From Decision Making to Decision Support
Summary and Keywords
In such a complex and well-researched domain as decision support systems (DSS), with a long history of authors making insightful contributions since the 1960’s, it is critical for researchers, especially those less experienced, to have a broad knowledge of the seminal work that has been carried out by prior generations of researchers. This can serve to avoid proposing research questions which have been considered many times before, without having consideration for the answers which have been put forward by previous scholars, thereby reinventing the wheel or “rediscovering” findings about the life of organizations that have been presented long before. The study of human and managerial decision-making is also characterized by considerable depth and seminal research going back to the beginning of the 20th century, across a variety of fields of research including psychology, social psychology, sociology or indeed operations research. Inasmuch as decision-making and decision support are inextricably linked, it is essential for researchers in DSS to be very familiar with both stream of research in their full diversity so they are able to understand both what activity is being supported and how to analyze requirements for developing decision support artefacts. In addition, whilst the area of decision support has sometimes been characterized by technology-based hype, it is critical to recognize that only a clear focus on the thinking and actions of managers can provide decisive directions for research on their decision support needs. In this article, we consider first the characteristics of human cognition, before concentrating on the decision-making needs of managers and the lessons that can be derived for the development of DSS.
Decision-making has been a key human activity since the beginning of time. As an area of research, it has received increasing levels of attention over the last 100 years when Max Weber and John Dewey wrote on the workings of organizations. Decision support—the deliberate attempt to help and augment the thought processes of human decision-makers through various means—has evolved over the last 60 years as researchers (e.g., Herbert Simon) realized the potential that certain technologies had to aid human decision-makers. Furthermore, as these technologies improved, the momentum behind the development of decision support artefacts increased with, in parallel, a realization that decision support is not only about computer systems but is a capability that organizations must develop, leveraging specific expertise, technologies, and data. Thus, the relationship between decision-making and decision support in organization settings today is often reciprocal; hence, we need to consider decision-making and decision support at the same time.
The ambition of this article is to deliver a structured set of observations about decision-making as a typically human activity (Damasio, 1994; Damasio, Damasio, & Christen, 1996) and to then consider what these observations mean for decision support—the provision of information and the formalization of models for diagnosis, predictions, and prescriptions to guide or alter decision-making (Silver, 1991). Decision-making is quite a prevalent activity in human societies, and it is specifically a human undertaking—insofar as (1) other mammals are mostly instinctive creatures and do not make decisions in the sense that we assigned to human decision-making as an expression of individual free will1 (Pomerol & Adam, 2008) and (2) the decision-making activity commonly associated with organizations, countries, or other entities is an abstraction that conventionally assigns decisions made by human agents to the entities to which they belong (Langley, Mintzberg, Pitcher, Posada, & Saint-Macary, 1995).
The fact that some decisions are collective to the extent that it can be difficult to identify who “made the decision” does not change the fact that people make decisions, not organizations or countries. For example, France did not decide to become a nuclear power. Rather, scientific progress in French research labs enabled political leaders to consider the proposition at a given point in time and to allocate funding and issue instructions to make it a reality. One man, Charles de Gaulle (president of France between 1958 and 1969) is often represented as having made the final call—at any rate, he shouldered the responsibility for making the decision based on his creation of the Commissariat pour l’Energie Atomique (CEA) in 1945.2 This personification of decisions is important for scenarios where the outcome of decision processes needs to be attributable to an individual, as is often the case where legacy or litigation matter or to record key decisions in history books.
While decision-making is very common in a variety of contexts, this article focuses on the type of decision-making that managers in organizations undertake. Other contexts, such as driving a car or making life choices, are also interesting examples of human decision-making, but they are not the primary focus here, which is centered on managerial decision-making as it occurs in organizations. This restriction is not very limiting because in research that focuses on decision-making, the contexts in which decision-making processes take place are critical in analyzing phenomena and their consequences.
Some decision-making scenarios have received more attention than others. For instance, Daniel Kahneman has studied decision-makers faced with risk and those facing win or loss situations; Gary Klein has focused on emergency responders such as doctors, nurses, and firemen; Herbert Simon has concentrated on administrators and managers, and Mintzberg studied strategic planning decisions. While generalizations are interesting and important in research on decision-making, observations that are specific to certain typical contexts and specific genres of decisions are critical in understanding how decision-makers can be supported by information technologies in the specific context in which they operate.
Thus, this article provides a broad overview of two important topics for modern managers: decision-making and decision support. The next section examines why the study of decision-making is specific and important across a variety of disciplines, as well as how it can be used by researchers of organizations to set precise boundaries for their studies. The discussion then considers the complexity of decision-making as a managerial activity and examines what we know about the decision-making process as it unfolds in organizations and the models that have been proposed by researchers over the last 100 years to capture its essence. Specific difficulties that arise when attempting to model decision-making are examined, including a consideration of the role of expertise in the problem domain of the decision situation. The discussion pays particular attention to the emergence of cognitive biases in human decision-making and the impact they can have on the outcome of decisions. The other side of the bias coin—the existence of heuristics to accelerate judgment in decision-making—is also considered, as is the role of intuition and the difficulty in understanding its impact on and contribution to decisions.
The discussion then moves toward the development of decision support artefacts, and the concept of decision support systems (DSS), as proposed by Gorry and Scott Morton in 1971. Tools and techniques for analyzing the requirement for decision support systems are discussed. The final section proposes key insights from the article and teases out their impact on DSS research and practice.
Why Study Decision-Making?
Many researchers have been attracted to the study of decision-making for a variety of reasons. Daniel Kahneman, one of the pioneering researchers in the field—and a Nobel Prize in Economics winner for his research on Prospect Theory (2002)—explained his very personal attraction for the area of human decision-making as follows:
It must have been late 1941 or early 1942. Jews were required to wear the Star of David and to obey a 6 p.m. curfew. I had gone to play with a Christian friend and had stayed too late. I turned my brown sweater inside out to walk the few blocks home. As I was walking down an empty street, I saw a German soldier approaching. He was wearing the black uniform that I had been told to fear more than others—the one worn by specially recruited SS soldiers. As I came closer to him, trying to walk fast, I noticed that he was looking at me intently. Then he beckoned me over, picked me up, and hugged me. I was terrified that he would notice the star inside my sweater. He was speaking to me with great emotion, in German. When he put me down, he opened his wallet, showed me a picture of a boy, and gave me some money. I went home more certain than ever that my mother was right: people were endlessly complicated and interesting.
(Kahneman, 2003a, p. 417)
While this quote reports observations of a very personal and historically charged nature, it is a very good illustration of the richness and importance of human decision-making as a key area of research. It also illustrates the complexity of researching it and the multiple perspectives that can be adopted in reference to it, as well as the many disciplines that have explored it, including economics, management, and psychology. The history of humankind can be explained, in a large part, in terms of making decisions, from the micro to the macro level, from the trivial to the most fundamental, from the individual to the societal. Thus, the methods and concepts that come under the heading of decision-making research can be applied equally to the fine analysis of the aerial disaster that unfolded in the German skies on July 1, 2002, when two aircraft collided,3 to the political analysis of the failed Bay of Pigs invasion on April 12, 1961 (see Janis, 1972, and Janis & Mann, 1977, for a superb analysis of this event), as well as the largely historical investigation of how Hannibal evaded attempts to prevent him from reaching northern Italy by deciding to follow a northern route and cross the Alps at one of its highest passes, suffering huge losses in the process, but crucially outmaneuvering his enemies’ armies even before engaging with them. These examples have corresponding scenarios in the management domain, with the Amazon case study explaining Steve Bezos’s decision to create his new venture based on the growth rates he had read about the Internet or the multiple case studies of Apple Corporation, the analysis of the collapse of Pan American airways, or the seminal American Airline’s SABRE case study.
Another leading researcher in the area of decision-making as it relates to management, Herbert Simon (1997), also winner of the Nobel Prize in Economics in 1978, first conceptualized the need for a science of management which would hinge on a study of the management of organization based on analyzing managers’ decisions and the actions that follow. In this he considered the manager as a decision-maker (Simon, 1977) and described himself as somebody “who has devoted his scientific career to understanding human choice” (Simon, 1991, p. xvii). In his book with his career-long colleague James March, he stated: “The central unifying construct of the present book is not hierarchy but decision-making, and the flow of information within organizations that instructs, informs, and supports decision-making processes” (March & Simon, 1993, p. 3).
Another fundamental contribution Simon made was to understand very early that new information technologies would radically and forever change the entire management landscape. In this he predicted the development of large-scale decision support tools based on computing technology such that “The computer and the new decision-making techniques associated with it are bringing changes to white-collar, executive and professional work as momentous as those that the introduction of machinery has brought to manual jobs” (Simon, 1977, p. x).
This prediction was an indirect reference to Taylor’s (1911) earlier work on his time and motion studies and their application to the organization and optimization of manual labor. Simon’s ambition was to understand organizations and their management as an aggregate of human choices, based, at least in part, on the behavior of actors in organizations.
Simon had started his career working in a local government authority where he began his observation of managerial behavior. He was particularly interested in how decision-makers acquire the necessary information, how they perform their calculations, or, more importantly still, whether they are capable of correctly evaluating the consequences of their decisions, based on multiple contributory events, as postulated by utility theory (Simon, 1997).
