human personality weeks 3-6

why is personality science uniquely positioned to predict real-life ouctomes?

• life outcomes are complex and multidetermined → many small causes

• personality science provides many variables (traits, facets, ABCDSs)

• measures capture real-world behaviour (high ecological validty_

• ABCDS cover broad psychological processes

  • can combine many small effects to predict ouctomes

• eval

  • correlations usually small

  • stronger for self-report ouctomes due to method overlap

what is the explanatory approach in psychology?

• goal: explain why someting happens

• uses:

  • simple (parisomnious) models

  • few variables

  • controlled/lab experiments

what are the problems with explanatory models?

• focuses on in-sample explanation, not prediction

• often overfit to specific samples and methods

• low ecological validity

  • do not generalise well to real life

what is the predictive approach in psychology?

• Goal: accurately predict outcomes

• uses:

  • many predictors

  • real world relevant measures

• focus on out of sample generalization

  • can predict out of the sample it was built on

what is the key tension between explanatory and predictive approaches?

• explanatory models:

  • simple and elegant

  • poor prediction

• predictive models

  • complex, many variables

  • better real-world prediction

• trade-off between parsimony vs accuracy

what are the main kinds of outcomes linked to personality traits?

• education

• SES

• relationships

• political, socially responsible behaviour

• health and mortality

relationship between dropping out of school and low self control? moffit et al. 2011

children with low self control were substantially more likely to drop out of school

big five and GPA? poropat et al., 2011

three of big five

• openness .12

• agreeableness .07

• conscientiousness .22

• main finding : children with higher conscientiousness are more likely to do well in school. Most associations pertain to self reported conscientiousness. 

what are the effect sizes and how do we meaninfgully interpret them?

• they appear to not be particularly strong or weak

• to intrepret menaingfully we would comapre to something we know to be a ribsut association with meaninfgul effect sizes

• such as IQ

how does IQ effect sizes and big five relate?

• intelligence is .25

• conscientiousness is .22

• pretty much the same, not far off

how do the associations maintain throughout stages for big five and GPA?

• associations are generally stronger in primary school

what are the associations between methods of gpa and big five?

• associations are there and stronger so they do generalise w

• conscientiousness .43

• openness .37

why could self reports be smaller?

• parents perceptions are tainted with school performance

• kids self reports are unreliable

• grades reflect how kids appear to others

how do you students self select ito academic majors? what traits with what degres? vedel, 2016

• neuroticism highest for arts and humanities, psychology

• extraversion highest for economics, law and medicine

• openness highest for arts, humanities, psychology, polities

• agreeableness lowest for law, business and economics

• conscientiousness lowest for arts and humanities

job performance and big five? salgado & de fruyt, 2005

Neuroticism	-.21
Extraversion	.10
Openness	.09
Agreeableness	.19
Conscientiousness	.33

• high c seems to be highets prefictor for job performance

job satisfaction and big five? salgado & de fruyt, 2005

• N-.29

• E - .25

• O- .02

• A - .17

• C - .26

• low N high E and C seem to be more satifies with jobs on averga e

socioeconomic success and big five? judge et al., (1999)

• children measured in childhood and in adulthood

• higher income?

	Income	Income
	Adult	Child
N	-.32	-.26
E	.24	.25
O	-.01	.10
A	-.11	-.01
C	.34	.16

relationship quality and big five?

• personality traits linked with relationship quality (meta-nalaysies)

• N	-.22
  E	.06
  O	.03
  A	.15
  C	.12

• personality traits predict likelihood of divorce (Roberts et al., 2007)

antisocial behaviour and big five?

antisocial behaviour

• Low A and Low C, high N.

agreesiveness

• high N

• low E, O, A and C

political converservtaism and big five?

• when times are hard, less open people are likley to be politically conservative,

• however not strong associations

N	-.03
E	-.01
O	-.18
A	-0.2
C	.10

environmentalism and HEXACO

	Pro environmental behaviour	Pro environmental attitudes
N	-.02	.02
E	.10	.09
O	.21	.22
A	.10	.15
C	.11	.12
H	.26	.21

big five and longevity? graham et al., 2017

• low C more likely to die in certain period of time

N	1.05
E	0.94
O	0.97
A	0.95
C	0.89

how are big five traits related ro a variety of self-rpeorted helath issues?

