Beware lumpers, splitters and slice globblers when you’re building outcomes models October 15, 2007
Posted by Paul Duignan in : Using the approach, Standards, Outcomes models, DoView , trackback
When you’re drawing outcomes models (program logics, strategy maps, means-ends diagrams, results chains etc), using DoView or other software there’re a few things which will come up on a regular basis. The first of these is the personality difference between lumpers and splitters who are present in the room. Lumpers, obviously want to lump and splitters, well, they just want to split. So if you’re working on an outcomes model which at a high level includes, say, social and economic outcomes, a lumper will want a single outcome which goes ‘complete social and economic whatever’. A splitter, on the other hand, will want to have two separate outcomes - a ’social whatever’ and an ‘economic whatever’ outcome. Neither of them is right or wrong, although in the first instance I often let the splitter have their way because outcomes or steps can always be combined at a later stage if needed.
Splitting is also more in accord with two principles of the outcomes guidelines I use when drawing models - 5. Use ’singular’ not ‘composite’ outcomes and 6. Keep outcomes short. No doubt some psychologists will one day discover that lumping and splitting can be traced back to early developmental experiences and neuropsychologists will find different fMRIs for lumpers and splitters, but in the meantime we will usually have both types working with us on outcomes models.
I find that talking about lumping and splitting often diffuses arguments which can occur between lumpers and splitters. From then on when you think that an outcome needs to be lumped or split you can just call on one of your lumpers or splitters to do it for you.
Another phenomena when building outcomes models is the ’slice globbler’. In DoView each diagram is called a slice (see my blog entry on slicing up the world to build outcomes models). There are some words used in some context which globble up large parts of your outcomes slice into the one box and they should be avoided. They are the type of word that lumpers are likely to like because they can let them efficiently put a lot of material into the one outcome or step.
Examples of slice globbers are: sustainable or good communication. Sometimes you can find that most of the other outcomes or steps in your model suddenly become redundant because you’ve introduced a new outcome or step called something like sustainable growth. Having such an outcome at a high level might be fine, but at a lower level it might just absorb large chunks of you model. Summarizing outcomes may be fine in some contexts, it all depends on the scale at which you’re drawing your model. The more detailed the part of the model you are working on, the more you usually need to stay away from slice globbling outcomes or steps.So use lumpers and splitters to help build good models and keep your eye out for slice globbers in your models.
Paul Duignan (outcomesblog.org)
Comments»
no comments yet - be the first?