Decision-Based Schemes for Setting Boundaries in Research in Management and Decision Support
Langley et al. (1995) have noted both the great advantages and the difficulties in using units of analysis related to decision-making, for instance conducting research focusing on specific instances of decisions. In particular, they highlight the ability of researchers to identify clear boundaries for the phenomenon being researched by reference to who was involved and what they did. The issue of boundaries in research is quite an important one as it provides firm directions to set the scope of investigation. In the real world, a fortiori the world as seen by social scientists, many common research objects have no ontologically independent boundaries (i.e., there are no specific unambiguous tangible demarcation lines around objects of study), and researchers need to propose clear definitions and rationales to explain how they have decided to characterize the objects of their studies. Even such entities as organizations are not easily distinguished from their environments. Starbuck (1976) has commented that:
Assuming that organizations can be sharply distinguished from their environments distorts reality by compressing into one dichotomy a melange of continuously varying phenomena. . . . One can sometimes say “Now I am inside” or “Now I am outside,” but one can never confidently say, “This is the boundary.” (pp. 1069–1071)
When setting clear boundaries as they operationalize their research questions, data collection, and analysis methods, researchers may be guided by a decision-focused scheme: a decision taken by managers in organizations or in broader society, or case studies of decision support provided by systems or by individuals. Such schemes allow them to actively include actors and/or entities that have played an active or supporting role in a given decision or stream of decisions over time or segregate other individuals and entities deemed not relevant. Laumann, Marsden, and Prensky (1983) concluded that researchers must consciously adopt strategies to actively set boundaries that reflect the goals of their research targets. Knoke (1994, pp. 280–281) identified four main strategies for boundary setting:
• Positional method where people and organizations occupying the key roles in the entity studied are interviewed
• Decisional methods where actors involved in making or influencing a given decision constitute the core element of the sample
• Reputational methods where actors who are widely believed by knowledgeable observers to occupy key positions are targeted
• Relational methods where the researcher interviews the key people named by other key people and the network is uncovered in a guided snowball fashion.
The interest in following a decisional method is evident in terms of deciding where the target population resides, the pertinent timeframe, and what data to collect. In this manner, Kadushin (1978) established that an elite group of about 50 people runs the city of Detroit.
However, Langley et al. (1995) have also noted that many researchers of decision-making processes have fallen into a number of traps, such as the temptation to assign to “garbage can”–type models (borrowing terminology from Cohen, March, & Olsen, 1972) anything that could not be explained rationally or to reduce decision-making to the moment of choice, instead of embracing the complexity and evolutionary nature of many decision problems, from the abstract representation that managers may have in their minds to the concrete and tangible actions that end up being implemented by organizations (Humphreys, 1989). Studying these processes as they unfold over time often requires longitudinal approaches—or at least good access to managers (as enjoyed by such seminal researchers as Henry Mintzberg or Sune Carlson) and the freedom to ask them many questions.
Beyond managerial decision-making, it is also our understanding of decision support that is at the core of this questioning. By decision support, we mean providing some degree of help to managers, directly or indirectly, with or without the use of technology and artefacts, as they identify problems in their organizations or their environments and seek to solve them by taking some action or deciding to do nothing (Silver, 1991).
Complexity and Decision Making—Finding One’s Way in the Jungle of Previous Research
Understanding and modeling human decision-making have been attempted by three generations of researchers since Fayol (1916) and Barnard (1938), all the way to the current day. The motivation behind attempting to do so stems in part from the need to underpin the education of would-be managers as well as from the need to find ways to study what is a very complex and diverse organizational activity. Numerous papers published in the Harvard Business Review throughout the 1970s and early 1980s reflect this ambition. Mintzberg (1975) and Kotter (1984) described the nature of managers’ jobs, and Mintzberg (1976) explored the cognitive efforts made by managers during planning and execution activities. Isenberg (1984) reflected on the networking underpinning the work of good managers, while Harriman (1974) discussed their communication needs.
In other forums, their contemporaries Keegan (1974) and Kefalas (1973) had explored the environment scanning activities of managers, while Pounds (1969) had specifically discussed problem finding as a key managerial activity. Earlier, Carlson (1951)4 had pioneered this type of “real world” research on the work of managers in his study of nine CEOs of Sweden’s largest companies.
This rich and varied research was absolutely essential to understanding the nature of management and furthermore to teach about it or research how to support it, with technology or otherwise. Bates and Sykes (1962) noted that, at the time this vein of research emerged, “a bad feature of management education . . . [had been] the over-frequent attempt to classify executive and management occupations in readily recognisable activities without much regard for the facts” (p. 44).
Schwenk (1984) outlined two sources of research pertaining to human decision-making: observations that arose from experimental research conducted in laboratories and observations from field studies that observed real-life processes. There is a degree to which both sources of observation converge, but equally, there is a doubt whether some of the experiments conducted in lab settings are really representative of actual managerial decision-making (Ungson, Braunstein, & Hall, 1981). If MBA students are asked to role-play in a simulation pertaining to new product development, their behavior is interesting but necessarily limited by their limited understanding of the scenario. Thus, subjects are often not experienced in the way top managers would be, and to use a common expression: “they have no skin in the game” in the way real managers would have. Similarly, trial-by-jury simulations deliver much harsher outcomes than real trials do because in a simulation, participants are not sentencing real people and apply sanctions rigidly.
Indeed, there is a difficulty inherent in learning from field studies of organizational decision-making where the number of variables at play make observations and developing inferences from these observations very difficult (Schwenk, 1984). Real decision-making, especially at the strategic level, also makes it very difficult to analyze outcomes and to judge whether managerial assumptions have been validated. The possibility that these outcomes are causally independent from managerial actions or result from chance cannot be discounted, and the accounts provided by participants may not be reliable (Kinder & Weiss, 1978). At any rate, the conclusions reached by researchers may be doubtful and unverifiable (Nutt, 1984). The most reliable manner to learn about managerial decision-making processes entails relying on both sources and to rigorously triangulate observations of actual and laboratory processes to propose adequate representations of decision-making processes in organizations.
This approach was broadly followed in the late 1960s and 1970s and became the basis for much of DSS research, centered on the need to understand management so as to be better able to support it. This research gave a rich account of the sheer complexity of the scenarios that management decision-making entails, as well as their variety. There was also an acknowledgment that DSS did not always have an impact. There is a degree to which, therefore, decision-making as a process has remained somewhat elusive, but fantastic contributions have been made to its understanding (e.g., by researchers from Herbert Simon, Daniel Kahneman, and Henry Mintzberg, with their colleagues and collaborators, James March, Amos Tversky, and Dan Lovallo). In fact, there have been so many interesting attempts to characterize decision-making that this presents a critical barrier for students of decision-making—this segment of the literature is at first impenetrable or at least very difficult to apprehend in its variety such that the risk of grabbing one piece without understanding how it fits in the overall landscape is very significant. When the literature on decision support systems, knowledge management, and their more recent offshoots analytics and big data is added, it is not excessive to describe literature on decision-making and decision support as a jungle.
The Decision-Making Process
A simple characterization of decision-making was proposed by Pounds (1969), who proposed that making decisions was about implementing a plan for reducing a gap between the current state of the world and a more desirable, different state of the world—increasing the company’s market share or extending the distribution of a firm’s products and services to a national or international level. Of course, this of itself is not as simple as it seems insofar as it entails assumptions about the real world and how readily observable and intelligible it is for observers that may not always be true. When knowledge of the future is involved, it is possible to make assumptions and perhaps attach probabilities to them, but little more.
One way to study decision-making and teach about it has relied on characterizing the decision-making processes of managers based on a number of typical phases (e.g., Mintzberg, Raisinghani, & Theoret, 1976; Schwenk, 1984; Simon, 1977). They vary in the degree to which they propose abstracted phases or seek to describe the observable reality of actual decision-making processes in their complexity. For the purpose of our discussion, let’s propose a simplified view of what managerial decision-making entails. First, the decision-maker must be able to carry out an evaluation of the current state of the world—this is the diagnosis part of the process. Second, the decision-maker must be able to express a set of preferences—either his or her own or those of a group or organization—this is the preferences part of the process. For instance, managers might decide to offer cheaper products, or they might prefer to be technology leaders but sell products at a premium price.
These two stages may be performed simultaneously, or at different times, but they may often be performed by different levels of management. Different managers may be more or less reactive to what happens in the environment, from spending more time on developing a vision of the future to spending more time evaluating what is happening in their industry in order to react to it. Thus, diagnosing problems and setting goals for the organizations will occur in view of each other, although not always in a tightly coupled manner. The causal linkage between the two may also evolve in dynamic fashion where intended strategies become realized strategies when emergent elements are added by the occurrence of new events (Mintzberg & Waters, 1985).
Third, the decision-maker must be able to design a path to the more desirable state of the world. This is the look ahead part of the process, which is the anteroom of the actual choice, the moment when a commitment is made to a specific decision—rollout of a strategy for distribution, acquisition of a small competitor to rebrand their products, etc. All three phases are problematic, and we explore why in the following sections. Figure 1 shows a model of individual decision-making proposed by Pomerol (1997) that will be examined in more detail in this section.
The evaluation of the current state of the world is not a simple matter inasmuch as not all information about it is available to human decision-makers. While more and more information is available to decision-makers, with the availability of the Internet of things and the growing array of sensors connected to it—delivering more information in a day than can be processed in a lifetime, it still cannot be said that human decision-makers are “fully aware” of what is happening around them.
Even where information is available and information overload can be addressed, there is still a large gap left in terms of properly analyzing what is and what is not, let alone how the situation occurred. Some diagnoses are very complex or elusive—for example, gauging customer opinions and preferences, especially future preferences. Generally, generating causal explanations for events is fraught with difficulty, and it has been recognized as a key managerial skill in constructing competitive advantage (Reid & Defillippi, 1990) where organizations develop or acquire competences that are causally ambiguous and make imitation of their strategies very difficult for their competitors (Lippman & Rumelt, 1982).
