Apr 112016

Ask any scientist how their research is going and, after they have stopped day dreaming about what it would be like to have a stable job, you’ll normally get an answer which alludes to “I’m dead inside and there aren’t enough drugs in my body for me to be able to lie to you” or put another way “fine thanks”.

Scientific progress is slow, painful (at least for those at the lab bench) and sometimes soul destroying.  Scientists will spend days/months/years repeating failed experiments or conducting iterations of the same experiment just so they can publish that next paper. Just like a slot machine, with enough persistence and luck, science will occasionally reward you with a payout. The payout takes the form of a piece of data that you can take to your supervisor and shout in their face “THERE IT IS, IS THIS WHAT YOU WANTED?” to which they will respond “why are you in my house so late and who let you in?”

Why then, if failure is one of the main themes of science, do we not build resilience into our budding scientists? Even undergraduate experiments are set up to always succeed. This does nothing to actually prepare young people for a career as a scientist. All this teaches prospective scientists is that science is fun and always works – ask any PhD student what they think of science and I am certain that the word fun will not feature. Science is rewarding, sure, but most of the time it is definitely not fun in the same way liquid nitrogen ice cream is. So, can we not change the message from science is fun to science is rewarding? We have done it with sport; we tell kids all the time that sportsmanship is about being a gracious loser and not chucking a tantrum or referring to the other team as stinky poo faces.

It is therefore very important that science outreach teams and undergraduate experiments feature failure in their activities. However, the failures shouldn’t be highlighted as something bad, but rather as a way of getting students to problem solve. That way, if they decide to take the red pill and enter the lab as a research scientist they’ll be ready for some of the harsh realities of scientific research and beat all those over stinky poo face scientists.


  2 Responses to “The importance of teaching failure”

  1. I like the sentiment of this post. It must be a basic human response to feel greater reward from confirmation than refutation, but it’s not even as though a properly constructed scientific experiment can truly ‘fail’, at least in the same sense that you can fail to win a sporting match. I aspire to feel satisfaction from being methodologically, rather than practically, vindicated; one of many reasons why I’m such good company.

    It is possible to fail to meet objectives imposed by, say, supervisors or funding bodies, which is a separate question to the feeling of failing ‘at science’. Is this a problem with the structure of science as a career? Are the majority of research objectives stated with certainty that they can be achieved, thereby punishing unexpected or null outcomes? Perhaps if research objectives were stated as questions we could mitigate feelings of failure, though the awarding of grants on this basis seems unrealistic.

    Adam Savage has a nice talk on failure, too: https://youtu.be/1825zkmJVuE which may be relevant.

  2. Slot machines and gambling and success in science are quite related. The random nature of rewards is a key feature of all these pursuits, and why they are addictive. The podcast on ABC interviewing the brains behind slot machines is worth a listen regarding the psychology of gambling (and therefore science, imho)

Leave a Reply

%d bloggers like this: