How many students do we have #4

Not quite!

In the previous three instalments, we’ve covered the?why,?what?and?how?of designing and building trusted, meaningful populations.??Which is a big step forward to answering the question we came in with. So, we’re done? Not quite, these new assets in our data landscape have zero value if no-one actually engages with them.

Why wouldn’t they? We’ve consulted widely, our methodology is sound, priorities agreed, and use cases identified. What links these activities together is they are all part of a start up/project phase. Implementation and evolution are far more aligned to business as usual.

Transitioning our ‘pops’?to this phase needs us to consider three components.

Trust – Data Quality?

The absolute core of populations is they are built on high quality data. We know where it comes from, what it means, how we count it, what processes we’ve applied to it, etc. Sure, we can caveat that with known issues we need to fix, but we must set and maintain the quality to quash the assumption that somewhere in our institution ‘better datasets exist’

Exception reporting and swift intervention are your two key tools here. Closely followed by documentation to explain how the populations are built, as we’ve shown in the previous two articles.

Skills – Data Literacy

Data consumers use the data in lots of different ways. That’s one of the great benefits of a well crafted population, it encourages a high level of utility. However, with that comes the responsibility to support of ‘citizen data community’?with the guidance, support and – sometimes – training to both make best use of this new asset, but also not to bend it way out of the shape to meet nefarious ends!

There are many ways to approach this; formal training, communities of practice, centres of excellence, 1:1 support, team data champions. What works best depends on how your university works! Bottom line is that this cannot be ignored and needs to be built into the implementation and evolution plan.

Enhancement - Data Accountability

All of these consumers will want slightly (or wildly!) different ‘stuff’?to enhance what’s been initially delivered. Realistically we cannot treat all these requests equally, nor can we service them all. Instead, we must evolve the asset in line with the university’s medium and long term objectives. We cannot go back to?’47 versions of the module evaluation report’.??

Developing and publishing a roadmap (with the rationale for priorities) is a great way of getting in front of a thousand suggested tweaks.?

The first and third of these components fall squarely into the realm of Data Governance. Without a recognised and respected roster of Data Owners and Stewards, a robust and transparent process for logging, triaging and resolving issues and a published shared vocabulary of terms through a business glossary, the end of the project will be the start of your problems!

We’ve seen universities spin up Data Governance capabilities as part of work around populations. This is a workable approach, but our view is the capability needs to be considered quite separate to any kind of project. Rather than go completely off-piste here to discuss why, we’ll come back to it in a new (I suspect series of) articles.

Whatever approach is taken, it should be clear that for both the creation and implementation of populations, visible ‘governance for the good of the university’?across our data is mandatory for sustainable management and use of these new data ‘products’. This must be supported by a move away from an instinctive silo culture. Of course, not all data must be created by the ‘centre’, but where it is, it must be the closest we have to that often requested but never delivered ‘single version of the truth’

For skills, well that’s a whole other article I’d summarise as ‘no bloke I’ve ever met considers himself a bad driver, and that’s statistically troubling’. Removing tongue from cheek, what I’m really saying is uplifting skills around data – especially moving into more analytical capabilities – is not something the university does to people. It’s something it does?with?them as it comes with a whole load of cultural and behavioural stuff that needs careful management. Having written that, let me upgrade our guidance on that to ‘quite a few articles’. Some hard learned lessons to share.

Reading back through the four articles, it’s easy to consider populations as ‘too hard’ or ‘not for us right now’. We’d respectively disagree! These are the very top of the tree in terms for making better decisions with trusted data curated from high quality data sources.??The question is not ‘how much will it cost us to build these?’, rather it is ‘how much will we lose if we don’t have them’?’

So, the next time you are asked ‘tell me again, how many students do we have?’, you can confidently respond with ‘I’m very glad you asked me that question’.

As we always say, we absolutely don’t call this stuff best practice. We’re happy to be challenged on the approach, and very much encourage you to do so. Do drop us a comment so we can keep this discussion going.?