This stage in the decision process is also characterized by some well-known cognitive biases, which are documented in a number of papers and books published by Tversky and Kahneman. Although there is some evidence that these biases—easy enough to observe in experimental conditions—are sometimes difficult to observe in the reality of actual decision-making processes, it is still useful to know about them from a theoretical viewpoint. We return to this discussion in later sections to discuss the impact of these biases on managerial decision-making.
Students of management and organizations often make an assumption that, unlike (their own) everyday decision-making, managerial decision-making is essentially rational and inherently evidenced based. In particular, they assume that managers and their organizations know what they are trying to achieve and rationally decide how to get there. In practice, this is incorrect for a number of very good reasons that are explored later in this article. As it pertains to preferences and the role they play in managerial and organizational decision-making, it is certainly the case that this assumption of rationality is incorrect: managers spend much time debating which direction to steer their organizations toward, and not only are the arguments they use in these debates sometimes based on subjective rationales and opinions rather than evidence, they are sometimes not even the real reasons. In other words, hidden agendas can make the search for organizational consensus very hard to track down for researchers studying organizations (Mintzberg & Waters, 1985).
Brunsson (1989) has gone as far as proposing a vision of organizations that is radically different from the normative view. He argues that organizations are composed of two kinds of individuals, the leaders—concerned with the legitimacy of the organization—and the led—concerned with the production of goods and services. The leaders are immersed in political games aimed at ensuring the support of the environment of the organization and at creating legitimate structures, ideologies, and processes, while the led attempt to improve the efficiency of the firm. Thus, these two groups constitute two suborganizations which interact as the leaders translate their ideas into decisions to be implemented by the led. In such a framework, the leaders use a wide array of control mechanisms to influence all the tasks carried out in the organization, and decision-making processes are among these control mechanisms ensuring the coupling and decoupling of politics and action.
As noted by Brunsson (1989), decisions are not a reflection of a unique set of ideas that govern the organization of action but the reflection of the dominant leaders who can impose their views on other leaders and on the led. As such, decisions are the transition between ideas and action. The decision-making process becomes a twofold process where leaders try to impose their values on the organization and struggle to verify whether their decisions are implemented in the way they intended. Often preferences end up being conflicting—it is not possible to be both leader on price and on quality or to be generous with staff and ultra-competitive on price. Hickson, Butler, Cray, Mallory, and Wilson’s (1985) model of strategic decision-making is broadly consistent with this view of organizations. It specifies that the first rule of the “game” is power and its ultimate holders, the legal and financial owners of the organization. These are the people who “establish and maintain the organization as a setting in which decision-making can take place . . . they set implicit or explicit limits on what may be considered and what may be decided” (Hickson et al., 1985, p. 120).
Thus, organizations may be regarded as political arenas where individual managers are involved in an intense game of winning and losing power rather than as strictly rational organizations involving rational actors making rational decisions. As this vision makes studying organizations and their decisions even more complex, it reinforces the need for and usefulness of normative models that provide simpler (even though inherenty incomplete) explanations for what happens in the real world (LeMoigne, 1995).
Look Ahead and Decisional Outcomes
The next activity of the decision-making process involves designing suitable courses of action leading to desirable states of the world, and this is quite problematic in a world that is characterized by so much uncertainty. The notion of evidence-based decision-making relies on the availability of evidence and the existence of agreement on desirable outcomes. Yet, in the real world, future events have substantial uncertainty attached to them; many business contexts and cause-and-effect relationships—the causal relationships that exist between actions and their consequences—are at best incompletely understood and therefore assumed rather than known with certainty.
Le Moigne (1990) argued that attempts to model complex situations in a way that refuses to account for their complexity are likely to move decision-makers further away from rather than closer to robust solutions (solutions that are resilient in the face of changes in the environment and generally viable in the long term). This is empirically illustrated by examples from the real world, where entrenched opposite positions are allowed to polarize debates about complex societal/political issues through the acceptance of trivial arguments and a blind refusal to admit other viewpoints, leading to stalemate and long-term conflict. There is a long list of human conflicts and debates on a global scale that illustrates this basic principle. No example is provided here in a bid to avoid polemic arguments.
Thus, complexity must be embraced, even where it means the acceptance of incomplete models and the persistence of uncertainty, requiring managers to use judgment and develop a vision of the future that they are willing to commit to. Thompson and Tuden (1959) proposed a 2 × 2 matrix which explores the relationship between uncertainty and decision-making (Figure 2). They distinguish between two types of uncertainty, one which focuses on preferences (see previous section) and one which concentrates on the cause-and-effect relationships between variables in the environment.
Figure 2 illustrates the genre of decision-making that should be promoted in organizations in the different areas of the matrix. In contrast, Thompson and Tuden (1959) argue that organizations may also develop decision-making pathologies when assumptions about the world are assumed to be true instead of being recognized as mere assumptions: decisions that require judgment are reduced to simple computational decisions, ignoring the possibilities of deviations from the expected scenario and making proposed solutions that are less resilient to or even blind to changes in the environment. Incorrect assumptions about consensus or conflict that may exist in organizations or in society will also result in arbitrarily reducing the need to consider compromise, each side of the debate focusing narrowly on their own arguments and proposing solutions that are only acceptable to themselves. Both scenarios indicate poor management and promise solutions that lack resilience to evolving circumstances and shifting political alliances.
The end of this phase of the decision-making process is the moment of choice where resources are committed to the implementation of the selected alternative. While this moment should not blind researchers in decision-making with respect to the underlying longitudinal processes that led to the choice, it is still a unique and important moment in the decision process. In particular, it signals the boundary between a planning cycle and an execution cycle in management. Typically, this is also a where different categories of managers may be called into action, as top managers pass on agreed-upon solutions to their subordinates for implementation (Humphreys, 1989; Mintzberg, 1973). At any (or all) points in the process, decision support experts may be called upon in support of any of the phases outlined in this section, from diagnosis, to conflict resolution, to choice, and, eventually, to implementation.
So far we have illustrated with broad strokes the complexity and richness of decision-making as a human endeavor. The next section explores the formal normative models that have been proposed to underpin the investigation of decision-making by researchers.
When attempting to describe and capture the fullness of a complex phenomenon, it is useful, notwithstanding the theoretical impossibility of the task, to propose models to facilitate two key activities: research into the phenomenon in question and teaching about the phenomenon. Normative models, even though they are not expected to be directly observable in reality, do serve both of these purposes, their usefulness being judged by the degree to which they capture the essence of phenomena and make it more easily apprehendable by non-experts. This gives rise to a clear trade-off where the complexity of the normative model may increase its ability to account for a very diverse reality, but at the expense of its apprehendability. This is sometimes captured in terms of the parsimony principle.
Karl Weick (1979) has described the Bonini paradox (by reference to Charles Bonini and his work on simulations of decision systems, published in 1963) to characterize the problems inherent in proposing models that are so complex in and of themselves that it is no easier to understand them than to understand raw observations in the real world. His description is based on observations made by William Starbuck (who studied for his PhD at the same time as Bonini at the Carnegie Institute of Technology in the early 1960s), which he summarized in 2004 as follows:
As a model grows more realistic, it also becomes just as difficult to understand as the real-world processes it represents. A researcher builds a model to gain or demonstrate understanding of a causal process, and the researcher states this model as a simulation to allow complex and realistic assumptions. The resulting program generates outputs that resemble those observed in the modelled situation. But the model itself is very complex, and the interdependences between subroutines are obscure, so the model is no easier to understand than the original causal process. (pp. 1237–1238)
One fundamental element in modeling human decision is the extent to which models based on an assumption of rationality make sense or are a good fit for available observations. Early studies, mostly from researchers trained as economists, are all based on this assumption of rationality—human agents seek to maximize their utility, and this function can be approximated such that rational outcomes can be computed (Savage, 1972). While exploring this is beyond the scope of this article, it is useful to remember that such utility can be computed based on expectancy of gain or loss multiplied by the estimation of the probability of occurrence of each outcome. This has remained a central tenet in teaching about decision-making for many years.
The theoretical representation of such a model is compelling and easily implementable from a computing viewpoint, as explained in Savage (1972). On the other hand, there are severe limitations to this kind of modeling, related to the extent to which discrete outcomes can be identified and the probabilities attached to them can be estimated. It is the essence of these limitations that shapes Thompson and Tuden’s model (see Figure 2) with regard to decision and uncertainty: in an uncertain world, the probability of specific outcomes is never known with any precision, neither are all outcomes known in advance. Modeling consumer responses to new products, for instance, does not lend itself to rational modeling inasmuch as they are driven mostly by subjective judgment, which rests on subjective criteria. There are in fact countless examples of business situations where a rationality-based view of human decision-making will not be useful, other than in a pedagogical sense, to serve to illustrate the limitations of a rationalist perspective on management. Drucker (2002) proposes that intuition and vision must also be used, especially in situations where some innovation is required. He singles out the case of the Ford Edsel to illustrate that even the most rigorous research into customer perceptions and customer needs can never guarantee commercial success.
Notwithstanding these limitations, there are specific circumstances where rational optimization models can support decision-making, and the discipline of Operations Research has explored them. Economic Order Quantity (EOQ) and Materials Requirement Planning (MRP) are illustrations of models that have been applied universally in certain activities and certain industries. An entire segment of the software industry—focused on DSS—has emerged to commercialize applications that embody such models, and, by contagion, these business logics have snowballed from customers to suppliers up and down entire supply chains within certain industries. On the other hand, MRP and its derivations are also a great example of the boundaries of the rationalist perspective—and one worth exploring in detail—because in many organizations, while it offers the possibility of greatly improving the ordering cycle and the scheduling of production (in the case of Manufacturing Resource Planning, or MRPII), it also exposes the dependency of such models on the existence of accurate forecasts and their frailty in the face of the vagaries of the real world (Carton & Adam, 2005, 2008), especially where Enterprise Resource Planning (ERP) applications are concerned.
