Week 6 Lec 11: Heuristics

Decision makers have limited capacity for processing information Depending on properties of the task environment and mental strategy, preferences will vary

Choice among options with many features/attributes

By taking the subjective value of an option as the weighted sum of choice relevant attribute values

Multi-attribute utility theory/technique

Structured methodology designed to handle the trade offs among multiple objectives Considers all relevant attributes of all decent alternates Uses a weighted linear model to calculate evaluations Result: Rank ordered evaluation of alternatives that reflect DM’s preferences

Public sector and policy issues, privatization of healthcare, college sports teams Power plant related decisions Military development of new weapons Ranking of college sports teams, degree programs

Step 1: Choose a set of relevant attributes Step 2: Assemble or create a set of alternatives Step 3: Score each alternative on each attribute

Step 1: Rank the attributes by its importance Step 2: Create raw weights Step 3: Standardize the weights by dividing each one by total weight

Multiply standardized weight by single attribute utility functions Choose alternative with highest evaluation

Yes, because it should lead to optimal choice

When there is time pressure/other constraints When it makes you pay too much attention to unimportant attributes For when we have trouble identifying important attributes

Adds attribute values regardless of importance (ignores attribute weights)

Model in which DM has pre-specified ordering of the importance of different attributes If there is a clear winner on first attribute, you stop Leads to extreme choices

Most people search sequentially and stop when a satisficing level of reservation utility is reached

• Set cutoffs on all important attributes • Consider alternatives one at a time • Pick the first alternative that exceeds all cutoffs

Alternative based strategy

Search theoretic choice experiment to study the impact of incomplete consideration on the quality of choices Asked how long do people search for the best option before they make a choice Results: Best option found less often in more complex situations, even when there is no time pressure

The minimum level of utility that must be guaranteed to make a contract acceptable

• Pick the most important attribute and set a cutoff • Pick the next most important attribute and set a second cutoff • Continue through all attributes till only one alternatives remain

Attribute based (consider all alternatives one attribute at a time)

Solves the problem of choosing extreme options

May lead to poor decisions Fails to ensure alternatives chosen are superior to those eliminated

Satisficing and EBA

May not choose best alternative Choice may be affected by irrelevant factors such as search order