Week 3: Self-fulfilling Prophecy, Correspondence Inference Theory, Covariation Model, Lau & Russell, Lau & Russell Review Sheet (new)


Self-Fulfilling Prophecy:

Process by which oneís expectations about another person eventually lead the other person to behave in ways that confirm these expectations

NOTE: Self-fulfilling prophecies must involve at least two people-- you cannot have a self-fulfilling prophecy with yourself

4 Steps to Self-Fulfilling Prophecy

1. Perceiver has expectations about how target will behave

2. Perceiver then behaves in a way that is likely to elicit the expected target behavior

3. Target indeed behaves in a way that confirms perceiverís expectations

4. Perceiver (Objective Perceiver) sees predicted behavior

Examples of Self-Fulfilling Prophecy

* You expect your new roommate to be shy so you donít speak much to him after he moves in, and he therefore does seem shy

* A coach expects his freshmen to be uncoordinated and unskilled so he does not play them often, and when he does they are rusty and do not perform well

* Your professor expects you to do well and she spends extra time with you preparing for the exam, so you get an A

NOT examples of Self-Fulfilling Prophecy

* You repeat "I think I can, I think I can" as you try to jog up a steep hill, and as a result you are able to make it

* You have a bad experience the first time you go skiing so you never go back again and never get better

* Because you get a bad grade on your first exam you give up and stop studying at all, leading you to fail the course

Points to Consider

* Can these expectations be accurate, as when teachers judge their students?

- Jussim and colleagues say yes

* Can perceivers do anything to avoid self-fulfilling prophecy?

- As textbook suggests, reconsider your expectations and behavior towards people

* Can targets do anything to avoid self-fulfilling prophecy?

- Swann & Ely find that targets often strive to disprove inaccurate/unappealing expectations

Review questions for self fulfilling prophecy

a) Jeffs wants to increase his grades this semester. He therefore posted inspirational posters and banners all around his room (e.g., "work hard", "Believe in yourself"). He ended up getting a 4.0 GPA.

No. No distinct perceiver and target involved (need 2 people)

b) You believe Tom is an introvert. You ask Tom whether he has ever felt left out from groups or felt uncomfortable in parties. Tom gave you a "yes" as an answer. You therefore conclude that Tom is really an introvert

No. No bhvr. From you or Tom. Instead, confirmatory hypothesis testing.

c) April went to a fortune teller, who told her that she might be involved in an accident this month. April did not tell anyone, but she worries about it. A week later, April hit another car when she was driving because she was not paying attention.

No, no "perceiver" involved to see expected bhvr.

d) A soccer coach believes Susan is a better player than Amy. She praises Susan more and spends more time with Susan at practice. At the end of the season, other players noticed that Susan has become a better player that Amy.

Yes. All 4 steps present and perceiver + others see results.


Internal Attributions: because personality, character, ability

External Attributions: due to situation


Correspondent Inference Theory

Purpose: To explain when we are able to make internal attributions about a person

Key Issues

- Did the person have a choice in this behavior?

- Was the behavior expected?

- What were the effects of the behavior?

When can we make Internal Attributions?

* When the behavior is freely chosen

- According to Jones & Davis, we sometimes fail to remember this rule

* When the behavior is somewhat unexpected

* When we only know a few things about a behavior/choice (i.e., it only has few effects)

Noncommon Effects

* Think of them as "unique effects": things that are different

* If Jeff has two choices for where to go to dinner, the fewer unique effects of that choice, the more plausible an internal attribution

- China Gate vs. Gandy Dancer

+ Many noncommon effects (price, location, food, ambiance)

- Arbor Brewing Company vs. Grizzly Peak

+ Few noncommon effects (same price, same location, both brew pubs, but ABC has live music tonight-- I conclude she chose it because she wanted to hear the band or likes live music)


Review Questions of Correspondence Inference Theory

Case 1

Internship option 1: HIV Testing Agency (nonprofit)

Work in NYC, $800 Stipend per month, meet interesting people, meaningful work, good for pre-med preparation.

Internship option 2: Pfizer Pharmaceutical Co.

Work in NYC, $3500 Stipend per month, interesting colleagues, meaningful work, medical related work.

Noncommon effect = $$$

Only 1 noncommon effect ? We can make inference about the person

Case 2

College option 1: UC Berkeley

Tuition=$4000, prestigious univ, high ranking, close to home, state university, close to the sea, large student body

College option 2: Cornell University

Tuition=$22,000, prestigious university, high ranking, on East coast, private university, far from sea, but close to mountains, small student body.

Noncommon effects:

Many! Cannot clearly say why she chose to go to UCBerkeley (Sea? $$? Size of U? home?)


Kellyís Covariation Theory

We make attributions based on three aspects of the event:

* Consensus

* Distinctiveness

* Consistency


* Consistency is the easiest to deal with

* If a behavior is low in consistency, we make a situational attribution

- For example, if Larry sometimes talks about how great "The Blair Witch Project" was, but sometimes says he didn't like it, I assume that specific circumstances are the cause of his behavior (who he's talking to at the time, what other movies are in the

news, etc.)

* If it is high in consistency, we look to...

