Session 4: Decision Making

Definitions

• Judgement: Defined as "the ways in which individuals make use of various cues (which may be ambiguous) to draw inferences about situations and events" (Eysenck & Keane, 2010, p. 499).

• Decision-Making: Defined as "how people choose what action to take to achieve labile [changeable] sometimes conflicting goals in an uncertain world" (Hastie, 2001, p. 657).

Judgement Theories

Bayes’ Theorem and Base Rate Neglect

• Core Idea of Bayes' Theorem: The formula presented is $P(H|D) = rac{P(D|H) imes P(H)}{P(D)}$, where:

  • $H$ = Hypothesis

  • $D$ = Data/Evidence

• This theorem is instrumental in calculating the probability of a hypothesis based on observed data. The significance of Bayes' Theorem in judgement is that it integrates prior probabilities (base rates) with new evidence. While ideal reasoning should incorporate both elements, human beings often neglect base rates, leading to outcomes that reflect Base Rate Neglect.

Base Rate Neglect and the Taxicab Problem

• This concept was examined by Kahneman & Tversky (1972) through a thought experiment involving an eye-witness of a hit-and-run accident. In the scenario, a taxi cab involved is described by the witness as a blue cab, yet the statistics reveal that only 15% of cabs are blue and 85% are green. The witness can identify blue cabs accurately 80% of the time.

• The underlying probabilities involved:

  • Prior odds: $P(H<em>a) = 0.15$ (probability that the cab is blue) and $P(H</em>b) = 0.85$ (probability that the cab is green)

  • Data given the hypothesis: $P(D|H<em>a) = 0.8$ (probability cab is blue when identified as blue) and $P(D|H</em>b) = 0.2$ (probability cab is green when identified as blue)

• Calculating the likelihood that the witness correctly identified the cab as blue yields:

  • Witness's identification of blue cabs (15 in every 100) would yield a correct identification of 12 cabs, while the incorrect identification from green cabs (85 in every 100) would yield 17. The total probability of the witness being correct is:
    $P(correct) = rac{12}{12+17} = 0.41$, or 41%.

Heuristics and Their Influence

• Definition of Heuristics: Heuristics are described as "rules of thumb that are cognitively undemanding and often approximate accurate answers" (Eysenck & Keane, 2010, p. 634).

• Reasons for reliance on heuristics include a preference for mental shortcuts over statistical or probabilistic reasoning. Common misconceptions include Base Rate Neglect, Conjunction Fallacy, and the Availability Heuristic.

Representativeness Heuristic

• In the scenario proposed by Kahneman & Tversky (1973), the characteristics of a 45-year-old man named Jack are provided, detailing his conservative and non-political lifestyle. Participants are asked whether Jack is more likely a lawyer or an engineer. The statistical breakdown reveals that 70 descriptions depict lawyers while 30 depict engineers. This leads participants to often incorrectly judge Jack as a lawyer based on representativeness rather than statistical likelihood.

• Representativeness Heuristic Definition: "Events that are representative or typical of a class are assigned a higher probability of occurrence" (Kellogg, 1995, p. 385).

Conjunction Fallacy

• In the Linda Problem proposed by Tversky & Kahneman (1983), Linda is depicted as a conscious, outspoken, and socially active individual. Participants are asked to determine her likely occupation among several choices, with many believing it more probable for Linda to be a feminist & a bank teller than simply a bank teller.

• Conjunction Fallacy Definition: "The mistaken assumption that the probability of a conjunction of two events (A & B) is greater than the probability of one of them (A or B)" (Eysenck & Keane, 2015, p. 549).

Availability Heuristic

• Inquiry into perceptions surrounding the probability of death by murder versus suicide (as investigated by Lichtenstein et al., 1978) revealed that many participants deemed murder more probable than suicide.

• Availability Heuristic Definition: "The rule of thumb that the frequencies of events can be estimated accurately by the subjective ease with which they can be retrieved" (Eysenck & Keane, 2015, p. 552).

Evaluation of Heuristics

Support for Heuristic Approaches

• Numerous research studies substantiate the considerable influence of heuristics in various contexts, highlighting the associated tendency for errors—even among experts. The heuristic approach has significant implications across psychology, economics, philosophy, and politics. Additionally, heuristics alleviate cognitive load (Kool et al., 2010; Fiedler & von Sydow, 2015).

Limitations of Heuristic Approaches

• The broad usage of the term 'heuristic' may lead to a dilution of its meaning. Many participants may not understand the problem posed rather than exhibit judgement errors. For instance, multiple individuals may interpret that saying "Linda is a bank teller" suggests a lack of feminist engagement.

• Additionally, the availability heuristic may be better explained by media influence rather than faulty cognition. Accurate judgments cannot always be captured by heuristic explanations, and much research is hindered by artificial conditions lacking ecological validity.

Fast & Frugal Heuristics

• Proposed by Gigerencer and colleagues, the Fast and Frugal Heuristics model suggests evolutionary adaptations enable humans to utilize straightforward and effective strategies for rapid, yet accurate, judgements. This includes an adaptive toolbox consisting of various heuristics, including the Take-the-best strategy.

