Decision-Making Lectures

Introduction to Decision-Making

• Decision-making involves information processing and is influenced by cognitive capabilities and environmental factors.

Mental Representations

• The brain processes internal representations of information, including prior knowledge and expectations stored in memory.

The Mind vs. The Brain

• The concept of internal representations is central to psychology, distinguishing between the 'mind' (conscious internal world) and the 'brain' (biological entity).

Rationality

• Classical view of rationality involves complete knowledge, a well-organised system of preferences, and the ability to calculate the highest attainable point on the preference scale.
• Bounded rationality, introduced by Herbert Simon, considers an organism's cognitive capacity and environmental structure in defining rationality.

Models

• Models in decision-making recognize an analogy between the brain/mind and a computer, considering the critical role of physical environments and the mental machinery.
• Models can be categorised into three levels of analysis:
  1. Computational level: what problem does the process solve?
  2. Algorithmic level: through what series of steps is input transformed into output?
  3. Implementational level: how are the transformational steps physically realised in the brain?

Decision-Making and Judgement

• A judgement involves discerning patterns in cues to form a conclusion about an unobserved state of the world.
• A decision is a choice among several possible action alternatives based on a judgement.

Decision-Making in Toy Problems

• Toy problems have well-understood cues, a correct answer, and often involve judgement and decision-making.
• Researchers use toy problems to develop theories of everyday decision-making, such as Prospect Theory, Heuristics and Biases, Fast-and-Frugal Heuristics, and Bayesian approaches.

Prospect Theory

• Prospect Theory is a highly influential theory within decision-making research, expanding on Expected Utility Theory.
• The theory considers the perceived usefulness of money and how it declines with the amount already possessed.
• Key concepts in Prospect Theory include:
  1. Loss aversion: the tendency to prefer avoiding losses over acquiring gains.
  2. The framing effect: the presentation of identical information in different ways, impacting decisions.

The Framing Effect

• The framing effect demonstrates how people can be influenced by the way information is presented, leading to different decisions based on gain or loss frames.

• The effect is reversed for small probabilities, with risk-seeking behavior in gain frames and risk-aversion in loss frames.### Prospect Theory Predictions

• Prospect Theory predicts loss aversion, reflected in a steeper expected value curve for losses.

• It predicts framing effects, reflected in a concave expected value curve for gains and a convex one for losses.

• Framing effects are reversed for small probabilities.

Rationality in Prospect Theory

• According to Kahneman and Tversky, Prospect Theory demonstrates that human decisions are rational in a classical sense.
• Loss aversion reflects the fact that pain is more urgent to attend to than pleasure.
• Risk aversion reflects the fact that organisms settle into habits.

Individual and Situational Differences

• Studies have demonstrated that risk preferences vary across individuals and situations.
• Individual differences in the framing effect have been observed (Abedellaoui et al., 2008).
• Situational effects on loss aversion have also been observed (Walasek & Stewart, 2014).

Criticisms of Prospect Theory

• Prospect Theory was developed to account for experimental findings, not to describe all of human decision-making.
• The theory says little about whether human judgments in two-choice tasks are rational.
• It does not explain how people might arrive at the two choices they have in judgments.

The Endowment Effect

• The endowment effect is the phenomenon where people who own a product tend to value it more and are less willing to trade it for another product.
• One explanation for this effect is loss aversion, which states that exchanging an owned product feels like a loss, and people are averse to losses.
• An alternative explanation is based on the concept of “psychological ownership,” where owning a product makes it part of one’s self-concept, leading to a more positive perception of the product.

Limitations of Prospect Theory

• Prospect Theory is a model of decision-making that extends on Expected Utility Theory, which sought to demonstrate human rationality.
• The theory says little about rationality and does not specify the purpose of the described processes in maintaining effective functioning in the environment.
• It does not explain why people deviate from the proposed processes in some circumstances.

