How can we tell if an app for autism works?

People often ask me something along the lines of “are iPads useful for kids with autism?” – a question I struggle to answer.  It is bit like saying “are pills good for headaches?”  The iPad (or tablet, smartphone, computer…) is just a delivery method, it all depends what apps you are using and how.  Yes, I can generalise a bit – for example, autistic people of all ages do often enjoy working with technology.  Similarly, yes, pills are good for headaches because you can carry them around in your bag, take them whenever you need them, don’t have to see a doctor.   But if you take a pill for indigestion, it won’t do anything for your sore head.

So, ultimately, it’s the apps that matter.  How can we tell which of them are most useful? Here’s a query posted recently on twitter:

This got me thinking, what evidence is it reasonable to expect for an app? With over 400 apps for autism on iTunes alone, it would be a huge job to carry out a randomised controlled trial for every one, though this would be the ‘gold standard’ scientific approach. Not to mention that by the time you got the data in, another 400 new apps will have been released. Our report on the trial of FindMe is currently in peer review and (hopefully) moving towards publication: this will be about two and a half years after the app was made and close to six years after we first applied for funding for the project.

There are other obstacles – most apps are made by commercial enterprises who may not have many staff with scientific research training. Besides, a good evaluation should ideally also be independent.  Even if they approach a University to request an independent evaluation, they may be met with suspicion.  Most academics don’t want to give away their expertise for free, and Universities are (justifiably) cautious about lending their implicit endorsement to commercial products. As this review of our own app indicates, a University backing lends an automatic respectability and reliability to a product.

So what are the alternatives?  Here are a few suggestions of what you might look out for when judging the evidence-base of an app, which are a bit more realistic than a full-scale independent trial:

  • theoretical grounding – is the app targeting a behaviour, or using a learning method which is based on an established theoretical model of autism, or of development and learning more generally?  Are the creators citing any academic literature to support their chosen target?
  • claims – how bold are the creator’s claims?  Anyone who says “your child can learn to talk with this app” is aiming very high, and possibly exploiting a major worry for parents.  This level of therapeutic claim should be supported by some real evidence – at the very least independent research studies from a University. A few personal stories do not cut it. If on the other hand the app says: “this app might help autistic kids learn new words” that’s a bit less dramatic.  The developers are saying something like: “this is what we’re trying to do – why don’t you give it a go?”.
  • price – likewise, apps which retail for £1.99 versus apps which cost £199.99 should be approached differently. The developer of the expensive app should be putting some of that profit into some evaluations of their product and if they haven’t, then why not?
  • expert consultation – this is something we can expect from almost any app developer who is creating something for an atypical and sometimes vulnerable population. There are developers out there who have clearly seen the commercial potential of the autism market and are creating apps labelled ‘for autism’ which have nothing ‘for autism’ about them. It isn’t much to expect a developer to have talked to autistic people, parents of children with autism, teachers, clinicians, therapists or whoever might be the most relevant to their app. If they have, they should be sharing this on their website.  Also, look to see if the consultants are also getting a profit share…
  • reviews – as a scientist, I don’t set much store by anecdotal evidence.  Any app developer worth their salt will be able to muster up a few satisfied customers (or just friends) to write positive reviews of their app online. But if an app has been widely reviewed, on iTunes, blogs, twitter and other outlets, and you’re reading a consistently positive message, that is not a bad sign!

The other issue is how you use your apps – iPads and apps aren’t a panacea, they need to be used wisely and combined with real-world, interpersonal learning. This week I am in Belgium, working with a colleague, Petra Warreyn, to develop a new research project idea. We’re hoping to create some new apps and combine them with therapist sessions and parent training. Our belief is that combining different approaches will lead to better outcomes for the children with autism who trial the intervention. The therapists will be flexible, highly-skilled, creative and social; the apps will be motivating, allowing child-led learning which is repetitive, structured and available all day; the parent training will bind the two together, providing opportunities for children to practice their new found skills during everyday activities.  Or at least, that’s the plan.  Wish us luck!