BETTER* DATA

Want to get to the heart of what’s really going on in your organisation? The answers you need could be right there in your data. “Data is our superpower,” says Dr Emily Harburg, Class of 2007 (1996-2000), CEO and co-founder of PairUp, a platform that connects and supports staff across organisations to get the answers and resources they need. “I think of data like knowledge, or intellectual property that you can utilise and tap into. It helps us understand trends. It tells the story.”

But with so much data available – it’s estimated that one person can create around 15.87 terabytes every day – how do we use it to get insights that matter? Will Kirkwood is educational data and technology coordinator at ZIS, and championing the effective use of academic data to inform teaching and learning is a central part of his role. “We have a plethora of data at our fingertips,” he says. “Every lesson at ZIS contains a million potential data points. But one of the challenges – in education, as in society – is trying to sort through that data and find the meaningful information. The stuff that will enable you to make an actual impact on student learning.”

Centralising and consolidating, he says, is key. ZIS has a long history of collecting a significant amount of academic data, but it wasn’t always accessible to everyone who needed it, and, critically, it wasn’t centralised. Today, there is a set process for what data is collected, when it is collected, and how it is shared. Historical data has been cleaned, meaning that all data can be easily visualised on the academic learning dashboard, allowing teachers to access a rich seam of information about their students – information they can use to inform practice.

“For example, say a teacher wants to see how a class did in maths and reading two years ago, they can now just click and see this information,” says Will. “We had that data in the past, but because it was in different little pockets, we couldn’t visualise it.”

Finding an effective way to present data is vital: there’s no point in having it if non-technical people can’t understand it. Last year, ZIS created a similar tool for parents, the Learning Growth Dashboard, which provides visualisations of key metrics for parents of students in at the Secondary Campus. “Parents are keen to know what good looks like,” says Will. “The dashboard shows a student’s results over time, in one location – and it’s the same information that the teacher sees.”

ZIS isn’t alone: organisations from all sectors are increasingly turning to data to gain deeper insight into areas as diverse as HR and communications, says Cornelius Carlsson, Class of 2017 (2013-17). Currently studying for a PhD at the University of Oxford, focusing on quantum computing and AI, he is a member of student-run finance and consulting society CapitOx, which harnesses student energy and insight to solve real-world business problems.

His team recently took on a project with pharmaceutical giant Roche, which asked them to come up with ways of using data to drive their internal communications strategy. “They had undergone a big restructuring and wanted to better understand their employee demographic, such that they could improve the tailoring and targeting of their communications,” explains Cornelius. “For instance, younger people have a delayed circadian rhythm, which means they may tend to wake up later and look at their emails later compared to more senior colleagues. We were given a lot of data that was highly unstructured. And our task was essentially to tell Roche what the data showed.”

There was indeed a lot of data: Roche has more than 100,000 employees, and each employee had around 10 variables – age, length of employment, location and so on. Cornelius and his team used a programming language to process and organise it, splitting the task so that one person examined employee geographies, another organisational level, and so on. “One of the biggest challenges is not necessarily acquiring the data but organising it in a way that can give you the most meaningful insights for your objective, whatever that is.”

The insights that the project was able to reveal are confidential, but Roche was very impressed with the team’s conclusions and will continue working with them. “They now want to get recommendations on how to implement what the data shows, because of course data is only as valuable as the actions you take in response to it,” says Cornelius.

His team, too, discovered the importance of visualising data in an accessible way. “We made some very nice visualisations but received the feedback from the company to stick to bar charts and pie charts, because that’s what people are used to seeing and therefore find quickest to process.”

Finding meaning like this in millions of data points is a key challenge for organisations, says Emily, but it’s also a big opportunity. PairUp, for example, seeks to cut through the noise, simplifying and cleaning up knowledge pools that have become vast and unwieldy. “There is a lot of benefit to having a lot of data and content. But it can become confusing, especially when there are conflicting pieces. How do you help quiet the noise and discern what the truth is?”

“For example, what’s out of date? Who needs more support? What documents have fallen out of use? It’s like having a house with junk everywhere.” Once they’ve identified what is needed, PairUp’s teams work with the organisation to sort, order and organise that knowledge, so it’s easily accessible.

But they have also found that much of an organisation’s valuable knowledge isn’t necessarily on any documents. It’s inside people’s heads. “And it’s vital to better document and share that,” says Emily. “Otherwise, if someone leaves, all that knowledge is lost with them. Our system ensures that knowledge doesn’t get lost – and that people get the right help and the right answers, based on their needs.”

As a social psychology and anthropology undergraduate at Harvard, Emily wasn’t sure where her career would take her. But a lecturer reassured her that the world would always need people who were human-centred to build technology – and that advice resonates with her now more than ever. “AI has allowed us to utilise data so much more powerfully,” she points out.

We need to use data with thoughtfulness – and with humans at the centre

“I think a lot of times in the past there was just data for data’s sake. Now, we have the ease of using that data in more powerful ways. There is great possibility – and also risk, which is why we need to build with thoughtfulness, and with humans at the centre. The more we have this approach to building, the more helpful our tools will be to people.”

This kind of human-centred way of thinking is vital at ZIS, too. Will is keen to point out that gaining insights from data is so much more than crunching test scores. “A test gives you rich results. But you also need to think about how that test fits into the complete picture: a student could just be having a bad day. That’s why we triangulate and cross-reference data points, too.”

Teachers get a read on students; they are aware of how they are feeling. That’s an important data point to bring to the table

And those data points aren’t just numbers, either. ZIS is increasingly adding weight to anecdotal notes and observations about a student’s behaviour, wellness and relationships. “For example, if you know that a student has been struggling with friendships, that’s another lens you can use to look through,” says Will. “Teachers get a read on students; they are aware of how they are feeling. That’s an important data point to bring to the table.”

Every organisation is data-rich, he says, but not every organisation is data-informed. That means having a process, reflecting on that information, actively collecting that information, and then acting on it, as we do at ZIS, to help every student grow.”