So what are some of the key issues and challenges when it comes to learning analytics and ethics in higher education.
This was the challenge I was set for a presentation at the University of Hertfordshire Teaching and Learning Conference on the 10th July.
When this was originally planned I was going to travel over to Hatfield and deliver the presentation in person, however with the Covid-19 pandemic, it was soon apparent that this wasn’t going to happen. The conference was going to go ahead, but using Teams, and presenters such as myself would deliver their presentations online.
I have done this a fair few times, so I know what to expect, however it’s still a little weird delivering a presentation in front of my PowerPoint slides and not seeing the audience, nor any kind of audience reaction. At the end their were a fair few questions in the chat pane, so got to answer more questions than I would at a face to face event.
I didn’t really know initially how it went down, but there was some positive feedback on the Twitter.
Echo that! It was soo good that I forgot about the ice cream! Thanks @jamesclay @HelenBarefoot @KarenAnneBarton for a very informative and forward looking debates! Fabulous closure for the #UHLTC2020 conference
— Zamzam Ahmed (@Zamy301) July 10, 2020
@jamesclay is taking us through the ethical issues of using student data & analytics. Privacy & consent is so important. Consent of the processing of data and what we do with it. Can we understand how data tells us story? how does this lead to action if any? #UHLTC2020 pic.twitter.com/FSqbmj8SOF
— Lucy Bamwo (@LucyBamwo) July 10, 2020
Importance of validity of data and appropriate understanding of what the data tells us #UHLTC2020 @jamesclay pic.twitter.com/CsgzeHlY5P
— Helen Barefoot (@HelenBarefoot) July 10, 2020
Great to have @jamesclay talking to us about how revolutionary-changes are changing society and how technologies coming together enable data analytics #UHLTC2020 pic.twitter.com/9AbECVz8CN
— Helen Barefoot (@HelenBarefoot) July 10, 2020
I took the audience on a journey, through the technological changes that have taken place and what the implications of this are for analytics in teaching and learning.
As with any presentation I talked about who I was and what I did and also who Jisc are and what they do.
Jisc’s vision is for the UK to be the most digitally advanced higher and further education and research nation in the world.
Jisc’s mission is to enable people in higher education, further education and skills in the UK to perform at the forefront of international practice by exploiting fully the possibilities of modern digital empowerment, content and connectivity.
I always think that it’s useful to set the scene when it comes to any kind of technology, so I wanted to set the audience on a journey about how we’ve had a number of technological revolutions, which have all had an impact on education and universities. To understand the future sometimes you need to understand the past. Steam was the power behind the first industrial revolution, allowing urbanisation and mass production.
The second revolution was electricity.
The third, which most of my audience would have lived through was the computing revolution. Over the last forty years we have seen the growth of the personal computer, networking and reliance on the internet for research, communication and collaboration.
So what of the fourth industrial revolution, what’s that about? Well the current thinking is that this is not just one thing, but bringing lots of things together.
A major part is data and analytics. The ability to gather a wide range of datasets, the computing power to crunch and analyse that data and the cloud servers to store and access data.
Another aspect is robotics. In regard to higher education, this isn’t about robots teaching classes, but how the software and artificial intelligence that allows robots (as in manufacturing robots for example) to learn about their environment, and learn how to react as their environment changes. I think this has potential in how an intelligent learning platform could be built.
This leads onto autonomous vehicles and there are many issues and challenges here, including ethical dilemmas.
Another technology that will help revolutionise society is 3D printing, thought the benefits for education are difficult to work out. There are some universities out there for example printing 3D fossils for distance learning courses, but you don’t see students with their own printers printing objects.
An area which will also have a dramatic impact, but one which is way out of my comfort zone is biotech. We know that the use of biotech in robots for example is one way in which different techological approaches are coming together.
Along with autonomous vehicles we are seeing a huge growth in the development and use of electric vehicles and the resulting technological advances that this can bring in charging and battery technologies. There is also the resulting impact on the reduction of pollution and reducing the impact on climate change.
Another technology which will add to this fourth industrial revolution is the improvements that technologies such as 5G will bring to connectivity and for devices to be able to quickly report activity, but also gather data from other devices and sensors more rapidly as well. This means much faster reactions from stuff like autonomous vehicles for example.
When we think about the fourth industrial revolution then one of the key aspects of this revolution is how all these different technologies are coming together to build new methods and ways of undertaking activities and experiences.
I then started to focus on the main part of the presentaiton on learning analytics and ethical considerations. We already have systems that gather data on student activity (and other university functions).
What learning analytics does is take that data from various sources and datasets and brings it together into a single hub. You can then use that data to discover insights about student activity, learning, the curriculum and other parts of the student experience.
Jisc has been at the forefront of the learning analytics revolution and I mentioned some stats I had.
• 750 million records in the learning data hub.
