Thinking, The Past and Where-to-Go
The History and Future of Judgment & Decision-Making Research
First of all, pardon me for the terrible pun in the title.
In my previous posts (Social Heuristics and Advice-Taking), I shared about some of the popular topics in judgment and decision-making (JDM) research, explaining some of the heuristics and biases in human decision-making. Ever since psychological experiments started showing that human participants did not conform to economic theories, the debate over human rationality has always been contentious. But is it really necessary to classify humans as only rational or irrational? Can’t we be both rational and irrational at the same time? What is the best way for JDM research to move forward? To answer these questions, I will first talk about the history of the field, the debates that are on-going, as well as the issues that current research is still not addressing.
Blast from The Past
Even before the advent of prospect theory (Kahneman & Tversky, 1979), JDM was already a topic of interest for philosophers and economists alike. Through rational choice theory of economics, individuals were thought to optimise their decisions by maximising their utility (Jevons, 1866). Logically, it makes sense that people should not want to settle for anything less than optimal if given a choice. But empirical studies have revealed that rational choice theory is not only often unable to account for the findings, the observed behaviours of individuals also do not match the theory’s predictions.
This is perhaps psychologists’ greatest contribution to the study of human JDM, as it added a realistic human element to the understanding of JDM, instead of just assuming that people are homo economicus with perfect information. The proposal of prospect theory by Daniel Kahneman & Amos Tversky (1979) showed that the expected utility theory was overly simplistic, and that individuals in fact make value judgments differently depending on how information is being perceived. The theory’s ability to model actual choices in real-life earned Kahneman a Nobel Memorial Prize in Economics in 2002.
Along with the discovery of prospect theory, Tversky and Kahneman (1975) uncovered a number of heuristics that people appear to use, which systematically led to certain biases. Since rational choice theory did not seem to describe how people make decisions, it consequently led to the proposition that humans are irrational. This propelled research interest in addressing whether humans are truly irrational, and how rationality is defined in the first place. Herbert Simon (1982) conceived the concept of bounded rationality, whereby humans can only be rational to the extent of the information, time and cognitive resources that are available to them. Although bounded rationality has helped to shape the scope of human JDM, the definition of full rationality is still open to interpretation.
In one school of thought, behaviours described in rational choice theory may still be considered as the gold standard of rationality, while actual behaviours are perceived as suboptimal strategies that are bounded by the best of human abilities; in another school of thought, the bounds of human abilities are seen as the highest achievable level of rationality, and comparison to rational choice theory is just irrelevant. Current research is still debating on the answer, where the former stance is taken by academics like Kahneman, who think that the heuristics and biases program has demonstrated that people are irrational; and the latter stance is adopted by researchers such as Gerd Gigerenzer, who argue that rationality needs to be understood under an ecological context, and believe that heuristics can be good decision tools in the correct context (Gigerenzer & Goldstein, 1996).
Same Story, Different Perspectives
A closer examination of the disagreement will reveal that the two schools of thought are just arguing from different perspectives. There are namely three perspectives of JDM: normative, descriptive and prescriptive.
- Normative theories, such as rational choice theory, outline the ideal decisions an individual would make, if the individual was fully rational. As shown by empirical findings, normative theories do not necessarily describe what is observed in real-life, but they can be thought of as what would be observed if conditions were perfect.
- On the other hand, descriptive theories, like prospect theory, have the role of accurately describing the actual behaviour of decision makers. A good descriptive theory would even be able to make accurate predictions of the behaviour that will be observed based on given conditions.
- The newest perspective amongst the three is prescriptive theories, which are not overly concerned about the ideal standards of rationality, but rather focus on how to get people to make better decisions by making the most out of the given conditions.
Research from the Kahneman camp mostly concentrates on developing descriptive theories, while showing how the findings are incompatible with normative theories. The Gigerenzer camp, however, interprets normative and descriptive theories through a different lens, with the intention of prescribing heuristics as an efficient way of making decisions. In other words, the two camps have quite different agendas and should not be in any direct conflict. They are both valid in their own respect and are merely two sides of the same coin.
Research in JDM has been caught up in the debate between these two camps over the last two decades, and many academics have been forced to intellectually choose one camp over another. While a battle of competing theories can help to generate new ideas that will ultimately determine the better descriptive theory, overly fixating on the battle has also resulted in a neglect of other aspects in JDM research that are potentially more beneficial to our society. Kahneman (2011) has never disagreed that heuristics could be useful when applied in the correct contexts. In fact, he often begins his explanations by describing heuristics as an evolutionary adaptation for making quick decisions, which is in line with Gigerenzer’s view. However, Gigerenzer has to also concede that when these heuristics are applied in the wrong contexts, errors are bound to occur. Hence, instead of fruitlessly arguing which theory is right, I believe a more fruitful direction is to adopt a constructive approach for the future of JDM research.
