Tag: bad science

torture the data and it will confess to anything

Published on by purplexedscienceLeave a comment

Torture the data and it will confess to anything, as they say at Guantanamo Bay.

Bad Science, Ben Goldacre

This quote is somewhat amusing but very much true of some research carried out. It’s not just limited to science either. How many times do less than reputable sources cherry-pick data and use it to support a conclusion that isn’t correct?

Particularly nowadays with the internet being an open house in terms of the information availability. Researchers must handle data with great care and avoid the temptation of moulding it to fit their hypotheses or narratives.

How to spot bad science

Published on by purplexedscienceLeave a comment

Here at Purplexed Science, we always try to provide references and sources for any claims or information, but with a wide range of information available thanks to the internet, it can be hard to know what to trust.

For example, our first Did You Know? post was about human stomach being able to dissolve razor blades. We discovered this fact via a basic Google search, and it appeared on several ‘science facts’ posts, but there was no citation to the original paper or any kind of research on the vast majority of these posts.

If you are a student or working in the industry, it can be incredibly frustrating to search for information only to be let down when it doesn’t measure up to an acceptable standard.

Luckily, Compound Interest published this infographic (initially in 2014, updated in 2015) detailing a ‘rough guide’ to spotting bad science that is still very much relevant today.

The most important sign is the first – it’s very easy to spot a headline and not read further. It’s something most of us do every day, but when it comes to science – it’s never just about a headline.