The diagram in Fig. 5.8 is a partially filled-in version of Fig. 5.1 that graphically summarizes this chapter, as well as outlining the next six chapters. We began this chapter by describing the generic characteristics of formative approaches to work analysis. As shown in Fig. 5.8, such frameworks focus on modeling classes of intrinsic work constraints that have a tight and direct connection to classes of design interventions. One particular formative framework, CWA, was introduced. This framework is based on an ecological approach and is deliberately geared to the unique characteristics of complex sociotechnical systems. Because of the open nature of such systems, the philosophy behind CWA is that the primary role of workers is to act as flexible, adaptive problem solvers. Thus, uncovering the kind of information support that workers need to deal with decisions that cannot be anticipated by designers is a primary concern. As shown in Fig. 5.8, CWA is based on five conceptual distinctions. Frequently, the end result of this analysis will be a distributed control architecture that gives workers the responsibility for finishing the design. Two innovative examples of this philosophy were illustrated. Finally, the connections between the characteristics of CWA and the all-important criteria of safety, productivity, and health were briefly described. The evidence and arguments presented in these first five chapters all suggest that CWA has a good chance of meeting the demands of complex sociotechnical systems.
Next, in chapter 6, we introduce a case study that is used in Part III to illustrate the phases of the CWA framework. Chapters 7-11 each describe a layer of behavior-shaping constraint (the first column in Fig. 5.8) in more detail, introduce a modeling tool for that layer of constraint (the second column in Fig. 5.8), and then show how that modeling tool can be applied in the context of a case study (the third column in Fig. 5.8). By describing these layers and case studies in detail, we hope to show you how CWA can directly and productively inform systems design (the final column in Fig. 5.8).