Before you use an app to train your mind, be sure to read the fine print on its label

Ten years ago I left IBM Research to start a company that will build the “technology of psychology”. Back then, most people didn’t quite understand what that was. The intersection between behavioral science and computer science was virtually nonexistent.

My first step as an entrepreneur was to approach Lew Goldberg at the Oregon Research Institute. At the time, his team was running a massive data collection effort that involved mailing paper-and-pencil questionnaires to a thousand local households. The technology I was planning to build would offer him hundreds of thousands of responses within days. So at least at the research level the benefit of marrying psychology with technology was clear, and Lew agreed to join and head the company’s science team. He suggested that we approach Dan Levitin at McGill, a pioneer in cognitive neuroscience, and the three of us became the founding team of the company: a neuroscientist, personality psychologist, and data scientist – no developers. The core DNA of our joint endeavor was science and scientific research.

Today, ten years later, I no longer need to explain the synergy between behavioral sciences and technology. During the past decade, several new scientific journals have been published that are dedicated to Internet interventions. Affective and cognitive computing labs employ joint teams of research psychologists and computer scientists, and data scientists employ machine learning algorithms to analyze big behavioral data of grand new data sets (like the one Lew, Dan and I ended up collecting). Digital behavioral health interventions are now commonplace and the market is buzzing with apps that claim to be using psychological research to make life better.

Yet this proliferation came at a cost. Apps range in their quality from mere pseudoscience to being truly scientifically-sound. If you are looking to harness the power of science to reduce stress, become more resilient, more centered, more present, or perhaps seeking to alleviate symptoms of depression and anxiety, you will surely find that “there’s an app for that”. How can you tell what truly works and is based on scientific research rather than falsely boasting science?

Determine how tightly an app is based on scientific research:

When it comes to their relationship with science, apps tend to fall into one of the following four categories:

  • Not science-based: Apps that are not science based do not point to any research that serves as their scientific basis. They may mention peer-reviewed scientific papers to show that their target problem (e.g. stress) is worth fixing, but what they do in their product does not appear to be supported by science.
  • Associated with science: Apps that are associated with science point to science, and mention peer-reviewed scientific papers to show that their target problem (e.g. stress) is worth fixing, but what they do in their product does not appear to be supported by science.
  • Science-oriented: these are apps that point to several peer-reviewed scientific papers on which certain specific aspects of the program are based. For example, they may point to publications showing the efficacy of Cognitive Behavioral Therapy (CBT) if they use principles of CBT in their program. Unlike apps that are only associated with science, these apps show some general relevance of their solution and not just the problem they attempt to solve.
  • Science-based: these are apps that cite the specific peer-reviewed research on which each individual intervention they offer is based. They point to independent, published studies demonstrating that these specific activities work, and the conditions under which they were tested. This allows user to determine for themselves how the specific activities they use have been tested to work, by whom, and when, providing full transparency. Such apps are usually endorsed directly by leading researchers in the field.
  • Scientific apps: while science-based apps use solid research to show that their interventions work in the lab, scientific apps point to peer-reviewed papers describing studies where the entire program was tested for efficacy and engagement as a whole. Companies that develop scientific apps not only enjoy the direct endorsement of leading scientists, but also collaborate with leading researchers to continuously test the company’s programs for validity and fine-tune and adjust it based on findings.

Behavioral health apps vary wildly on these categories. Companies who are foreign to the research world may misrepresent their scientific validity and claim scientific merit even if their offerings are only loosely associated with science. This could result from being reluctant to make the investment in true scientific research, or simply from being foreign to the research community and its standards.

In summary:

If you are looking to use a behavioral health app to improve your well-being or the well-being of your organization, go for one that is truly science-based. Search the company’s website for the pages describing the science used, and the scientific advisors and collaborators. You should see:

  1. Peer-reviewed studies are cited for each individual intervention.
  2. The app is endorsed by leading researchers – run a Google Scholar search: collaborators and endorsers should have at least a few 100s of citations for some of their work, as a sign of impact in the scientific community. World-class scientists will have citations in the 1000s.

And most importantly, don’t be blinded with a sticker of science on the package. You wouldn’t take a pill or subject your body to treatment that’s not been tested to work; why subject your mind to it?