It's Thursday. The second Thursday in the month. Well, it is for us, but not for our bin men, or apparently the NHS, both of whose schedules appear to have been disrupted by the bank holiday.

Our data processing elves did loaded new data this morning, but it's not looking at all like an ordinary second Thursday... the elves can rest up and get ready for next week, which perhaps will be a bit busier for them.


Here's a camera phone shot of David Spiegelhalter from a talk he gave a few weeks back, where he showed these quotes from Omara O'Neill, a philosopher by trade. I grabbed the shot because although our motivation is to pay our bills, these principles are pretty much what I thought was important when we created Public View.

We're software people, not philosophers, but Omara's thoughts made a lot of sense:

  • There was always a lot of public NHS data about, but most of it isn't remotely accessible, it's tucked away in PDF, Excel and other formats spread across the internet. We thought we could easily fix that by automating the process of collecting it all together into one simple data warehouse.
  • Much NHS data we've dealt with over the years hasn't been particularly intelligible, or at least not to people without Spiegelhalter type educational backgrounds, which excludes almost everyone. The data itself is just a few numbers, and it's not collected by statisticians, so we felt that we ought to be able to do better than presenting data which required customers to hire an "analyst" to understand it.
  • We reckoned that the people who need to see PublicView data are NHS managers, not statisticians or analysts. So we didn't build a "report generator" and expect NHS managers to attend a training course in order to write some SQL or other query to work out what they should be focusing on. We rolled up our programmer sleeves and built those "reports", and provided something usable.
  • On assessiblity, we've always ensured that our output can be checked by end users - the source data is one click away from any metric, and our "cooked" version of that data can be downloaded in a single click too.

Let's hope next Thursday is a busier, or I'll have to find another philosopher who cares about data.