Frankl for autism
The first Frankl project aims to make autism research more open, more robust, and more useful
Frankl is a platform for open science. It makes it easier for scientists to share their methods and data as transparently as possible — and rewards them for doing so.
The concept for Frankl is one that’s evolved out of my own experience as a cognitive scientist, investigating the psychology and neuroscience of autism.
By its nature, autism is a difficult condition to investigate. If someone tells you that they “know” what causes autism, it’s a sure sign that they’re not as expert as they might think that they are. The same applies if they give an unqualified recommendation for a particular intervention. There are no simple answers.
Autism researchers also face challenges that are shared with other scientists across a wide range of disciplines. The scientific community is slowly waking up to the fact that many published findings don’t stand up to scrutiny, are hard to replicate, and are even harder to build upon.
There’s a growing recognition that science should be conducted and reported more openly and transparently, that it needs to be more collaborative, and that stronger connections need to be built between scientists and the broader community so the research we conduct translates into real world benefits.
Frankl exists to address these challenges. Ultimately, we want Frankl to be used by scientists of every flavour. But our approach is to focus on solving specific problems. And then work with scientists in other fields to repurpose those solutions.
And so our first project focuses on autism. We think Frankl can make a real difference to autism research and to the lives of autistic people and their families. But it’s also a template and a testing bed for ideas and technology that can be applied well beyond autism.
As a scientific concept, autism is complicated and messy
Before introducing the Frankl autism project, it’s important to sketch out where I think we are as a scientific community in terms of understanding the condition.
The first thing to stress is that the reality of autism is far more complex than you might imagine based on popular media or introductory textbooks. According to the latest estimates, autism affects 1 to 2 percent of the general population. But the diagnosis of autism (or autism spectrum disorder) captures a very diverse group of people.
To receive an autism diagnosis, an individual must demonstrate difficulties in engaging with other people. But this can manifest in very different ways — from what might be described as social awkwardness all the way through to extreme avoidance of social interactions.
Autistic individuals also vary enormously in terms of their cognitive and linguistic abilities. Some people have excellent language skills and are obviously highly intelligent. But around a third have little or no spoken language.
Until recently it was assumed that these “minimally verbal” individuals were also intellectually impaired. But it has become apparent that many of these children and adults actually have good language comprehension and high levels of intelligence that were effectively hidden because they were unable to communicate in conventional ways or complete traditional cognitive assessments.
Increasingly, researchers accept that there are many different types of autism — we just don’t know how best to delineate them. There’s also a recognition that the autisms (plural) overlap with other conditions that were traditionally thought of as being quite separate. These include developmental language disorder, ADHD, schizophrenia, intellectual disability, as well as a host of genetically defined conditions such as Fragile X syndrome. This overlap is not just in behavioural features but in the underlying genetics and neurobiology.
From a scientific point of view, autism is complicated and messy. At least for now, it is a description, not an explanation.
The future of autism science is to better understand individual people
In my view, the way forward for researchers has to be better characterisation of individuals so we can go beyond the label and understand the variability within autism. We can’t go on simply comparing people with an autism diagnosis to those without.
In practical terms, this will require much larger sample sizes to draw out subgroups and patterns across individuals. In turn, that means more collaborative research and pooling of data from independent studies. This will require infrastructure for data management and sharing that preserves the privacy of the people taking part in research.
It also means greater openness in methods so we can understand how different studies looking at the same questions may reach different conclusions. And it means that we need to move towards standardization of test procedures so that data collected by different researchers can be pooled in a meaningful way.
Alongside the move to greater openness within the scientific community, autism researchers also need to conduct studies that reflect the immediate needs of autistic people and their insights derived from lived experience. Many of the wrong turns that autism research has taken in the past have come from non-autistic people misinterpreting the behaviours of autistic people.
Finally, recognising the diversity within autism also means adapting research methodologies to accommodate all diagnosed individuals, including those needing the highest levels of support. We can’t go on testing only the most cognitively able individuals and then extrapolating our results to other people who may have very little in common except the diagnostic label.
Our first step is to build iPad applications that improve cognitive assessment
Our first step on the Frankl journey is to develop iPad-based cognitive assessments that allow researchers and clinicians to better characterise the abilities of autistic individuals.
An iPad interface plays to the strengths of people with autism — particularly those who are minimally verbal — and so provides a more accurate measure of true abilities than conventional “pen and paper” cognitive tests that many autistic kids fail to engage with. It reduces (and potentially removes) the need for the child to interact with a researcher or clinician administering the test. And it allows the tests to be “gamified” so they are intrinsically fun and rewarding.
There are other benefits. Test administration and scoring can be partly or wholly automated, increasing efficiency and reducing errors in transcribing results. Automation also improves consistency of data collection across different personnel and sites. It broadens the user-base and allows tests to be administered remotely in places that may not have access to trained psychologists.
Frankl apps will incorporate a number of features aimed at facilitating and rewarding open science
The user interface is really only half the story. The apps we are building will also serve as a vehicle for features that advance the cause of open science.
Built-in data management and archiving: Current practices in cognitive assessment typically involve typing test scores into an excel spreadsheet and sharing with collaborators via email or a service such as DropBox. With Frankl apps, data will be automatically pushed to a secure repository. This facilitates data sharing within collaborative studies. It also means that the data are automatically archived and that anonymised data can be easily shared beyond the research team if and when that is appropriate.
Blockchain integration: Another key feature of Frankl is our use of blockchain technology to provide secure, immutable (non-deletable), public records of scientific data. Like many others, we’re very excited about the potential of blockchain to transform scientific practices. But we’re starting with a very simple innovation. When Frankl applications push data to a repository they will also write the metadata (that is the description and location of the data) to the Ethereum blockchain. This will facilitate data management, collaboration, and meta-analysis, and will reduce selective (and inevitably biased) reporting of data.
Digital currency tokens: Access to Frankl apps will be via the Frankl cryptocurrency token. The sale of tokens is the mechanism by which we’re initially funding the project. But there are many other good reasons for adopting a token. The use of cryptocurrency facilitates micropayments that avoid transaction and currency exchange fees. The token also provides a mechanism to reward scientists for engaging in open science practices and for Frankl to engage in strategic partnerships with research organisations.
Open science marketplace: Initially, the Frankl team will build the applications, but we will work with researchers who have developed their own tests and make them into Frankl applications. We will also release the code to make it possible for anyone to put their own tests on the platform and receive tokens each time their test is re-used. Frankl will effectively become a marketplace for data collection and management applications.
The big picture
We’re really excited about this autism project. But it’s only the beginning. Cognitive assessments that work for autistic kids will also work for typically developing kids and kids with other conditions such as Down syndrome or dyslexia. Similar applications can be used with adults. We’re already working to build assessment apps for dementia research and clinical assessment.
Frankl is also positioned to take advantage of new developments in blockchain tech and applications as they become established. To give one example, there are numerous projects working on blockchain-based medical records that give patients “sovereignty” over their own data. Building these capabilities into Frankl apps would, for instance, give parents of autistic kids the ability to share the assessment results with relevant professionals — or contribute their data to research.
And while our initial focus is on cognitive science, the infrastructure we’re building will be modular and multi-purpose. The key elements of Frankl — integrated data management, blockchain, a token economy — can be re-used, recycled and repurposed to facilitate and incentivise open science across the scientific spectrum.
At Frankl, our mission is to make open science easy and rewarding for scientists. If you’d like to know more, you can read our whitepaper, check out our website, follow us on Facebook and Twitter, or join our Telegram channel.
We love feedback (positive and negative), so please let us know what you think — write a response or just hit the ♥ button and share this post with friends and colleagues.