In recognition of such observations about the uncertainty inherent in the real world, and particularly of its future states, Simon (1955) proposed the concept of limited rationality, which he later relabeled bounded rationality—the term that is currently used to refer to his work on management decision-making and rationality. It is for this fundamental contribution that he received the Nobel Prize in Economics in 1978. Simon (1955) specifically described the underpinnings of bounded rationality in terms of:
• The difficulty in correctly assigning probabilities to events or even simply to enumerate all possible events and their combinations
• The lack of rationality in the preferences of organizational decision-makers
• The spread over time and space of many decisions in organizations and the existence of chains in which, over time, subdecisions taken at different times and levels, using a variety of justifications and criteria, become imbedded into each other, multiplying in their effects and outcomes
• The confusion at times between facts, values, and objectives, which contradicts the theory of rationality where these are assumed to be readily distinguishable
• The impact of available information on decision-making and the inherent unevenness of managers’ access to it, which drastically affects their choices
• The impact of the many cognitive biases affecting human judgment, as well as the limited attention and cognitive limitations of human decision-makers
Simon also proposed that managerial decisions and actions deliver not optimal but “satisfycing” decisions (Simon, 1955, 1956) and that, given the above limitations, the decision process stops when decision-makers reach an acceptable solution that satisfies them within what appears to them to be the most probable hypothesis. He proposed an analogy with the familiar expression “finding a needle in a haystack” and likened satisficing with finding any needle that will do the job, as against finding the sharpest one (Simon, 1987).
Bounded rationality is a key contribution to our understanding of human, and specifically managerial, decision-making, but Simon is also recognized for having proposed a normative model to capture the essence of the process of decision-making. He proposed that decision-makers go through four stages involving: problem finding, intelligence, design, and choice. A fifth stage is sometimes added—the review phase—where managers debrief events and examine whether their solutions delivered the intended outcomes. On the other hand, observation of organizations suggest that this is rarely undertaken in real life.
There are limitations to the usefulness of normative models in practice inasmuch as they are not expected to be directly applicable to or observable in practice. It remains that normative models such as Simon’s model of decision-making are extremely useful for theorizing and teaching about decision-making and for analyzing decisions retrospectively, even where one never assumes to be able to completely map reality against a normative process. Mintzberg et al. (1976) made the point that “decisions processes are programmable even if they are not, in fact, programmed . . . there is strong evidence that a basic logic or structure underlies what the decision maker does and that this structure can be described by systematic study of his behaviour” (p. 247).
As far as this universal underlying structure is concerned, researchers in management have traditionally distinguished the period preceding the announcement of a decision and the period that follows it with the actual choice being the most important and privileged moment of the decision-making process (Fischer, 1953). As early as 1910, Dewey (1933) suggested that the decision-making process was in fact made up of a number of phases, namely, suggestion, intellectualization, development of hypotheses, reasoning or mental elaboration of these, and testing of the hypotheses.
The advantage of breaking down decision processes in this manner is that it enables a progressive study of what is otherwise an intractable process. It also emphasizes that “choice” is only one phase in this complex process where the first two phases in Simon’s model—intelligence and design—act as boundaries—or as constraints as noted in Simon (1977)—for the response of organizations. Obviously, alternatives that have not been considered from the outset are less likely to be selected than courses of action that have been well documented. In addition, good decisions are unlikely to be reached if the search for assumptions has not been carried out properly. In Victims of Groupthink, Janis (1972) illustrates in detail how the Bay of Pigs decision made by the Kennedy administration turned out to be one of the worst fiascos in recent history because “all the major assumptions supporting the plan were so completely wrong that the venture began to founder at the outset and failed at the early stages” (p. 14).
Janis also made the point that if the American president and his advisors had imagined that the “nightmarish” scenario that actually unfolded5 would materialize, or even if they had simply considered that it could happen, they would surely have rejected the CIA’s plan of invasion outright (Janis, 1972). Based on his empirical study of the decision-making process, Janis (1968) described five stages individuals go through to reach a suitable decision. His approach is different from Simon’s insofar as it takes into account the progressive changes in the decision-maker’s mind as new information becomes available about alternative courses of action. Table 1 shows the five stages envisaged by Janis and the major concerns associated with each stage.
Table 1. Janis’s Five Stages of Decision-Making
1. Appraising the challenge
How serious are the risks if I don’t change? If significant, go to Stage 2.
2. Surveying alternatives
What are the alternatives?
Is a specific alternative an acceptable means for dealing with the challenge?
Have I sufficiently surveyed the available alternatives?
3. Weighing alternatives
What alternative is best?
Does the best alternative meet the essential requirements?
4. Deliberating about commitment
Shall I implement the best alternative and make my decision known?
5. Adhering despite negative feedback
How serious are the risks if I do change? Repeat the stages if appropriate.
Source: Janis (1968).
Many researchers have adopted phased models to describe the decision-making process. The fundamental problem with such an approach to organizations is the extreme complexity that has resulted from cumulative research in the decision-making process. Witte (1972) found that the 233 decisions he studied consisted of a number of operations occurring in sequence with an average of 38 and a maximum of 452! Mintzberg et al. (1976) identified no fewer than seven standard routines, three sets of supporting routines, and six sets of dynamic factors, which they took as being the basic elements of strategic decision-making. This leads to the increasing sophisticated complexity and, ultimately, to the Bonini paradox discussed earlier, not only for researchers of decision-making but also for decision-makers themselves and for decision support systems developers who try to help them. For this reason, all stage models of the decision-making process are inherently normative or at least abstracted to various degrees.
Role of Expertise and Special Cases of Decision-Making
In reaction to the criticisms leveled at normative models and more generally at proposed models which, arguably, breached the parsimony principle of research, a branch of investigation emerged to learn from actual processes of decisions made by human agents in complex situations under severe time pressure, which became known as naturalistic decision-making. This stream has been led by Gary Klein (2008) who formalized his observations of fire crews and emergency response personnel into his Recognition Primed Decision-making (RPDM) model.
Observing that personnel placed in these very demanding circumstances did not have the time or luxury to follow complex processes and still seeking to bring a strong theoretical basis in the area of emergency response, Klein (2008) proposed that experienced actors who have become experts at dealing with certain types of events (triaging patients or attending to fires in urban areas) follow a very efficient reasoning process which delivers both rapid response and an acceptably low failure rate. Thus, RPDM rests on two key phases:
• Rapid recognition of previously known patterns and implementation of a solution designed to match this pattern and solve the problem
• Real-time estimation of the response of the event to confirm the accuracy of the initial recognition and pursuit of the implementation or correction of the diagnosis and alignment of the solution with the new diagnosis until the problem is solved
The essence of RPDM is that the experience of the decision-maker allows them to carry out a mental simulation to imagine what events have unfolded to deliver the observed state of play and anticipate what actions could offer remedy to the problems that require immediate correction (Klein, 1993). The role of expertise in this vision of decision-making is a central one, and it is obvious that novice operators/decision-makers would doubtless perform extremely badly if they sought to address an emergent problem by simple trial and error in the absence of specific support either from a human expert or from a decision support system. In the medical area, for instance, the value attached to experience is evident and the entire organization of healthcare systems relies on the progression of medical experts toward excellence throughout their careers. The same applies to different degrees to certain business environments and business domains of expertise (e.g., accounting, consultancy, software development, project management).
One conclusion, among other important conclusions, that one can legitimately draw from this discussion is that the nature of the decision problem tackled by managers has a significant bearing on both the practice of decision-making and on the research and analysis pertaining to it. Many questions pertinent to the analysis of decision-making processes and how to support them can best be answered in terms of “it depends.” This is the expression of fundamental contingencies that are inherent in any rich description of a decision problem.
The degree of structuredness of a problem (Simon, 1977), the extent to which it is new or recurrent, the extent to which it is perceived to be reversible by managers, the absence or existence of a consensual view in the organization facing the problem, the nature of the expertise involved, among other contingencies, mean that generic research in managerial decision-making is likely to be limited in terms of the degree of prediction, let alone prescription, which it can deliver. This is critical for students of decision-making processes who must understand the nature of the decision-making activities they wish to investigate, not in a general sense—decision-making is decision-making—but in its intricacies—clinical decision-making, for instance, is not the same as strategic decision-making, just as individual decision-making is not the same as group decision-making or the decision-making, which can be said to be made by organizations. Furthermore, decision-making is not the same at the different levels of hierarchy in the organization (Anthony, 1965; Mintzberg, 1973).
Another critical observation is that the role of expertise and experience cannot be overstated in our understanding of decision-making. Specifically, the novice decision-maker and the expert decision-makers are definitely two very different varieties of decision-maker. This is critical when considering the issue of decision support, and again, this is a critical consideration for both research and practice in the area of decision support.
There is an important research agenda that pertains to how decision-makers need to be trained so that novice decision-makers can be turned into expert decision-makers as rapidly as possible. How quickly this can be achieved matters for organizations who must either train tomorrow’s managers or hire costly experts from other organizations, from within their industry, or, when a new frontier of innovation is reached, from other industries which are ahead of the curve in comparison with theirs. For instance, financial services organizations, banks, and insurance companies have hired engineers and mathematicians to help them develop novel online services and mobile applications and to create new types of high-end financial products. In relation to this, expertise comes at two levels, in that managers can be experienced but be faced with new categories of problems or new technologies for solving them, just like managers can have experience of deploying certain type of solutions, but only in other contexts, being newcomers in the industries that hired them.