• most consistently conscientiousness

alzhiemer’s and the big five?

• high N seems to be predictive of future of Alzhemier’s

• Low C seems to be predictive

• effects comparable to established risks (low education, diabtees)

	AD (OR)
N	1.33
E	0.95
O	0.86
A	0.88
c	0.76

what can explain the associations?

• personality traits are linked to SES whihc might have some helath relvane

• health related behaviours

  • adherence ro doctors advice and medical treatment

  • higher C linked

• stress

  • employee burnout

    

• accidents

  • low C and high N are more likely

general conscientiousness patterns?

• High C

  • better education

  • higher income and job performance

  • better relationships

  • better health and longevity

    • healtheir behaviour

    • better adhernece to medical illness

    • fewer chronic ilness

• Low C

  • more antisocial behahviour

  • worse health

    • risky behaviours (smoking alcohol)

    • more accidents

    • higher mortality risk

• high C is consistently positive across most ouctomes

general neuroticism patterns?

• High N

  • worse education and work ouctoes

  • poorer relationship quality

  • worse health

    • higher stress and anxiety

    • mor emental health problems

    • increased risk of physical illness

    • linked to alzhimers’s risk

  • more agression and antisical behvaiour

  • lowee socioeconomic stress

• Low N

  • better eductaion and job ouctomes

  • higher SES

  • better relationships

  • better health

extraversion general patterns

• High E

  • better job performance and satisfation

  • better SES

  • better relationships

  • Pro environmental behaviour and attitudes

  • better longevity of life

• Low E

  • higher aggressiveness

agreeableness general patterns?

• HIgh A

  • better relationships

  • less antisicoal behvaiour

  • health:

    • better cooopertaion with doctors

    • less stress

• Low A

  • more agreesion

  • worse health via hoistility

openness general patterns

• High O

  • better education

  • pro environmental behaviour and attitudes

Low O

• more agressiveness

• political conservative views

  • not strong associations

what are the typical effect sizes across all outcomes?

• generally small to moderate

• however the are similar in size to other important predictors

what explains the associations between personality and these ouctomes?

• health behaviours

  • e.g - high c → better diet, exercise and adherence

• stress and emotional processes

  • e.g - high N → more stress, burnout

• indirect pathways

  • personality → education/income → health

• risk behaviours

  • low c / high N → more accidents, poor decisons

what happens when you aggregate trait outcome association?

• when you move from individual associations → aggregate patterns:

  • effects become substantially stronger

  • small correlations accumulate across traits and outcomes

• examples:

  • combined trait predictions = r = .28-.43

do personality traits still predict outcomes when controlling for SES and cognitive ability?

• personality traits still predict life outcomes even when controlling for factors like SES and cognitive ability, effects typically modest but not too different

• many cases, evidence supports the conclusion that personlaity traits predict these ouctomes better than SES

• perosnlaity addes incremental preictive validty beyond traditonla predictors

what is the personlaity coefficient and why could it be unhelpdul?

• the personality coefficient refers to the idea that personality traits typically show small effect sizes (low correlations) with outcomes

• it is unhelpful as it can mislead people into thinking personality doesnt matter

• focuses only on single correlations, ignoring broader patterns

how replicable are personality trait-ocutome associations?

soto (2019)

• most trait outcome associations successfully replicate

• replication rates are higher than in many psychological fields

how does the replication compare to other fields?

• many areas of psychology have poor replication rates

• classic findings often fail to replicate

• personality research shows greater robustness and more consistent patterns across studies

• personality is considered methodologically stronger and more stanle than otehr subfields

which level of trait hierarchy tends to be most strongly linked with outcomes ?