• 24 HEIs signed up for using the LA service – all at various stages
• 7 FEIs signed up for piloting LA in 19/20
• 40,000 + students have used the Study Goal App this month (at least once)
• Almost 7000 staff have used Data Explorer Dashboard last month (at least once)
The real power of learning analytics is bringing dataset together to understand the narratives behind those data stories. It’s not just about looking at data, but using those insights from that data to discover the true underlying narrative.
I discussed the algorithms and recipes that could be used to analyse data, but be aware that algorithms are subject to bias, but you could also use algorithms to remove bias.
I wanted to explain why the use of data could be useful in the world of learning and teaching. I wanted to discuss how universities should harness the power of their data and use analytics to tackle some of the big strategic challenges within the organisation
It could help us to understand why and how we could transform teaching. Could we use analytics to help us in transforming teaching to improve the overall student experience, taking advantage of the full potential of big data, AI and other technologies in enabling, enhancing and improving the university and college experience. What does the teaching look like when it has been transformed? What does the curriculum look like when teaching has been transformed? We need to understand the initial steps and activities universities and colleges need to undertake to start the process of transforming teaching.
Could analytics and data allow for personalised adaptive learning. An individualised approach that takes learner diversity, performance and behaviour into account. A learning journey that adapts as the learner travels along. What does learning look like when it adapts to meet individual learner needs? We need to understand the initial steps and activities universities and colleges need to undertake to enable them to build a personalised adaptive approach for learning.
Can we create a credible, authentic, valued and accurate assessment model. Could data, AI, digital experiential learning and micro-credentials replace high-stakes tests? Could this help deliver true lifelong learning? Could this improve employability and student outcomes? Do we need to “fix” assessment, or do we need to re-imagine assessment? Understanding the reasoning and processes behind current assessment models, alongside what is the purpose of assessment, will allow us to build new a new paradigm for assessment and assessing learning in new ways.
Could we improve the student experience through creating a seamless experience between the digital and physical estates. Making digital and physical estates work together which are responsive to student journeys and interactions. How do we move from the smart estate to the intelligent estate? Can we describe our estates from a digital perspective? How do we do this, how could we do this?
Using data and analytics for data informed decision making is a journey
It’s not a straight line journey, there are barriers and unseen dangers
We need to start lay down the foundations that will enable a future that includes data informed decision making.
Universities and colleges will need to begin the process of implementing the use of analytics across their organisation. They will need to reflect on the capabilities and skills needed to undertake this, as well as infrastructure, tools and services to make this happen.
However today often much of our data is in silos, which results in siloed decision making
Our data is often locked down in proprietary systems.
We need to plan for the future, we need to break free of our traditional thinking about data.
One way to start thinking differently, is for example, a data lake, a single store of data, one true picture of data. Data is no longer stored in individual systems or services and those systems draw data from the data lake or provide new data for the lake. There is a single source of truth.
How do institutions build their data estate, manage this estate, secure the estate, use the estate. Institutions will need to audit their current data “estate”.
We could build dashboard that allow us to visualise and see what we know. We can then decide if we need to make an intervention, based not just on the data, but what else we know about the situation or the student.
However those insights and data informed decisions, require some level of data literacy and capability. There is no point in having dashboards that nobody uses.
With learning analytics there are some key questions that need to be asked before you even begin to do any kind of analysis. One aspect is consent. You need consent to gather the data, content to process the data and importantly consent to act on that data. You need to do all that within a legal and ethical framework.
There is a real need to understand that gathering data will impact on an individual’s privacy. Even anonymised data is never truely anonymised.
Then there is the validity of the data and assumptions about the data. Often correlation is confused with causation. Just because every student who got a first spent an hour a day on the VLE, doesn’t mean that every student who then spends an hour a day on the VLE will get a first. With more data can you reduce that ambiguity?
Then there are the ethical concerns, are we in danger in moving to a surveillance culture because we think we are making the lives of our students better. As Ian Malcolm once said “… they were so preoccupied with whether they could, they didn’t stop to think if they should.”.
Learning analytics is one part of improving and enhancing the overall student experience.
Data and analytics is not about replacing human decisions, but about supporting and enhancing decision making. Allowing people to be informed from the data.
Data and analytics is also about improving the campus as well. Making better use of our spaces and again improving the student experience.
In March 2020 the campuses of universities across the UK emptied of students, as the coronavirus pandemic hit the country.
This created a shock to the system and cohorts of students and staff found themselves working from home on different tools and systems.
We don’t know what September will being, apart from uncertainty. We do know that we need to plan for the future. What we also know is that there is likely to be less face to face interaction with students. This will make it harder to spot those students who are facing challenges, personal or academic. Does data and analytics have a role to play? We don’t really know.
I then started to summarise my presentation. Using data and analytics for data informed decision making is a journey. One that academics, students and professional staff need to take together.
It takes time for this journey, though in the current landscape, we may not have the time.
But we mustn’t forget that digital is not about the data, nor the technology, digital is about people.
Are there any questions?
3 thoughts on “Learning Analytics and Ethics”