Analytics in Education : An Endless Number of Possibilities

Data runs the modern world. Targeted advertising has made relatively new companies like Google and Facebook corporate behemoths. Online retailers like Amazon and Target can predict your next purchase before the though even forms in your mind. Data is everywhere and deeply affects our lives in the 21st century.

Another thing that also deeply affects our (and everyone else's) lives in the 21st century is Education. It has, however, remained somewhat aloof of the frenetic disruption that data-driven technologies have caused elsewhere. Education certainly isn't lacking in data; in fact it is one of the most data-intense domains where performance is tracked and evaluated as an end in itself. However, precious little has been done to leverage this veritable gold mine in order to improve the quality of education students receive.

The Start of Analysis

Schools are slowly adopting the latest and greatest in technology. Once a domain only for the ultra-exclusive schools for the rich or the so-called “alternative education” practitioners, opportunities for producing data-driven solutions are increasing day by day. Modern technology allows for tracking fine metrics and accurate data analysis.

Analytics in education can work for the betterment of both teachers and students. Learning from the pitfalls of No Child Left Behind program in the US (viz. connecting teachers' pay to student performance, among others), we can use the data rich environment afforded by schools to improve both how teachers teach and how students learn.

Window shopping

Because of the sheer scope of data that is tracked by schools, analysis can lead to insights in not only education but also how that affects and is in turn affected by socio-economic factors. Tracked variables include, but are certainly not limited to, location, health records, parental income, parental status, past learning experiences, educational variables like grades and scores, attendance and truancy records, etc. It is paramount that this sensitive data be properly anonymised and randomised in order to prevent biases and data-hacking to influence results favourable to vested interests.

Once the proper precautions are taken, though, even very simple analytical tools can be used to derived valuable insight, from drawing attention to laggard students to drawing connections between academic performance and behavioural patterns such as truancy or late submissions of assignments.

A new stratagem

In order to leverage this analytical muscle, education must move to new age media that supports tracking. This means an increased focus on online learning, education software, and social network integration. The conventional methods of paper books must be abandoned at the institutional level. This will not only reduce overhead in converting data-on-paper into digital but also allow tracking behavioural patterns that otherwise could never have been discerned and analysed.

The way Google tracks login behaviours to detect and predict fraudulent activity, educational analytics can track the time spent on a particular chapter or paragraph to model not only each individual student but also the student body as a whole vis-a-vis learning ability and instruction assimilation. Tracking how students interact with each other on social media and other platforms of discourse can also be effective in evaluating the value of each individual's contribution and further diversify the data pool from which we can draw insight.

This situation has also spurred further research and development into Learning Management Systems (LMS) that are environments tailor made to facilitate and prioritise learning in every interaction while minimising the possibilities of entertainment or distraction that conventional social networks facilitate at every level.

From personalising courses to individual ability to providing new means of evaluating progress that are not just standardised testing, analytics really hold the key to unlocking the next paradigm shift in education.

There are however, concerns. Privacy is paramount among them, as the invasive data collection can be used to affect discrimination in the wrong hands. Also concerning is the question of ownership of the data generated by these mechanisms; does it belong to the students or the institution? What if a student wishes to exercise the equivalent of The Right to be Forgotten in this context? Who can access the results of the analysis? Should students be able to check their own profiles? There are no easy answers here and significant teething problems are to be expected.

Treading carefully

For all the importance we assign to education, we still have no clear idea about the underlying mechanisms of education. Even after all these years, we do not know which standard educational practices affect favourable outcomes in students, if any. We do not know which practices need to be curtailed. It is mostly flying blind and the Silicon Valley approach of Move Fast and Break Things is as appropriate in this environment as an ice cream cake near a crackling fireplace.

Analytics also has its limits. The variables we track are structurally defined by the medium we observe. As such, there is no way to analyse the impact of an encouraging teacher or a supportive circle of friends and co-learners. We must tread carefully and not lose sight of the goal that is to improve education in learners, not just as statistics.

We know that we need to reform our education system but we cannot predict its effects with any degree of certainty. Analytics can certainly ameliorate this impasse by providing dependable insight on how to best treat the pain points in modern education systems. The blinds must be lifted before we can chart a course to the destination. And only then will we be able to enjoy the view from the windows.

Author: Subhom Mitra