Search This Blog

Friday, February 21, 2014

Liberating Science

Okay, this post is going to be a sort of follow-up to the last one (which was about scientific research if you were wondering). Last time I talked mainly about how scientists (ought to) do research. But scientists (& research) are also influenced by factors outside of the lab (quite a shock, I know). To illustrate, let me sketch a picture of how research would work if it existed in a complete and idealist vacuum.

A scientist spots a phenomenon that interests them. After reviewing existing literature on the subject and integrating that with their own observations, they formulate a theory as to what makes this particular phenomenon occur in a certain context. They think of a way to variate what makes the phenomenon occur in a controlled environment, what should be measured, and which results would support their theory and which wouldn't. Then they run the experiment, collect the data, analyse the data, and report their theory, predictions, methods and results to their peers.

Obviously this is a great simplification and depending on the field and type of research things might work a little different, but I believe this example is functional enough to illustrate my point. Now let's take a look at how research might work in the real world.

A scientist spots a phenomena that interests them, either in the natural world, the lab, or somewhere in existing literature. Or they receive a grant to perform research in a specific direction (which may be broad or very specific). If not, they will either need to find funds or already have access to them. To make the research seem more valid or acceptable to peers, or simply to not have to reinvent the wheel, they might contrive their questions and/or predictions and/or methods from existing literature. Because the scientist has a job, the decision whether or not to actually execute the experiment will probably also be influenced by their perception of how publishable their report will be. If they run the experiment, they again collect the data, analyse the data, and report their theory, predictions, methods, and results (in the last post I went into some of the things that can go wrong or right in these last few steps). These are sent to a journal, which in turn sends the report to peer-reviewers (at least if it's a reputable scientific journal, I guess).
Now from what I've heard from some researchers, I'm using the term "peer" pretty loosely here. Because there are often competing explanations for any given phenomenon in scientific research, reports are often reviewed by researchers who support the other competing explanation, because they are assumed to be more critical of the research. Which is probably a valid assumption. But sometimes this practice leads to reports going unpublished or having to add or remove text, methods, experiments, only on the basis of methodological disagreements that are being fought out in the rather private setting of review instead of in the open in journals. Now maybe this is a phenomenon exclusive to the behavioural sciences, but I'm guessing other researchers are just as human as behavioural researchers. It might also be that I've been primed to this sort of information through previous lectures and such, or that this type of information is reported more often than all the successful publications or stands out more. But I digress.

If you compare these two situations, there may be quite some unwanted factors influencing what is being researched and how, and what is being published. Researchers are stuck in constraining science, where not scientific principles are dictating what to research as much as individual beliefs and career ambitions, organisational requirements and regulations, and publication requirements.  There have to be funds to do research, and also "consumers" of said research in the way of journals and readers of those journals. I understand the need for some constraints, but I believe the ones in place may be the wrong ones.

Now I want to take a look at another business where information is (should be) the main currency (or main service might be a better word), the news. For many years, news reached the people in a top-down matter. Journalists travelled around the world in search of news and reported it when they found it. What I mean by working in a top-down matter, is that what was news and what was not was dictated by journalist and editors, and assuming they wanted to sell copies, their views of what the public wanted to know. Nowadays, everyone pretty much travels everywhere and is equipped with the means to report on any news they may perceive to be important. And, through the internet, if a lot of other people also find it important (or interesting, or amusing), it will be seen by millions all over the world. This has created a situation where news has a fair probability of reaching people in a bottom-up fashion. That is, actual participants in any kind of situation can report on it instead of mainstream media, and the mainstream media now often report on these situations if they attract enough attention online.

My hopes for scientific research lie in the addition of these bottom-up methods to the top-down methods already in place in science (that is, where money, journals and editors dictate what gets published, maybe more so than researchers do). And it seems I am not the only one!

Two great initiatives are (from my perspective) implementing a more bottom-up strategy.
Open Science Framework is a place where researchers can create and share research. Detailed information on all aspects of research can be shared, which increases transparency and thus (if it's solid research) validity. It possibly also makes it easier to replicate studies, and hopefully will stimulate discussion with interested peers at all stages of research without the usual restrictions of being in the same institute or being acquainted already.
PLOS ONE is a collection of online peer-reviewed journals that operate under the Creative Commons Attribution License, and on principle publish research from all disciplines of science as long as it's found to be methodologically sound. Now I have to add that there is a fee required to be published, but all this information is available to be read and shared for free! This is not only a good thing for scientists but for anyone interested in critically evaluating scientific research for themselves.

The former of the two deals with some of the problems I described in my previous post, while the latter ensures that how publishable a paper is is based on working with sound scientific methods instead of results.
Now all we need is some form of crowd-funding for scientific research. This may be my idealistic side again, but imagine that instead of donating money to (questionable) charities dedicated to finding a cure for a certain disease, or sponsoring one child to get a cure, there could be public fora where medical researchers place research proposals, other scientists publicly discuss the validity and such of said research proposal, and one could invest in these types of research, and maybe even the distribution of medicine if one is successful!

I accept that this view may be a bridge too far and may not be entirely realistic. But I still feel that these developments may one day make it easier to have research that is more independent, more unconstrained by the ideas of a few people of what science should be.

By liberating science from the top-down headlock it is in, we may start practising liberating science as opposed to constraining science.

No comments:

Post a Comment