LANGUAGE AND THOUGHT

1
LANGUAGE AND THOUGHT

• MAKING CAUSAL ATTRIBUTIONS

2
Explaining the Causes of Behaviour

• How do we decide how an event is caused?
  (Attribution Theory).
• Two main versions of the theory.
• Correspondent Inference Theory (Jones and Davis,
  1965)
• Applies centrally to intentional actions
• Covariation theory (Kelly, 1967)
• Applies centrally to spontaneous reactions

3
Errors in Explaining Causes of Behaviour

• What errors do we make in attributing causes?
• Actor/Observer Divergence
• Fundamental Attribution Error
• Self-serving Bias.

4
Attribution Theory Overview

• Do we make similar attributions in the social and
  non-social worlds?
• Importance of attributions - tell us
• What to expect of the world.
• How to influence the world.
• Application of attribution theory to depression.

5
Intuitions about Causal Reasoning

• Occurrence December 3rd, 1979, Cincinnati. 8,000
  Who fans surged forward, crushing 11 fans to
  death (see Brown, R, 1986).
• Question WHY? (to what should the cause be
  attributed?)

6
What Caused the Crush?

• Do you think that the crush was caused by
• Something about Who fans?
• Something about the situation?
• Answer given will influence action taken (e.g.,
  cancel future concerts vs. have greater security
  precautions.)

7
Something about Who Fans?

• An internal, dispositional attribution
• Mayor of Providence cancels future appearance.
• Unfortunately there were thousands of impatient
  dope heads who are more concerned about a good
  seat than a human life. Cincinnati Post.

8
Something about the Situation?

• An external, situational attribution.
• Insufficient security guards.
• Confusion about ticket sales.
• Erroneous opening of doors.
• Previous crushes at Cincinnati Coliseum.
• At about 7.15 a door opened.... The people in
  the back thought all the doors were open and
  started pushing through a usual pace. The people
  at the front started falling over, getting
  crushed.... If you went down, you saw what
  happened - you hurt damn bad. Eyewitness.

9
Which Account is Correct?

• Later concerts in tour did not cause crushes.
• Evidence suggests situational account is correct

10
Kelley's Covariation Model (1967)

• Lay theory of causal attributions.
• Person as intuitive scientist.
• We use intuitive laboratory method to discover
  and test causes.

11
Attribution An Example

• Observation
• Claire laughs at a comedian. WHY?
• Is it something about Claire?
• Is it something about the comedian?
• Causes covary with effects, so must determine
  when effect occurs/does not occur in
  presence/absence of hypothesised cause.

12
Making Attributions3 Kinds of Useful Information

• Consensus Information - does this stimulus
  produce same effect in other actors? (Does
  everyone laugh at this comedian?)
• Consistency Information - does stimulus always
  produce this reaction in this actor? (Does Claire
  always laugh at comedian?)
• Distinctiveness Information - does only this
  stimulus cause this effect in this actor? (Does
  Claire laugh at only this comedian?)

13
External and Internal Attributions

• Kelley's theory says
• We infer external (situational) causes from
• HIGH distinctiveness, HIGH consensus, HIGH
  consistency (caused by entity, i.e. person or
  object in situation).
• HIGH distinctiveness, LOW consensus, LOW
  consistency (caused by circumstances).
• We infer internal (actors disposition) causes
  from LOW distinctiveness, LOW consensus, HIGH
  consistency.
• Applies to Claire - also to Who concert.

14
Proportions of Internal Attributions for Various
Types of Information
15
Kellys Attribution Theory

• BEHAVIOUR CONSENSUS CONSISTENCY DISTINCTNESS ATTR
  IBUTION
• Low High Low Internal
• Only Claire Claire always Claire laughs It
  must be laughs at laughs at at
  everything something
• the comedian the comedian about Claire
• Claire that makes
• Laughs her laugh
• At the
• Comedian High High
  High External
• Everyone laughs Claire always Claire laughs It
  must be
• at the comedian laughs at the at only
  this Something
• comedian comedian about the
• comedian that makes
• Claire laugh

16
Kelley's Pocket Calculus

• Attribution Consensus Distinctive Consistency
• Actor-stable L L H
• Situation-stable H H H
• Actor-unstable L L L
• Situation-unst. H H L
• A-S inter-stable L H H
• A,S-stable H L H
• A-S inter, uns. L H L
• A,S -unstable H L L

