The Structure of Concern Project compares many theoretical models from many disciplines to the Adizes PAEI model, arguing that they must all be reflecting the same underlying phenomenon. One concern structure model is described below.
Ned Herrmann was the head of management education at General Electric in the 1970's and 80's. His background was in physics, and he was also active in artistic and cultural circles, which gave him an appreciation of different styles of creativity. During the late 1970's he undertook a reform of GE's management training programs to make them more reflective of individual differences in learning and thinking style preferences (Herrmann, 1989).
Herrmann's initial categories emerged out of a factor analysis of 500 survey forms filled out by subjects participating in his thinking-styles research. The survey forms were revised and administered to a second group of 300 participants, and correlated with the original data. Based on those results, an initial thinking style assessment instrument was created. The research and assessment instruments underwent 19 cycles of revision and refinement over the course of their development, but the revision was never wholesale, and items persist in the contemporary instrument that were composed during the initial revision (Herrmann, 1989).
Herrmann's research was energized by his understanding of the different processing specialties of the left and right hemispheres of the human isocortex, in the wake of Sperry's studies with split brain patients in the 50's and 60's. He was also influenced by Paul MacLean's triune brain theory, and by a general appreciation of the limbic system's role in emotion, cognition and memory. (Herrmann, 1989) These high-level biological subdivisions served as the framework for his theorizing; resulting in a model with four quadrants, one for each major system of the brain as he then understood it:
A: Upper-left (cerebral) hemisphere - Person favors activities involving analysis, logic and fact-finding - left isocortical dominance.
B: Lower-left (limbic) hemisphere - Like type A but more action-oriented, impatient and distrustful of abstract considerations, intensely focused and persistent - left limbic dominance.
C: Lower-right (limbic) hemisphere - A sensitive and receptive people-reader and mood-minder, evaluates issues in terms of their emotional significance - right limbic dominance.
D: Upper-right (cerebral) hemisphere - Wild and original, motivated only by novelty, possibility, variety, oddities and incongruities, can be impersonal and fears structure - right isocortical dominance. ((Herrmann, 1989) p. 79-85)
Rearranging these categories into PAEI order gives us:
P: [B] Lower-left (limbic) hemisphere - Like type A but more action-oriented, impatient and distrustful of abstract considerations.
A: [A] Upper-left (cerebral) hemisphere - Favoring activities involving analysis, logic and fact-finding.
E: [D] Upper-right (cerebral) hemisphere - Wild and original, motivated only by novelty, possibility, variety, oddities and incongruities.
I: [C] Lower-right (limbic) hemisphere - A sensitive and receptive people-reader and mood-minder.
In the early stages of his research, Herrmann took the hemispherical assignations in this model very seriously, trying to tie survey results very closely to anatomical brain regions. He later abandoned this approach, using the anatomical designators as metaphors for the four thinking styles he was measuring. While the instrument is still called the Herrmann Brain Dominance Indicator, the quadrants are now referred to by the letters A, B, C, D, rather than by anatomical regions of the brain.
Scores on the HBDI are presented on a radar diagram. A circle is divided into quadrants A-D, and two diagonal axes are drawn like an "X" through the circle as well. The diagonal axes are graduated to indicate scores from the indicator, and a profile can be drawn on the diagram by connecting the scores together with lines. This will lead to an uneven polygon that "points" in the direction of one's dominant style or styles. If someone was equally dominant in all four quadrants, the polygon would be a perfect square. Interestingly, Herrmann does admit "Quadruple Dominance" as a possible cognitive profile, but one that carries costs of greater internal conflict and longer decision-making processes. A sample radar diagram is reproduced below. Note that the placement of the quadrants around the circle matches the position of the metaphorically associated brain region.
Profiles are also rendered as ranked quadrants, so that the Adizes profile paEI would be rendered "ABCD" = 3-3-1-1
Substantial effort was made to validate the constructs underlying this model and the instrument for assessing it, internally and externally, including six different factor analysis studies. These studies found that four stable and discreet clusters of preferences did exist, that scores from the instrument were valid indicators of these clusters, and that scores permitted valid inferences about a person's preferences and avoidances for each cluster, among other findings.