Participants write a short case on a real problem: the problem in a paragraph, the strategy for a meeting to begin solving it, then, on a split page, the conversation they would have or had in the right-hand column (what they would say, the others' replies and their responses, for two or so pages) and, in the left-hand column, the thoughts and feelings they would not communicate. The cases give directly observable data from which theories-in-use and defensive reasoning can be inferred, hold the writer responsible for the meanings produced, and become the material for group analysis, redesign and practice; Argyris reports that almost no one writes a Model II case unaided.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.