Home   About Kurtosis   Courses   Course Calendar   Booking Information   Ideas   Technique   Clients   Contact

The D-Word

What does the word "dashboard" tell us about the way we see data?

Dashboard. How many of you knew that the word “dashboard” originally meant a barrier of wood or leather fixed at the front of a horse-drawn carriage or sleigh to protect the driver from mud or other debris "dashed" (thrown) up by the wheels and horses' hooves?

But that was then. And now of course a dashboard is a control panel placed in front of the driver of an automobile, housing instrumentation and controls for operation of the vehicle.

Anyway, Wikipedia's etymology apart, I think the word “dashboard” is a good word to explore the way we feel about data. After all, it’s a commonly-used analogy in the NHS. And the reason it’s widely-used is because it has resonance and relevance. Dashboards display the data you need in order to steer a safe course.

But let’s see if we can dig a bit deeper.

The first observation is that dashboards are for the here-and-now. People who like the dashboard analogy are people who like looking at data that relates to the present and the immediate future. And not the past. In fact, one thing I’ve observed over the years is that some dashboard-loving managers are so consumed by the hunter-gatherer here-and-now that they express actual disdain for historic data about the past (which they characterise as “rear-view mirror” data, and which gets labelled as irrelevant). So: dashboards aren’t really for seeing what happened yesterday or last week or last month or last year. Dashboards are for measuring what’s happening right now and next.

My second observation follows on from the first. Dashboards pre-suppose that we understand the landscape we’re driving through. Dashboard measurements only really make sense in the context of other knowledge that we've already gained about the workings of the system. The speedometer tells us we’re doing 40 mph and we interpret that information in the knowledge that we’re in a 30 mph zone (se we need to slow down), and the reason why it’s a 30 mph zone is because we’re driving through a built-up area and 30 mph is the safest maximum speed that we should be doing in a built-up area. Similarly, the fuel gauge is telling us we’ve only got quarter of a tank left, and we interpret that (we'll need to stop for fuel before we get back on the motorway because the next leg of the journey is 150 miles long and we’ll run out of petrol unless we put more fuel in).

But in the NHS this important other knowledge about how the system works is often missing. And of course one reason why it’s missing is that we are so obsessed with the here-and-now that we don't have time to do the un-sexy work involved in gaining that knowledge of the system. In fact it’s often the case that the only way we can understand the system is to look closely at much-maligned “rear view mirror” historical data. We need to mimic the football managers who watch video replays over and over again to find out why their team is always conceding goals from corners so that they can understand why, and then take corrective action in the next match. Only when the mechanism is understood (through exposure to historical data) can we then think about designing the dashboard measure which will help us avoid encountering problems in future.

My third observation follows on from this. Most dashboard measurements confirm what your senses should already be telling you. The speedometer, for example: most people can tell from vision and “feel” roughly how fast they are going. Sure, the absolute knowledge that you’re doing 35mph instead of “roughly” 30mph can help you modify your speed to the legal limit, but that ought to be just fine-tuning. The tachometer confirms what most of us can already hear from the noise the engine’s making. The two smaller gauges on my dashboard are—I concede—the bearers of data that you can’t see or feel. The fuel gauge tells you how much fuel is left in the tank, and yes, that is invisible otherwise. And the engine temperature gauge fulfils a similar function. But it’s instructive that these two dials tend to be the smaller of the four dials that you commonly see on a car dashboard. The bigger ones are confirming things rather than telling you new stuff.

And I think this reveals something about our attitudes to dashboards.

I have a theory: too many NHS managers think they don’t need data in order to manage the NHS. They think they can get by just by using the information they get from their senses. If they wander around, seeing, listening, getting the feel of the place, that’s enough.

And sometimes it is. If you’re a charge nurse in Ward 15, why would you need data to tell you that 25 of your 28 beds are full, when you can see it with your own eyes?

But if you need to know about things you can't see, you need data. Data is supposed to fill the gaps left by your senses. This is what data does. It tells us stuff that we can’t sense. It makes the invisible visible. As such, it’s almost as if it’s the opposite of experience or anecdote. And it’s the opposite of most of the stuff you see on dashboards.

So we can make a distinction between, on the one hand, “dashboard” data that confirms what we can sense already, and, on the other hand, historic, “rear-view mirror” data that—used properly—helps us understand the system mechanisms. And recognising this distinction can help us understand why we so often fail to communicate effectively the results of data analysis in the NHS. It’s not because we design our data badly (though we often do design our data badly). It's not because we provide substandard narratives or no narrative at all (though we are guilty of that as well). It’s not even because we publish inaccurate data (again, guilty).

No, it's because we haven’t appreciated the true extent to which NHS managers don't think they need data. They instinctively trust their senses and their lived experience more than they trust data. They prefer dashboards to slow-motion video replays. And we’re trying to sell them videos. And what we’ve got to do is explain to them that unless they understand why they knocked over three pedestrians when they drove through a red light at a Pelican crossing 30 seconds ago, there’s not much point in even having a windscreen never mind a dashboard.

[5 July 2013]

© Kurtosis 2013. All Rights Reserved.