Reference:Principles of STS Analysis and Design/Work Design

The next set of principles relates to the work content within a work team. They create the conditions for effective problem solving within the group.

Principle 19: Variance Control

If variances cannot be eliminated, they should be controlled as near to their point of origin as possible.

A variance is any disturbance, deviation or unplanned event that can have a negative effect on the throughput of a productive organization (Berniker, 1983). Cherns (1976) calls this principle the "sociotechnical criterion". Variance control is the central task of the work group and the key to its productivity and effectiveness. The longer a variance moves through a production system, the more costly it becomes.

Design for effective variance control requires (1) that variances occur or are observed within the group's boundary; (2) that the work group has the resources to measure and control the variance ( Davis and Wacker, 1982); (3) that the work group has the requisite response variety and information to control the variance; and (4) that the group has the authority and responsibility to take the required actions for control. This logic is so self-evident that we are tempted to take it for granted. Yet, many traditional organizations distribute these elements of effective action among different organizational units.

Principle 20: Multi-Functionalism and Requisite Response Variety

Design work groups with flexible work roles enabling members to increase their competence, master multiple skills and gain the requisite response variety to solve problems.

There is no substitute for competence. The challenges of complex, automated, and expensive technical systems increase the competence requirements of operators. Interdependencies and scale require cooperation and coordination for effective management of such technical systems.

Designed like simple machines, traditional organizations assign each person highly specialized tasks (Cherns, 1976). Jobs are fragments of work; tasks assigned to individuals as careers (Berniker, 1985). Such a work group can be deployed in exactly one way, each worker executing his assigned task.

The uncertainty and complexity of modern technical systems require flexible deployment of diverse competencies. To meet these needs, design work groups so individuals have the opportunity and motivation to master multiple roles and increasing response repertoires. Multi-functionalism argues that workers can master each other's roles and maintenance, administrative, and social system tasks. Groups, with multi-skilled members, would have many possible deployments of their skills exhibiting 'equifinality', the flexibility and adaptability that is an important characteristic of living organisms (Bertalanffy, 1980).

What are the skills needed within a work group? Requisite Response Variety (RRV) is the set of unique worker responses to technical system variation and organizational problems necessary to sustain effective operation (Berniker, 1985, 1986). It is a measure of required work group competence. Work groups need both social and technical skills to function effectively. Social system skills, including communication, listening, teamwork and organizational skills and technical system skills are developed through appropriate training.

Disturbances, defined as variances in STS terminology (Berniker, 1983), often disrupt production systems. The control of these variances requires human intervention. Ashby's (1956) "Law of Requisite Variety," states that, for a given variety of environmental disturbances, only an organism's greater response variety can reduce the variety of outcomes to acceptable limits. It is the response variety within the group that constitutes its capacity to act as 'failure managers' of a technical system.

Principle 21: Information Flow

Information should flow to people who can take effective action. Information used to control performance cannot be used for self-regulation.

Typically, information flows upward to managers who are expected to make decisions, instruct employees, and provide feedback to employees. The need for feedback renders such a flow indirect and less effective (Berniker, 1985). Better to design information flows so that employees have immediate access to information needed for effective action.

Control and steering are conflicting uses of information. Information used by management to measure performance and administer rewards and punishments is being used for control. Such controls lead the controlled to respond to the measurements instead of acting on emergent problems in the production process. They also will manipulate information to protect themselves. Under such conditions, managers become responsible for corrective action and the performance of remote technical systems. The result is less effective variance control.

Alternatively, information routed to action takers enables steering. Individuals and groups can respond directly with responsibility for both corrective actions and results.

Experience suggests that control and steering are mutually exclusive information functions. Zuboff (1988) discusses in great detail the conflicts that arose from attempting to use information for both steering and control at a paper mill. Information may support action or the control of the actions of others. Information used for self-steering focuses on errors and deviations highlighting problems and challenges. We want information used by others to measure our performance to highlight success and conceal problems. The philosophy and values of STS practice argue that effective action take precedence over administrative control.