Representativeness and Availability Kahneman

1
Representativeness and AvailabilityKahneman
Tversky

• Umut Öztürk

2
RepresentativenessDefinition

• - Assessing the likelihood of an events
  occurrence by the similarity of that occurrence
  to stereotypes of similar occurrences.
• - The more X is similar to Y, the more likely we
  think X belongs to Y.

3
Insensitivity to Sample Size

• The size of a sample greatly affects the
  likelihood of obtaining certain results in it.
• People, however, often ignore sample size and
  only use the superficial similarity measures.
• For example, people ignore the fact that larger
  samples are less likely to deviate from the mean
  rather than smaller samples.

4
Misconception of Chance

• People expect that random sequences are
  representative even in small samples.
• - E.g. they consider a coin-toss run of HTHTHT to
  be more likely than HHHTTT or HHHHTH
• Gamblers fallacy A deviation from a stable
  equilibrium generates a force that restores the
  equilibrium. (misconception of the fairness of
  the laws of chance)
• The laws of chance Deviations are not canceled
  as sampling proceeds, they are only diluted.

5
Misconception of Chance

• E.g. After a run of reds in a roulette, black
  will make the overall run more representative
  (self correcting process?)
• Even experienced research psychologists believe
  in a law of small numbers (small samples are
  representative of the population they are drawn
  from)

6
Example on Gamblers Fallacy

• The mean IQ of the 8th graders in a city 100
  (known)
• Random sample of 50 students
• The first students IQ150
• Expected mean IQ for the sample?
• Correct Answer 101
• Answer for large number of people100
• Why? Belief in self-correction

7
Insensitivity to Prior Probabilities

• The base(population) rate of outcomes should be a
  major factor in estimating their frequency.
  However, people often ignore it.
• Bayes Theorem When we make a decision, we should
  take the prior probabilites into account unless
  we are absolutely certain about the decision.
• Is Representativeness Heuristic in accordance
  with Bayes Theorem?

8
What is Toms Major

• High intelligence
• Need for order
• Neatness
• Dull and mechanical writing
• Little sympathy for other people

• Not enjoying interacting with others
• Self centered
• Deep moral sense

9
Graduate Mean Mean
Specialization judged base similarity
area   rate (in ) rank

Business 15 3,9
Administration
IT 7 2,1
Engineering 9 2,9
Humanities 20 7,2
Law 9 5,9
Library Science 3 4,2
Medicine 8 5,9
Physical Sciences 12 4,5
Social Sciences 17 8,2
10
What is Toms Major

• Kahneman Tverskys questions
• - What percentage of people in different
  majors?
• - How similar is Tom to each major?
• More than 95 of the respondents jugded that
  Tom is more likely to study IT than humanities,
  although they were surely aware of the fact that
  there are many more graduate students in the
  latter field. (ignoring the base rates)

11
Conjuctive Fallacy

• A B can not be more probable than just A or B.
• Example Sarah is 40 - single, outspoken and
  bright. She majored in philosophy and was
  interested in social equality as a student.
• Is Sarah
• a) a sales representative or
• b) a sales representative who is active in
  feminist movement?

12
How to AVOID representativeness bias?

• Dont be misled by detailed scenarios.
• Pay attention to the base rates.
• Dont forget the Gamblers fallacy. (chance is
  not self-correcting)
• Seperate representativeness from probability.

13
Availability Heuristic

• Availability involves...
• Assessing the frequency, probability, or likely
  causes of an event based on the degree to which
  occurrences of the event are readily available in
  memory.
• - People inadvertently assume that readily
  available instances, examples or images represent
  unbiased estimates of statistical probabilities.

14
Availability BiasesEase of Retrievability

• Samples whose instances are more easily
  retrievable from memory will seem larger.
• For example, judging if a list of names had
  more men or women depends on the relative
  frequency of famous names.

15
ExampleList of Names

• Read the list once.

• Michael Jordan
• Sandra Grey
• Barbara Walters
• Maria Schulz
• George Bush
• Kim Melcher
• Indira Gandi
• Jack Smith
• Madonna
• Gill Williams

16
ExampleList of Names

• Are there more men or women on the list?
• Judging if the list of names has more men or
  women depends on the relative frequency of famous
  names.

17
Experience Antecedent of Availability Bias

• A successful executive who attended Yale is
  likely to remember fellow alums he encounters in
  his business circle and his social life. Because
  of his special, circumscribed range of
  experiences he is likely to overestimate the
  relative proportion of successful Yale graduates.
• Thus, range of experiences can cause the
  availability bias.

18
Salience Antecedent of Availability Bias

• Unemployed executives are likely to overestimate
  unemployment among executives, whereas employed
  executives are likely to underestimate
  unemployment among executives. For each
  executive, employment estimates are biased by the
  vivid salience of their personal situation.
• Vivid salience can cause the availability bias.

19
Ease of Recall

• Events more easily recalled from memory, based
  upon recency, are regarded to be more numerous.
• Ex Managers appraisals of employees
• Ex Watching an accident
• Ex Loud repeated advertising

20
Effectiveness of a Search Set

• We often form mental search sets to estimate how
  frequent some occurrences are. However, the
  effectiveness of the search might not relate
  directly to the real frequency.

21
Effectiveness of a Search Set

• Consider the letters K,L,R,N,V.
• Are they more likely to appear in
• - the first position?
• - the third position?
• Result Among the 152 subjects, 105 judged
  the first position to be more likely for a
  majority of the letters, even though in reality
  the third position is more frequent.

22
Ease of Imaginability

• The difficulty of imagining instances is used as
  an estimate of their frequency.
• - E.g. number of combinations of 2 out of 9
  people, versus 7 out of 9 people.
• - Number of combinations of 2 people is seen
  more at first glance, it is more disctinctive and
  easier to visualize, even though the number of
  both combinations is the same.
• Thus, imaginability might cause overestimation
  of likelihood of vivid scenarios, and
  underestimation of the likelihood of difficult to
  imagine ones.

23
Example

• Estimate the result of the following operation
  within 5 seconds!
• One group (87 people) is given 8x7x6x5x4x3x2x1
• The other group (114 people) is given
• 1x2x3x4x5x6x7x8
• The median estimates
• 2250 for the first group and 512 for the
  second one. The correct value 40320

24
Example

• Why?
• Because the results of the first steps of
  multiplication are larger in the descending
  sequence than in the ascending one, the former
  expression is judged larger than the latter.