Big Data has been a bit of buzz word for a few years now. As a tool for collecting and processing information en masse, Big Data is supposed to be the solution to marketing problems. By using computers to processes information at a level and rate humans simply cannot, computers can provide companies with detailed information about ourselves, including where we shop, what we search for on the internet and our all our Facebook likes.
It sounds great if you’re the director of a big corporation looking to make the company’s marketing strategy more effective. But it has ramifications beyond the corporate sector. Some in edtech realised Big Data could be applied to education and jumped on the band wagon. By collecting lots of information about learners and getting a computer to process it, the end result might turn out, thanks to Big Data, to be greatly insightful to the learning institution.
So, has it turned out like that? In short, not quite. Keep reading to find out more…
Big Data in Education
The Big Data movement came along and it was supposed to be the solution to all our problems. By collecting and analyzing huge volumes of information about learners and their academic performance, teachers would be in a better position to provide more individualised Learning Programmes.
Imagine how amazing it would be if learners could complete exercises, their answers quickly processed, and then be told by a computer which areas they need to revise and focus on further.
Although it sounds nice, the very nature of this scenario is that it can only lead to one thing: more testing for learners.
Tests, Tests, Tests
As Valerie Strauss points out in this Washingtong Post article, to find out where learners are in their Learning Journey, computers require tangible input. The easiest way to get this input is from tests.
A test doesn’t necessarily mean the traditional fortnightly sit down, row by row, exam-style test. Any exercise which is based on recently taught knowledge and has a right and wrong answer will fall into the category of test for the sake of Big Data. A typical classroom example is a multi-choice quiz or a True/False questions.
While Big Data can provide lots of insightful information about the learner and their strengths and weaknesses, it is ultimately flawed by the basic requirement for computable input i.e. test results.
By giving learners lots of tests – both in the traditional sense and in the meaning of a multi-choice exercise – not only are we wearing them out but we’re also putting quantity over quality.
Education isn’t quantative: each learner is unique, every factor is variable. Learning is qualitative: better learning results in greater understanding and improved performance in applying knowledge. However, this is incredibly difficult to quanitify.
Making It All about Small Data
Strauss, in the article mentioned before, introduces the idea of Small Data. It’s a term I had never heard of before. However, it is clearly a logical step from Big Data: if quantity isn’t producing the results, go in the opposite direction and go for quality i.e. Small Data.
Small Data is all about handing over the decisions to the teachers in the classroom: let them decide what is best for their learners based on the qualitative results of the learning that is taking place in the classroom. Move away from making institution-wides decisions based on trends and averages in tests and start to focus on empowering teachers to do the best they can for their learners.
If Small Data means a greater appreciation of the qualitative aspect to education over the quantative, then I’m all for it. What about you?