• associations are quite facet specific

• for aggression

  • angry hostility facets of N seems to predict more aggressiveness whereas self consciousness predicts none

• for prediction of BMI in independent samples (vainik et al., 2019)

  • facets based predictions were 4x more accurate

  • single items (nuances) explained ~30% more variance

what is the overall pattern (“architecture”) of personlaity-outcome associations?

• associations are pervasive (traits linked to many life outcomes)

• at a broad level:

  • good traits → good outcomes

  • bad traits → worse outcomes

• at specific levlels:

  • associations are facet or nuance specific

• broad general pattern + specific underlying detail

what is the correlation/causation point?

• correlation does not inevitably imply causation

• just becase traits and ouctomes are linked

• does not mean one causes the other

• traits → outcomes

• outcomes → traits

• something esle cause both

  • example - genetic s

how does trait hierarchy relate to bandwith-fidelity dimenna?

• facets improve prediction but increase complexity and interpretation difficulty

• trade-off between simplicity and predictive accuracy

why should associations be independent of how traits are measure?

• if a trait is real and meaningful, then its association with outcomes should be stable regardless of how it is measures

• example

  • neurotisicms should predict ouctomes regradless of what facets are used

what is associations depend on measurement? what should we do ?

• associations change on depending on which facets/items are included

• implications

  • effects may reflect specific components, not whole trait

  • we should:

  • interpret results at the facet level or nuance level

  • avoid overgeneralizing to broad traits

what are reflective personality traits?

• traits that summarise patterns of behaviour, thoughts and feelings

• they are not distinct causes, but descriptions of consistent tendencies

how could traits and outcomes overlap?

• two ways

• conceptyal overlap

  • trait and outcome describe similar things

  • e.g. conscientiousness and academic success

• measurement overlap

  • same behaviours appear in trai questionnaires

  • outcome measures

why is personality science unqiuely positioned to map demographic variations?

• personality science is uniquely suited because it uses standardised trait models (e.g., Big Five) that allow direct comparison across individuals, groups and cultures (Costa & McCrae, 1992).

• traits are quantitative, enabling calculation of effect sizes, correlations, and variance explained

what else does personality science use to position it as suited approach for mapping demographic variation?

• also uses multiple methods (self + informant ratings) → helps separate real differences from bia (Vazire, 2010)

• large cross-cultural datasets → allows testing universality (McCrae et al., 2005)

what are the main patterns of gender differences? who is source?

• costa et al., (2001)

• across cultures

• women are higher in

  • Neuroticism (d ≈ .48)

  • Agreeableness (d ≈ .45)

  • Slightly higher Conscientiousness (d ≈ .11)

• men higher in

  • assertiveness (facet of extraverison)

  • risk taking

  • some aspects of openness

do gender differences match intuitions?

• match common stereotypes

• they also replicate across observers

• same pattern is observed for big 5 facets and overall big five

• differences between genders are small to medium

are gender differences large?

for individual traits

• gender differences are small to moderate

  • for example Neuroticism d ~=.40

    • schmitt et al., 2008

• when traits are combined

  • differences become larger

  • overall differences = 1.07 SDs (large)

what happens when you compound gender differences across a range of traits?

• helps more intuitive interpretations

• random man vs woman:

  • 77% chnace difference is in epxpected direction

• men vs women differ

  • ~30% more than two people of same gender

is big five the best level?

• big five is too broad

• as facet level rveals clearer differnces betwen men and women

• women are higher in warmth on extraversion

• whereas men are higher on assertivness

  • same trait but differenct facets

are gender differences similar in self - and informant ratings? source?

• McCrae et al., (2005)

• yes broadly consistent

• evidence

• self report

  • neuroticism = d =.48

• informant

  • neuroticism, = d.43

• same directions and similar magnitude

• BUT

  • informant effects sometimes smaller (e.g aggreeableness .45 → .25)

  • suggets some inflation in self reports

  • but differences are real

how do gender differences vary across cultures? source

• there is cross-cultural universality of gender differences - universality

• however gender differences are larger in more developed, egalitarian countries

• poorer or less equal countries → smaller differences

does this match intutions?

