PSY220 - 3. Mental representation + Heuristics Flashcards
Social cognition
Stimulus→Cognitive, affective, →Behavior
motivational mechanisms
as computers became popular
memory, retrieval
describe mental operations ppl perform as they saw their problems
measurable + systematic methods for cognition
Social cognition
borrowing heavily from cognitive psychology has become dominant paradigm in social psych in the past 20 years
(Schema) Concept
unit of knowledge (category)
Building block of cognition
Can be objects/intangible
Need not be accurate, Great deal of subjectivity
Measure regularity of cognition + consequences for everyday life
What do concepts do?
reason we can even make sense of all info
1.Reduce amount of processing we need to do when too much info available
Rather than scrutinizing every feature, we can activate concept while using cognitive effort elsewhere
Takes time to learn category – experience
Stereotypes: develop ideas about how ppl behave
What do concepts do?
May not always be true, but have functional value – allows us to devote cognitive resources where its more needed 2.Add info when is too little available.
Going beyond info given – meeting strangers – signaling they belong to a category – let our info on category fill in missing info
3.Guide attention, interpretation
Bransford & Johnson (1973)
Concept of laundry disambiguates ambiguous stimuli
Category guides your interpretation + aids in comprehension
Concept activated at time of encoding leads to memory advantage
After – lost ability to encode with info
When available at time of encoding help facilitate meaningful encoding of ambiguous stimuli
Bransford & Johnson (1973)
# ideas recalled No instructions: 2.8 Washing clothes (before reading): 5.8 Washing clothes (after reading): 2.7
Different concepts applied to the same stimulus input can lead to dramatically different interpretations & behavior
a) features of a house from the perspective of homebuyer vs. burglar. burglar – safety (locks) , homebuyers – luxury aspects (fireplace, moulding)
b) Man crying (did a loved one die/did he win the lottery?) sadness/joy
c) Ambiguous behavior of African-American vs. Caucasian person.
Duncan (1976)
White subjects watched videotape of two men in a discussion. discussion gets heated. begin shouting. shoves the other.
At this point, the tape is stopped, and subjects are asked to characterize what just happened.
Duncan (1976)
What subjects didn’t know: 2 versions of the tape: “shover” was Black/“shover” was White
Black guy was seen as an aggressive
African american disambiguated situation as hostile + aggressive due to activation of african american concept
Duncan (1976)
Incoming info assimilated into concept that is activated.
disambiguation of ambiguous situation in the direction of the activated concept
cognitive processing limit – mind has created shortcuts, autopilot mode
Duncan (1976)
severe costs of using concepts and social categories (stereotypes, “lazy” thinking) + clear benefits – that’s why they’re so prevalent:
- solidify/reify ambiguous information.
- conserve cognitive resources.
Macrae, Milne, & Bodenhausen (1994).
- Participants asked to read information about several social targets presented on a computer screen while at listening to an audiotape playing completely unrelated material.
- Either “John” then list of traits (3 seconds each) or “John-skinhead” “John-artist” “John-doctor” then traits.
- 10 traits presented – five were stereotype consistent (“aggressive” “creative” “caring”). 5 traits were neutral.
Macrae, Milne, & Bodenhausen (1994)
- In headphones, passage about geography + economy of Indonesia (no one would have prior knowledge about).
- DV’s: Cued recall task-each target name written on top of paper and they were to recall and correctly attribute as many traits as they could. Also: Given a written quiz about Indonesia to test whether they were listening to the passage
Macrae, Milne, & Bodenhausen (1994)
How much info about indonesia penetrated
Better memory for traits (caring, knowledgable) if they knew beforehand that it was John the doctor
6.Results: Subjects for whom a stereotype was provided recalled twice as many traits as those without:
Stereotype Present Stereotype Absent
Consistent 4.42 2.08
Neutral 1.83 1.33
Macrae, Milne, & Bodenhausen (1994)
7.those for whom a stereotype was activated should have more resources available for the listening task.
DV: questions anwered correctly:
Present Absent
8.75 6.66
saves us the trouble of thinking too much about the group + can slide our attention to learning about indonesia
allows us to economize resources
WHAT ARE THE PRINCIPLES OF CONCEPT ACTIVATION?
Often concepts arent given easily
How do we know which ones will be activated
1. Salience – surface features powerfully connote specific concept Black man most likely to be noticed among whites, than among men with 1 woman
most unusual thing becomes the most salient
What variables make a concept more likely to be activated?