17
Evidence for the use of Kellys theory

• McArthur (1972)
• Subjects given target Sue is afraid of the
  dog.
• Plus distinctiveness and consensus information,
  e.g.
• Sue is afraid of most dogs.
• Most people are afraid of this dog.
• LL pattern causes dispositional (actor)
  attribution
• HH pattern causes situational attribution.
• Results support Kelley' s model

18
Errors in Attribution

• Kelley' s model is normative model of causation.
• What is correct descriptive model?
• Two important aspects of attribution that do not
  fit the model.
• Actor/Observer Divergence (Jones Nisbett,
  1971)
• Fundamental Attribution Error (Jones Harris,
  1967)

19
Actor/Observer Divergence(Jones Nisbett, 1971)

• People see their own behaviour as caused by
  situations, but others' as caused by
  dispositions.
• Evidence
• Subjects given 20 trait pairs (quiet/talkative).
• Told to rate self, friend, father, TV
  personality on each trait pair
• Quiet Talkative Depends on situation.
• Subjects used Depends much more for self than
  others.
• But failures to replicate effect consistently.
• Interpretation complicated by Storms' study.

20
Storms (1973) Study

• Two people take part in a conversation
• Two other people watch
• Conversation is videotaped from each of the 4
  peoples perspective
• People then viewed the videotape, from
• Their original point of view
• The other (actor vs. observer) point of view
• OR, not at all

21
Storms (1973) Study Results

• People who saw no videotape or a videotape from
  their original point of view showed the standard
  actor-observer divergence.
• People who saw the tape from the other point of
  view showed the actor-observer divergence
  relative to the new point of view.

22
Fundamental Attribution Error(Jones Harris,
1967)

• We tend not to make sufficient allowances for
  external causes of others' behaviour. (We take
  their behaviour at its face value as telling us
  something about them).
• Consistent with actor/observer divergence.
• We dont make this mistake with our own behaviour

23
Fundamental Attribution Error Example

• Ross, Amabile Steinmetz (1977)
• Undergraduate subjects randomly put in 2 groups.
• Questioners
• Answerers
• Questioners made up hard but not impossible
  general knowledge questions.
• Answerers tried to answer (got about 4/10 right).

24
Ratings of General Knowledge Compared to Others
25
Why does the Fundamental Attribution Error Occur?

• Possibly related to ignoring base rates.
• Remember Tom W studies (Kahneman and Tverksy).
• Subjects underweight base rate information
  (numbers of computer science and humanities
  graduate students), overweight causal information
  (description of Tom W).
• Similarly, subjects in Rosss experiment
  underweight base rate (knowledge that
  statistically questioners and answerers are from
  same population) overweight what they see
  (answerers failing to answer questions that
  questioners know the answers to).

26
Why does the Fundamental Attribution Error Occur?
(Cont.)

• Possible common mechanism
• Representativeness
• How like a CS graduate student is Tom W?
• How like an intelligent/stupid person is the
  questioner/answerer (on basis of knowledge of
  answers)?
• People dont always neglect base rates. Zukier
  Pepitone (1984) - gave subjects Tom W.
• Subjects told to reason like a clinical
  psychologist (ignored base rates) or like a
  scientist (used base rates).
• BUT, Kelly likens attribution to intuitive
  science!

27
Self-serving Bias in Attributions

• People tend to make internal, stable attributions
  for successes, but external, unstable
  attributions for failures.
• Is this due to motivational bias?
• Miller (1976) - subjects given test of social
  perceptiveness.
• Half were told they did well, half badly
  (actually random).
• Half were told test was reliable (ego-involved),
  half told it was unreliable (ego-uninvolved).

28
Self-serving Bias in Attributions (Cont.)

• Ego-involved subjects made more external
  attributions for failure and more internal
  attributions for success.
• Results support motivational not (purely)
  cognitive view.
• But is the self-serving bias universal?

29
Depressive Attributions

• Attributional style seems to be a personality
  trait.
• Attributional Style Questionnaire (ASQ) measures
  this trait
• e.g. You have been looking for a job
  unsuccessfully for some time.
• Is this your fault?
• Or is it the fault of other people or the
  situation?
• Will the cause be present in future or not?
• Does the cause influence all areas of your life
  or not?
• Normal subjects show self-serving bias.

30
Depressive Attributions (Cont.)

• Depressives show opposite bias stable, global,
  internal attributions for negative events
  -unstable, local, external attributions for
  positive events.
• One one view, depressives have learned that they
  can't influence world to their own good. Leads to
  low self-esteem. Seligman (1975) used the term
  learned helplessness to describe this
  condition.
• So is the self-serving bias adaptive? As a
  counter to depression?