• not really because you would expect that as men and women became more equal, they should become more similar however data showed that more equality resulted in more personality differences

what might explain the size of these differences?

• from Schmitt et al., 2008

• social role theory (eagly, 1987) would have been a reaosn if match intuitions but it is not supported

  • social role theory says gender differences come from social roles

  • if roles became equal → differences should shrink

  • but reuslts showed the opposite

  • study challenges social role explanations as the main cause

• evolutionary explanations (Buss, 1997)

  • in freer sociaties, people can express natural tendencies more

  • in wealthy countries people have more freedom, less survival pressure and more oppurtunity to express personlaity

  • this might allow innate differences to show more strongly

• measurement/artifcat explanations

  • richer countries just measure personality better

  • better data quality in developed countries

  • but they found little support for measurement errors

age : what forms of stability and change in personality traits can you distinugsih and what are they?

• rank-order stability

  • peoples relative position compared to others

  • person A B C - high medium low at age 20

  • person A B C - high medium low ar age 40n

  • the order stays the same

  • this = high rank-order stability

  • even if te numbers change but order remians

• mean level change

  • average score of a group

  • is everyone chnaging overall ?

how do they relate to each other?

• personlaity is

  • stable in rank -order

  • and chnaging in mean-level

what are the main patterns of age differences? stable? unstable?

• rank order stability - personality traits show high stability across many years and even decades

• long tems correlations of personality across years can be around ~.60 over 15 years

  • even higher when measurement error is controlled

  • not completely fixed but not also constantly changing

• mean - level change

  • even though individuals sty relatively consistent compared to others, the average level of traits in the populations chnages with age

  • typical life-span pattern

  • ⬇Neuroticism 

    • Decreases with age (people become emotionally stable) 

  • ⬆Agreeableness 

    • Increases (more cooperative/kind)

  • ⬆Conscientiousness 

    • Increases (more responsible, organised)

  • ⬇extraversion 

    • Often decline slightly in later adulthood 

what kind of changes are they?

• gradual

• small

• most noticebale from teens → adulthood

what is cumulative continuity?

• personlaity becomes more stable with age, especially from adolescence to adulthood

why does stability increase with age ?

• life become more consistent

• people choose environments that fit their triats

• roles stabilise

• personlaity maturation

what age does personlaity become stable overall ?

• about age 30

are age differences large?

• overall, age differences in personality traits are small in magnitude, especially compared to gender differences

• evidence:

  • changes are roughly ~1 T-score point per decase

    • terraciono et al., 2005

  • changes with age are also small, e.g.:

    • Neuroticism: r ≈ −.14

    • Agreeableness: r ≈ +.08

    • Conscientiousness: r ≈ +.03

      • Chopik & Kitayama, 2017

• changes are gradual, supports the idea of gradual maturation

are age differences similar across methods?

• broad patterns are consistent across methods

  • self-reports

  • informnat reports

• magnitude differences as some effects are stronger in self and weaker in informant

• timing differences as changes may appear earlier in one method and later than another

what level of the trait hierarchy captures age differences best?

• facet-level is better than big five level

• some facets openness to feelings increases but opennens to values decreases

• studies have shown that different facets within the same trait follow different developmental patterns, this creates inconsistency across measures

how large are cohort effects?

• very small

• evidence for this includes

  • cross cultural simialrty

  • longitudinal vs cross sectionla agreement

• most differences reflect development, not generation

why is personality stable?

• genetics

  • twin studies show:

  • genes explain about half of personality stability

  • genetic influence becomes fully stable by audlthood

• environment

  • life experiences can affect personality

  • but many life events show little long-term personality change amd effects are often small otr temporray

  • challenges the idea that personlaity constantly responds to life events

  • rubber band (stretaches from life events) but snap backs

why could personality change?

• social invetsment theory

• people chang because they take on adult roles

• become workers, partners, parents

• adapt to social expectations

what are the main findings of cross cultural differences?

• country differences are: small

• rankings:

  • moderately reliable (r ≈ .50–.70)

do traits