Personality Structure and Stability week3
Structure Change & Stability of Personality
Personality Structure: Classifying Traits
Which Traits to Measure?
Aim for completeness without redundancy.
Classify all personality traits into a few large “groups.”
Each “group” consists of correlated traits.
This allows measuring personality thoroughly and efficiently.
Factor Analysis (FA)
Data reduction technique: aims to determine the number and nature of "underlying factors" that explain a pattern of correlations between large numbers of variables
Generates hypotheses
Statistical technique for model testing
Tests hypotheses
Ashton's EFA Example (Physical Fitness Tests & Physiological Measures):
Vertical jump (height of jump from a crouching position). vertical jump height (VJ)
40-yard dash (time to sprint 40 yards or 36.5 m). 40-Yard (36-m) dash time (dash)
Standing triple jump (distance of hop, step, jump from standing start). Standing triple jump distance (TJ)
12-min run (distance run in 12 min). 12-min run distance (run)
2-km row (time to row 2 km, or 1.25 miles, on rowing machine). 2-km (1.25-mile) rowing time (row)
20-min cycle (distance cycled in 20 min on standard exercise bike). 20-min cycle distance (cycle)
Percent fast-twitch muscle fiber (from tissue sample taken from thigh). Fast-twitch muscle fiber type percentage (type)
Body fat percentage (measured by skin-fold calipers). Body fat percentage (fat)
Correlation Matrix:
The correlation matrix displays the relationships between the variables, with values ranging from -1.00 to 1.00. Key variables include:
Vertical jump height (VJ)
40-Yard (36-m) dash time (dash)
Standing triple jump distance (TJ)
12-min run distance (run)
2-km (1.25-mile) rowing time (row)
20-min cycle distance (cycle)
Fast-twitch muscle fiber type percentage (type)
Body fat percentage (fat)
2-Factor Solution & Factor Loadings:
Factor 1: Jumping & Sprinting Ability
Factor 2: Endurance
A Personality EFA Example
Self-Report Questionnaire: How accurately do the adjectives below describe you? (1 = not at all, 5 = very much)
Anxious
Shy
Tense
Sociable
Active
Assertive
Correlations
Matrix showing correlations between personality traits like anxiousness, shyness, tenseness, sociability, activity, and assertiveness.
Factors and Factor Loadings
2 Factors, uncorrelated (orthogonal rotation)
Factor 1
Factor 2
Item Loadings in Factor Space
r(Factor 1, Factor 2) = 0
Neuroticism (N) Extraversion (E)
Eysenck’s early theory of personality (1947) had only two broad factors:
Extraversion-Introversion
Neuroticism-Stability
Note. Later, a third factor (Psychoticism) was added.
Summary 1
Personality models should aim for completeness without redundancy
Classify all relevant personality traits into a few large “groups” (broad factors) consisting of correlated traits
This allows to measure personality thoroughly and efficiently
The main statistical method used to create such models is factor analysis
How many traits?
2? 3? 11? 15? 16? 20? 22? 30?
See Ashton (2018, Chapter 2, pp. 46, Box 2-2)
The Idea of the Lexical Approach
The Fundamental Lexical Hypothesis
“The most important individual differences in human transactions will come to be encoded as single terms in some or all of the world’s languages.” (Goldberg, 1993, p. 26)
Original Idea: Galton (1884)
Early Attempts
Baumgarten (1933), in German
Allport & Odbert (1936), in English
18,000 English words
Traits: approx. 4,500
States: approx. 4,500
Evaluations (good/bad): approx. 5,200
Miscellaneous : approx. 3,600
Cattell:
First factor analysis of 35 traits/synonyms
Found 12 factors + added 4 later =16PF (1949)
Cattell's 16PF
Shows the different factors and how low and high score show up in terms of certain actions.
Factor Analyses
Tupes & Christal (1961, 1992): 5 factors
Norman (1967): 5 factors
Costa & McCrae (1985, 1992a): 5 factors
Goldberg (1990): 5 factors
Different samples, ages, languages, and nations → 5 factors
How many traits? Converging Evidence
Same 5 Factors with different measures:
Trait descriptive adjectives: "quite vs. talkative"
Statements describing feelings, cognitions, and behaviours: "I really enjoy talking to people"
Self-ratings vs. observer ratings
Nonverbal measures (e.g., cartoons [cf. Ashton, 2018, p. 48]; puppet interviews with young children [Measelle et al., 2005])
Five-Factor Model (FFM) of Personality - or - “Big Five”
Neuroticism (N)
Extraversion (E)
Openness to Experience (O)
Agreeableness (A)
Conscientiousness (C)
OCEAN !
