Irreproducible results

We’re all guilty. We make statements and draw conclusions based on insubstantial evidence. But that’s okay!

In normal conversation. But not in science. The scientific method, and the advancement of scientific knowledge, depend on well-founded observations and accurate reporting of results, which should be reproducible. Otherwise they’re one-off and probably meaningless. Anyone with common sense knows you can’t make a trend from one data point…

So when in 2012 researchers at the LHC claimed they had discovered the Higgs Boson, the particle that gives everything in the universe mass, they backed up the claim with evidence from millions of observations that gave a statistical significance of “5 sigma“. But when in 2014 Haruko Obokata “discovered” STAP cells, which could have revolutionised healthcare and biotechnology, it turned out that her results couldn’t be reproduced. And rightly, her papers were retracted and other scientists could stop following the false leads that her work had led them down. Because of the high-profile nature of stem cell research, there was a lot of other harmful fall-out too, which I won’t go into here…

In my research there’s also a risk of falling into that trap. If I have a hunch and one experiment seems to agree with it, what’s stopping me reporting it straight away without performing other experiments to confirm it or, worse still, ignoring other conflicting evidence? After all, I hear about this regularly from my colleagues – it’s real and it happens. But when I’m writing a paper I want to write it in the most positive way possible, making sure the important results stand out; I want to tell a good story, not just report facts.

But, importantly, the story should be based on all the facts, it shouldn’t use the facts as and when it sees fit. And that means that I have to be clear how I got my results and what my conclusions are based on. After all, science isn’t politics!*

*perhaps we need our politicians to understand science a bit better…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s