Right Brained Analytics (Inclusive Analytics)
Ive been thinking about how the human brain thinks, between the left and right brain, and the way we approach analytics. As I see it, the left brain approach, which is called audio-sequential, is to undertake a cognitive process by gather the different pieces of information, and one by one putting it together to build up the scenario / decision. While the right brained approach, which is called visual spatial, is to already have the scenario, and then dissect it apart to its various components that makes the decisions, etc.
As I see it, the left brained approach is the current way that analytics works, which is to reduce the noise, stream line the process, and have as little data as possible to achieve an economy in computing time. However, I wonder that we are missing in, what we are throwing out, ignoring, etc, as part of the discovery process.
So the type of analytics I am talking about is where, you have as much data as possible, and you use which ever tools you have available to search for as many different hierarchies, relationships, connections, patterns, and correlations that you can find, to help explore and explain the dataset, and the problem at hand.
So by this right brained processes, I content that you undertake as many transformations as possible, build up the dataset to be as big as it can be, and that you explore the hell out of it. I believe that transformations, such as reshape, compression, stratification, conversions, recode, to base scales, etc are just as important to the modeling process as the building of the model itself.
for example, if you have a money variable that has a value of $4.80. Its representation in the dataset should be.
1 - $4.80
2 - 480 (as numeric)
3 - 480 (as text)
4 - 450 ( as stratification numeric)
5 - 450-550 ( as stratification text)
6 - 48 (percentile as numeric)
7 - 48 (Percentile as numeric)
By treating a numerical value as a numeric, text, and stratification, you can increase the number of different statistical tests and model types you can apply. By undertaking other transformations and approaches like fuzzy logic, system of systems, ensembles, text mining, context analysis, 3d visualizations, and digital fingerprinting (hash's), you can create a number of different environments to explore the data, where with the left brained approach, you are left with one or two, etc.
I have started up a right brained analytics group for people who are interested in pursuing this idea with me.
IS there a relationship between right brain processing people and EIQ?
You do realize that no competent neuroscientists believe that old left/brain canard any more?
Tony, this is very interesting indeed. While I have some sympathy with your "complaint" about the incidence of left-brain systemised approaches to thinking, the right-brain model you extol is something we have been teaching and encouraging for many years. Of course in some agencies and units, unless managers understand what is possible, then there is sense of safety in pursuing rote approaches to analysis as in your left-brain model. But in the enlightened areas of the intelligence community, where quality of analysis is paramount instead of quick productivity and the illusion of certainty (left-brain model claims), it is the right-brain model that leads the way. As an example, production of European threat assessments on serious and organized crime back in the beginning of time (about 1998) were very much left-brain driven. As time has gone on and, dare I say, teaching and acceptance of strategic thinking and analysis has taken root, then the current crop of annual reports are definitely swinging to right-brain approaches. One prevalent issue is that there's no single approach that will suit all agencies and units. Each jurisdiction has its own stresses of resources and priorities, and this can often lead to unhealthy and ridiculous demands for urgent results without proper thinking to underpin them. In such an environment, small wonder indeed that the left-brain quickie approach is regarded as "best practice" to meet their perceived needs. I find around the world that the larger and more serious are the problems, so is the potential for a more serious approach to right-brain driven analysis to address the multiple complexities and the search for longer-lasting solutions.
I too have long been interested in the Right Brain/Left Brain paradigm, since first introduced to it in the late '80s doing an art course which used "Drawing on the Right Side of the Brain" (Betty Edwards) as a text. I remember the difficulty I had in getting into "right-brain" mode, until I realised that this was my normal problem-solving mode. Further, on reading books by Professor Temple Grandin, who is fully autistic, I was introduced to her classification of autistic modes of problem solving as (paraphrasing): verbal, thinking in pictures; abstract pictures and her own form of total recall and recombination. I tend to solve mathematical problems via abstract pictures/pictures, although I was also able to learn to be quite good at algebra/differential equations, which is a sort of verbal (left-brained) mode for problems that are not too difficult. When I was studying for my PhD, I came across some Russian work on statistical quantum electrodynamics where the algebra seemed fiendishly difficult to follow. However, an American (Richard Feynman) invented another approach through formalised diagrams (Feynman diagrams), which was much easier for some people to follow, and these approaches were later shown to be equivalent. The (Australian?) author, Stephen Few, has written a series of books based on using the superior power of the visual cortex to process information to teach a form of analytics to adults who had been left behind by mathematics, but needed to be able to understand at least basic maths/stats. However, he does not really examine the possibility of a richer structure of envisioning data other than visual/verbal. In the same way that it is claimed (by some) that almost anyone can be taught to sing in tune, I expect that the same can be said for understanding quantitative information, provided that the correct conceptual approach is applied. This has immense implications for both teaching, and for communicating quantitative information to executives, who may well have been selected for traits that are not right-brained. I am attempting to understand this currently at the University of Canberra.