Making tradeoffs & valuing outcomes (Week 9) Flashcards
What is the decision matrix?
- Attributes (conflict)
- Alternatives
- Outcomes
What is the Multi-Attribute Utility Theory (MAUT) / Weighted Additive Rule (WADD)?
- Choose relevant ATTRIBUTES & ALTERNATIVES; arrange them in matrix
- Assign a “UTILITY” score to each attribute (i.e. satisf. level, not actual attribute) for each alternative
- Assign a WEIGHT to each attribute
- Calculate the SUM OF WEIGHT*UTILITY for each alt.
- Choose the alt. with the HIGHEST SCORE
What are the types of choice heuristics?
- Satisficing (SAT)
- Lexicographic (LEX)
- Elimination-by-aspects (EBA)
- Majority of confirming dimensions (MCD)
- Frequency of good & bad features (FRQ)
- Equal weight (EQW)
What is satisficing (SAT)?
- Set a MINIMUM THRESHOLD for each attribute e.g. 1
- Consider options in the order they came
- Choose the first item that EXCEEDS he threshold for ALL the attributes
What are some evaluations about satisficing?
- Simplified method, DOES NOT DIFFERENTIATE between attributes (assumed to be all equal), less info
- For making quick decisions
- Tradeoff: Might miss a better option
What is the lexicographic (LEX) choice heuristic?
- Choose the alt. that MAXIMISES the MOST IMPT attrib.
- Break ties by looking at next most impt attrib (rest of options already excluded)
- Repeat #2 as needed
What is the weakness of the lexicographic (LEX) choice heuristic?
Ignores info about less impt attributes
What is the elimination-by-aspects (EBA) choice heuristic?
- Choose a MINIMUM THRESHOLD for each attribute
- ELIMINATE the alt. below the cutoff for the MOST IMPT ATTRIBUTE
- Repeat for the next most impt attrib, until 1 alt remain
What are some evaluations about elimination-by-aspects (EBA)?
- Combination of SAT & LEX
- Assumes that there are big diffs. among all attributes
- Removal of options
- If unable to eliminate/no options remaining: Change decision rule, min. req., or increase no. of attributes
What is the majority of confirming dimensions (MCD) choice heuristic?
- For a given PAIR of alternatives, choose the one that has the HIGHEST NO. OF “WINNING” ATTRIBUTES
What are some evaluations about majority confirming dimensions (MCD)?
- To make PAIR COMPARISON (only 2 options!)
- Ignore differences b/w the 2 scores
What is the frequency of good & bad features (FRQ) choice heuristic?
- Determine the REFERENCE POINT for each attribute that separates “good” from “bad” (e.g. 2.5)
- Choose the alt. that maximises “good” or minimises “bad” (greater no. of pluses and fewer minuses)
What are some evaluations of the frequency of good & bad features (FRQ) choice heuristic?
- Extension of MCD
- If tied number of pluses: Increase no. of attributes. But note that new attributes have to be of RELATIVELY EQUAL IMPTANCE
What is the equal weight (EQW) choice heuristic?
- Similar to WADD, but assigns EQUAL WEIGHT TO ALL ATTRIBUTES
- Sum up all outcomes to generate final score (no need to multiply by weights like in WADD)
What are some notes about choice heuristics?
- There is a TRADEOFF between the EFFORT required and the expected ACCURACY of a decision rule
- Contingencies for choosing decision rule
