Snap judgments of personality
Review from last time
How are subjective and objective impressions differentiated?
What traits predict male CEO’s company success?
What traits predict female CEO’s company success? Power as well
Leadership judgments predict CEO profits but not revenues. Why?
What is role congruity theory?
Why can leadership success be predicted from university photos?
What is the difference between consensus and accuracy?
What brain structure is associated with CEO success?
What is the amygdala responding to?
What traits predict CEO success in japan?
What facial features predict CEO success? What else affects it?
What features predict the success of black CEOs? Why?
What is personality?
Measurement models of personality
Brunswick’s lens model and personality
We infer personality before we know it!
Effects of viewing angle
What are some of the valid cues?
Do people care about personality?
Enduring patterns of thoughts, feelings and behaviours
The key assumptions are that personality is:
a) Causal – causes behaviour, thoughts, feelings
b) Stable – does not change much over time
c) Systematic – does not change across environments
o The traits that are important become a part of the natural language =
appear in the dictionary
Researchers pulled the person descriptors from the dictionary and
attempted to find general underlying
Traits are stable, states are temporary
Trait: openness to experience, conscientiousness, extraversion, agreeableness,
State: afraid, talkative, worried, tense, depressed, anxious, full of energy What are the big five traits
Openness to experience
Big five inventory (BFI) = 44 item questionnaire that is commonly used to assess
individual differences in personality
Be aware that BIG in BFI was assigned to represent how broad and inclusive the
factors were, rather than eliminating the possibility that other traits existence
o Big = broad
o Big ≠ important (per se)
Statistical measurement mode: BFI
Latent variable is not measured in the model: it is an arbitrary yet theoretical concept that
researchers want to examine
E.g. extraversion, happiness
Observed variable is something we can measure to estimate the latent variable. It is
theoretically assumed to
Observed variables: talkative, outgoing, sociable, quiet
Latent variable: extraversion
Uses of BFI in person perception
The researchers examine how accurate someone is at inferring personality; ask targets to
fill out a questionnaire: they use the scores as a criterion
Brunswik’s (1956) lens model
So, what are the variables again?
How much the observers agree with each other on a judgment
Estimated by calculating the interrater correlation
Correlation between the observer ratings and the actual criterion
Correlate the average (crosstarget) ratings for the trait with the criterion Cue validity:
Cue validity is estimated by correlating the criterion with the cues.
While you probably can utilize the correct cues from the face to infer that Adam
lambert is a gay man, the cues to his personality are even more ambiguous
Researchers code the features and correlate the ratings of these features with the
selfratings of the targets to estimate cue validity
How well can you use the correct cues?
Achieved through correlating the observer judgments to the cues
Different from the use of correct cues (i.e. the correlation between cue validity and cue
Inferring personality from the face
What are the minimal conditions under which personality can be inferred?
Note. Here, the study did not look at accuracy
245 university students:
128 preliminary study: no time constraints
117 main experiment
Pretest: independent sample
Rated pictures on multiple dimensions of personality. These ratings were used as a
criterion judgment, in this case the control condition
100 ms, 500 ms, 1000 ms
Participants rated: is this person [trait]?
How confident are you in your judgment?
Randomized across condition, exposure time intermixed
Self paced vs. other exposure times
100 ms judgments are very similar to those made under no time constraint: no difference Again, no improvements in R^2 across time points
Judgments became more negative: fewer people attributed positive qualities
People attributed traits faster; they had more time to think since the reaction times
were recorded as the stimulus disappeared
People became more confident in their judgment
The judgments became more differentiated as people had more time to see the stimulus
(increasing their deliberation)
Conclusions: across time
1. 100 ms is sufficient to make judgments of personality from faces
a. Inferred more negative traits
b. Became more confident in their judgments
c. Reacted to faces faster
3. Note: judgments tended to stabilize after 500 ms of exposure, suggesting that 500
ms is enough time to make a decision
4. People started to differentiate more between traits, as suggested
Do our judgments agree at different viewing angles?
What happens if the exposure time is limited?
Ecological validity: usually we see other people’s faces from different viewing angles and
just for a few brief moments
Study 1: Selfpaced
Counterbalanced design: participants saw only 1 viewing angle for each target – 3
Ratings of personality:
Rating of physiognomy:
Attractiveness Facial maturity
The ratings of the faces were consistent across viewing angle as indicated by significant
The magnitude did not change as a function of viewing angle
Study 2: 50 ms
The ratings of faces were consistent between full frontal and ¾ view but not in the
other conditions for personality traits
For physiognomic features the judgments were consistent across viewing angle
1. If given unlimited time, personality is inferred consistently from all 3 viewing
2. If presented for a few milliseconds, the profile view does not produce the same
judgments of personality as the ¾ or full frontal views
3. Judgments of physiognomic features are consistent across time and angle
4. Overall, because judgments are consistent across time for full frontal and ¾
views, fata suggest that there are features on the face that we are able to extract
and utilize to arrive to our judgments repeatedly
What are some of these cues?
Carre and McCormick (2008)
Study 1: 37 men and 51 women
Study 2: 21 varsity hockey players
Study 3: 113 Canadian NHL players
Facial width to height ratio was measured for each individual
Independent of body size
Related to testosterone concentration in development
Like having authority over others?
Point subtraction aggression paradigm
Correlated with selfreported measures of aggression Three response options against an opponent”
1. Press 100 times = +1 point
2. Press 10 times = steal a point from a partner
a. Led to believe that they are in a condition where the points
stolen do not add to the points they have
3. Press 10 times = protect your points from partner
DV: the number of times Ps selected each option
o Selecting option 2 was used as an indicator of aggression: Ps stole points
without benefit to self and lost opportunity to increase their own score
Study 1 result:
Regressions predicting aggression from dominance and fwhr
Men: dominance is not a predictor of aggression (p = .27)
fewh is a significant predictor
Dv: number of penalty minutes per game
Used a proxy measure of aggression
Just aggressive penalties: slashing, crosschecking, high sticking, boarding, elbowing,
checking from behind, and fighting
fwhr explained 29.2% or variance in penalty minutes per game
The greater the fwhr the more minutes
DV: number of penalty minutes per game
Again, the correlation w