A couple of reflections on two aspects of data that came to mind this week.  Firstly tools to manipulate data from that growing range of datasets; and secondly some reflections on data for decision-making.

Data wrangling
Standford University have released an alpha version of a nifty tool to manipulate data, the Data Wrangler  It’s a tool that allows you to take some data, paste it into their tool and then play around with reformating it.  Stanford Data Wrangler screenshot It seems to be a really powerful tool that lets you select data in a cell then uses what it knows about that data to offer suggested transformations.  They demonstrate selecting a US state name and it highlighting all the state names and showing them in a separate column. 

When you play around with the tool you see that when you select something within the data it offers you a series of suggestions for extracting the data into a separate column.  It lets you go through a series of steps to transform the data into something much more usable. Stanford University Data Wrangler screenshot 2It looks like a really good device to help tidy up a dataset into something that can much more easily be used in a spreadsheet or turned into a visualisation.

There’s a really good introductory video here.  At the moment you can only use the tool by going to their website and pasting in some text but the intention is to ultimately release the tool as open source.  One definitely to keep an eye on.

Data and decision making
Just before the summer we had our annual library awayday (a sort of stayawayday as it was on campus this year) and we all split up into teams and spent the day playing a business management game.  So we all ran a series of leisure/fitness centres, made decisions on fees/charges, staff, marketing etc etc and went through of a series of six-monthly business cycles to see the effect of our decisions. 

The key to the game was that we were all presented with a series of sheets of data, covering balance sheets, market intelligence, profit and loss accounts etc.  And each round we had an updated set of data to use in the next round.  The game was interesting and at the end of several business cycles there was a wide variation in how successful each of the businesses had been.  If I remember correctly, anything from a million pound profit to a million pound loss.

That made me think about data.  There’s a business mantra that points to the importance of ‘facts and data’, and yes, I agree that facts and data are important when managing any business or service.  But it’s only half the story.  The data has to be relevant, accurate and meaningful.  And even if the data is accurate it has to be interpreted properly and the correct business decisions made.  So in our game everyone started with the same facts and data, but made different decisions, leading to radically different outcomes.  Which seems to me that the critical thing is making sure that the right decisions are being made on the data.