Interestingly, both observations are connected, in that the nature of the problems faced by a decision-maker has a bearing on the impact of accumulated experience. In repeatable situations, where there are clear patterns of decision-making and the variables are often the same, experts can learn from previous instances and consolidate their diagnosis as well as their solution design skills to become better decision-makers. In highly variable settings characterized by low structuredness, managers may not become any more efficient at designing robust solutions (Hogarth, 1975; Nisbett & Ross, 1980). The ability to deal with highly variable situations may, in itself, become the core competence needed in such contexts. One key aspect of this is whether decision support tools and data analytics can actually promote greater opportunity for decision-makers to learn in highly unstructured decision situations. Evidently, the more repeatable decision situations will also provide better opportunities to codify expertise in a decision support system than those involving unstructured or very variable decision problems.
One critical consideration where decision support and analytics are concerned is that human decision-makers have been found to be imperfect decision-makers in the sense of the theory of rational decision-making. Recurrent errors of reasoning occur, and decision-makers are affected by cognitive biases, which distort either their perception of reality and/or their ability to process the data available to them.
The list of recognized cognitive biases literally runs into pages under a number of headings pertaining to memory, cognition, perception, attitude to risk, social biases, although many of these are somewhat redundant. Generally, the identification of these biases indicates a human propensity to deviate from rational and logical solutions for a variety of reasons, either conscious or unconscious.
A compelling example of a cognitive bias is Kahneman’ attribute substitution (see Kahneman, 2003b), which illustrates the cognitive shortcuts that human reasoning sometimes follows when faced with a complex evaluation task, if there is a simple way to get to a solution. Attribute substitution bias can affect human cognition in varied and substantial manners, including in situations that are relevant to business settings. For instance, the perception by recruiters that candidates in an interview are or are not suitably qualified by reference to their appearance rather than their competences—where tangible visual attributes that are easy to see are used to judge a candidate instead of properly assessing if a candidate has the requisite qualities for a given job description—is a common occurrence. This is quite simply because assessing the suitability of a candidate for a complex job is very difficult and therefore a shortcut may be taken that relies on tangible, easily observable attributes. Attribute substitution results in interviewers unconsciously forming a quick opinion based on an easily observable, and often irrelevant, attribute (e.g., age, gender, height, race) instead of making the effort to determine the suitability of candidates based on complex criteria, evaluated through questioning and careful listening and observation.
This coin has two sides however, inasmuch as human decision-makers often suffer these biases because of the need to make rapid decisions and rely on heuristics and other decision-making shortcuts in difficult situations (Schwenk, 1984; Tsversky & Kahneman, 1974). It is arguably a moot point to discuss whether it would be better not to suffer from these biases at the expense of slower intuitive ability.
As human beings, we perhaps should not be too prompt to dismiss the advantages of our cognitive shortcuts—labeled system 1 thinking in Kahneman (2003)—in contrast with system 2 thinking, which involves deliberations and computation and is therefore inherently much slower, although possibly also more accurate, and where quality information can be collected (Kahneman, 2011). Intuition may be fallible, but it is still a critical skill for managers, especially top managers who are expected to deliver the vision that will guide their organizations into the future (Dane & Pratt, 2007). Dane and Pratt (2007) observed a number of barriers in the way of researching the role of intuition correctly. They cite definitional issues and identification issues, which mean that it is often difficult to recognize when managers use intuition and what impact it has on their decision-making. They contend, however, that in certain situations, managerial decision-making is likely to be facilitated and improved by reliance on intuition. Arguably, given the balance of things that are known with certainty and things that are uncertain in the world of managers, intuition (the ability to develop a vision of where to go without the benefit of specific forecast or clear organizational preferences) is likely to prove very useful as well as, in many respects, decisive—otherwise the personal computer might never have been invented, let alone released on the market. Neither would Apple have ever emerged as a dominant brand in the global market. Man might not have walked on the moon either—definitely not at a time when computers were about as powerful as a modern-day washing machine.
Insofar as many cognitive biases—notably those that are connected to misapprehension of probabilities and risk (Adam & Pomerol, 2008)—are often predictable, it may be wiser to educate managers as to the existence of these biases and how to deal with them (inasmuch as knowledge of some of the biases is not enough not to suffer from them because, of course, they are unconscious). This can promote the emergence of human decision-makers who can perform as intelligent and intuitive organizational actors capable of “educated” free will in a broad variety of complex decision-making situations, rather than decision automatons who blindly apply generic rules to guide their decision-making. As already stated, what is of importance from a decision support viewpoint is to understand how managers operate and how to further their cognitive and decision-making reach rather than consider them to be inherently deficient decision-makers. Silver (1991) makes a convincing case for such an approach.
Contemporary to the start of the DSS movement, Raiffa had proposed that “The human brain can be a magnificent synthesizer of disparate pieces of nebulous information, and often formal techniques and procedures thwart and inhibit this mysterious mechanism from operating efficiently” (1968, p. 272).The key issue is whether decision support and, in particular, recent development in the area of analytics change the extent to which formal decision support can enhance these “mysterious mechanisms” or just inhibit them through formalization, standardization, and incorrect rationalization (Earl & Hopwood, 1980). Carrying out research that attempts to understand how to support managers with advanced tools and techniques is both as important and as difficult as it was back in the ’60s, with one caveat formulated by Alter (1992, 2004): the objective is to make better decisions, not to develop artefacts for the sake of it. This is considered in the next section.
Developing Decision Support and Decision Support Systems
Many years after the seminal Gorry and Scott Morton (GSM) DSS framework was proposed in 1971, the exploitation of emerging technologies and concepts—graphical user interfaces (GUIs), data modeling, business intelligence (BI), business analytics (BA), information cockpits or dashboards, mobile computing, real-time decision-making (Burstein, Brezillon, & Zaslavski, 2010; Davenport, Harris, & Morison, 2010; Hopkins & Brokaw, 2011)—have offered the renewed promise to finally deliver effective support solutions to decision-makers (Davenport et al., 2010; Piccoli & Watson, 2008). An expanded DSS framework (Power, 2002) identified communications-driven, data-driven, document-driven, and knowledge-driven DSS as an expansion of the traditional model-driven DSS concept.
However, current evidence indicates that managers’ decisional needs are still not met in many organizations, despite large-scale IT investments (Daly & Adam, 2011). Few organizations, let alone traditional organizations, reach the ideal insights promoted by Davenport (2006) in Competing on Analytics. The evidence suggests that the problems inherent in managerial decision-making and the provision of information to support it are of a fundamentally intractable nature. Empirical reports on the impacts of BI, BA, and other decision support tools are sometimes inconclusive, especially where managers have dealt with highly uncertain or equivocal situations (Daft & Lengel, 1986; Earl & Hopwood, 1980; Speier, 2006; Speier & Morris, 2003; Williams, Dennis, Stam, & Aronson, 2007) or in the case of novice managers. In particular, they have failed to provide clear directions for conclusive, long-term integration of decision support in the decision-making processes of organizations (Marjanovic, 2010).
High levels of uncertainty combined with low frequency of repetition yield decision problems for which no predetermined and explicit set of ordered responses exists in organizations or in society (Le Moigne, 1995 Mintzberg et al., 1976). Where decision-makers, individually and as groups, do not have a model to underpin potential solutions, sometimes not even a partial one (Humphreys, 1989), they must endeavor to understand the problem and develop an ordered response, long before a programmed system can be considered (Filip, 2008; Simon, 1977). In these cases, managers rely mostly on intuition (Dane & Pratt, 2007), a thought process difficult to support. By contrast, BI/BA tools have been reported to be high-impact-in-high-volume decision scenarios that are repeatable to some degree and where a partial or complete model of the decision problem can be arrived at (Hopkins & Brokaw, 2011). The gap between the decision-making scenarios must be bridged both in the theory and the practice of decision-making and decision support.
Managerial Decision Support
As early as the 1970s, researchers began to wonder about the impact of the kind of decision aids that were becoming available. Slovic, Fischoff, and Lichtenstein (1977), Jungermann (1980), and Aldag and Power (1986), for instance, found no evidence of improvement in the quality of decision-making where decision aids where used. This basic question has remained problematic (Lohman, Sol, & de Vreede, 2003), and Arnott and Pervan (2005, 2008) have observed that the DSS field had a major practical credibility gap, where researchers seem to be focusing, at least in part, on issues of no practical relevance to managers. On the other hand, recent developments in the area of data analytics/business analytics seem to suggest a new enthusiasm for the decision support concept, as do the emergence of social and mobile computing and their impact on DSS research (Hosack, Hall, Paradice, & Courtney, 2012). These debates come at a time where expenditure in IT, from enterprise systems to data analytics projects, probably surpasses the scale of any IT investment to date.
Typical large-scale IT applications, such as SAP, match the scenario of large diversely connected multinationals running global operations (Lee & Lee, 2000; Lee, Siau, & Hong, 2003; Leidner, 2010; Markus, Tanis, & van Fenema, 2000) and confer some decision-making benefits (Holsapple & Sena, 2005). However, their use as decision support tools remains limited, notably because they are essentially operational systems (Carton, Jayaganesh, Pomerol, & Adam, 2007; Holsapple & Sena, 2005; Lee & Lee, 2000; Lee et al., 2003). In today’s world, managers increasingly administer processes that they cannot see and sometimes don’t control (Marjanovic, 2010; Zhao, Liu, & Yang, 2005). Therefore, the tasks facing today’s managers are increasingly difficult, yet the core focus for IT expenditure has not been focused on decision support. In other words, the informational architecture of firms is becoming increasingly complex and costly, but its focus is primarily operational, and its need for resilience, often based on standardized business processes, entails a rigidity which runs contrary to the needs of managers facing complex scenarios (Carton & Adam, 2010; Carton et al., 2007).
