Week 1 of #EL30 addressed the topic of Data. Within that, two core conceptual challenges relating to eLearning were explored, “first, the shift in our understanding of content from documents to data; and second, the shift in our understanding of data from centralized to decentralized.”
All of this exists within the backdrop of “what is now being called web3, the central role played by platforms is diminished in favour of direct interactions between peers, that is, a distributed web”. The topic of data is relatively new to me and I am figuring much of it out as I go.
Our data exists online across multiple distributed nodes and each of us embodies the unique identifier that links all of this data together. In Stephen’s week 1 data summary article he highlights how digital data is beginning to permeate many aspects of our lives – “We are beginning to see how we generate geographic data as we travel, economic data as we shop, and political data as we browse videos on YouTube and Tumbler. A piece of media isn’t just a piece of media any more: it’s what we did with it, who we shared it with, and what we created by accessing it.” The traces of data we leave behind of where we’ve been online creates a depiction of us for those that can see it, an online identity, from breadcrumbs in the digital woods.
Activity – Conversation with Shelly Blake-Plock
- using data in actionable ways to understand learners, to improve instruction and content and to manage data systems that support learning,
- the Experience API (xAPI) specification,
- the xAPI enterprise learning ecosystem,
- Learning Record Store (LRS),
- data ownership and management,
- identity management applications,
- the privacy trade-off of these systems.
There was good discussion around Experience API, commonly abbreviated to xAPI, a modern specification for learning technology that helps to turn learning activities, experiences and performance into data. Shelly was the Managing Editor of the IEEE Learning Technology Standards Committee Technical Advisory Group on xAPI (TAGxAPI) who created a technical implementation guide for xAPI.
Essentially, xAPI was created as a way of tracking learning experiences and performance that extends beyond the bounds of our traditional Learning Management Systems (LMS) and the content and activities that learners launch from within them. It facilitates an individual’s learning to be recorded and moved more freely from siloes such as the LMS, as long as it in xAPI format or can be converted to it. The notion is that learning occurs everywhere, it’s not simply confined to the LMS or to the classroom, and now it’s possible for the data generated from learners’ experience and performance (online and offline) to be tracked and sent via x API statements (signals) from a range of different origins such as mobile apps, simulations and games, and the physical world through wearable technology, sensors and online games.
With this data it becomes possible to analyse and understand how learners are learning and potentially improve the content and activities that they receive. xAPI statements about learning experiences can then be hooked up via a number of launch mechanisms to a Learning Record Store (LRS) to collect reams of data about how the learner interacts with their learning environments. Analysis of this data can be automated through machine learning algorithms depending on what type of information is being sought.
Most of us have likely become familiar with the term ‘surveillance capitalism’ as the purported business model employed by many web2 corporations and platforms. Online data generated by each of us (our digital footprint) is already bought and sold to online advertising and marketing agencies. We unwittingly and nonchalantly give our ‘consent’ to it by clicking agree to the terms and conditions of the seemingly ‘free’ online platforms and services we sign up to.
The ‘business model’ is explained early in this presentation by Laura Kalbag of Ind.ie:
When viewing all of this through a critical lens, talk about tracking and gathering learner data for analysis immediately brings with it the need to talk of a range of considerations around ownership, ethical use, privacy, security, and data governance. I’ve noted similar sentiments from many of my fellow #EL30 participants.
The use of learning analytics to support the student experience could afford valuable insights, but there are ethical implications associated with collection, analysis and reporting of data about learners.
According to Rebecca Ferguson (2012), “Learning analytics is “the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs.”
JISC UK’s Code of practice for learning analytics, authored by Niall Sclater and Paul Bailey, provides very helpful guidance in this regard beneath 8 key headings identified to help institutions (and possibly other organisations) understand and carry out responsible, appropriate and effective analysis of the data that they gather:
- Transparency and consent
- Enabling positive interventions
- Minimising adverse impacts
- Stewardship of data
Niall Sclater also compiled a literature review of the ethical and legal issues for this code of practice, in which he collates some critical ethical questions from a diverse literature authorship in relation to many of the areas identified in the code of practice. Here’s a snapshot of some of the thought-provoking questions posed in that review:
|Ethical questions||Code of Practice area|
|1||Does the administration let the students/staff know their academic behaviours are being tracked? (Hoel et al., 2014)||Responsibility|
|2||Does an individual need to provide formal consent before data can be collected and/or analysed? (Campbell et al., 2010)||Transparency and Consent|
|3||How transparent are the algorithms that transform the data into analytics? (Reilly, 2013)||Validity|
|4||Who can mine our data for other purposes? (Slade & Galpin, 2012)||Stewardship of data|
|5||Who is responsible when a predictive analytic is incorrect? (Willis, Campbell & Pistilli, 2013)||Privacy|
|6||Does [a student profile] bias people’s expectation and behaviour? (Campbell et al., 2010)||Minimising adverse impacts|
On the #EL30 course I’ve read a bit about IndieWeb, a community based on the principles of owning your own domain and owning your own data. IndieWeb attempts to make it easy for everyone to take ownership of their online identity and believes that people should own what they create. https://opencollective.com/indieweb#about I definitely want to explore this further in light of the next generation of learning technologies.