Knowing Where to Go
In order to move towards a constructive approach of JDM research, two specific areas can be explored.
- More research should focus on collaborations to reconcile the differences between the various schools of thought. This is perhaps the most important step, because if reconciliation cannot be achieved, then the ghosts of their differences will continue to haunt future research.
- Although research on the biases in human JDM should not be discontinued, it will be more useful to find out what helps to eliminate those biases and improve decision making. By searching for solutions together with the problems that they discover, researchers may be able to learn more about the underlying processes of those biases.
I will explain in greater detail how reconciliation and improvement can be achieved.
Adversarial collaboration for any discipline is never an easy endeavour. In an interview, Kahneman explained that “it is not easy because people who think poorly of your work and of your ideas get on your nerves and so you have to overcome that” (Big Think, 2012). Nonetheless, by putting aside differences for the sake of advancing scientific knowledge, an adversarial collaboration can bring about a reconciliation. In an admirable example of such a collaboration, Kahneman and Gary Klein, pioneer of Naturalistic Decision Making (NDM), published a paper together on expert intuition (Kahneman & Klein, 2009). Although Kahneman and Klein agreed that intuitive judgment is most likely a System 1 process that is often fast and automatic (Evans, 2007), their views on the usefulness of such judgments differ.
Researchers in NDM generally consider intuition to be a skill that is acquired from experience, but academics from the heuristics and biases (HB) program are more concerned about intuitions resulting from oversimplified heuristics that are not experienced-based. For skilled intuition to be acquired, Klein (1993) proposed a recognition-based model where two conditions need to be satisfied: not only must the learning environment contain sufficient cues related to the task, there should also be regular opportunities for the individual to pick up these cues. While Kahneman agrees with these conditions, he cautions that professionals in some fields like the financial industry do not work in environments that are constant enough for feedback cues to be valid (Einhorn & Hogarth, 1978).
Eventually, the collaboration helped them to realise that they actually agree on most things, but the biased perspectives of both camps led them to fixate on different aspects of the problem. Hence, adversarial collaborations can help different schools of thought to reconcile and focus on more constructive JDM research.
In another heartening proposal of alliance, the fast-and-frugal heuristics (FFH) program of the Max Planck Institute for Human Development has also reached out to the NDM community for a collaboration. Keller et al. (2010) noted that in agreeing that human decision making should not be viewed as irrational, the fundamental principles of FFH and NDM are actually similar. Although both schools of thought believe that the environmental context is important for studying more realistic decision making, NDM approaches it through naturalistic methods such as cognitive task analysis (Militello & Hutton, 1998), and FFH attempts to replicate those contexts in laboratory settings. However, Keller et al. (2010) criticized NDM for abandoning theoretical development in the course of studying specific situations through case-by-case fieldwork, which does not help in generalizing findings and making predictions. They suggested that FFH’s theories and methods may help to improve NDM’s models of decision making, while NDM’s models can guide FFH in navigating applied domains that FFH researchers are not familiar with.
It is rather thoughtful of the FFH program to be interested in conducting research with the NDM community. If Kahneman is able to work with Klein and the FFH program sees commonalities with the NDM community, perhaps there is a possibility for the HB and FFH programs to reconcile and produce more constructive JDM research together.
Besides working on reconciliation research, more effort can also be put into thinking of ways to improve JDM. As noted by Lilienfeld et al. (2009), of the 1300 over articles on cognitive biases in the PsycInfo database, only 158 were on debiasing, indicating the sparse literature on improving decision making. Although the likes of Kahneman believe that biases arise from a reflexive System 1 that is difficult to control (Shariatmadari, 2015), some have suggested to solve this problem by designing solutions that uses System 2 to override System 1 (Milkman et al., 2009). A pessimistic mindset about the possibility of eliminating biases is unconstructive to begin with, and smothers any opportunity to conduct research in this area.
In this regard, Gigerenzer’s ABC group of the Max Planck Institute has taken a more constructive approach by developing FFH and decision toolboxes that aim to improve decision making (Gigerenzer et al., 1999). Similarly, credit has to be also given to the work done in behavioural economics and nudging, but the idea still revolves around tapping onto the biases that people are already committing to shape their decisions (Thaler & Sunstein, 2009). A challenge for HB researchers to consider is to try and eliminate the biases they discover in their experiments. In the process of doing so, they will probably be able to learn more about the underlying mechanisms of the biases in question, which should be of interest to them anyway. After all, it is certainly more constructive to present solutions along with the problems they find.