- accessibility
Temporary vs. chronic accessibility
Condition: had to solve puzzle where solution was (reckless vs. adventurous)
In reckless condition – rated him more reckless
In adventurous condition – rated him more adventurous
What variables make a concept more likely to be activated?
Ambiguous stimuli disambiguated in direction of the concept activated before hand
Priming: unconscious – no idea their interpretations was affected by puzzle
Irrelevant task but concepts were activated that influenced interpretation
Subliminal: looks like a flash of light, unconsciously able to read word
Superliminal: have some idea what’s going on, but don’t know how they’re connected
What variables make a concept more likely to be activated?
temporary: priming in the lab
chronic: more predisposed to concept
doesn’t matter if from lab or just carry concept activated around with us, still produces assimilation effects
subliminal messaging works at a limited level – not gonna get up to buy a pepsi, but may choose it over coke at supermarket
Stereotypes
influence how we understand different traits.
“aggressive” means one thing when referring to the category “lawyers” (verbal)/football players (physical).
Stereotypes can influence how we understand other stereotypes. “Harvard-educated carpenter.”
Jerome Bruner and “The New Look”
Bruner & Postman (1948)—poor vs. rich children asked to estimate (drawing) the size of a quarter.
Results: Poor kids drew much bigger circles
Add own info: nonmaterialistic to reconcile concepts
because they value quarters more, so they perceive it as larger
Jerome Bruner and “The New Look”
3rd element to disambiguate
what we perceive systematically influenced by accessible cognitions
values can influence our perception
Is Colin Fenton famous?: Jacoby et al. (1989):
- Subjects merely pronounced list of 40 nonfamous names.
- Either immediately afterwards or 24 hours later, subjects asked to determine from a large list of names (that included the previous, pronounced words) who was famous and who was not.
Is Colin Fenton famous?
More errors in the nonfamous→famous direction for pronounced words. (after delay only)
Misattribute feeling of familiarity to the wrong source
Forgetting needs to occur to create ambiguous
Colin fenton has become more accessible than nonfamous name
How incoming info is encoded: hamilton, katz, leier.
a. subjects read 30 behaviours describing target person.
b. Half explicitly told to form an impression. Half told to memorize list of behaviours
c. After a delay, recall as mamy behaviours as possible
Encoding
Counterintuitive results: impression > memory
Creating concepts + assimilating subsequent info to make it more meaningful
Try to come up with coherent themes for textbook social phenomenon
Asch: are there lawful principles that govern formation of impressions about ppl?
What are the ways in which ppl form a single unified impression of someone based on knowing many diverse bits of info?
How do ppl combine + integrate such info?
Approach: make models that are simpler, yet reflective messy real life
Asch
Elegant control + manipulation. From one study to the next, he made minute changes in the paradigm + eventually certain regularities or laws were uncovered
Insight: in impression formation, the whole is different from the sum of the parts.
Asch
Evidence: primacy effect
Impressions have a life of their own: ppl are able to remember their impression of someone long after specific behaviours are long forgotten
Subsequent info is disambiguated by the concepts activated by impressions
Hastie + Park
a.online (over time) or memory based conditions
b.online told to form impression as they went, updating. Memory based asked after reading sentences
c.mem based judgement correlated with recall. Online judgement not correlated with recall
order of info mattered for online based – early info has more weight than later info
for mem based order doesn’t matter – weren’t asked to form impression explicitly, less weight for early sentences
The Rationality Assumption:
Are people truly rational actors?
Ppl tend to choose most beneficial to us
Computer might make same decision + true for many cases
But in many cases, humans deviate from rational behaviour
Pioneers of research of human’s lawful nonrationality:
Amos Tversky and Daniel Kahneman
Human decision-making often deviates from what a computer would do, but lawful deviations
Systematic deviations
Understanding of probability
THE REPRESENTATIVE HEURISTIC
judge likelihood of an event based on prototypicality.
Bill enjoys adventure + aeronautical engineering. spent 10 years in US Air Force. Bill more likely to be a taxi driver/an astronaut?
Astronauts Taxi drivers
c. 50 c. 25,000
1% of 25,000=250
50 < 250
Base rate
overall #s of a given occurrence in pop
fail to take into account base rates
Base rate: Tversky & Kahneman:
Jack generally conservative, careful + ambitious. no interest in political + social issues + spends most of his free time on home carpentry, sailing and mathematical puzzles. Is Jack a lawyer or an engineer?
drastically under-use base rates in decision-making
Half of the subjects told that the sample consisted of 70 e’s and 30 l’s, half told 30 e’sand 70 l’s
Assume he’s an engineer regardless of condition – failed to take into account base rates
A related heuristic fallacy – the “conjunction fallacy
Cannot be greater probability than bank teller alone, maybe equal if all bank tellers were all feminists. Active info seems so representative of linda.
