Diagnosing a learning disability can prove to be a challenging job: every learning disability is different and so is every learner. In the UK, over 1 million people live with a learning disability. Of this growing number, 2 out of 5 are not appropriately diagnosed in childhood.
The traditional teaching model relies on a classroom of students being uniformly delivered the same message. The presence of a student with a learning disability poses a challenge to this model, as they may struggle with focus, the material, the speed, or several other aspects. As a result, these individuals often feel left behind.
Fortunately, a range of assistive technologies exist to aid learners with physical and mental handicaps, although these often have isolated functions and find limited success.
Detecting and diagnosing learning disabilities
Diagnosing learning difficulties in a child can take a trained specialist a significant amount of time and the accuracy of each diagnosis may vary considerably. Conditions such as dyslexia and dyscalculia often escape appropriate diagnosis, whilst autism and ADHD take many forms.
Despite being an area guaranteed to attract debate, there have been several promising studies conducted in recent years, testing the suitability of artificial intelligence in this area. These studies collected a wealth of data. Typically, they found that the application of AI would be a great asset in identifying a struggling learner’s particular difficulties – an indicated success rate of 88.674% when tested with children aged 6 to 11.
The first step in knowing how best to approach a particular students’ educational needs is to ensure the most accurate diagnosis possible. Using data on problem-solving, spatial reasoning, vocabulary, listening and memory, an algorithm has proven capable of detecting a broad range of disorders in children, including those not previously diagnosed with a specific disorder.
However, it is one thing to make a diagnosis and another to teach for that diagnosis. So what else can technology do?
Students struggling with the same subject may experience difficulty for very different reasons. But AI can help to determine this. So like a fingerprint, an AI diagnosis could be as unique as the individual in question. Could this approach compound the challenge of teaching individuals with learning difficulties by giving such nuanced instructions for handling their education?
Of course the challenge of meeting their needs already exists; having a more detailed diagnosis simply shines a light on their specific needs. However it does indicate several shortcomings where time, resources, and understanding are concerned. Often, these needs cannot be articulated – especially in children. A teacher in a classroom can only dedicate so much time to each individual learner and may not always be able to meet – or identify – their exact needs.
These varied abilities in learners often stretch the capabilities of human educators. Meanwhile, the same Artificial Intelligence responsible for pinpointing a learner’s difficulties could deliver on the educational remit. Perhaps unsurprisingly, machine learning algorithms demonstrate far greater adaptability.
Assisting learning, recording feedback
The notion of seeing your high school teachers replaced by robot counterparts is still some way off, but over the last few years the EdTech industry has increasingly made its presence known. Whilst educational technology should be used to complement good pedagogy, rather than replace it, the tech’s use in special education may see it take on an increased significance. As platforms and programmes become more advanced, the gaps in understanding can be lessened to the benefit of the students who need this tech most.
Currently, software such as Content Clarifier, which uses IBM’s Watson deep learning technology, is able to summarise or simplify text, including breaking down figures of speech into plainer terms. An EU accessibility project ‘Able to Include’ plans to release open source simplification software after trials, and companies such as YouTube and Facebook are incorporating intuitive image and sound descriptors for the benefit of hard of hearing or partially-sighted users.
A host of learner benefits have already been indicated, such as increased attention span, better understanding and improved retention. Furthermore, the ability of the AI to record data has already been established and this can be a huge benefit when it comes to gathering feedback. As special education often deals with learners who have trouble expressing their thoughts, accurate feedback on what does or does not work can be an issue.
This burgeoning tech suggests it could both implement a more adaptable model of education, and monitor how multiple learners adjust to it. Although this particular algorithmic technology has yet to fully make it out of university research departments, and developing a virtual assistant or intuitive teaching tool may take some time yet, the outlook for assisted learning is promising.
Children with learning disabilities often become adults with learning disabilities. Afforded sufficient investment towards design and distribution, the needs of special needs classrooms could be one day met in a comprehensive platform. If the technology is truly able to provide bespoke teaching assistance, it could help to better equip individuals for later life, and reduce impairments related to old age.
In stark contrast to a one-lesson-fits-all classroom, perhaps the main strength of artificial intelligence lies in its ability to treat each learner as an individual.