Decision making & forecasting (Week 2) Flashcards
What are the steps in the decision-making process?
- Identifying problems & objectives
- Identifying alternatives
- Modeling the problem structure
- Choosing
- Implementing & learning
What are the steps in consumer decision-making?
- Need cognition & problem awareness
- Information search
- Evaluation of alternatives
- Purchase
- Post-purchase evaluation
What are the initial steps of rational decision-making?
- Identify the problem
- Seek multiple perspectives; question your assumptions - Understand your objectives
- Ask yourself “Why is this important to me?” - Generate alternatives that satisfy those objectives
- Ask yourself “How can I achieve my objective?”
What are some common mistakes in problem structuring?
- Wrong interpretations of problem
- Low motivation to cope with problem
- Inattention to problem
What is omission bias?
People judge outcomes more harshly if they derive from action rather than failure to act (inaction)
=/= Risk aversion
What are some advice for problem structuring?
- Actively seek out decision opportunities
- Make sure that you are asking the right questions and using an appropriate context
- Actively question what values you are trying to serve; generate alternatives that satisfy those values
- Evaluate the process and outcomes of your decisions
What is the main lesson from single variable regression toward the mean?
When 2 variables are imperfectly correlated, your prediction of the second (in std. units) should be less extreme than the first (in std. units)
Measured score = True score + Error
What are the perspectives of single variable forecasts?
Normative perspective: Regression toward the mean
Descriptive perspective: Prediction by evaluation
- Using their evaluation as their prediction
- Sensitive to predictive validity
- Make predictions based on their first impression –> invent false explanations
What is the main lesson from regression toward the mean (an intuitive approach)?
When you make a prediction of one variable based on another variable, stop an consider the correlation between variables
The lower the correlation, the more regressive the prediction you should make (i.e. closer to mean)
What are some intuitive errors in forecasting?
1) Illusion of validity
- We assign too much weight to salient, vivid info, especially if it is representative of the outcome
- We overweight causal variables
2) Noisy, inconsistent
- Prediction is not generated in a systematic way
What are proper linear models?
- Translate each variable into a standard (z) score
- Weigh each variable to minimise error variance
- Predict using these optimal weights
Why do we not use proper linear models always?
1) Too few observations per predictor to calculate optimal weights
2) No opportunity to measure outcome variables
What is conditional probability?
= The probability of an event (A), given that another (B) has already occurred
P(A|B) = P(AnB) / P(B)
What are the types of positive illusions?
People tend to be overly optimistic about:
- Accuracy of their forecasts/prediction, i.e. overconfidence
- Their attribute, abilities, & motives, i.e. self-enhancement
- Their ability to complete plans in a timely manner, i.e. planning fallacy
- Their ability to control outcomes, i.e. illusion of control
What are some remedies for positive illusions?
- Overcome MEMORY limitations by facilitating learning
- Keep records of your past performance
- Adjust present estimates based on past tendencies - Overcome MOTIVATIONAL biases by seeking disconfirmation
- Play devil’s advocate
- Solicit opinions of more objective outcomes - Overcome EGOCENTRIC biases by taking an “outside view”
- Select a reference class, assess distribution of outcome
- Assess your position in the distribution and adjust cautiously