Five-Factor Model (FFM) of Personality - or - “Big Five” (was first, FFM later)
Neuroticism (N) = Emotional Stability, reversed
Extraversion (E)
Openness to Experience (O) = Intellect/Imagination
Agreeableness (A)
Conscientiousness (C)
Q: What’s the difference? Very similar, but the FFM factor names are now used much more frequently!
Neuroticism (N)
Description:
Assesses maladjustment vs. emotional stability
Identifies individuals prone to …
psychological distress
unrealistic ideas
excessive cravings or urges
maladaptive coping responses (Pervin et al., 2005, Tab. 8.1)
*Sample Items: Shows how Very relaxed, at ease, stable, contented and unemotional shows up on the other extreme as tense, nervous, unstable, discontented and emotional.
Extraversion (E)
Description:
Assesses quantity and intensity of …
interpersonal interaction
activity level
need for stimulation
capacity for joy (Pervin et al., 2005, Tab. 8.1)
Sample Items: How silent, unassertive, unadventurous, unenergetic and timid relates to talkative, assertive, adventurous, energetic and bold.
Openness (O)
Description:
Assesses …
proactive seeking and appreciation of experience for its own sake
toleration for and exploration of the unfamiliar (Pervin et al., 2005, Tab. 8.1)
Sample Items: Items such as unimaginative, uncreative, uninquisitive, unreflective and unsophisticated are related to imaginative, creative, curious, reflective and sophisticated.
Agreeableness (A)
Description:
Assesses the quality of one’s interpersonal orientation along a continuum from compassion to antagonism in …
thoughts
feelings
actions
(Pervin et al., 2005, Tab. 8.1)
Low A High A sample items show that unkind, uncooperative, selfish, distrusful and stingy correlate to kind, cooperative, unselfish, trustful and generous.
Conscientiousness (C)
Description:
Assesses the individual’s degree of
organisation
persistence
motivation in goal-directed behaviour
Contrasts dependable, fastidious people with those who are lackadaisical and sloppy
(Pervin et al., 2005, Tab. 8.1)
Sample Items: The items disorganized, irresponsible, impractical, careless and lazy correlate to organized, responsible, practical, thorough and hardworking.
Why are the Big Five so important?
For example,
They help us sort out the confusion of different models and measures of personality that suggest 2 -30 factors/dimensions.
They provide a “framework” (or mapping system) to understand where in the personality space to place specific traits we want to know more about.
Example: Perfectionism
Hewitt & Flett (1991) Three different forms of perfectionism:
Self-Oriented Perfectionism
Other-Oriented Perfectionism
Socially Prescribed Perfectionism
Question:
Using the Big Five as a “framework”, how can we map – and thus better understand – the three different forms of perfectionism?
Self-Oriented Perfectionism
(Hill et al., 1997) shows correlation between N, E, A and C.
Other-Oriented Perfectionism
(Hill et al., 1997) shows correlation between N, E, A and C.
Socially Prescribed Perfectionism
(Hill et al., 1997) shows correlation between N, E, A and C.
Big Five Facets NEO Personality Inventory (NEO-PI)
Costa & McCrae (1985)
Derived from analyses of Cattell's 16 PF ð 3 Scales: N, E and O
NEO-Five Factor Inventory (NEO-FFI)
Costa & McCrae (1991)
Adopted five-factor personality model and added A and C ð 5 Scales with 12 items each: N, E, O, A and C
NEO Personality Inventory-Revised (NEO-PI-R)
Costa & McCrae (1992)
Measures 6 facets of each of the Big Five dimensions
ð 30 Scales with 8 items each: N, E, O, A and C domain and facet scores
Development of the NEO-Five Factor Scales (see John & Srivastava, 1999, for references)
Neuroticism (N) NEO-PI-R Facets:
N1: Anxiety
N2: Angry Hostility
N3: Depression
N4: Self-Consciousness
N5: Impulsiveness
N6: Vulnerability
Extraversion (E) NEO-PI-R Facets:
E1: Warmth
E2: Gregariousness
E3: Assertiveness
E4: Activity
E5: Excitement Seeking
E6: Positive Emotions
Openness (O) NEO-PI-R Facets:
O1: Fantasy
O2: Aesthetics
O3: Feelings
O4: Actions
O5: Ideas
O6: Values
Agreeableness (A) NEO-PI-R Facets:
A1: Trust
A2: Straightforwardness
A3: Altruism
A4: Compliance
A5: Modesty
A6: Tender-Mindedness
low high
Conscientiousness (C) NEO-PI-R Facets:
C1: Competence
C2: Order
C3: Dutifulness
C4: Achievement Striving
C5: Self-Discipline
C6: Deliberation
Example: Self-Pity
Big Five Dimensions and Facets Question:
Using the Big Five’s dimensions and facets as a personality model, how can we map – and thus better understand – individual differences in self-pity?
