By Alastair Montgomery, Co-founder and Managing Director of APT
I have read a number of articles around this subject over the past few weeks, as this year forces an alternative approach to students’ end-result of their courses.
Typically, the argument appears to be between teacher-led assessment versus examinations in their current state. I don’t believe either are the best way forward. Having worked for years with adaptive testing methods – assessments that adapt to the candidate’s abilities – I believe these could be applied in a much wider context. I appreciate this would not be applicable in the short term, but whilst we are discussing the matter; here’s a proposal.
How about regular low stakes adaptive testing? Resulting in highly detailed academic profiling of individuals over time. I would recommend not calling it testing at all. But computer-adaptive profiling would work. I would suggest that such profiling exercises could be done at home, as homework, at the end of a topic – or the start. All feeding into a long-term picture of academic performance.
It could be done on a national scale, and thus it would become a hybrid of our current licensed formal examinations and teacher-led assessments to create something entirely focused on the learner.
Computer marked and testing a wide range of competencies including skills, understanding and knowledge, Adaptive tests can check forward (beyond what has been explicitly taught in class) and check back on what is expected to be known at a given level. This would challenge the most able and identify those students with gaps, leading to better differentiation and support for the students at risk of being left behind.
The effect over time would be official academic reporting on progress, strengths, weaknesses, potential. A full transcript would be all a future employer, university admissions team, or selective school would need to see. Lock it all up in blockchain technology and you’ve got a very interesting way forward.
I have bearly even thought about the some of the big data implications, but the opportunities in this area are intriguing.
The technology is there. The levels, skills, curricula, and expectations are all set. Nothing wildly new from that point of view. Of course I am conscious of caveats around legitimacy, security, branding a learner because of a certain profile, etc – all things that I believe could be countered and worked with. I would love to hear your thoughts on it.