enkerli's picture

Open Data

Big Data in Education

By Alexandre Enkerli. The 28 November 2014


Data Everywhere: Even in Education

Big Data is a big topic, these days. The tools needed to collect, process, and display data are widely available and the computing power necessary to accomplish such tasks is now within the reach of organizations of any size. Including colleges.

Some of that data may go through textbook publishers, learning management systems, registrars, or even productivity software. Among methods used to collect such data, eye tracking might be an eye opener (no pun intended). However, the field includes more casual data gathering techniques, including grade reports or observation of students’ friendship ties.

Are we all ready for data-driven education? Will access to data cause a new rift between haves and have nots?

Nothing New Under the Sun?

Five years ago (according to this strongly-worded op-ed), it was already clear that pedagogical use of data was an old story. Surely, the emergence of new approches to educational information is far from a revolution. This change may, however, represent a true shift in paradigm for learners, teachers, professionals, and administrators.

Colleges and other institutions of higher learning have clear roles in the construction, transmission, and evaluation of human knowledge. Changes in both education and technology are blurring the lines around these roles. Are colleges the appropriate platform for such work? Part of the drive behind the use of data in education relates to this question, which becomes ever more pressing in the context of budget cuts and political discourse.

A Pressing Concern?

Two series of events have pushed Big Data to the forefront: large-scale privacy breaches and disclosures about global surveillance. Partly due to widespread public interest, these events have shifted the discourse (and, in some cases, heightened the conversational pitch) about data. From work done in relative isolation, gathering data about learners has entered a much wider debate about privacy and security.

Some parts of this debate may help us understand what is at stake, when we talk about Big Data. For instance, as part of the discussion on surveillance, the notion that organizations only collected metadata was met with useful (and sometimes humourous) discussions of data usage. Simply put, metadata can serve as well as any type of data. Similarly, though it may sound counterintuitive, even aggregate data affords the targeting of individuals.

In such a context, colleges’ transparency about proper safeguards for students’ data appears essential.

Moving On

Despite legitimate (and, sometimes, not so legitimate) worries about privacy and safety, Big Data will very likely remain relevant for the college sector in the foreseeable future. Learners, educators, professionals, administrators, and other stakeholders might as well find ways to make data useful.

For instance, some place high hopes in “Early Warning Systems” which may help in student retention and drop out prevention. As should be expected, others question the effectiveness of such systems. The point remains that learners’ data could help in solving a tricky problem affecting education. Given their high drop out rates, MOOCs may represent an ideal context for action research in data-driven student retention.

College administrators and other decision-makers may also find value in education data. Some technology companies almost sound like they may be courting school administrators. Though critical pedagogues may perceive increased instrumentalization seeping through such an approach, an important discussion about the effects of teaching could reach a wider public, based on work done around educational data.

Appropriating Data

According to some, the most positive outcomes from educational data come from students’ self-tracking. Those familiar with metacognition may recognize a common pattern leading to learners’ empowerment. Issues affecting student-driven education abound, yet the dream of helping engaged students accomplish their own goals remains in many of our minds.

The tricky issue of data ownership rears its ugly head, once we start delving into learners’ own use of data. However, solutions to such problems may come from longstanding discussions. Digging through archival documentation on “Educational Information System Requirements”, doctoral candidate Kyle M. L. Jones found this useful quote, from 1967:

Some of us advanced the proposition that student records should be thought of as records that the institution holds in trust for the student. We said that if we proceed from that premise, then the institution should reveal information in its records only when the student consented. – Arthur S. Flemming, in (Loughary and Tondow, 1967)

Such a model goes beyond the educational sector. Learning rarely happens in isolation. Indeed, legal considerations may serve as a backdrop for discussions of data ownership. In learning contexts, special care might be taken to identify diverse stakeholders’ concerns.

More than Technicalities

Much work on learners’ data revolves around the technical requirements for the collection, analysis, management, and transmission of information about students. Though this work can get quite involved, it proves essential to understanding the field.

Coming out of SCORM, the well-known collection of standards used in online learning, the Experience API (also known as “xAPI” and “Tin Can API”) provides an important specification for the implementation of services tracking learning experience. What sounds like a technical detail takes a deeper meaning once we explore the situation further.

The xAPI emphasis on activity streams brings learning to social web initiatives. Those of us wishing for improved social experiences in online learning may find it too early to rejoice. Still, the acknowledgement that social ties add meaning to students’ experiences could hardly come from a less likely source than the World Wide Web Consortium (W3C).

In addition, the adoption of new standards may accompany a move to “microlearning”. When significant trends in pedagogy and technology begin to converge, we know that time for experimentation is near.

Swim with Us in Big Data

Surfing waves of change may prove satisfying, but VTE’s involvement in technology and pedagogy tends to go below the surface, into deeper (and more involved) immersion. Issues described in this article are part of exploratory work we are currently doing. Depending on people’s needs, we may transition from exploration to concrete experimentation.

To help us assess interest in work on educational data, feel free to either leave a comment here or contact us.

The following questions may help us break the ice together.

  • In your own work, have you found uses for learners’ data?
  • Do you know people interested in the use of data in education?
  • Would you be interested in projects surrounding data-driven education?
  • Have worries about data safety and privacy been expressed in your milieus?

The time is right to tackle Big Data in education.

You should be a member to comment. - Subscribe

Netiquette rules

Enter your username.
Enter your password.