Self-Pity: Big Five (Overview)
(Stöber, 2003, Study 1) showing correlation between N, E, O, A and C.
Self-Pity: N (In-depth)
(Stöber, 2003, Study 2) N1: Anxiety N2: Angry hostility N3: Depression N4: Self-consciousness N5: Impulsivity N6: Vulnerability
Five-Factor Model of Personality
*Is the Five-Factor Model the final answer?
*Seems to capture Western personality language dimensions better than Eastern
*Does not always replicate, not even in Western cultures:
*“Big 5 plus or minus 2”
*But it provides
*A very useful taxonomic system
*A common language to communicate traits
*A “framework” to help map and better understand specific personality characteristics
What's missing?
Honesty–humility = the 6th factor? (Ashton et al., 2004)
HEXACO Model
Honesty–Humility
Emotionality
Extraversion
Agreeableness
Conscientiousness
Openness
*HEXACO-PI-R Facets:
*Sincerity Fairness • Greed-avoidance • Modesty
Big Five Neuroticism (N) vs. HEXACO Emotionality (E)
Neuroticism (N) Emotionality (E)
Anxiety Anxiety
Angry Hostility Fearfulness
Depression Dependence
Self-consciousness Sentimentality
Impulsiveness
Vulnerability Facet Differences
The HEXACO model is …
Important to know!
Makes a valuable contribution to research on individual differences and personality
Explains some selected personality characteristics better than the Big 5/FFM (see Ashton, 2018, p. 78, 2nd paragraph)
How I see the HEXACO …
How I see the HEXACO … (cont’d)
However, …
The differences to the Big 5/FFM factors are often confusing and not helpful (see, e.g., Ashton, 2018, Fig. 3.1 and Chapter 3.5.2, 2nd paragraph)
Ashton is overstating the importance of the HEXACO model:
Most personality researchers today work within the Big 5/FFM framework
Summary 2
It is important to know and understand …
how factor analysis (EFA!) works
the idea of the lexical approach and the fundamental lexical hypothesis
the five-factor model of personality
the differences between broad personality factors and factor facets
the HEXACO model of personality
Developmental Change and Stability of Personality
Defining Change and Stability
Two different questions:
How the average person changes across their lifespan → mean differences
How stable are differences between people in their levels of traits → correlations between time points
Hypothetical Example
Illustrates concepts of mean-level change and stability using personality scores of 4 individuals (A, B, C, D) over time. ** Let’s say we have:
4 individuals (A, B, C and D) &
their personality scores (possible range = 1-8)
Mean-level change, yet perfect stability (r = 1)
Demonstrates a scenario where all individuals increase their scores by the same amount, resulting in mean-level change but perfect rank-order stability.
1 2 3 4
M = 2.50
A B C D
Time 1
4 6 8 2
M = 5.00
A B C D
Time 2
Mean-level change, and zero stability (r = 0)
Illustrates a scenario where individuals' scores change in a way that alters their relative positions, resulting in mean-level change and zero rank-order stability.
1 2 3 4
M = 2.50
A B C D
Time 1
6 6 4 4
M = 5.00
A B C D
Time 2
NO Mean-level change, and zero stability (r = 0)
Presents a scenario where individuals' scores change, but the mean remains the same, and there is no rank-order stability.
1 2 3 4
M = 2.50
A B C D
Time 1
4 4 1 1
M = 2.50
A B C D
Time 2
NO Mean-level change, and perfect stability (r = 1)
Shows a scenario where individuals' scores remain in the same relative positions, resulting in no mean-level change but perfect rank-order stability.
