly accessible; and (2) the researchers themselves sometimes/often? hold back information or misinterpret the results, perhaps because they are biased and have an agenda to further.  

Now, as a pragmatist, there are a number of issues here.  For example, making available raw columns of tab-delimited numerical data and, e.g., matlab code, won't give a nonscientist the technical know-how to do analysis properly, or to know what models to apply, etc.  Things really get tricky if the "data" consists of physical samples (e.g., soil, or ice cores, or zebrafish)....  Yes, scientists that are publicly funded have the responsibility to make their research results available to the public, and to explain those results and their analysis.  As a practical matter, scientists are not obligated to make any interested citizen into an expert on their research.

While this is an interesting topic, I'd rather discuss a related issue:  How much public funding triggers the need to make something publicly available?  For example, suppose I used NSF funding to buy a coaxial cable for $5 as part of project A.  Then, later on, I use that coax in project B, which is funded at the $100K level by a non-public source.  I don't think any reasonable person would then argue that all of project B's results should become public domain because of 0.005% public support.  When does the obligation kick in?  Just an idle thought on a Saturday morning.

5 comments:

Y.H.N. said...

Here are my 2 cents

I believe that public funding or not that the data should be made available in all cases. Having access to the Linux Kernel source code does not make one an operating system expert. In a similar manner having access to data from an experiment does not make a expert in analysis. What open access does accomplish however is to show folks that:

- science is hard - it takes a lot of work just to understand the questions.

- barriers to entry are not insurmountable - there is no secret handshake or external licence that can be withheld.

- anyone can be a scientist - by opening the tools of science for examination we can put these tools in to the hands of the general public.

Too often we struggle for support from the general public because we are viewed as a separate cast. The public come to view themselves as passive recipients of knowledge beyond even their ability to comprehend. No wonder we have to fight for every dollar in funding we get.


http://brownian-notions.blogspot.com/2011/04/open-science.html

sylow said...

Well, the sourcecodes are available in some cases. In density-functional theory, authors state what program they used (gaussian, socorro, wien2k, siesta etc.) and what basis set. Hence, in general anyone should be able to reproduce those results but 50% of published papers cannot be reproduced. Everybody knows it so how much of it is junk? That is the 60 million $ question, right?

DanM said...

What really interests me is that Doug has apparently admitted to using his NSF funds to support a non-federally funded project. Heresy! I think we should form a mob and storm his office. Who's with me?

(disclosure: his office contains a coffee machine, so I may have an ulterior motive.)

Homesure Services said...

As a practical matter, scientists are not obligated to make any interested citizen into an expert on their research.

Web Design Tampa said...

scientists that are publicly funded have the responsibility to make their research results available to the public, and to explain those results and their analysis.