The Science of Stripes

12/04/2012 22:33 BST | Updated 12/06/2012 10:12 BST

In the drive to show that science isn't just for scientists BBC Radio 4's Material World programme offers amateurs like me the chance to put our theories to the test. Most of us would expect that claims about medicines, educational theories or beneficial foods would be backed up with some scientific evidence but... fashion?

I want to find out whether it's true that vertical stripes are more flattering than horizontal or whether Steve Jobs and Coco Chanel had it right and black is best. Material World agreed that this is a question people would be interested in and I'm now planning the experiment.

You may remember some research a year or so ago that seemed to answer that question. This was based on the idea that the Helmholtz illusion would work for people as well as squares. This is the visual illusion that makes the square with the horizontal stripes appear taller and narrower.

That study used pictures of people so there is an opportunity to take the research to the next stage and use real people. Dr Peter Thompson of the University of York is the specialist in the psychology of perception who conducted that research and he agreed to act as my mentor.

When I said I was doing this experiment the responses I got fell into two groups. Some engaged with the details of the experiment (does the width of the stripe make a difference? Will you be doing diagonals?). And some questioned the premises (this is so subjective, how can you test it? Doesn't it depend what you already think? It's a bit frivolous isn't it?).

The answers to the first type (yes and no respectively) are less interesting than the second, so let's have a look at those.

The question of subjective judgement is one that experiments in psychology often face and it comes down to a question of experimental design - being very clear what you are testing and what you need to exclude.

For our experiment we'll show participants a series of videos of people wearing the different patterns; we'll then ask them to gauge the height and size of the person. Because we shall be asking large numbers of people we'll be able to remove the individual subjectivity so if there is a significant difference we'll be able to say with some confidence that pattern affects perception.

However, even though there will (with luck) be lots of participants, they will all be the sort of people who come to the Edinburgh Science Festival on April 13-15, so there may be shared preconceptions which influence what they see. That means that even if we see a difference we won't know whether that difference is due to a preconception or a visual illusion. To deal with that we'll ask people after the experiment what they expected to see and I'm particularly looking forward to finding out whether people who expected a certain effect did see it.

Of course what we find won't be conclusive. Any experiment is a compromise between scientific validity and usefulness in the real world. In an experiment you have to hold as many factors constant as you can to be sure that the effect you see is real. In our case that means choosing models of the same age, height and colouring and showing them on video rather than in the flesh, so it's not completely lifelike.

Our participants will also know what the experiment is about before they start, which is not ideal. And the students at the University of the Creative Arts who made and modelled the clothes may have had their own preconceptions. Perhaps they stand taller and walk with more confidence when they're wearing certain styles?

Finally, is it all a bit frivolous? Yes. So what. Why does science have to be worthy?

For more information go to this Radio 4 page:

Or follow our progress on