Decision support for rational decision making processes in service systems

It is proposed that the decision makers’ typical requirements for decision support for the rational decision making processes in different work systems within complex service systems are influenced by the characteristics of their associated decision making context and its typical decision situations. It is viewed that decision support for the rational decision making processes in different types of Cynefin framework decision making contexts should focus on providing support for different aspects of the decision making process and can be based on different types decision support technologies. The proposed relationship between different types of decision support technologies and the Cynefin framework decision making contexts is represented in Figure 24.

Figure 24. Decision support for rational decision making processes

The simple and complicated Cynefin framework decision making contexts represent an ordered world and the domain of fact based management, where rational decision making processes are drawing on known facts about the decision making context and its decision situations and build on the exploitation of explicit knowledge that is created through the combination cycle of organizational knowledge creation (Nonaka et al. 2000). It is viewed that effective decision support within these decision making contexts can typically be based on data, representing the facts about the decision making context and its decision situations, and various models that embed the necessary explicit knowledge needed for processing the data into useful information that supports the decision makers’ perception and understanding of the decision making context and its decision situations and provides the basis for their judgment. Typical types of DSSs in the simple and complicated decision making contexts are therefore viewed to include various Data-, Model- and Knowledge-Driven DSSs (Power 2002, p. 13). Furthermore, it may be possible to automate decision making in the highly structured and repetitive decision situations within the simple decision making context with various Decision Management Systems (DMS) (Taylor 2012) that are able to emulate the decision making capability of a human decision maker.

In the complex and chaotic decision making contexts, that represent an unordered world and the domain of pattern based management, the decision making processes are drawing more on the decision makers’

perception and understanding of the decision making context and its decision situations and build on the sharing and exploration of tacit knowledge that is created through the socialization cycle of organizational knowledge creation (Nonaka et al. 2000). It is viewed that effective decision support within these decision

making contexts needs to focus on facilitating sense making and collaboration to allow the decision makers to increase their perception and understanding of the decision making context and its decision situations, and develop their tacit knowledge through sharing and exploration to provide the basis for their judgment.

Typical types of DSSs in the complicated and complex decision making contexts are therefore viewed to include various Communications-Driven and Group DSSs (GDSS) (Power 2002, p. 14) that may be based on different decision support technologies, and other collaboration and communications technologies.

According to French (French et al. 2009, pp. 82-85; French 2013) different types of DSSs and decision support technologies can be further categorized based on the level of decision support provided to the decision makers, ranging from minimal analytical support through acquisition and presentation of data to full judgmental support through evaluation and ranking of alternative courses of action. The proposed levels of decision support are presented in Table 6.

Table 6. Levels of decision support (adapted from French et al. 2009, pp. 83-84; French 2013) Level Decision support provided

Level 0 Acquisition, checking and presentation of data, directly or with minimal analysis, to decision makers; supporting their perception and understanding of the external environment.

Level 1 Analysis and forecasting of the current and future environment; supporting their perception and understanding of the evolution of the external environment.

Level 2 Simulation and analysis of the consequences of alternative courses of action;

determination of their feasibility and quantification of their benefits and

disadvantages; supporting their capability to identify and analyze alternative courses of action.

Level 3 Evaluation and ranking of alternative courses of action in the face of uncertainty by balancing their respective benefits and disadvantages; supporting their capability to reach judgment under conflicting preferences and uncertainty.

It is characteristic for level 0 to level 2 DSSs that they mainly relate to supporting the evolution of decision makers’ perception and understanding of the external environment, including the relevant aspects of the decision making context and its typical decision situations (French 2013), but do not necessarily themselves recognize that the decision makers are facing a decision situation (French et al. 2009, p. 84). It is only at level 3 that DSSs fully relate to the decision makers’ understanding of their preferences and evaluation and ranking of alternative course of action in a decision situation in order to reach a judgment (French 2013).

Different levels of decision support can be provided by different types of DSSs and decision support technologies, but besides the level of decision support provided also the characteristics of the decision making context and its typical decision situations influence the types and characteristics of DSSs and decision support technologies (French et al. 2009, pp. 82-85; French 2013). Different types of DSSs and decision support technologies can therefore be associated with different Cynefin framework decision making contexts, according to the level of support provided (French 2013). The proposed relationship between the different types of DSS and decision support technologies and the Cynefin framework decision making contexts, according to the level of decision support provided, is represented in Figure 25.