Consensus & Distinctiveness

Consensus: what do other people do/think?

Distinctiveness: does this behavior generalize to other similar stimuli?

* If behavior is low on both, personal attribution

- If no one else raves about "Blair Witch," and Larry raves about many other films, then something about Larry causes the behavior

* If behavior is high on both, stimulus attribution

- If everyone raves about "BW," and Larry rarely raves about films, then the movie causes the behavior

Covariation Summary

* First check out consistency of the behavior

- If it's low, attribute to passing circumstance

- If it's high...

* Then check out consensus & distinctiveness

- If they're low, then the person is probably the cause of the behavior

- If they're high, then the stimulus (or other person involved) is probably the cause of the behavior


Review Questions for The Covariation Model

"Joe became sick riding the commuter bus."

Internal attribution: Joe has a weak stomach

External attribution: the commuter bus ride was rough

Consensus: do other people behave in the same way to the stimulus?

Do others get sick riding the commuter bus?

High consensus: others also get sick

Low consensus: only Joe got sick

Consistency: does the actor always behave in the same way to the stimulus?

Does Joe always get sick riding the commuter bus?

High consistency: yes, he always does.

Low consistency: no, itís only today

Distinctiveness: how does the actor behave to other stimulus?

Does Joe get sick riding other busses? How about cars? Vans?

High distinctiveness: Joe only gets sick riding the commuter bus

Low distinctiveness: Joe gets sick whenever he gets in a motor vehicle






(J has weak stomach)


(Only J gets sick)


(J always sick on bus)


(J gets sick in all vehicles)


(commuter bus ride rough)


(others get sick on bus)


(J always gets sick on bus)


(J usually not sick on busses)




(J not always sick on bus)



or just use: SenDC

Consensus Information: actorís behavior=otherís behavior?

Distinctiveness Information: actorís bhvr. same for different stimuli?

Consistency Information: relationship btwn. Actor and stimulus is same across time and circumstances.

Internal attribution = SenDC = LLH

External attribution=SenDC=HHH

Situational attribution=peculiar circumstances H/L H/L L


Lau and Russell (see below class notes for article review sheet)

Aim: to investigate whether people have self-serving attributions, and why they do it.

-Archival study of sports page article

Hypotheses 1 and 2:

1. CH: People have Hedonic Bias: Internal attribution for own success; External attribution for failure

OH: Coaches, athletes, sportswriters will make more internal attributions win they win and more external attributions when they lose


2. CH: When there is a need to enhance / protect self esteem we engage more in hedonic bias

OH: so coaches & Athletes will show more hedonic bias than sports writers b/c their self-esteem is at stake


-Coaches, athletes, sportswriters all make more internal attributions whether win or lose

-players/coaches overall make more internal attributions for success and more external attributions for failure than sportswriters


Hypothesis 3:

  1. CH: unexpected success/failure leads to external attributions and expected lead to internal attributions

OH: teams unexpected to win/lose will explain outcome by situation (luck); expected wins/losses explained by team ability

Hypothesis 4:

  1. CH: Unexpected outcomes lead to more causal attributions in general; expected outcomes lead to stable causal explanations

OH: more attributions for unexpected wins/losses in articles

-supported; supported (but didnít reach level of significance)


Lau and Russell Review Sheet


1. What is the central motivation underlying the attributional process, according to the authors, and what does it tell us about how we should empirically study attributions? (p. 22)

2. Why did the authors choose to study attribution in the sports pages? (p. 22)

3. To what tendency does the term "hedonic bias" refer, and what are two possible explanations for that bias? (pp. 22-23)

4. What hypothesis do the authors put forth regarding when attributions are likely to occur (i.e., the frequency of causal attributions in general)? (p. 23)

5. The authors also present a hypothesis regarding attribution frequency and whether or not the outcome of an event was expectedówhat is it? (pp. 23-24)


1. How were articles selected for inclusion in the study? (p. 24)

2. On what two dimensions were quotations coded? For the second dimension listed, give a brief example of two quotations and how they were coded on that dimension. (pp. 24-25)

3. To what extent did the two coders for the study agree on their decisions, and how were disagreements resolved? (p. 25)

4. How was the expected outcome of the event measured? (pp. 25-26)

5. What type of methodology is being used in this study (experimental, correlational,

observational, etc.)? How can you tell?


1. What results were obtained regarding internal vs. external attributions and winning vs. losing? How were these results different for players/coaches vs. sportswriters? (focus on Table 1, p. 26)

2. Did whether or not the outcome was expected influence internal vs. external attribution tendencies? (focus on Table 2, p. 27)

3. Did whether or not the outcome was expected influence the frequency of attributions in general? (p. 27, and Table 3, which is a bit confusing)


1. Did these results support the authors' hypotheses?

2. What is your opinion of the external validity of this study? (p. 28)

3. What difficulties are presented by the fact that the authors could only study those attributions the sportswriters saw fit to include in their articles? (p. 28)

4. What other difficulties did the authors encounter in using this type of methodology? (pp. 29-30)

5. Can we draw any causal conclusions from this study why or why not?

6. What other implications can you think of for these findings?