Five stars is not enough!

 Puzzling over brains, that's us    Courtesy of  Stephenhamshire

Puzzling over brains, that's us

Courtesy of Stephenhamshire

One of our aims in this project is to develop products that genuinely help people and to do this we need feedback from users. However, even if we get all the likes, tweets and stars the internet has to offer that won't be enough for us. The problem is: we could get 100% great feedback saying that we cured everyone's phobia, but that actually doesn't prove we have done a good job. It would be great and a good start but it isn't good science.

Powerful placebos

The problem is the placebo effect; people generally feel better it they believe they are doing something that is going to help them. Taking a sugar pill, with no painkiller in it when you have a headache will probably make the pain less. It even works when you are told you are taking a sugar pill. Take a look at this study in irritable bowel syndrome if you don't believe me.

So the problem is just because an app like this gets good feedback doesn't mean it works. It could be a 'sugar pill'. We want to be sure it isn't. The best way would be to do what is called a double-blinded, randomised, placebo-controlled study or RCT. This is the 'gold standard' way of proving a treatment is effective. It aims to remove bias from the study and ensure there is more than a placebo effect.

This is how the RCT sausage is made


  • Two 'treatments': the one you are testing and a 'placebo control'
  • A big group of people with the condition


  • Randomly select half to have the treatment and half the placebo
  • Have a test which measures the condition you are trying to treat
  • The people receiving the treatment and those asking them questions or looking at tests and analysing the results should be unaware if they are getting the treatment or placebo
  • Crunch the numbers and out comes the answer if your treatment works

Sounds simple? it isn't!

One problem for non-drug treatments is creating a placebo,. The placebo control should be inactive, but the same as the treatment, lacking just the part you believe to be the key ingredient. For a tablet this can be relatively easy to do: just make the placebo the same size, colour and taste as the active tablet. As you can imagine it is much trickier to do for the kind of psychological process we are creating. 

What is often used is a non-specific psychological treatment. So for us it could be a similar general app for anxiety rather than a specific phobia. This isn't a true placebo---as it has some of the 'key ingredient' itself---but this is good because it makes it harder for our app to prove it has an effect. As an alternative we could get someone to play Angry Birds for an equivalent amount of time to the Phobia Free app but we think that's too easy.

There are many other practical problems, how do you 'blind' the participants, accounting for people not doing the app, and 'measuring' fear to name but a few.

Both me and Andres have been involved in designing and carrying out medical studies and it is really tough because you want to find a new way to help people, but you have to design the study to try to prove the opposite.

We hope that Virtually Free takes off because one of the things we would do is exactly this kind of study. Unfortunately it takes a lot of time and money so, not yet. 

We are still trying to be scientific; we have based the app on a treatment that has really good evidence and the techniques we are using have shown to work under other circumstances. This isn't proof our app works, but we think it has a really good chance to be effective. We could be doing like other groups and getting 10 people at a time tested until we have evidence and spend years and years in development. We preferred to take this route. Launch now with our best effort, be transparent about it, and get people like you to help us figure out how to make it as effective as possible. We will get the science bit done and we will gradually improve our formula as we go along.

One of our goals on kickstarter (thank you for backing us if you have and if you haven't go on and do it; it's for a good cause!) is to use fMRI (functional magnetic resonance imaging) to assess how effective Phobia Free is in helping people. The first step would be checking that Richard and his team have done the job in creating spiders that cause fear in phobic people. As a first step we intend to once again use me! After that we will look at whether the 'fear signal' disappears as the treatment progresses. We will do other projects with and without fMRI. We will keep you posted on how to get involved!