Heuristics and Biases Model

• The heuristics and biases model explains how people make decisions under uncertainty
• The model proposes that people use mental shortcuts (heuristics) to make decisions, which can lead to biases (systematic errors)

Representativeness Heuristic

• The representativeness heuristic is a mental shortcut that judges the likelihood of an event based on how closely it resembles a typical case
• Rule: “A is representative of B to the extent that A resembles B”
• Example: Linda is 31 years old, single, outspoken, and very bright, which is more likely: 1) Linda is a bank teller or 2) Linda is a feminist bank teller?
• Biases due to representativeness:
  • Conjunction fallacy
  • Gambler’s fallacy
  • Base rate neglect

Confirmation Bias

• Confirmation bias is the tendency to seek information that confirms one’s existing beliefs and ignore information that contradicts them
• The positive test heuristic: when testing a hypothesis, ask questions that will yield an affirmative response if the hypothesis is true
• Example: Wason Selection Task, where participants are asked to determine the rule governing the sequence of cards
• Confirmation bias is adaptive in certain situations, such as when the causes and effects being judged are rare in the environment

Anchoring and Adjustment Heuristic

• The anchoring and adjustment heuristic is a mental shortcut that relies on an initial value (anchor) and adjusts it to reach a final answer
• Example: Plous (1993) study, where participants were asked to estimate the probability of a disease given a positive test result
• Epley and Gilovich (2005) study, where participants were given either self-generated or experimenter-provided anchors to estimate quantities

Criticisms of the Heuristics and Biases Approach

• Criticisms include the approach being too simplistic and not stating when heuristics produce accurate judgments
• The confirmation bias is exempt from this criticism, as research shows that the positive test heuristic produces accurate judgments in certain situations
• Another criticism is that the approach does not explain why biases are reduced or disappear under certain conditions

Dual Systems Perspective

• The dual systems perspective proposes that there are two systems in the brain: System 1 (fast, unconscious, and unintentional) and System 2 (slow, conscious, and intentional)
• Criticisms of the dual systems perspective include:
  • Brain imaging research suggests that different parts of the brain work together in a more integrated and iterative way
  • The approach focuses on identifying rare cases where people make errors, but does not explain what makes those cases different from others

Introduction to Fast-and-Frugal Heuristics

• The concept of fast-and-frugal heuristics is introduced as an alternative to rational decision-making
• It is based on the idea that people make decisions quickly and efficiently using simple rules of thumb, known as heuristics
• These heuristics are adapted to the structure of the environment and are not necessarily optimal, but are accurate enough to be useful

Definition of Fast-and-Frugal Heuristics

• Fast-and-frugal heuristics are defined as simple rules of thumb that allow for quick and efficient decision-making
• They are based on a few cues, which are processed in a specific order to make a decision
• Examples of fast-and-frugal heuristics include the Take-the-Best heuristic and the Recognition heuristic

The Take-the-Best Heuristic

• The Take-the-Best heuristic is a fast-and-frugal heuristic that involves selecting the best option based on the most important cue
• It involves searching through cues in order of their validity, stopping when the first cue that discriminates between alternatives is found, and choosing the alternative with the "yes" value of the stopping cue
• The Take-the-Best heuristic is shown to be accurate in experiments, even when compared to more complex decision-making strategies

Evidence for Fast-and-Frugal Heuristics

• Studies have shown that people are highly attuned to their environments and are able to select cues and heuristics that are most appropriate in various environments
• Gigerenzer's research has shown that people are able to use fast-and-frugal heuristics to make accurate decisions in a wide range of situations
• The use of natural frequencies has been shown to be an effective way to elicit accurate judgments from people

The Less-Can-Be-More Principle

• The Less-Can-Be-More principle states that more information is not always better, and that there is a point in the decision-making process when doing more computation or considering more information becomes detrimental to making a good decision
• This principle is related to the concept of ecological rationality, which emphasizes the importance of understanding the environment in which decisions are made

Applications of Fast-and-Frugal Heuristics

• Fast-and-frugal heuristics have been applied in a wide range of fields, including business, medicine, and education

• They have been used to develop decision-making strategies that are quick, efficient, and accurate

• The use of fast-and-frugal heuristics has been shown to be effective in making decisions in a variety of contexts.### Heuristics and Biases

• Take-the-best is a fast and frugal heuristic that uses only the most valid feature to predict the outcome.