Take-the-Best Strategy

• An example utilized involves determining the city with the larger population between Herne and Cologne by applying the take-the-best strategy. Key elements include:

  • Search rule: identification of cue (name recognition, cathedral presence, etc.)

  • Stopping rule: termination of search upon finding an indicative trait

  • Decision rule: selection of the identified outcome.

• This concept was further assessed through tasks involving pairs of cities. Research conducted by Goldstein & Gigerenzer (2002) demonstrates participants achieving higher scores with recognition heuristics compared to familiar cities.

Evaluation of Fast and Frugal Heuristics

Support

• Studies affirm the existence of fast and frugal heuristics, suggesting they enable individuals to arrive at prompt judgements even under cognitive or time constraints.

Limitations

• Contrarily, it has also been observed that heuristics may not be employed as extensively as hypothesized (Newell et al., 2003; Oppenheimer, 2003). Additionally, heuristics may exhibit complexities overlooked in foundational descriptions, as the take-the-best strategy, for example, entails hierarchical cue organization. Heuristics do not elucidate complex decision-making scenarios, such as relational choices (e.g., marital selection) and may lack predictive and explanatory validity.

Dual-Process Theory

• Dual-Process Theory, articulated by Kahneman (2003), posits that while individuals predominantly rely on heuristics in judgement, they occasionally resort to more sophisticated analytical processes.

• The two systems are delineated as follows:

  • System 1: Fast, intuitive, and often heuristic-based thinking.

  • System 2: Slow, deliberate, analytical, and rule-based thinking.

Evidence for Dual-Process Theory

• De Neys (2006) provided empirical support through experiments indicating such a duality. In one experiment, participants who took longer to resolve conjunction-fallacy problems were more likely engaging System 2 cognition, as evidenced by a correlation between answer accuracy and response time.

• Additionally, a central executive task where participants tapped a sequence had a noticeable reduction in accuracy on the conjunction-fallacy tasks, dropping from 17% correct with problems alone to 9.5% with tapping included.

Evaluation of Dual-Process Theory

Strengths

• The Dual-Process Theory encompasses both intuitive and analytical cognitive functions, reinforced by evidence from reaction time studies and cognitive load experiments (e.g., De Neys, 2006). It offers insights into biases and decision-making errors, illuminating thresholds for when deliberate reasoning may be required.

• Moreover, it integrates findings from heuristics research and cognitive load theory.

Limitations

• Despite its contributions, the dichotomy between System 1 and System 2 theorizes a distinction that may overlook intermediate or hybrid processing mechanisms. Moreover, it fails to explicate how or when System 2 overrides System 1 and does not adequately accommodate individual differences in personality, expertise, or cultural influences. Lastly, it may present challenges in empirically differentiating these systems during real-world cognitive tasks and lacks predictive and explanatory power.

Decision-Making

How Do We Make Decisions?

• Two prominent theories underpin decision-making frameworks: Utility Theory and Prospect Theory.

Utility Theory

• Utility theory asserts that individuals endeavor to make choices aiming to maximize expected utility (subjective value) rather than only monetary considerations.

• The underlying principles include:

  • Expected Utility: Calculated as the product of Probability and Subjective Value.

  • Individuals assess outcomes based on their utility rather than their absolute worth, which helps elucidate preferences for smaller secure gains versus larger uncertain ones.

• An example is presented with two options:

  • Option A: £100 guaranteed

  • Option B: A 50% chance of £250
    The explanation posits that individuals tend to favour options that lead to assured gains.

Prospect Theory (Kahneman & Tversky, 1979, 1984)

• Central to Prospect Theory is the observation of a reference point, where the value function reflects an asymmetrical perception of gains and losses. Gains increase gradually, meaning that winnings do not represent proportional subjective value increases (e.g., winning £200 does not double the perceived value over £100). Conversely, losses escalate rapidly as they increase, indicating a more significant cognitive impact.

Sunk-Cost Effect and Loss Aversion

• Definition of Sunk-Cost Effect: This phenomenon describes the inclination of individuals to persist in investments or commitments despite irrationality due to previously incurred costs. This effect is intertwined with Loss Aversion, which indicates a psychological tendency to prevent losses more so than to pursue equivalent gains, ultimately influencing decisions related to sunk costs.

Evaluation of Prospect Theory

Support

• Prospect Theory accounts for the psychological underpinnings governing decision-making mechanisms, distinguishing itself from normative theories by describing a value function where more weight is allocated to losses than gains, and thus interpreting behavioral phenomena accurately.

Bounded Rationality and Satisficing (Simon, 1957)

• Unbounded Rationality: This perspective proposes that all relevant information is readily available for the decision-maker, enabling optimization for the best choices.

• Bounded Rationality: This concept posits that decision-makers devise workable answers to issues within their restricted processing capacities; thus, they utilize heuristics (satisficing) as simplified decision strategies.

• Definition of Satisficing: "The selection of the first choice meeting certain minimum requirements; formed from 'satisfactory' and 'sufficing'" (Eysenck & Keane, 2010, p. 527).

• An example is illustrated in car purchasing, where a consumer selects the first option that meets specific basic criteria rather than seeking out the perfect vehicle. This reflects the reality of decision-making, emphasizing that perfection often yields to satisfactory, functional solutions that facilitate progress.