Towards a Constructive Future
In summary, a fruitful direction for the future of JDM research is to adopt a constructive approach that reconciles differences with the other schools of thought and works on improving decision making. Increasing collaboration with NDM may not only promote reconciliation between the HB and FFH programs, it can also encourage JDM research to look beyond laboratory experiments, and study some of the most difficult and important decisions in professional expertise. Although NDM has already been working on these areas, it is certainly useful for JDM research to validate the fieldwork of NDM, and attempt to explain the phenomena observed in NDM’s findings. Also, for JDM research to be more beneficial, studies in debiasing and techniques to improve decision making should be encouraged. By doing so, perspectives on rationality will become less pessimistic, and opportunities to discover useful solutions will likewise increase.
It is not uncommon for competing theories to exist in academic disciplines, and JDM is not an exception. While JDM may not be old compared to other scientific disciplines, the battle between the different schools of thought has been going on long enough without much conclusion. If a fruitful direction for the future is more than just satisfying academic egos, then perhaps it is about time we push for a more constructive approach to JDM research.
If you found this article interesting, you might also enjoy these:
Social Heuristics: The Pros and Cons of Gut Feelings
In the previous post on Machine Learning vs Human Learning, I briefly mentioned that the heuristics used by humans…
Variety Over Selectivity: A Guide to Better Advice-Taking
In the previous post, I used the Galton’s ox-weight estimation and IMDb’s polling system to illustrate how the wisdom…
- Big Think (2012, October 9). Daniel Kahneman: Adversarial Collaboration. Retrieved from: http://bigthink.com/videos/daniel-kahneman-adversarial-collaboration
- Einhorn, H. J., & Hogarth, R. M. (1978). Confidence in judgment: Persistence of the illusion of validity. Psychological review, 85(5), 395.
- Evans, J. S. B. (2007). Hypothetical thinking: Dual processes in reasoning and judgement (Vol. 3). Psychology Press.
- Gigerenzer, G., & Goldstein, D. G. (1996). Reasoning the fast and frugal way: models of bounded rationality. Psychological review, 103(4), 650.
- Gigerenzer, G., Todd, P. M., & ABC Research Group, T. (1999). Simple heuristics that make us smart. Oxford University Press.
- Gonzalez, C., & Meyer, J. (2016). Integrating trends in decision-making research. Journal of Cognitive Engineering and Decision Making, 1555343416655256.
- Jevons, W. S. (1866). “Brief Account of a General Mathematical Theory of Political Economy” by William Stanley Jevons Journal of the Royal Statistical Society, London, XXIX (June 1866), pp. 282–87. Journal of the Royal Statistical Society, 29, 282–87.
- Kahneman, D. (2011). Thinking, fast and slow. Macmillan.
- Kahneman, D., & Klein, G. (2009). Conditions for intuitive expertise: a failure to disagree. American psychologist, 64(6), 515.
- Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica: Journal of the econometric society, 263–291.
- Keller, N., Cokely, E. T., Katsikopoulos, K. V., & Wegwarth, O. (2010). Naturalistic heuristics for decision making. Journal of Cognitive Engineering and Decision Making, 4(3), 256–274.
- Klein, G. A. (1993). A recognition-primed decision (RPD) model of rapid decision making (pp. 138–147). New York: Ablex Publishing Corporation.
- Lilienfeld, S. O., Ammirati, R., & Landfield, K. (2009). Giving debiasing away: Can psychological research on correcting cognitive errors promote human welfare?. Perspectives on psychological science, 4(4), 390–398.
- Militello, L. G., & Hutton, R. J. (1998). Applied cognitive task analysis (ACTA): a practitioner’s toolkit for understanding cognitive task demands. Ergonomics, 41(11), 1618–1641.
- Milkman, K. L., Chugh, D., & Bazerman, M. H. (2009). How can decision making be improved?. Perspectives on psychological science, 4(4), 379–383.
- Shariatmadari, D. (2015, July 18). Daniel Kahneman: ‘What Would I Eliminate If I Had a Magic Wand? Overconfidence.’ The Guardian. Retrieved from http://www.theguardian.com/books/2015/jul/18/daniel-kahneman-books-interview.
- Simon, H. A. (1982). Models of bounded rationality: Empirically grounded economic reason (Vol. 3). MIT press.
- Thaler, R. H., & Sunstein, C. R. (2009). Nudge: Improving decisions about health, wealth, and happiness. London: Penguin Books Ltd.
- Tversky, A., & Kahneman, D. (1975). Judgment under uncertainty: Heuristics and biases. In Utility, probability, and human decision making (pp. 141–162). Springer Netherlands.
Originally published at: https://learncuriously.wordpress.com/2019/02/01/thinking-past-and-where-to-go