• In contrast, rational strategies consider all available features to make a decision.

Response Patterns

• Research by Bergert and Nosofsky (2007) showed response patterns across five unseen pairs where Take-the-best and rational strategies make different predictions.

Fast and Frugal Framework

• The fast and frugal framework proposes that heuristics can be adapted to an environment, allowing for fast and accurate decision-making.
• Key components of the framework include:
  • A definition of heuristics as containing three components
  • The less-can-be-more principle (ecological rationality), which states that speed and accuracy of decision-making do not need to be traded off against each other
  • The notion of an adaptive toolbox, from which people draw out appropriate heuristics due to evolutionary hard-wiring, individual learning, and social processes
  • Evidence of reversal of some of the biases identified by Kahneman and Tversky when the question is presented in a natural frequency format

References

• Key references include Barbey and Sloman (2007), Bergert and Nosofsky (2007), Cosmides and Tooby (1990), Gigerenzer (1991), Gigerenzer and Goldstein (1996), Goldstein (2018), Kahneman (2011), and Lee et al. (2017)

Bayesian Models of Decision-Making

• Bayesian models improve on the 'fast-and-frugal' approach by considering multiple cues and their relationships.
• Bayes' rule is a method for deriving rational answers to problems, involving the mixing of evidence with priors to obtain posteriors.
• The taxi cab problem is a classic example of applying Bayes' rule to derive a rational answer, involving the calculation of posterior probability.

Bayes' Rule: Solving the Taxi Cab Problem

• The problem involves a hit-and-run accident with two cab companies (Green and Blue) and a witness who identified the cab as Blue.
• The witness's reliability is 80% correct and 20% incorrect.
• Bayes' rule is applied to calculate the posterior probability that the taxi was Blue, given the witness's identification.

Applications of Bayes' Rule

• Bayes' rule can be used to derive rational answers to toy problems, alternative to frequentist statistics for data analysis, and to derive rational answers to more complex problems.
• It is a basis for approaching decision-making and human perception as a task that involves applying scientific common sense.

Bayesian Model of Decision-Making in the German Cities Problem

• The German cities problem involves a search-and-stop model, where participants search for cues to make a decision.
• The model takes into account cue validity, cue discriminability, and the order of cue search.
• The results show that the search-and-stop model had the highest level of correspondence with actual participant answers (89.5%).

Other Bayesian Models

• Not all Bayesian models rely on graphical models.
• Some models assume that people simply apply Bayes' rule involving distributions when predicting durations and quantities.
• Other models explain why people adjust insufficiently from provided anchors, especially when they know little about the subject.

Machine Learning

• Machine learning involves the use of algorithms to analyze data and make predictions.
• Ethical issues arise when the data used to train the algorithms is biased, leading to inequality and reinforcement of existing biases.

Metacognition

• Metacognition involves the ability to monitor and control one's own cognitive processes.

• It is a critical component of rational cognitive and memory performance.

• Confidence is not always a valid indicator of accuracy, and metacognition plays a role in evaluating the accuracy of one's own judgments.### The Dunning-Kruger Effect

• The Dunning-Kruger effect is a phenomenon where people who perform poorly in a task tend to overestimate their performance.

• This is because they are unable to recognize their own errors and assume they make fewer errors than they actually do.

Metacognition and the Dunning-Kruger Effect

• Metacognitive ability (monitoring) correlates with task performance.
• Poor performers make more errors about which they can be mistaken (considering them successes) in a task without immediate feedback on errors.
• The Dunning-Kruger effect is not due to differences in metacognitive ability between high and low performers.

Researchers and Their Contributions

• Kruger and Dunning (1999) argued that the lowest performers in a task suffer a dual burden: poor performance and a metacognitive deficit that impedes their ability to distinguish accurate from inaccurate performance.
• McIntosh et al. (2019) demonstrated that the Dunning-Kruger effect is not due to differences in metacognitive ability, but rather due to poorer performers making more errors.

Other Relevant Theories and Concepts

• Prospect Theory (1979) deals with decision-making under uncertainty.
• Heuristics and biases (1974) refer to mental shortcuts that can lead to errors in judgment.
• Bayesian models (1994) are used to describe decision-making processes.
• Machine learning and Bayesian mathematical computing are related fields of study.