Doctors overweight prototypical symptoms not taking into account base rates – misdiagnose
Financial analysts can lead to financial collapse
conjunction fallacy
RH is so pervasive is that people don’t understand difference between big + small numbers
Equally likely, small set anything can happen
Previous outcome doesn’t affect the subsequent outcome
Gambler’s fallacy
conjunction fallacy
could hot streak have happened by chance
streaks were not common than chance
regardless of skill, still no increase in shooting percentage if they’re on a streak
lebron is the only one who knows if he’s hot
we don’t have access to subjective experience, only observable behaviour
Factors that can improve people’s probability judgments – make them more aware of base rates
1) Hi knowledge of domain in question.
2) More simply, clearly stated (Ginossar & Trope, 1987 turned lawyer/engineer problem into a fun and engaging card game and improved subjects’ performance.) – more engaged
Factors that can improve people’s probability judgments – make them more aware of base rates
3) When choices more clearly distinguished
4) Hi self-relevance
5) Contextual cues – increasing the salience of chance factors
THE AVAILABILITY HEURISTIC: (should be called the “accessibility heuristic”)
Outcomes more quickly + readily pop into mind thought to occur more frequently
“In the English language are there more words that start with the letter R or have R as the third letter?”
correct answer: 3x more words with r as the third level
THE AVAILABILITY HEURISTIC: (should be called the “accessibility heuristic”)
Ease of retrieval assumed to reflect that actual state of affairs (think of R-starting words more easily because there probably more R-starting words.”). ease of retrieval is often unrelated to what’s actually “out there” in the world
Egocentric biases
Who does more housework? (Ross & Sicoly found that husbands’ and wives’ estimates add up to greater than 100%)
Assessments to contributions to team, groups – ppl have more access to their own behaviour
Egocentric biases
Tend to think of your own work rather than your partner’s
False consensus effect: ppl tend overestimate amount of ppl that agree with them
We are surrounded by likeminded ppl – we have a bias sample – we extrapolate to whole population
Ease of retrieval mistaken for prevalence
What variables contribute to AH?
A key variable: salience.
Taylor & Fiske (1975), Arrangement of room
Subjects 1 and 2 only saw back of D1’s head, but saw D2 in full view. Subjects 3 and 4 only saw back of D2’s head but saw D1 in full view.
All subjects asked: “Who contributed more to the discussion?”
Results: discussant whose face you see – have more rich info – salience – mistaken extrapolation that they were more influential
ILLUSORY CORRELATION (Hamilton & Gifford, 1976)
Group A Group B
Positive 18 8
Negative. 9 4
(Replicated hundreds of times—robust and repeatable.)
Why? Salience + AH
Group B-Negative cell has the least amount of behaviors, most rare occurrence of the 4 possibilities.
- Occurrences rare tend to leave stronger impression.
- Events that leave a stronger impression tend to be easier to remember.
Why? Salience + AH
Group B-negative behaviors are easier to remember, ppl mistake ease of retrieval for fact + rate Group B overall more negatively than Group A.
Salience → accessibility → easier to retrieve → IC
Why? Salience + AH
Bias against group b because of a byproduct
Negative behaviour of minority group is more salient
Outside of conscious awareness
Misrecall negative behaviour of group A as those of group B
every bad behaviour minority does leaves a bad impression
Why? Salience + AH
- 3 examples of being extraverted/10 examples
those who thought of 3 rated themselves as more extroverted, difficulty of retrival leads them to conclude they aren’t as extroverted
ease of retrievals drives ppl’s self views – accessibility
diff to think of examples because I’m not extroverted
easiest to remember rarest things – metacognitive feeling of easy to remember is associated with prevalence
ANCHORING AND ADJUSTMENT HEURISTIC
How many homicides do you think were committed in the state of Iowa in 1999?
More or less than 25/400: in 400 condition had higher guesses
ANCHORING AND ADJUSTMENT HEURISTIC
In the face of an uncertain estimation, ppl will use any number provided, even if arbitrary, to calibrate their scale.
when people “anchor” on a certain figure, they fail to sufficiently adjust. Their metric is off: calibrated to the number they’re given
BELIEF PERSEVERENCE (Anderson, Lepper, & Ross, 1980)
Once an idea has been invalidated it takes a long time for it to disappear. Even if struck from record it’s difficult to forget it.