More than ever, it is left to DSS/BI/BA tools to bridge the gap between an increasingly diverse business world and a portfolio of large IT applications, increasingly standardized and consolidated, associated to highly specialized legacy applications that are proving equally difficult to replace or to integrate. Furthermore, management decision-making is based on much more than computer-generated outputs and also relies on softer processes and “back-of-the-envelope” calculations. It is the achievement of evidence-based management (Pfeffer & Sutton, 2006), beginning with the enlightened selection, by managers and DSS staff, of the performance indicators most useful to the business, including those weak signals that provide real insight into the future, that are critical to business success (Ackoff, 1967; Courbon, 1996; Keen & Scott Morton, 1978; Negash & Gray, 2008; Rockart & DeLong, 1988; Watson & Frolick, 1993). The intricacy of this objective explains managers’ continued reliance on simple spreadsheet tools to build up their perception and drive key aspects of the business (Panko, 2006; Pemberton & Robson, 2000). We conclude from these observations that, despite the passing of time, the need to understand how support systems can make an impact on decision-making and decision-making processes is as crucial as ever.
The Gorry and Scott Morton Framework
In their seminal article “A Framework for Management Information Systems,” Gorry and Scott Morton (197) developed a framework that was to become seminal in DSS research as the “GSM” framework (Figure 3). The GSM framework brings together Anthony’s three levels of managerial activity and Simon’s work on programmed and non-programmed decisions. Gorry and Scott Morton used the Anthony framework to take into account a basic observation of the work of managers: information requirements are different at different hierarchical levels and so is the type of tasks assigned to managers (see Mintzberg, 1973). They used Simon’s categories of decision problems to take into account that the issues faced by managers can also be inherently different in terms of their structuredness or programmability.
Gorry and Scott Morton had in mind that structured decision systems were those developed to solve fully structured problems where algorithms and/or decision rules proposing alternative solutions could be derived from manager’s knowledge and accumulated expertise. By contrast, decision support systems were much harder to develop where unstructured or semi-structured problems existed where little was known about the problem situation or how to solve it.
As Kirs, Sanders, Cerveny, and Robey (1989, p. x) indicated, “The Gorry and Scott Morton framework is perhaps the best known, most durable and most frequently cited in the IS field.” Keen and Scott Morton (1978) agreed that the Gorry and Scott Morton framework made a real and continuing contribution. Generally, the GSM framework was largely used from the 1970s right through to the current times.
However there has been some criticism of their work. Stabell (1994) noted that it was a simplification to regard the degree of structure of a given decision as an inherent characteristic: the structure of the environment of an organization and the nature of the relationship between means and ends—the important variables—within a particular industry affect the degree to which a problem is or is not complex. Moore and Chang (1983, p. 174) further argued that
DSS scholars have attempted to apply this structured/programmed distinction as a criterion for classifying potential DSS problem areas—the conventional wisdom being that a DSS is appropriate for semi-structured decision problems. . . . As we see it, a problem can only be considered more or less structured with regard to a particular decision maker, or group of similar decision makers, and at a particular point in time.
Based on this observation, it may be both incorrect and misleading for DSS designers to speak of structured or programmed problems in general when approaching a DSS design situation, instead of focusing on the specific situation in which the decision-makers they are trying to help operate. This is considered in the next section.
Analyzing Complex Decision Problems to Develop DSS
The decisions made by managers and the actions that follow are underpinned by their level of understanding of the decision problems they face. This, however, is variable, across firms and over time, as managers fine-tune their experience and use various heuristics to chart their way through complex environments by way of trial and error (Klein, 2008; Pounds, 1969). Thus, managers’ understanding of the world is the basis of their decision-making, allowing them to generate causal explanations for events and construct competitive advantage (Reed & DeFillippi, 1990). It is complemented by other intangible factors, such as intuition (Dane & Pratt, 2007). This perception of the world underpins a dialogue between managers and DSS developers as they strive to develop applications that provide key data and key support for the organization (Murphy, 1994). Where managers cannot express their understanding in tangible terms, or where key factors that bear on the decision problems are ill-defined, the quality and completeness of decision support will consequently be limited. Hence providing decision support in situations where managers cannot provide structured methods for computing outcomes is more difficult, and designers run the risk of imposing overly rational solutions on problems that have none (Earl & Hopwood, 1980; King, 1985).
Humphreys (1989) analyzed managerial understanding based on his concept of representation levels. His five representation levels theorize the evolution of managers’ thinking as they learn about the reality that surrounds them, based on the degree of abstraction of the representation of the problems they have in their minds and the degree of formalization of the representations of the proposed solutions that can thus be proposed. The evolution process described by Humphreys is a top-down process where the structuration of the concepts investigated by managers is refined from one level to the next over time. The process takes place at the level of individual managers but ultimately leads to a debate on the organizational stage, where shared meaning is constructed (Phillips, 1984). Thus managerial teams can “make decisions” by developing a single, agreed-upon representation of “the problem,” then “relentlessly employ(ing) the logic of choice to progressively strengthen the constraints on how the problem is represented until only one course of action is prescribed: the one which ‘should be’ actually embarked upon” (Humphreys & Jones, 2006, p. 3).
Levels 5 and 4 are viewed as strategic levels handled by top executives (problem defining), whereas the bottom three levels correspond to more operational and tactical levels (problem solving). Although all levels of management could theoretically span the 5 cognitive levels, it is clear that lower-level managers are more likely to be given problems already well formulated to work on, such that their thinking is mostly geared toward levels 1 and 2 (Levine & Pomerol, 1995), which are essentially the implementation levels. By contrast, at level 5, the decision-maker has some freedom to decide on a direction to follow within the constraints set by high-level constructs.
Humphreys and Jones (2006) noted that the process is characterized by a decrease in decision-maker discretion, as managers develop a set of constraints bearing on the representation of each problem, until a perception of reality emerges. The implementation of solutions then becomes logical, such that it can be delegated to lower-level management. As noted by Le Moigne (1995), complex problems rarely have objective solutions, and only a process such as that described by Humphreys can deliver a negotiated outcome that participants will agree to be realistic, viable, and acceptable (Phillips, 1984).
Humphreys’s (1989) concept of representation levels theorizes about processes that are mostly beyond description and offers opportunities to learn how to support managers in the situations corresponding to the different levels of the framework. The Humphreys framework is a sturdy and useful “walking stick” for both practitioners and researchers who seek to explore what decision support is available to managers in support of their decision-making processes in an organization and understand how further developments could help improve support in existing areas or open new grounds for decision support.
This article has sought to explore the bridge between what we know about managerial decision-making and the concept of decision support, particularly the concept of decision support systems, which has emerged over the last 50 years as a result of the availability of technologies for collecting, storing, manipulating, and displaying data and information. It has illustrated the diversity of domains that have contributed to this very rich vein of information systems research and provided a warning to novice researchers in the area that extensive reading of seminal material is required prior to carrying out research in decision support. Failure to do so is likely to lead to reinventing the wheel, or at least analyzing as new or novel issues and ideas that already received much attention in the past.
The two areas of decision-making and decision support are inextricably linked, and an understanding of both is required for both practice and research that seeks to break new ground in the area of decision support, whatever label is used: business intelligence, business analytics, or even big data analytics. Yet BI/BA and BDA research must continue to consider first and foremost the impact of decision support interventions on the quality of managerial decision-making rather than the technologies embedded in DSS artefacts.