Confirmation Bias and Misdiagnosis

• Confirmation bias is the tendency to test hypotheses by asking questions that will yield an affirmative response if the hypothesis is true.
• The bias arises from the application of the positive test heuristic, which involves asking questions that will yield a "yes" response if the hypothesis is true.
• In medical settings, confirmation bias occurs when health professionals:
  • Attend to symptoms that are present without considering symptoms that are absent.
  • Ignore evidence that disagrees with a leading diagnosis.
• This bias can result in misdiagnosis, overtreatment, and unnecessary use of diagnostic tests.

Examples of Confirmation Bias

• A study by Christensen-Szalanski and Bushyhead (1983) found that physicians' predictions about the probability of pneumonia were sensitive to the presence of symptoms but not the absence of symptoms.
• A case study by Mohmed et al. (2021) showed that a patient's diagnosis was overlooked due to confirmation bias, leading to unnecessary treatment and discomfort.

Reducing Confirmation Bias

• Systems promoting awareness of the bias, tool adoption, and a team-based approach can help reduce confirmation bias.
• Tools supporting reflection, such as specially designed forms, can help individuals and teams to:
  • Identify clinical findings that favour or contradict the initial diagnosis.
  • Consider alternative diagnoses and their corresponding findings.
• A "fresh pair of eyes" can be instrumental in revising the diagnosis.

Prospect Theory and Health Promotion

• Prospect Theory proposes that people make decisions based on perceived gains and losses rather than absolute values.
• The framing effect, a phenomenon of Prospect Theory, causes people to:
  • Be risk-averse in the gain frame, choosing the more certain outcome.
  • Be risk-seeking in the loss frame, choosing the more uncertain outcome.
• The framing effect reverses for small probabilities.

Applications of Prospect Theory in Health Promotion

• Health messages can be framed in a gain or loss frame to promote preventative or detection behaviours.
• A study by Gray and Harrington (2010) found that participants rated messages in a gain frame to be more effective, and reported greater intentions to exercise.
• However, the effects of outcome uncertainty and outcome severity are still debated, and there is no scientific consensus on the best frame for health advertising.

Fast and Frugal Heuristics

• Gigerenzer's fast and frugal heuristics propose that people select cues and heuristics that are most appropriate in various environments.

• Heuristics identified by Gigerenzer and his colleagues can be used to present health-related information in a more effective way.### Heuristics and Biases

• The ability to select appropriate cues is shaped by evolutionary hard-wiring, individual learning, and social processes such as imitation and explicit teaching.

• Heuristics include the Take-the-Best heuristic, the Recognition heuristic, the Fluency heuristic, and Tallying.

The Fast-and-Frugal Model

• Gigerenzer provided evidence that people are highly attuned to their environments, demonstrating reversals of biases when questions are asked in a natural frequency format.
• Research by Hoffrage and Gigerenzer (1998) showed that the natural frequency format is more effective for communicating with doctors about the predictive value of diagnostic tests.
• The fast-and-frugal approach has had a widespread impact on information presentation, influencing the design of infographics relating to healthcare and other topics.

Applying Heuristics

• The Take-the-Best heuristic involves following a decision tree.
• Green and Mehr (1997) developed a decision aid for emergency room doctors, applying the decision tree to allocate patients to a specialist heart unit or regular bed.
• Application of the tree reduced the number of patients at risk of heart attack being wrongly sent to a regular bed and reduced the load on the coronary unit.

Bayesian Model of Impulsive Decision-Making

• Bayesian models enable complex processes to be simulated to check whether observed data are well-approximated by assuming that different kinds of decision-makers adopt complex strategies.
• Research by Banca et al. (2015) used Bayesian modeling to demonstrate that individuals who binge drink have difficulty thinking ahead and are more impulsive in their decision-making.

Health Promotion Strategies

• Confirmation bias can contribute to misdiagnosis, and reducing it can be achieved through strategies informed by research on heuristics and biases.
• Prospect Theory can help motivate health-related behavior by understanding how people make decisions under uncertainty.
• The fast-and-frugal heuristics model has shaped the way health-related information is presented to practitioners and the public.
• The Bayesian approach has improved our understanding of impulsive decision-making in people with addictions.