HYPOTHESIS TESTING
Humans as “intuitive scientists.” “test” our hypothesis against incoming “data.”
Ppl as intuitive scientists
Vulnerable to inherent biases
Wason (1960)
3 7 D K
“Every card that has a D on one side has a 3 on the other side.”
Which two cards do you have to turn over to verify the truth of this statement? Answer: 85% turn over D + 3. But you need to turn over D + 7 because you need to look for evidence confirming + disconfirming hypothesis.
Confirmation Bias
focus on hypothesis confirmation, at expense of hypothesis disconfirmation when both are needed.
Fail to find disconfirming examples when you think back
What improves performance on the Wason Task?
Expertise. For example:
Drinking a beer 16 years old Drinking a Coke 22 years old
college students, when asked to verify the rule that “one must be over 21 to drink beer,” did much better (29/40 got correct answer) than when asked to do the same task with letters and numbers (0/40).
What improves performance on the Wason Task?
spending a lot of time thinking about diff contingencies + outcomes makes the task a snap.
Adding meaningful context makes task easier + less susceptible to confirmation bias
Information-seeking bias
- Half told that task was to determine whether person waiting in other room was an extravert, half told introvert.
- allowed to select which questions to pose toward the other person from a pool of 26 questions: 11 extravert questions, 11 introvert questions and 5 nondescript questions.
Information-seeking bias
Results: hypothesis affected questions ppl asked, failed to ask questions that disconfirmed hypothesis
Snyder & Swann, Study 2
- Had subjects actually pose questions they selected to a real person.
- Each interview was taped.
- separate set of subjects (who were blind to the hypotheses, of course) were asked to judge whether the person being interviewed was an introvert or an extravert.
Snyder & Swann, Study 2
Results: judged interviewee as intorverted when hypothesis was introverted. The biased questioned confirmed the hypothesis. Gives misleading impressions based on questions.
SELF-FULFILLING PROPHECY
originally false belief about person leads to that person acting in accord with that belief (so that the belief no longer appears false).
Leading questions/simple conversation bias. Hypothesis becomes confirmed. Perceiver unaccounts for selective nature.
“The Pygmalion Effect” – expression of character desires
Rosenthal & Jacobson (1968)
1 led teachers to believe that some students in their classes were “late bloomers” – destined to show dramatic increases in IQ over the school year.
2.students selected at random.
blooming label had greater increase in IQ. Expectations precipitating students performance. Explanation for performance gaps.
Jane Elliott: Blue vs. Brown Eyes - prejudice
Three steps to the SFP
(1) perceivers must develop erroneous expectancy;
(2) perceivers’ expectancy must influence how they act toward targets;
(3) target must react to the perceiver’s behavior in manner that confirms the original false expectancy
How do “teachers” respond when they hold high vs. low expectancy of target
(1)warmer emotional support; more open, forgiving
(2)more time and attention;
(3)provide more opportunities for high expectancy target to perform + learn difficult material; lower expectancy less opportunity to excel
(4)clearer + more constructive feedback
unconscious
What are some moderators
On the perceiver end…
(1) Perceiver’s goal: goal is to form a stable + predictable impression, more likely- if goal to form an accurate impression, less likely;
(2) rigidity of perceivers’ belief: if highly rigid, more likely - if less rigid, provisional less likely.
On the target end
(1) Unclear self-concept (need others to tell me what I’m like— social comparison), more likely - clear self-concept, less likely;
(2) age (stronger among younger children than older – except: 7th grade – susceptable to SP in adolescence.
On the situation end
(1) new situations (another form of unclarity)—(helps explain the 7th grade blip, a transition to the new situation of junior high school).
Hindsight bias
knowledge of how event ultimately turned out influences memory for your thoughts on event before it happened
Is it really a bias in memory?
(People may just say they knew it all along…without actually believing they knew it all along. If they actually believe they knew it all along…that’s very interesting: suggests that it is difficult to combat the hindsight bias.)
Hindsight bias
instructor introduces concept + student respond to themselves “duh”
only obvious after they tell you
greater obviousness after you tell them
in the past they were lacking critical information
in present discounting information you have now
1972: Richard Nixon visits China after decades
overestimated original estimates
assimilation in the direction of the outcome
during memory flashback, when asked to recall they increase percentage of their probability guess
Rosenhan:
Gathered ppl from diff jobs
Check in mental institution with same symptoms – voices
Voices stopped after checked in
How long did they stay there?
Confederates took notes on experiences
A nurse saw it as bizzarre note taking behaviour: disambiguated situation in direction of crazy (powerful situational pull)
