Recently, a friend was describing how he is teaching his pre-schooler to perform addition. The child did not respond to flash cards or verbal word problems. So, my friend used blocks: "you are playing with 3 blocks; here are 2 more blocks; how many do you have now?" Instantly, the child counted to five by pointing successively at each of the blocks. Presto, math by direct interaction with the problem.
Symbology is hard. That's probably why it's the exclusive domain of human beings. We can deal with concepts and symbols which are representative abstractions of tangible reality -- but there's a lot of mental work, as the brain continually swaps back and forth between the symbology and the root reality. This swapping back and forth is our way of validating that our representations are true analogies, because we're all just a little uncomfortable dealing purely in the abstract.
Direct manipulation depends upon simpler brain function. Paradoxically, it requires less mental work/energy/activity despite involving more physical work/energy/activity. That's because there's less (or no) swapping-for-validation required. Solving the problem IS solving the problem.
I've said before that the secret to great business analysis is remembering that all data are representations of underlying physical processes in the real world, and focusing tenaciously on making changes to/in that real world. (more here) Now, I'm pointing out that there is a strong relationship between the degree of abstraction in the description of the problem and the amount of brain power required to grasp the problem.
In addition to requiring less power, direct manipulation is also stickier. Retailers have known for a long time that if you can get a prospect to handle your product, there is a much higher probability he will buy your product. There is similar qualitative evidence in our research which suggests that direct manipulation increases the solver's patience with the problem -- keeps them on-task longer than if they work on the same problem purely in the abstract.
The power of direct manipulation is of vital importance for many people in many roles. As software designers, we need to find ways to allow the users to reach more directly into the problem and manipulate the problem itself (data and linkages as simulacra) -- rather than manipulating levers and abstractions and buttons and scripts. As analysts, we need to find ways to communicate our findings that are tied more closely to simulacra of the physical processes we intend to address. I propose a tweak on Occam's Razor: all things being equal, use the least abstract model possible -- put eyes and hands as directly onto the problem as possible.




