The question mark in the title of my Digifest talk is the key point, because I wonder whether data is the wrong place to start. In our current digital landscape, we’re all too used to hearing ourselves described as “silkworms”, donating “new oil” to “surveillance capitalists”; even the term “data subject” has a dehumanising feel.
All of these reflect a model where our activities are analysed, the results of that analysis go to benefit a corporate entity, and that corporate entity may share some benefit back to us. That may take the form of personalisation, “better service”, or simply access to the service without payment. Here the very best we can hope for is to get back as much benefit as we give away. But in education, I think, the flow of benefit is significantly different: here the activities of students and staff are analysed, and the benefits fed directly back to the individuals concerned. This may be hints to help them study or research better, pointers to relevant campus services, etc. The universities and colleges that perform the analysis benefit indirectly, from happier students, better learning and research outcomes, etc. The benefits to students should be greater than those to the institution.
So, rather than thinking about data, we should probably be thinking about benefits. How can we help students and staff, then which of our data can assist in that. There’s a useful categorisation of benefits in a paper by Aion et al: Learning, Health, Social, Management, Environment and Governance. This was written for intelligent campus data, but seems to cover most of our activities. And a HEPI study suggests that the majority of students would agree to their data being used to provide these kinds of secondary benefits so we are starting from a position of some trust.
So how to think about benefits? The first step is to identify them, and a good place to start that seems to be asking students and staff themselves. What might we do using data to make their lives better. It should definitely be an early warning sign if they don’t like our pet project at this stage. Consulting leads naturally on to transparency: being open about what we are proposing, open to comment and willing to change. Such engagement is radically different from what we are used to from commercial data users and should, itself, build trust.
Then we need to ensure the benefits are actually delivered by the systems, processes and data as we develop them. That means thinking about risks to individuals in the design stage, doing frequent temperature checks as systems operate, and responding quickly to any discomfort. Particular warning signs to look out for include individuals changing behaviour to avoid or obfuscate data collection, any reduction in the voluntary disclosure of information and any increase in the use of external, rather than internal services. Then review, whether the system itself delivered the benefits we expected, and also what we can learn from each design-operate-review cycle.
Three Jisc tools aim to help in different areas of benefit: Learning, Wellbeing and Intelligent Campus. These highlight key areas to consider: Governance and Transparency; Purpose Compatibility, Accountability and Impacts; Choice of Purpose, Choice of Data, and Ethics.
In all cases our thinking should be guided by three key points:
- “Should we do this?”, not “Can we do it?”
- “How does it feel?” (Big Brother isn’t a good look)
- Individuals as part of the team, not part of the product