Lecture on Models of Decision-Making
Course Overview
- Last two chapters of the course:
- Monday: Models of decision-making
- Wednesday: Evaluating decisions
Evaluating Scientific Hypotheses
- Learned to evaluate scientific episodes to support hypotheses (causal, statistical, etc.)
- Strong evidence suggests new information, but does not prescribe behavior
Decision-Making Framework
- Decision Theory: Framework for applying scientific information to decisions
- Related to Game Theory
- Uses games as models due to similar real-life decision factors
Components of Decision-Making Model
- Options:
- Must have at least two options
- Example: Betting on marble colors
- Options should be mutually exclusive and exhaustive
- States of the World:
- Possible outcomes (e.g., marble color drawn)
- Must be separate from options and also exclusive and exhaustive
- Outcomes:
- Combination of options and states of the world
- Matrix of possible outcomes
- Need to assign values to outcomes
Assigning Values to Outcomes
- Ordinal Ranking: Simple ranking from most to least preferred
- Numerical Values: Assign utilities or monetary value
Decision Strategies
- Certainty:
- Know exactly the state of the world
- Choose option with highest valued outcome
- Complete Uncertainty:
- No information about the state of the world
- Use matrix to determine better or worse options
- General rules:
- Eliminate worse options
- Choose best option if available
- Satisfactory Options:
- Set a satisfaction level (minimum acceptable outcome)
- Choose options meeting satisfaction level
- Play it Safe Strategy:
- Choose option with greatest security level (minimize losses)
Decision-Making with Risk
- Have some information about probabilities of states of the world
- Use Expected Value (EV) to guide decisions
- EV is the weighted sum of outcomes’ values by their probabilities
- Expected Value Strategy:
- Choose option with the greatest expected value
Examples and Applications
- Ramsey’s Method: Assign utilities to decision problems
- Use ethically neutral propositions for assigning value
- Convert ranked outcomes into numerical values
Example Case: Kurt’s College Decision
- Rank preferences for college and playing football
- Assign utility values using Ramsey’s method
- Calculate expected utility to decide on college
Conclusion
- Different strategies for decision-making depend on certainty/uncertainty and personal values
- Importance of using scientific and probabilistic information in decision-making
- Encouragement to review materials and seek further clarification if needed
These notes provide a summary of the decision-making models and theories discussed in the lecture. For more detailed understanding, revisiting the textbook figures and additional resources is recommended.