Chapter 1 Intuition in Decision Making – An Investigation in the Delivery Room
Frédéric Adam
Business Information Systems, University College Cork
INFANT Research Centre, University College Cork, Cork, Ireland
Eugene Dempsey and Brian Walsh
Department of Pediatrics and Child Health, University College Cork, Ireland
INFANT Research Centre, University College Cork, Cork, Ireland
Mmoloki Kenosi
Department of Pediatrics and Child Health, University College Cork, Cork, Ireland
Contents
- 1.1 Introduction
- 1.2 Researching Human Decision Making
- 1.2.1 Rational or Not
- 1.2.2 Thinking Slow and Thinking Fast – Where Intuition Fits
- 1.2.3 Unit of Analysis
- 1.2.4 Decision-Making Process – From Recognition to Intervention
- 1.2.5 Recognition Primed Decision Making
- 1.3 Research Method and Case Sampling
- 1.3.1 The Delivery Room Decision Making Scenario
- 1.3.2 Our Empirical Study
- 1.3.3 Data Coding and Analysis
- 1.4 Description of the Five Cases
- 1.4.1 Case 1
- 1.4.1.1 Overview of the Case and Care Dispensed
- 1.4.1.2 Analysis of Inputs and Outputs in the Decision Process
- 1.4.2 Case 2
- 1.4.2.1 Overview of the Case and Care Dispensed
- 1.4.2.2 Analysis of Inputs and Outputs in the Decision Process
- 1.4.3 Case 3
- 1.4.3.1 Overview of the Case and Care Dispensed
- 1.4.3.2 Analysis of Inputs and Outputs in the Decision Process
- 1.4.4 Case 4
- 1.4.4.1 Overview of the Case and Care Dispensed
- 1.4.4.2 Analysis of Inputs and Outputs in the Decision Process
- 1.4.5 Case 5
- 1.4.5.1 Overview of the Case and Care Dispensed
- 1.4.5.2 Analysis of Inputs and Outputs in the Decision Process
- 1.5 Discussion
- 1.6 Conclusions and Further Research Steps
- References
DOI: 10.1201/9781003030966-2
1.1 Introduction
The practice of decision making arguably involves some of the most demanding skills that human beings can acquire. As an activity, it is mostly exclusive to us as a species, insofar as animals are not considered to be able to display the free will that is required to underpin the kind of decision making that we associate with managers, medical doctors, or government ministers, as well as generally with individuals in their personal lives. Damasio et al. (1996) captured this point in their statement that “decision making is, in fact, as defining a human trait as language.”
As a result, decision making has received considerable attention from researchers in a variety of disciplines, from psychology to sociology, to management and information systems. In the IS discipline, decision making has been studied in terms of the decision aids or decision supports that can be developed to help decision makers make decisions without replacing them – i.e. in a way that leverages and enhances their decision-making skills. Separately, other domains of research have sought to pursue the artificial intelligence agenda proposed by Simon and Newell as early as 1972, where software artifacts take over specific segments of decision-making processes.
Research in human decision making, however, is complex because the mental activities involved in deciding and the circumstances facing decision makers are themselves complex and varied. It is also difficult because these mental activities are not directly observable for a researcher, not in the way that a physical practice may be visible, and even the closest observation of decision makers can yield ambiguous and incorrect conclusions. In management, we are following in the footsteps of such seminal researchers as Carlson (1951) or Mintzberg (1973 and many other dates) who have explored how to observe managers in a way that can make their reasoning intelligible. Researchers have leveraged this research because developing useful decision supports can only be undertaken on the basis of a stable understanding of human decision making – a science of decision making that delivers some certainties in relation to the mental and organisational processes involved in decision making (Pomerol and Adam, 2008). Without such a scientific base, much of the practice of decision support systems development would need to rely on trial and error, where potential systems are pushed onto managers with varying levels of success. The executive information systems (EIS) period provides examples of this trial-and-error process, with many systems failing to survive the turnover of managers in given positions, leading to systems dis-adoption (Elam and Leidner, 1995; Singh et al., 2002).
However, much progress has been made in relation to the establishment of a science of decision making, with major contributions by such leading scholars as Herbert Simon and Daniel Kahneman who, across a long period and many publications, have proposed some of the theories that still guide us today. Crucially, even they have admitted that we do not understand enough about human decision making to make conclusive observations about its intricacies. As a result of the complexity of the mental processes involved, observation and lab experiments have been the dominant scientific method to learn about decision making, and a rich history of reports on decision making, notably managerial decision making, are available for us to learn from.