Ackoff, R. L. (1967). Management misinformation systems. Management Science, 14(4), 147–156.Find this resource:
Adam, F., & Pomerol, J.-C. (2008). Developing practical decision support tools using dashboards of information. In F. Burbstein & C. W. Holsapple (Eds.), Handbook on decision support systems (Vol. 2, pp. 151–173). Berlin, Heidelberg: Springer.Find this resource:
Adam, F., Fahy, M., & Murphy, C. (1998). A framework for the classification of DSS usage across organizations. Decision Support Systems, 22(1), 1–13.Find this resource:
Aldag, R. J., & Power, D. J. (1986). An empirical assessment of computer-assisted decision analysis. Decision Sciences, 17(4), 572–588.Find this resource:
Alter, S. (2004). A work system view of DSS in its 4th decade. Decision Support Systems, 38(3), 319–327.Find this resource:
Alter, S. (1992). Why persist with DSS when the real issue is improving decision making? In T. Jelassi, H. Klein, & W. M. Mayon-White (Eds.), Decision support systems: Experiences and expectations (pp. 1–11). North Holland, The Amsterdam: Elsevier Science.Find this resource:
Anthony, R. (1965). Planning and control systems: A framework for analysis. Cambridge, MA: Harvard University Press.Find this resource:
Arnott, D., & Pervan, G. (2005). A critical analysis of decision support systems. Journal of Information Technology, 20(2), 67–87.Find this resource:
Arnott, D., & Pervan, G. (2008). Eight key issues for the decision support systems discipline. Decision Support Systems, 44(3), 657–672.Find this resource:
Barnard, C (1938). The functions of the executive. Cambridge, MA: Harvard University Press.Find this resource:
Bates, J., & Sykes, A. (1962). Studies of executive behaviour and their significance for management education. Management International, 2(6), 44–53.Find this resource:
Bonini, C. P. (1963). Simulation of information and decision system in the firm. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:
Brunsson, N. (1989). The organisation of hypocrisy (p. 242). Chichester, U.K.: John Wiley.Find this resource:
Burstein, F., Brezillon, P., & Zaslavski, A. (2010). Supporting real time decision-making: The role of context. Springer.Find this resource:
Carlson, S. (1951). Executive behaviour: A study of the work load and the working methods of managing directors. Stockholm: Stronbergs.Find this resource:
Carton, F., & Adam, F. (2005). Understanding the impact of enterprise systems on management decision making: An agenda for future research. Electronic Journal of Information Systems Evaluation, 8(22), 99–106.Find this resource:
Carton, F., & Adam, F. (2008). ERP and functional fit: How integrated systems fail to provide improved control. Electronic Journal of Information Systems Evaluation, 11(2), 51–60.Find this resource:
Carton, F., & Adam, F. (2010). Towards a model for determining the scope of ICT integration in the enterprise: The case of enterprise resource planning (ERP) systems. Electronic Journal Information Systems Evaluation, 13(1), 17–26.Find this resource:
Carton, F., Jayaganesh, M., Pomerol, J. C., & Adam, F. (2007). A primer for studying why ERP systems don’t work well as source of Data for supporting decision making. ICDSS2007 conference, Calcutta, India, January 2007.Find this resource:
Cohen, D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organisational choice. Administrative Science Quarterly, 17, 1–25.Find this resource:
Courbon, J. C. (1996). User-centered DSS design and implementation. In P. Humphreys, L. Bannon, A. McCosh, P. Migliarese, & J.-C. Pomerol (Eds.), Implementing systems for supporting management decisions (pp. 108–123). London, U.K.: Chapman & Hall.Find this resource:
Daft, R. L., & Lengel, R. H. (1986). Organisational information requirements media richness and structural design. Management Science, 32(5), 554–571.Find this resource:
Damasio, A. R. (1994). Descartes’ error. New York: Putnam’s.Find this resource:
Damasio, A. R., Damasio, H., & Christen, Y. (Eds.). (1996). Neurobiology of decision-making. Berlin: Springer.Find this resource:
Daly, M., & Adam, F. (2011). Understanding organisational decision support maturity: Case studies of Irish organisations. International Journal of Decision Support System Technology (IJDSST), 3(22), 57–78.Find this resource:
Dane, E., & Pratt, M. (2007). Exploring intuition and its role in managerial decision making. Academy of Management Review, 32(1), 33–54.Find this resource:
Davenport, T. H. (2006). Competing on analytics. Harvard Business Review, 84(1), 98.Find this resource:
Davenport, T. H., Harris, J. G., & Morison, R. (2010). Analytics at work. Boston, MA: Harvard Business School Press.Find this resource:
Dewey, J. (1933). How we think. Boston, MA: Heath.Find this resource:
Drucker, P. F. (2002). The discipline of innovation. Harvard Business Review, 80, 95–104.Find this resource:
Earl, M. J., & Hopwood, A. G. (1980). From management information to information management. In H. C. Lucas, Jr., F. F. Land, T. J. Lincoln, & K. Supper (Eds.), Information Systems Environment: Proceedings of the IFIP TC 8.2 Working Conference on the Information Systems Environment (pp. 133–143). North-Holland, The Netherland: IFIP.Find this resource:
Fayol, H. (1916). General and industrial management. London: Pitman.Find this resource:
Filip, F. G. (2008). Decision support and control for large-scale complex systems. Annual Reviews in Control, 32(1), 61–70.Find this resource:
Fischer, P. H. (1953). An analysis of the primary group. Sociometry, 16, 272–276.Find this resource:
Gorry, A., & Scott Morton, M. (Fall, 1971). A framework for management information systems. Sloan Management Review, 13(1), 55–70.Find this resource:
Harriman, B. (1974). Up and down the communication ladder. Harvard Business Review, (September/October), 94–102.Find this resource:
Hickson, D. J., Butler, R. J., Cray, D., Mallory, D., & Wilson, D. C. (1985). Comparing 150 decision processes. In J. Pennings (Ed.), Organisational strategy and change (pp. 99–113). London, U.K.: Jossey-Bass.Find this resource:
Holsapple, C. W., & Sena, M. P. (2005). ERP plans and decision-support benefits. Decision Support Systems, 38(4), 575–590.Find this resource:
Hogarth, R. M. (1975). Cognitive processes and the assessment of subjective probability distributions. Journal of the American Statistical Association, 70, 271–289.Find this resource:
Hopkins, M. S., & Brokaw, L. (2011). Customers matchmaking with math: How analytics beats intuition to win customers. MIT Sloan Management Review, 52(2), 35–41.Find this resource:
Hosack, B., Hall, D., Paradice, D., & Courtney, J. F. (2012). A look toward the future: Decision support systems research is alive and well. Journal of the Association for Information Systems, 13(5), 315–340.Find this resource:
Humphreys, P. (1989). Intelligence in decision support—A process model. In G. I. Doukidis, F. Land, & G. Miller (Eds.), Knowledge-based management support systems (pp. 22–51). Chichester, U.K.: Hellis Hovwood.Find this resource:
Humphreys, P.., & Jones, G. (2006). The evolution of group decision support systems to enable collaborative authoring of outcomes. World Futures, 62(3), 193–222.Find this resource:
Isenberg, D. (1984). How senior managers think. Harvard Business Review, 62(6), 81–90.Find this resource:
Janis, I. (1972). Victims of groupthink. Boston, MA: Houghton Mifflin.Find this resource:
Janis, I. (1968). Stages in the decision making process. In R. P. Abelson, E. E. Aronson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook (pp. 810–818). Chicago: Rand McNally.Find this resource:
Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict, choice, and commitment. Free press.Find this resource:
Jungermann, H. (1980). Speculations about decision-theoretic aids for personal decision making. Acta Psychologica, 45(1), 7–34.Find this resource:
Kadushin, C. (1978). Small world—How many steps to the top? Detroit News, 106(26), 19–A.Find this resource:
Kahneman, D. (2003a). Daniel kahneman—Autobiography. Les Prix Nobel. The Nobel Prizes 2002.Find this resource:
Kahneman, D. (2003b). A perspective on judgment and choice: Mapping bounded rationality. American Psychologist, 58(9), 697.Find this resource:
Kahneman, D. (2011). Thinking, fast and slow. New York: Farrar, Straus, and Giroux.Find this resource:
Keegan, W. (1974). Multinational scanning: A study of the information sources utilised by headquarters executives in multinational companies, Administrative Science Quarterly, 19, 411–421.Find this resource:
Keen, P. G., & Scott Morton, M. S. (1978). Decision support systems: An organisational perspective. Reading, MA: Addison-Wesley.Find this resource:
Kefalas, A. (1973). Scanning the business environment—Some empirical results, Decision Science, 4, 63–74.Find this resource:
Kinder, D. R., & Weiss, J. A. (1978). In lieu of rationality: Psychological perspectives on foreign policy decision making. Journal of Conflict Resolution, 22(4), 707–735.Find this resource:
King, W. R. (1985). CEOs and their PCs. MIS Quarterly, 9(3), xi–xii.Find this resource:
Kirs, P., Sanders, G., Cerveny, R., & Robey, D. (1989). An experimental validation of the Gorry and Scott Morton Framework. Management Information Systems Quarterly, 13(2), 5.Find this resource:
Klein, G. (2008). Naturalistic decision making. Human Factors: The Journal of the Human Factors and Ergonomics Society, 50(3), 456–460.Find this resource:
Klein, G. A. (1993). A recognition-primed decision (RPD) model of rapid decision making. In G. A. Klein, J. Orasanu, R. Calderwood, & C. E. Zsambok (Eds.), Decision making in action, models and methods (pp. 38–147). Norwood, NJ: Ablex.Find this resource:
Knoke, D. (1994). Networks of elite structure and decision making. In Galaskiewicz, J., & S. Wasserman (Ed.), Advances in social network analysis: Research in the social and behavioral sciences (pp. 274–294). Thousands Oaks, CA: SAGE.Find this resource:
Kotter, J. (1984). What effective managers really do. Harvard Business Review, 60(6), 156–167.Find this resource:
Langley, A., Mintzberg, H., Pitcher, P., Posada, E., & Saint-Macary, J. (1995). Opening up decision making: The view from the black stool. organization Science, 6(3), 260–279.Find this resource:
Lauman, E., Marsden, P., & Prensky, D. (1983). The boundary specification problem in network analysis. In R. S. Burt & M. J. Minor (Eds.), Applied network analysis (pp. 293–339). London, U.K.: SAGE.Find this resource:
Lee, Z., & Lee, J. (2000). An ERP implementation case study from a knowledge transfer perspective, Journal of Information Technology, 15(4), 281–288.Find this resource:
Lee, J., Siau, K., & Hong, S. (2003). Enterprise integration with ERP and EAI. Communications of the ACM, 46(2), 54–60.Find this resource:
