Developing expertise in management decision-making.

Author:Franklin, Carter L.


The need for effective decision-making and high quality decisions is common to many fields. The requirement is particularly acute in business where the quality of management decision-making directly affects the performance of the enterprise and how well it meets its objectives. Expertise in decision-making is an essential skill for a manager and an understanding of the acquisition of expertise is necessary to its development.

Two elements are necessary to the development of expertise in management decision: one is experience and the other is a model or schema that relates experience to the reality of the organization and its environment. Expertise in other areas, e.g., surgery, playing a musical instrument, driving a racecar, typically involves coordinating mental processes with physical ones. While physical attributes are largely irrelevant to the development and exercise of managerial expertise, the process of developing managerial expertise nevertheless has many elements in common with the general process for developing and applying expertise.

The point of departure for a discussion of expertise in decision-making is the decision process known as the "classical" or "rational" process. Following a brief review of this process, we suggest a revision that addresses several of its shortcomings. We next turn our attention to expertise, particularly as applied to management decisions. This development begins with schema of decision types that reveals where expertise may be applied. We note that some decisions are best made using a (revised) rational process and others using expertise. Finally, we illustrate how the rational process is the basis for developing expertise in management decision making.


The apparent variety and complexity of management decisions motivates the use of a classification scheme to bring order to a discussion of decision-making. Classes include decisions under certainty, risk, and uncertainty; single and multistage decisions; group or individual decisions; decisions with monetary or non-monetary objectives; and so forth.

Independent of classification, the decision-making literature consistently presents a "rational" process for decision analysis generally involving five steps. This process is:

* First: recognize and define the decision situation

* Second: Identify alternatives

* Third: Evaluate alternatives

* Fourth: Select the best alternative

* Finally: Implement the chosen alternative

This particular statement is from a basic text in management (Griffin, 2008) and reflects content from a specialized source on management decision-making (Harrison, 1999).

There are many variations on this basic theme, all with similar characteristics. A popular definition (Bazerman, 2002) is illustrative. In outline, the process is:

  1. Define the problem

  2. Identify the criteria

  3. Weigh the criteria

  4. Generate alternatives

  5. Rate each alternative on each criterion

  6. Compute the optimal decision.

    Other authors have presented additional variations from both the behavioral sciences and the quantitative disciplines. These are notable more for their similarities than their differences. Harrison summarizes them succinctly (1999, p. 37-38):

    There are various views on the process of decision making. Simon (1960) assigns three major elements to the process (1) finding occasions for making a decision; (2) finding possible courses of action; and (3) choosing among courses of action. Witte (1972) advances the notion of decision making as a total process involving discernable and separate activities: (1) information gathering, (2) development of alternatives, (3) evaluation of alternatives, and (4) choices. The process espoused by Schrenk (1969) focuses on three elements: (1) problem recognition, (2) problem diagnosis, and (3) action selection. Janis (1968) envisions a decision-making process with five stages: (1) recognition of a challenge, (2) acceptance of the challenge, (3) meeting the challenge through a choice, (4) committing oneself to the choice, and (5) adherence to the choice. Eilon (1979) advances a comprehensive process composed of eight stages, which begins with information input and culminates in a choice. Mintzberg and his associates (1976) offer an incredibly complex formal structure derived from twenty-five "unstructured" decision-making processes that are then organized into a general model of interrelated strategic decision processes. Fredrikson (1976) proposes a method for organizing noneconomic criteria in a decision-making process that includes four stages: (1) developing a criteria set, (2) posing criteria questions, (3) scaling the responses, and (4) choosing among alternatives. Nutt (1989) advances a decision-making process made up of: (1) exploring possibilities, (2) assessing options, (3) testing assumptions, and (4) learning. The salient point is that the "rational" process does not vary with the characteristics of the decision. Given the diversity of decisions and decision environments, it is an open question whether a sequence of five steps, more or less, will be universally successful in addressing the scope and variety of management decisions. We further observe that sequential processes, such as described here, are characteristic of a mechanistic world view (Ackoff, 1981) and note that the antecedents of "rational" decision analysis are the perfect information assumptions of competitive market theory. This theory is the product of a mechanistic worldview.


    The rational process described above (along with its several variants) is inappropriate as a procedure for making decisions. In summary form, the assumptions that render it ineffective are:

    1. The process is sequential; once a step is complete, it is not revisited.

    2. The process is complete; all alternatives are identified.

    3. The process is deterministic, in the sense that the decision-maker understands the relation between alternatives and outcomes. The relation may involve probabilities (risk) in which case the decision may involve an expected value.

    4. Perfect information about outcomes, and their probabilities, for each alternative is known.

    5. There is unity of purpose. The analyst (decision-maker), if not an individual, is a group of undifferentiated individuals with a unitary purpose

      A superficial comparison of these attributes to the real-world management decision environment reveals the assumptions as unrealistic. A decision process based on them will necessarily be ineffective.

      The nature and variety of management decisions renders the formulation of a completely effective decision procedure problematic. However, an improvement on the existing rational process is not difficult to conceive. There is a prima facie argument that a decision process with the following characteristics will produce superior decisions.

    6. Systemic

    7. Iterative

    8. Adaptive (to new information)

    9. Self-correcting

    10. Active (in seeking new information and innovative approaches)

      The formulation of a revised process incorporating these characteristics involves synthesizing several well known, but seemingly disparate, conceptual structures and research results.


      We propose a revision to the "rational" process that exhibits the above characteristics and consists of seven elements: (1) Understanding the decision opportunity, (2) Formulating the correct goals for the decision, (3) Identifying and involving the correct participants in the decision, (4) Framing the decision correctly (including understanding relevant alternative framings), (5) Generation of alternatives, (6) Choice: selecting an alternative and (7) Learning, to improve future decision-making.

      The star-shaped graphic of Figure 1 presents the relation among the elements in the process. Characteristics of the process are:

    11. Elements are systemically interactive. Each point of the star connects to all other points, indicating the joint and reciprocal influence of all elements in the process.

    12. The process is iterative. Consideration of any element of the decision may prompt a re-evaluation of any other element to accommodate newly developed or revealed information or perspectives.

    13. The process concludes only when all issues related to the elements at the points of the star are resolved. Selection of an alternative occurs only when the process is stable: there is no revision to any element.

    14. The process accommodates a broad range of perspectives and approaches. Multiple framings of the decision situation are required.

    15. The process facilitates the development of expertise, in situations expected to recur. Conducted properly, the process will reveal the crucial relations extant in the decision situation.

    16. The process encourages post hoc analysis as the basis for individual, and organizational, learning. Statements of expected outcome statements are an essential part of formulating alternatives.

      Each of the elements of the decision (points of the star) makes a distinct contribution to the decision. Each element controls, and is controlled by...

To continue reading