Dr Paul Duignan on Outcomes:  Being obsessed with visualization and outcomes I often find myself reflecting on how things are visualized in other domains. At the moment, I'm dabbling in music theory as part of an attempt to improve my guitar playing. Any evaluators who were subjected to my rendition of the Rolling Sones Sympathy for the Devil at a our recent national evaluation conference dinner will have some sympathy for this endeavor.

Anyway, my initial impression is that musical notation could do with a little optimization in the visualization department. Just one example is in regard to musical rests - small pauses that one makes when playing. These are of different lengths and the symbols that are used to represent them are shown above (source Wikipedia http://en.wikipedia.org/wiki/Rest_(music)).

In our work in visualizing outcomes and designing DoView outcomes software we have worked from a set of principles so as to optimize DoView as a visualization tool. The most basic is, of course, 'keep it simple' and a related one is 'consistency of representation'. This means that if you are representing the same basic thing, you should always try to represent it in a similar way. 

The symbols for the quaver onwards in the sequence above (quavers, semiquavers, demisemiquaver and hemidemisemiquavers (I will not be getting into the language they use in music here!) conforms to this principle of consistency in representation. However the first four (long, breve, semibreve and minim) use an entirely different system and are somewhat arbitrary in the way they vary the use of the underlying symbolic form - the use of a simple block of black. 

No doubt there's a long history to these symbols which a music historian could tell us all about and I don't hold out any hope of reforming musical notation. I'm sure that if I looked on the web I would find many schemes for rationalizing the way music is written.

However, in contrast to a field like music, in emerging fields such as outcomes visualization we still have the luxury of being able to explore a range of visualization possibilities. This means we have the time to ensure that we make our visualizations as simple as we possibly can while still communicating what we need to communicate. 

On a related note (excuse the pun) if you're interested in thinking about the level of complexity that should be represented in outcomes work check out the outcomes theory principle of Representing complexity in outcomes models - 'simple but not simplistic'.

Paul Duignan, PhD. You can follow me on Twitter.com/PaulDuignan or contact me here.

Back to the DoView Blog.