Leidner, D. E. (2010). Globalization, culture, and information: Towards global knowledge transparency. Journal of Strategic Information Systems, 19(2), 69–77.Find this resource:
Le Moigne, J. L. (1990). La modélisation des systèmes complexes. Paris, France: Bordas, Dunot.Find this resource:
Le Moigne, J. L. (1995). La modelisation des systèmes complexes. Paris: Dunod.Find this resource:
Levine, P., & Pomerol, J.-C. (1995). The role of the decision maker in DSSs and representation levels. Paper presented at the Proceedings of the 29th Hawaii International Canference on System Sciences, Hawaii.Find this resource:
Liberatore, M. J., & Luo, W. (2010). The analytics movement: Implications for operations research. Interfaces, 40(4), 313–324.Find this resource:
Lippman, S. A., & Rumelt, R. P. (1982). Uncertain imitability: An analysis of interfirm differences in efficiency under competition. Bell Journal of Economics, 13(2), 418–438.Find this resource:
Lohman, F. A., Sol, H. G., & de Vreede, G. J. (2003, January). The illusion of effective management information: A critical perspective. Proceedings of the 36th Hawaii International Conference on System Sciences (HICSS’03).Find this resource:
Mahoney, L. S., Roush, P. B., & Bandy, D. (2003). An investigation of the effects of decisional guidance and cognitive ability on decision-making involving uncertainty data. Information and Organization, 13(2), 85–110.Find this resource:
March, J. G., & Simon, H. A. (1993). Organisations (2nd ed.). New York, NY: Wiley.Find this resource:
Marjanovic, O. (2010). The importance of process thinking in business intelligence. International Journal of Business Intelligence Research, 1(4), 29–46.Find this resource:
Markus, L., Tanis, C., & van Fenema, P. (2000). Enterprise resource planning: Multisite ERP implementations. Communication of the ACM, 43(4), 42–46.Find this resource:
Mintzberg, H. (1976). Planning on the left side and managing on the right. Harvard Business Review, 54(4), 120–130.Find this resource:
Mintzberg, H. (1975). The manager’s job: Folklore and fact. Harvard Business Review, 53(4), 49–61.Find this resource:
Mintzberg, H. (1973). The nature of managerial work. New York: Harper & Row.Find this resource:
Mintzberg, H., Raisinghani, D., & Theoret, A. (1976). The structure of “unstructured” decision processes. Administrative Science Quarterly, 21, 246–275.Find this resource:
Mintzberg, H., & Waters, J. A. (1985). Of strategies, deliberate and emergent. Strategic Management Journal, 6(3), 257–272.Find this resource:
Moore, J. H., & Chang, M. G. (1983). Meta-design considerations in building DSS. In J. L. Bennett (Ed.), Building decision support systems (pp. 173–204). Reading, MA: Addison-Wesley.Find this resource:
Murphy, C. (1994). Decision support systems—Putting emphasis on personnel rather than technology. In W. M. Mayon- White, S. Ayestaran, & P. C. Humphreys (Eds.), Proceedings of IFIP WG 8.3 Conference. San Sebastian, September 1994.Find this resource:
Negash, S., & Gray, P. (2008). Business intelligence. In F. Burstein & C. W. Holsapple (Eds.), Handbook on decision support systems (Vol. 2, pp. 175–193). Berlin, Heidelberg: Springer.Find this resource:
Nisbett, R., & Ross, L. (1980). Human inference. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:
Nutt, P. C. (1984). Types of Organizational Decision Processes. Administrative Science Quarterly, 29(3), 414–450.Find this resource:
Panko, R. (2006). Spreadsheets and Sarbanes-Oxley: Regulations, risks, and control frameworks. Communications of the Association for Information Systems, 17(1), 29.Find this resource:
Parikh, M., Fazlollahi, B., & Verma, S. (2001). The effectiveness of decisional guidance: An empirical evaluation. Decision Sciences, 32(2), 303–332.Find this resource:
Pemberton, J., D., & Robson, A. J. (2000). Spreadsheets in business. Industrial Management & Data Systems, 100(8), 379–388.Find this resource:
Pfeffer, J., & Sutton, R. (2006). Evidence-based management. Harvard Business Review, 84(1), 62–68.Find this resource:
Phillips, L. D. (1984). A theory of requisite decision models. Acta Psychologica, 56(1), 29–48.Find this resource:
Piccoli, G., & Watson, R. (2008). Profit from customer data by identifying strategic opportunities and adopting the “Born Digital” approach. MIS Quarterly Executive, 7(3), 113–122.Find this resource:
Pomerol, J. C. (1997). Artificial intelligence and human decision making, European Journal of Operational Research, 99(1), 3–25.Find this resource:
Pomerol, J. C., & Adam, F. (2008). Understanding human decision makdecision-making—A fundamental step towards effective intelligent decision support. In G. Wren (Ed.), Intelligent decision making: An AI based approach (pp. 3–40). London: Springer.Find this resource:
Pounds, W. F. (1969). The process of problem finding. Industrial Management Review, 11(1), 1–19.Find this resource:
Power, D. J. (2002). Decision support systems: Concepts and resources for managers. Westport, CT: Greenwood/Quorum Books.Find this resource:
Raiffa, H. (1968). Decision analysis: Introductory lectures on choices under uncertainty. Reading, MA: Addison Wesley.Find this resource:
Reid, R., & DeFillippi, R. J. (1990). Causal ambiguity, barriers to imitation, and sustainable competitive advantage. Academy of Management Review, 15(1), 88–102.Find this resource:
Rockart, J., & DeLong, D. (1988). Executive support systems: The emergence of top management computer use. Irwin: Business One.Find this resource:
Savage, L. J. (1972). The foundations of statistics (2nd ed.). New York: Dover.Find this resource:
Schwenk, C. R. (1984). Cognitive simplification processes in strategic decision-making. Strategic Management Journal, 5(2), 111–128.Find this resource:
Silver, M. S. (1991). Decisional guidance for computer-based decision support. MIS Quarterly, 15(1) 105–122.Find this resource:
Simon, H. A. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 69, 99–118.Find this resource:
Simon, H. A. (1956). Rational choice and the structure of the environment. Psychological Review, 63, 129–138.Find this resource:
Simon, H. A. (1977). The new science of management decision. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:
Simon, H. A. (1987). Satisficing. In J. Eatwell, M. Millgate, & P. Newman (Eds.), The new Palgrave: A dictionary of economics (Vol. 4, pp. 243–245). New York, NY: Stockton Press.Find this resource:
Simon, H. A. (1991). Models of my life. Cambridge, MA: Basic Books.Find this resource:
Simon, H. A. (1997). Administrative behavior. New York: The Free Press.Find this resource:
Slovic, P., Fischhoff, B., & Lichtenstein, S. (1977). Behavioral decision theory. Annual Review of Psychology, 28, 1–19.Find this resource:
Speier, C. (2006). The influence of information presentation formats on complex task decision-making performance, International Journal of Human-Computer Studies, 64(11), 1115–1131.Find this resource:
Speier, C., & Morris, M. G. (2003). The influence of query interface design on decision-making performance. MIS Quarterly, 27(3), 397–423.Find this resource:
Stabell, C. B. (1994). Towards a theory of decision support. In P. Gray (Ed.), Decision support and executive information systems (pp. 45–57). Engelwood Cliffs, NJ: Prentice Hall.Find this resource:
Starbuck, W. H. (1976). Organisations and their environments. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of industrial and organisational psychology (pp. 1069–1123). Chicago: Rand McNally.Find this resource:
Starbuck, W. H. (2004). Vita contemplativa: Why I stopped trying to understand the real world. Organization Studies, 25(7), 1233–1254.Find this resource:
Taylor, F. (1911). The principles of scientific management. New York: Harper Bros.Find this resource:
Thompson, J., & Tuden, A. (1959). Strategies structures and processes of organisational decision. In J. Thompson, P. B. Hammond, R. W. Hawkes, B. H. Junker, & A. Tuden (Eds.), Comparative studies in administration (pp. 195–216). Pittsburgh: University of Pittsburgh Press.Find this resource:
Tversky, A., & Kahneman, D. (1974). Judgement under uncertainty: Heuristics and biases. Sciences, 185, 1124–1131.Find this resource:
Ungson, G. R., Braunstein, D. N., & Hall, P. D. (1981). Managerial information processing: A research review. Administrative Science Quarterly, 26, 116–134.Find this resource:
Watson, H. J., & Frolick, M. N. (1993). determining Information Requirements for an Executive Information System. MIS Quarterly, 17(3), 255–269.Find this resource:
Weick, K. E. (1979). The social psychology of organizing. New York: McGraw-Hill.Find this resource:
Williams, M. L., Dennis, A. R., Stam, A., & Aronson, J. E. (2007). The impact of DSS use and information load on errors and decision quality. European Journal of Operational Research, 176(1), 468–481.Find this resource:
Witte, E. (1972). Field research on complex decision-making processes—The phase theorem. International Studies of Management and Organisation, 2(2), 156–182.Find this resource:
Zhao, X., Liu, C., & Yang, Y. (2005). An organisational perspective on collaborative business processes. Lecture Notes in Computer Science, 3649, 17–31.Find this resource:
(1.) It might make sense from a zoological viewpoint to set aside certain categories of mammals from the point of the intelligence that they can display (e.g., perception of self), but this is beyond the scope of this article.
(2.) In the intervening period, another leader—Pierre Mendes France when he was prime minister of France—gave instructions to restart research on developing an atom bomb in late 1954, but this fact is not as visible in history as the decisions made by De Gaulle, who was an advocate of France as a nuclear power all along.
(3.) This aerial disaster was caused by a number of human errors, such as the fact that one pilot decided to follow the emergency evasion instructions from its on-board Traffic Collision Avoidance System, while the other decided to listen to the instructions of the air traffic controller. If both had listened to the same source of advice, either the TCAS or the air traffic controller, the collision would likely not have occurred.
(4.) Interestingly Professor Sune Carlson pronounced the award ceremony speech at the Royal Swedish Academy of Science when Simon received his Nobel Prize in Economics on September 16, 1978.
(5.) The attempted invasion of Cuba by 1,400 Cuban guerrilleros trained and equipped by the CIA turned into a military disaster when all invaders were killed or captured and the United States could not deny their support of the attempt. Other consequences were the rapprochement between Cuba and the USSR, an international outrage against the United States and the loss of credibility of the Kennedy administration both inside and outside the country.