Cognitive Perspectives on Judgment and Decision Making

The Nature of Decision Making

  • Introduction to Decision Making: Decision making involves determining how to behave or act based on feelings, associations, and cognitive evaluations. It is distinguishably different from judgment, although the two are closely linked.

  • Implicit and Explicit Factors:

    • Implicit Bias: We may not always be aware of why we decide things due to positive or negative associations between concepts that are not consciously recognized.

    • Attribution of Values: Decisions require assigning values to different options. This can be explicit (e.g., creating a pro and con list) or implicit (e.g., gut feelings about social interactions).

  • Social Interaction Example: A common decision-making scenario involves social reach-out (e.g., "Should I call or text someone I am interested in?"). Individuals weigh prospective outcomes:

    • Pro: Reaching out might lead to a connection.

    • Con: Reaching out might make the person look "desperate."

    • Alternative Con: Not calling might result in appearing as if they are ignoring the other person.

Prescriptive vs. Descriptive Theories

  • Prescriptive Theories: These theories establish how humans should decide to achieve the most rational or optimal outcome.

    • Based on economical or mathematical models.

    • Focused on achieving the "best" possible result.

    • Example: An objective friend suggesting you wait several days to call someone back to ensure the decision is conservative and calculated rather than impulsive.

  • Descriptive Theories: These theories outline how humans actually behave and decide.

    • Often shows how human behavior departs from prescriptive, rational choices.

    • Decisions are frequently influenced by internal states (e.g., excitement) or external stimuli (e.g., watching a movie like Sleepless in Seattle and impulsively making a phone call).

    • It highlights the competition between endogenous (internal goals) and exogenous (external triggers) factors.

Decision Trees and Value Assignment

  • Decision Trees: This is the most straightforward depiction of the decision-making process, often involving "what-if" scenarios and cost-benefit analyses.

  • Example: Roger Federer and Tennis: A decision tree for whether to play tennis based on external factors:

    • Weather Outlook: Options include Sunny, Overcast, or Rainy.

    • Rainy: One might immediately decide "No," although some models check if it is windy before deciding "No."

    • Overcast: Usually leads to a "Yes."

    • Sunny: Leads to a secondary check for Humidity. High humidity might lead to a "No," while lower humidity leads to a "Yes."

  • Subjectivity in Values: Value assignment is relatively arbitrary and based on individual priorities.

    • Social Value: One person may value eating out highly due to being social; another may assign it low value because they dislike public settings.

    • Universal Values: High agreement generally exists on the value of basic needs like food, water, and shelter.

    • Contested Values: Significant disagreement exists regarding the value of complex societal issues like taxes, healthcare, or recycling.

Rationality Principles and Framing Effects

  • Principles of Rational Decision Making: There are two key principles that humans consistently ignore:

    • Transitivity: If the same relation holds between option 11 and 22, and between option 22 and 33, it must also hold between option 11 and 33. Written as: If A > B and B > C, then A > C.

    • Framing Invariance: Two ways of asking the same question should result in the same answer.

  • The Tversky & Kahneman Study (1981): This study demonstrated how framing changes decisions even when mathematical outcomes are identical.

    • Scenario A (Positive Framing):

      • Option 1: 200200 people will be saved.

      • Option 2: A 13\frac{1}{3} probability that 600600 will be saved and a 23\frac{2}{3} probability that no one will be saved.

      • Result: Most people choose Option 1 because they prefer the certainty of seeing lives saved.

    • Scenario B (Negative Framing):

      • Option 1: 400400 people will die.

      • Option 2: A 13\frac{1}{3} probability that no one will die and a 23\frac{2}{3} probability that 600600 will die.

      • Result: Most people choose Option 2. Even though the math is the same as Scenario A (400400 dead means 200200 live), the image of "400 dead" is unpalatable, leading to risk-seeking behavior.

Prospect Theory and Bounded Rationality

  • Prospect Theory: Actions are determined by mental representations of situations rather than the situations themselves. The way a situation is "framed" dictates the choice.

    • Example: Convincing parents to pay for a concert.

    • Positive Representation: "This will be a lifelong memory/flashbulb memory."

    • Negative Representation: "I want to blow $400\$400."

  • Bounded Rationality: Proposed by Herbert Simon (Nobel Prize winner), this theory suggests people are as rational as their cognitive limitations allow.

    • The Working Memory Platter: Humans can only process a finite amount of information at once. Decisions are limited by the capacity of this "processing platter."

    • Example: Environmental Consciousness: A person may care about trash in the ocean (high value) but failed to recycle everything today or chose to buy a bag of SmartPop popcorn (creating non-biodegradable trash) instead of kernels. This happens because the logistical details of perfect recycling may not fit on the working memory platter at that moment.

Judgment and Heuristics

  • Nature of Judgment: Judgment is the process of thinking, "I believe this will be good/bad for me," based on information available.

  • Heuristics: These are mental "rules of thumb"—efficient strategies that usually lead to good outcomes.

    • Example: Using a side entrance of a parking lot to find a spot faster because fewer people use it.

  • Key Terms in Judgment:

    • Frequency Estimate: An informal assessment of how often something happens (e.g., saying something "usually" happens).

    • Base Rate: The actual, objective frequency at which something occurs (e.g., planes flying overhead exactly once every 20minutes20\,\text{minutes}).

    • Attribute Substitution: Relying on easily assessed information already in one's head rather than seeking out actual data (e.g., estimating flight frequency based on what you see out the window rather than checking a flight database).

    • Representativeness Heuristic: Assuming all instances of a category will resemble the prototype (e.g., dog, swimming pool). This relies on characteristic features as defining principles (as seen in the "Justice/Judge" association with Ruth Bader Ginsburg).

The Good Judgment Project (NPR Case Study)

  • Overview: An experiment funded by the intelligence community involving 3,0003,000 average citizens making probability estimates on geopolitical events.

  • The Experimenters: Psychologists Philip Tetlock, Barbara Mellors, and Don Moore.

  • Super Forecasters: Individuals in the top 1%1\% of accuracy. They are often teammates who outperform intelligence officers with access to classified information by approximately 30%30\%.

  • The Participant - Elaine Rich: A pharmacist from suburban Maryland with no professional background in international affairs.

    • Forecasting Process: She uses basic Google searches and Wikipedia rather than classified docs.

    • Questions Asked:

      • Will North Korea launch a new multistage missile before 05/10/201405/10/2014?

      • Will the government of Syria and the Syrian Supreme Military Command announce a ceasefire?

      • Will Russian armed forces enter Kharkiv in Ukraine by May 10?

    • Anonymity: Rich notes that her anonymity gives her the freedom to make true, unbiased forecasts without a professional reputation at stake.

The Wisdom of Crowds (Scientific Basis)

  • Origins: Concept first discovered by British statistician Francis Galton in 19061906.

  • The Ox Experiment: At a fair, 800800 people guessed the weight of a dead ox.

    • Individual Guesses: Most were very poor (too high or too low).

    • Crowd Average: The average guess was 1197pounds1197\,\text{pounds}. The actual weight was 1198pounds1198\,\text{pounds}.

  • Mechanism: Prediction involves a "true signal" surrounded by "noise" (statistical random variation). When many predictions are pooled, the random variations on either side of the signal cancel each other out, leaving the true signal.

  • Application: Jason Matheny from the intelligence community notes that this project has significantly improved the accuracy of geopolitical forecasts. While it may not replace traditional intelligence, it serves as a powerful complement to existing methods.