Decision Making in Management (MGMT 4023) Notes Chapter No.5 PDF

UNIVERSITY OF BELIZE FACULTY OF MANAGEMENT & SOCIAL SCIENCE COURSE: Decision Making in Management (MGMT 4023) LECTURER: Dr. Romaldo Isaac Lewis (DBA) Chapter No.5; Developing holistic models with qualitative methods of decision analysis OBJECTIVE After studying this chapter, you should be able to un...

UNIVERSITY OF BELIZE FACULTY OF MANAGEMENT & SOCIAL SCIENCE COURSE: Decision Making in Management (MGMT 4023) LECTURER: Dr. Romaldo Isaac Lewis (DBA) Chapter No.5; Developing holistic models with qualitative methods of decision analysis OBJECTIVE After studying this chapter, you should be able to understand: 1. Understand and apply Irrationality in Management Decision Making Distribution. 5.1) Irrationality in Management Decision Making Distribution The preceding chapters have stressed the role and function of quantitative data in developing models of management decision making. From chapter 1 and RAT models, this was defined as being the rational approach to decision making. Through chapters 2, 3 and 4 however, the reliance upon measurable data to define the decision context has weakened as the manager relies on more statistical relationships of data to anticipate future organizational needs; hence our focus in these subsequent chapters is to consider the ‘irrational’ models of decision making, where we are specifically concerned with the management of human decision values. To address this concern we can review the different approaches to incorporating the human input into the decision making process, at both an individual (Chapter 6) and collective level of analysis (Chapter 7). To achieve this, this chapter and subsequent chapters are focused upon decision making from an individual bias, collective bias, prejudice in decision making and interpreting data and in discussing systems approaches to modeling the decision making process which allow these human values to be considered. We, in particular, focus on mode 1 and mode 2 of Peter Check Land’s Soft Systems Methodologies in this chapter, as a vehicle to bring together both the human values in decision making and the rational use of measured data. Throughout this chapter and the presented context, the reader is encouraged to recognize the value of a contingent approach to management decision making, and as such, we will initially consider the individual human input to decisions. We are aware from the first chapter that humans are subject to ‘bounded rationality’ (Simon, 1951), where we are limited in our process abilities and capacity to identify and evaluate all relevant information necessary to achieve an optimal decision outcome. Instead our decision outcomes are satisficing. Similar early studies on decision making in public organizations identified that in traditional (tall, stable, large) organizations – the practice of incrementalism was observed in decision making and decisions reached by ‘muddling through’ (see for example Lindblom, 1959). This approach to decision making stresses the satisficing view then and that preferred outcomes reflect previous similar historic patterns of decision making and where possible decisions taken are minor variations of past outcomes. Cohen, March and Olson’s (1972) ‘Garbage Can’ model of decision making, is a famous interpretation of organizational decision making where satisficing decisions are the expected outcome for organizations. 1|Page This arises where the necessary decision body and context inputs of staff, resources, timing and context are rarely convergent to support an optimized decision outcome. This is perhaps not a surprise, given that public institutions and organizations tend to be risk averse and where there is a focus upon time for gaining knowledge and commitment (organizational learning), which relies on existing organizational knowledge of the problem and ‘muddling through’ using knowledge of similar problems, so that decisions are simpler combined insights of those solutions. This methodology was improved by assigning a goal to the ‘muddling through’ which is often described as logical incrementalism decision making. This approach to decision making has also been core to political decisions, where it is has been described as neo-functionalism and emerged as a dominant area of study of political negotiation and in particular political integration in the EEC in the 1970s. It is helpful to reconsider the concept of a three phase model of decision making introduced in Chapter 1, to view the qualitative contribution of the human input into decision making (Jennings and Wattam, 1998). This three phased model reflects a more realistic view of how individuals make decisions which are satisficing, rather than optimizing, by presenting this process as an interdependent 3 stage activity of: 1) problem identification – 2) solution development – 3) solution selection; however, the latter two stages are presented as being concurrent in this process. Adopting the three phase model is also helpful as it provides a structure to analyze and understand decision making without proposing a normative framework typical of RAT models. The three phase model therefore allows consideration of the strategic context, goals, ethics and risk in decision making in particular – which will shape the decision context. Clearly, developing a realistic interpretation of decision making in organizations will also require that such models consider organizational hierarchy (and changes to that hierarchy, such as ‘delayering’) and time – where for example, smaller organizations often report limited strategic engagement with decision making (with ‘firefighting’ decision making being their norm). It also means recognizing that there is a large cognitive aspect of strategic decision making for organizations – with a need to consider the subjective construction of individual reality (see for example the discussion of the Cognitive School of Thought of Strategy in Mintzberg, 1998). 5.2) The Monte Carlo Simulation The Monte Carlo simulation, is an approach to decision making which does not rely upon measurable input data (unlike for example those discussions in chapter 2), subjective estimates of likelihoods of given events occurring (such as those presented in chapter 3) nor on assumed patterns in success or failures in the environment of the organization (such as those presented in the poison distribution and queuing theory of chapter 4). Instead, the method uses randomness to assign weightings to potential outcomes and in so doing, allows the manager to make recommended decisions without recourse to the context of those decisions. For example, assume the following investment expectations and likelihoods: Table 5.1 2|Page If we assign random numbers to each ‘cash in ‘possibility in the same ratio as the forecast probability and then repeat for ‘cash out’, we would generate the following: Table 5.2 If we then generate random numbers and assign them to each input and output, e.g. Cash in = 46, Cash Out = 81 and repeat this with many simulations – the more likely combinations of cash flow will occur more often than the less likely combinations. With sufficient simulations (usually