One key lesson from this history is that the context and conditions in which a decision takes place and the position the decision maker finds themselves are very important. The more precise the research question, the more useful it is to study a specific type of problem or scenario, rather than attempt to propose generalities that risk providing little actionable understanding of decision makers and their mental processes. Thus, in this paper, we concentrate on a very specific type of medical decision making to try to decipher how a diagnosis is reached, how interventions are designed, and how decision makers could be supported in a very specific scenario: the environment of the delivery room where neonatologists look after the most vulnerable type of patients: premature babies born before 32 weeks of gestation. Our focus on such decision making is justified by the potential to improve health outcomes for these patients with targeted decision support that leverages the expertise of neonatologists and, at the same time, reinforces the evidence-based nature of decision making in a scenario that is characterised by the need for rapid and flawless interventions with very severe consequences. We find that the delivery room is a space where critical decision making takes place under the leadership of an expert and that, although intuition plays a critical role, there is a need to increase the objective nature of some of the observations made during the decision process. We conclude that certain technologies could help in this very special decision-making setting.
The next section reviews relevant observations about the science of human decision making and how it applies to the specific scenario of the delivery room. The paper then presents the methodology we followed, our analysis of five video recordings of deliveries, and our observations and conclusions.
1.2 Researching Human Decision Making
What we know about decision making has mostly been written in the last hundred years, originating in the research conducted by such seminal scientists as Foley, Barnard, Simon, Carlson, Mintzberg, Kahneman, or Drucker, to name only a few of them. Two of them, Simon and Kahneman, received Nobel Prizes for their work on human decision making; this illustrates the excellent research conducted in this domain. In earlier times, Plato had begun our journey of discovery with his observations that the human mind was a tool that could be used to uncover truths and use them to implement positive changes in human society. Thus, the history of research on decision making has been paved by great writers and researchers.
The other side of that coin is more problematic: if the science of decision making has received a lot of attention, from a variety of seminal researchers, this has meant that the decision-making domain is one of the most difficult to penetrate for new researchers within the IS discipline. The volume of material and its diversity are hard to grasp and even more difficult to apply to the analysis of decision making in practice. Yet, this is a requirement if we are to progress in our decision support systems endeavours.
In the context of this research project, the difficulty in studying intuition in decision making comes from a number of specific issues, as discussed in the next sections.
1.2.1 Rational or Not
If we decided to split the science of decision making between two camps, artificial though it may seem, we could oppose those who believed that human decision makers should strive to be essentially rational – that they should seek to optimise their decisions and their outcomes – and those who are more interested in exploring human decision makers as not particularly rational in their decisions, seeking to understand to what extent and why that might be and where and why they might not. The former will consider that deviations from rational orientations are “flaws” in decision making, and the latter will prefer to see these deviations as heuristics and as evidence that human decision makers use their cognitive abilities to adapt to the environment where they operate.
Thus, “classical” economics theories have proposed that human decision makers are rational, seek to maximise their utility by reference to the information they possess about the world around them, and design interventions that maximise the likelihood that their preferences will be achieved. Such theories have been put forward as prescriptions of how human decision makers should behave (Beach and Lipshitz, 2017). Other researchers have provided a rich and ample body of observations and experiments to show that human behaviour often deviates from what rational reasoning may suggest. They have sought to explain why we display such “deviant” behaviour and, furthermore, how we benefit from it as decision makers.
This debate is interesting, but it may be a distraction from the real objective of research on decision making: to understand how decisions are made and, further, how to support them. Clearly, based on seminal research on human decision making, our cognition is characterised by what have been sometimes called “biases” (Samuelson and Zeckhauser, 1988) and sometimes “heuristics” (Busenitz and Barney, 1997). The use of either term reflects a belief that, on the one hand, human decision makers are imperfect (bias) or, on the other hand, that they are well adapted to their environments). Reflecting on the nature of the environment and noting that it was characterised by so much uncertainty, Simon (1986 for instance) referred to Rationality (with a capital “R”) as Olympian (by reference to the mountain supposed to the home of the gods in Ancient Greece), and proposed that bounded rationality was the best we could aspire to as decision makers. The future is unknown, the present is a puzzle, and our preferences are often confused: there is much more in human decision making that is not known than that is known. By analogy with t...