Figure 25. Decision support technologies according to level of support and decision making context characteristics (adapted from French 2013)

Decision support at level 0 focuses on the presentation of data, or according to the distinction between data and information, on the presentation of information that is relevant to the decision makers to help them perceive and understand the decision making context and its typical decision situations. At this level various DSSs typically extract the relevant data from databases and other information sources and present them to the decision makers with minimal analysis. Different types of DSSs based on databases and data mining techniques and representational models of data can typically provide level 0 decision support in all the Cynefin framework decision making contexts, although their characteristics and capabilities may differ depending on the characteristics of the decision making context and its typical decision situations. These types of DSSs traditionally include Management Information Systems (MIS) and Executive Information Systems (EIS), which provide graphical and tabular summaries of information, and Geographic Information Systems (GIS) related to the presentation of spatial, temporal and factual data. In the complex decision making context, various Problem Structuring Methods (PSM), or soft Operations Research (OR) methods (Mingers 2011), can help the decision makers in exploring and structuring the decision making context and its decision situations in order to better perceive and understand them. (French et al. 2009, pp. 84-85;

French 2013)

Level 1 decision support provides the decision makers with forecasts that help them perceive and understand how the decision making context and its typical decision situations are likely to evolve in the future. At this level various DSSs typically combine the relevant data with expert knowledge, which may be

either expressed directly or through the use of one or more models. Different types of DSSs based on time series of data and various quantitative forecasting and statistical models can typically provide level 1 decision support in the simple and complicated contexts, and include, for example, economic forecasting systems and market share predictions. In the complex decision making context, various qualitative forecasting methods that rely on expert knowledge, such as the Delphi method (Linstone and Turoff 1975) can help the decision makers gain different perspectives and reach a consensus on the likely evolution of the decision making context and its decision situations. (French et al. 2009, pp. 84-85; French 2013)

Level 2 decision support provides the decision makers predictions about consequences of alternative courses of action in a decision situation, but does not support the process of judgment that the decision makers must undergo to reach a decision. DSSs providing level 2 decision support in the simple and complicated decision making contexts can be typically based on data and various quantitative techniques and models, including simulation models (Law 2014), and different Operations Research (OR) (Taha 2007) and Operations Management (OM) (Slack et al. 2010; Krajewski et al. 2010) techniques and models for standard categories of decision problems, such as linear programming, inventory models and maintenance models. Most of these techniques and models, however, assume too much structure and repeatability to be practical in the complex decision making context, where various qualitative techniques, such as scenario planning in the domain of strategic decisions (Schoemaker 1995) are instead more appropriate to help the decision makers identify potential contingencies and make predictions about the consequences of alternative courses of action. (French et al. 2009, pp. 84-85; French 2013)

Level 3 decision support provides the decision makers support with evaluating the consequences of and ranking alternative courses of action in a decision situation, and helps them to explore their preferences and evolve their judgment to reach a decision. DSSs providing level 3 decision support in the simple decision making context can potentially be based on Artificial Intelligence (AI) techniques (Russell and Norvig 2009) and include, for example, Expert Systems (ES) that emulate the decision making capability of a human expert, but are typically only suited to highly structured and repetitive decision situations that are characteristic for the simple decision making context. In the complicated and complex decision making contexts various Decision Analysis (DA) and Multiple-Criteria Decision Analysis (MCDA) techniques (Clemen 1996), such as, for example, decision trees and influence diagrams at different levels of detail, and the Analytic Hierarchy Process (AHP) (Saaty 1988) can instead help the decision makers in structuring and modeling the decision situation, evaluating and ranking alternative courses of action based on a number of evaluation criteria and weighing together the decision makers’ conflicting preferences and balancing potential benefits and costs with key uncertainties, and evolving their judgment to reach a decision.

(French et al. 2009, pp. 84-85; French 2013)

Different types of DSSs and decision support technologies can provide support for the aspects of different work systems’ real world decision making processes that are related to rational decision making in their characteristically different types of associated decision making contexts and their typical decision situations. It is viewed that they often structure the decision making processes according to the phases of the prescribed intelligence, choice and design model of organizational decision making (Simon 1977, pp. 40-41), and are often focused on finding solutions on individual typical decision situations, or decision problems, existing within the decision making context. It is therefore viewed that individual DSS and decision support technologies cannot by themselves holistically support all the aspects of the different work systems’ real world decision making processes, but providing holistic decision support just for the aspects of their decision making processes related to rational decision making may require combining a

variety of different types of DSS and decision support technologies to address different types of individual typical decision situations, or decision problems, either separately or together.