Applications of Decision-Making Theories

• The lecture focuses on applications of decision-making theories in other domains, including the legal system, gambling, and health-related information.

The Post-Identification Feedback Effect

• The post-identification feedback effect is a phenomenon where confirming feedback after an identification inflates witnesses' recollections of their confidence at the time of the identification.
• This effect is particularly harmful when identifications are false, as it produces confident but inaccurate eyewitness evidence against an innocent suspect.
• A real-life example of this effect is the case of Ronald Cotton, who was wrongly convicted of a crime due to post-identification feedback.
• Laboratory evidence from two meta-analyses shows that participants who receive confirming feedback report higher confidence in their identification.
• The post-identification feedback effect occurs due to selective cue integration, which is a form of confirmation bias.
• To avoid this effect, system variables can be put in place, such as conducting the identification procedure out of earshot and view of others, and using a double-blind administrator.

The Gambler's Fallacy

• The gambler's fallacy is traditionally considered a consequence of applying the representativeness heuristic, but it is best explained as the result of people having limited short-term memory spans.
• Memory limitations are universal, and truly random sequences are encountered in many parts of the world, making the gambler's fallacy a universal guiding belief for other gambling errors.
• Research suggests that educating people with gambling disorder about the trap of their own short-term memory can help address motivations to gamble based on the gambler's fallacy.
• Supernatural beliefs around the world have been identified, including the belief in the power of luck and the cyclical nature of luck.

Presenting Health-Related Information the Fast-and-Frugal Way

• A Bayesian model of impulsive decision-making can be used to present health-related information in a fast-and-frugal way.
• The model takes into account the cognitive limitations of decision-makers and provides a framework for making quick and accurate decisions.

Lecture Summary

• The lecture has explored applications of decision-making theories, including the post-identification feedback effect and the gambler's fallacy.
• These theories have been translated into concrete guidelines for obtaining eyewitness identification statements and educational messages for cognitive behavioural therapy for gambling disorder.

Applications of Decision-Making Theories

• The lecture focuses on applications of decision-making theories in other domains, including the legal system, gambling, and health-related information.

The Post-Identification Feedback Effect

• The post-identification feedback effect is a phenomenon where confirming feedback after an identification inflates witnesses' recollections of their confidence at the time of the identification.
• This effect is particularly harmful when identifications are false, as it produces confident but inaccurate eyewitness evidence against an innocent suspect.
• A real-life example of this effect is the case of Ronald Cotton, who was wrongly convicted of a crime due to post-identification feedback.
• Laboratory evidence from two meta-analyses shows that participants who receive confirming feedback report higher confidence in their identification.
• The post-identification feedback effect occurs due to selective cue integration, which is a form of confirmation bias.
• To avoid this effect, system variables can be put in place, such as conducting the identification procedure out of earshot and view of others, and using a double-blind administrator.

The Gambler's Fallacy

• The gambler's fallacy is traditionally considered a consequence of applying the representativeness heuristic, but it is best explained as the result of people having limited short-term memory spans.
• Memory limitations are universal, and truly random sequences are encountered in many parts of the world, making the gambler's fallacy a universal guiding belief for other gambling errors.
• Research suggests that educating people with gambling disorder about the trap of their own short-term memory can help address motivations to gamble based on the gambler's fallacy.
• Supernatural beliefs around the world have been identified, including the belief in the power of luck and the cyclical nature of luck.

Presenting Health-Related Information the Fast-and-Frugal Way

• A Bayesian model of impulsive decision-making can be used to present health-related information in a fast-and-frugal way.
• The model takes into account the cognitive limitations of decision-makers and provides a framework for making quick and accurate decisions.

Lecture Summary

• The lecture has explored applications of decision-making theories, including the post-identification feedback effect and the gambler's fallacy.
• These theories have been translated into concrete guidelines for obtaining eyewitness identification statements and educational messages for cognitive behavioural therapy for gambling disorder.