Heuristics and Biases

Question 1

What are heuristics?

Answer 1

Rule of thumb
Heuristics produce systematic errors (biases)
Take shortcuts

Question 2

Why do heuristics exist?

Answer 2

Perfect decisions are impossible
Imperfect information (limited and ambiguous)
Limited resources (time and cognitive resources)

Question 3

What are the design features of a good heuristic?

Answer 3

Efficient
Quick
Robust
Mostly right = produce a net benefit overall

Question 4

Outline the link between heuristics and biases

Answer 4

Heuristics produce systematic errors/biases because we adopt a shortcut so there are patterns in the errors we make.
By observing systematic errors we can try and understand the strategies people use to make choices

Question 5

Identify 3 heuristics which are responsible for a range of biases and errors

Answer 5

1. representativeness
2. availability
3. adjustment and anchoring

Question 6

Outline and explain the heuristic of representativeness

Answer 6

Comparing a single event to how we imagine that event typically looks like
Evaluating probabilities according to the match between an event and the features of its parent population or generating process

Question 7

Describe an example of representativeness using base rates

Answer 7

A test for a dreaded disease is 99.9% accurate
The DD is rare so only 1 in a million people have the disease. Get tested and it comes back positive, is it likely that you have the disease?
1 in 1,000,000 is the base rate
Unlikely you have the disease because there will be 1000 false alarms for every 1,000,000 tested due to the test accuracy.

Question 8

Outline and explain the heuristic of availability

Answer 8

Probabilities are assessed by the ease with which instances come to mind.
E.g. male celebrities and female non-celebrities names in a list, more likely to say there were more male names
Example of a systematic error

Question 9

Outline and explain the heuristic of adjustment and anchoring

Answer 9

People will begin with an initial estimate and then adjust.
Irrelevant information can bias the initial estimate
Adjustments can be insufficient
E.g. 1x2x3x4x5x6x7x8x9 and 9x8x7x6x5x4x3x2x1 even though are exactly the same produce significantly different estimates.

Question 10

Describe susceptibility to framing effects

Answer 10

Gamble to avoid losses

The way the choice is framed skews what seems like a reasonable option

Question 11

Outline Prospect Theory

Answer 11

Maps the objective outcome to subjective value
Kahneman and Tversky (1979) noted that the biases in people’s decision making often resulted from the fact that we do not treat gains and losses equally. Generally, we treat losses as more important than gains (loss aversion).
Losing £100 is much more impactful than gaining £100.
The framework also assumes diminishing returns; a gain or loss of £100 matters a lot if we have a balance of £1000 but it matters very little if we have a balance of £100,000.

Question 12

Explain what is meant by heuristics exploit environmental structure

Answer 12

If something is consistently associated with a good option, you can use it as a shortcut
E.g. take advice from someone who is usually right, pick the special in a restaurant
This means that even if a heuristic gives the wrong answer it may exist for a good reason

Question 13

Outline Optimal vs Adaptive decision making

Answer 13

Is the optimal strategy for a thinking machine to get the logically correct answer 100% of the time no matter how long it takes?
OR
To get an answer quickly which is usually right? (Adaptive)