1 2 3 4
M = 2.50
A B C D
Time 1
2 3 4 1
M = 2.50
A B C D
Time 2
Developmental Changes in Mean Levels of Personality Traits
Two approaches:
Cross-sectional studies
Longitudinal studies
Cross-Sectional Studies
1 time point (e.g., 2020/21) & different individuals
Pros: Efficiency (easy to do)
Cons: May reflect “cohort effects” (aka “generational effects”)
Longitudinal Studies
Different time points (spaced x years apart) & the same persons
Pros: Reflects changes in the same individuals across time
Cons:
Difficult/expensive to do
May reflect “historic effects” (aka “period effects”)
Personality Change and Stability
During adulthood
During adolescence & adulthood
During adolescence & young adulthood
In infancy & childhood
From childhood to adulthood
Changes in mean levels of traits over the life course
Roberts, Walton, & Viechtbauer (2006)
Meta-analysis of 92 samples!
*What’s a meta-analysis?
*Statistical technique for combining findings from independent research studies … and often thousands of participants
*Helps see overall trends
*Helps analyse where differences between studies come from
*Helps overcome the “file drawer problem” of unpublished studies
Mean Level Changes During Adolescence & Adulthood
Big 5 changes in mean levels of traits over the life course Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of emotional stability (N, reversed) Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of social dominance (E facet) Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of social vitality (E facet) Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of openness to experience (O) Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of agreebleness (A) Roberts, Walton, & Viechtbauer (2006)
Changes in mean levels of conscientiousness (C) Roberts, Walton, & Viechtbauer (2006)
Big 5 mean-level changes (Roberts et al., 2006)
All six trait domains show significant changes past the age of 30
Personality traits change most during young adulthood (age 20 to 40), not adolescence
Mean-level changes reveal a more complicated pattern than analyses of stability
HEXACO?
Remember …
Additional 6th factor = Honesty–Humility
HEXACO Emotionality ≠ Big 5 Neuroticism
HEXACO Agreeableness ≠ Big 5 Agreeableness
HEXACO
Emotionality (E) (Ashton, 2018, Fig. 4.2)
Extraversion (X) (Ashton, 2018, Fig. 4.3)
Openness (O) (Ashton, 2018, Fig. 4.6)
Agreeableness (A) (Ashton, 2018, Fig. 4.4)
Honesty-Humility (H) (Ashton, 2018, Fig. 4.1)
Conscientiousness (C) (Ashton, 2018, Fig. 4.5)
Summary 1
Regarding change and stability, you need to understand …
mean differences vs.
correlations between time points
and that they are independent and carry different information
the difference between cross-sectional and longitudinal studies
how mean levels of the Big 5 and HEXACO factors change across the life course
Why do we change?
Five Factor Theory (McCrae & Costa, 1999) Social Investment Hypothesis (Roberts, Wood, & Smith, 2005)
Humans have a species-wide genetic predisposition to develop in certain directions Investment in universal tasks of social living (e.g., getting a job, being married)
Hard-wired to become more socially dominant, agreeable, conscientious, emotionally stable, and less open to experience Similar tasks supported in most cultures
Normative experiences associated with increases in personality traits of social dominance, agreeableness, conscientiousness, and emotional stability aka Social Roles Hypothesis
Stability of traits
Shows question marks to represent instability across certain traits and phase of life.
Stability During Adulthood
Stability of traits in adults
*Costa & McCrae (1992): self-reports of adults, few weeks apart ð r » .80
*Costa & McCrae (1988): self-reports of adults, 6 years apart ð r » .70
*Costa & McCrae (1992): observer reports of adults, 6 years apart ð r » .70
*Costa & McCrae (1992): self-reports of adults, 24 years apart ð r » .65
*regardless of which trait is considered
*[Note. “»” means circa and on average.]
Stability During Adolescence & Young Adulthood
Stability of traits in adolescence and young adulthood Somewhat lower correlations!
Robins et al. (2001): 4 years in college ð r » .60
Roberts et al. (2001): between age 18 and 26 ð r » .55
McCrae et al. (2002): between age 12 and 16 ð r » .40 (.50 for C)
Self-esteem across life-span (Robins & Trzesniewski, 2005)
Mean-level changes, but high stability!
Self-esteem: Can relationships change personality?
(Neyer & Lehnart, 2007; see also Ashton, 2018, Box 4.1)
Personality in Infancy and Childhood:
Measurement & Structure
What about infants and toddlers?
*Researchers mostly examine “temperament” traits, not complex personality traits!
Caspi et al., 2005; Rothbart & Bates, 1998:
Activity level
Irritability
Fearfulness
Positive emotions
Attention span/persistence
Soothability
“Rhythmiticity” (preferring a clear daily rhythm/routine)
Measurement Problems
Only from 10 years of age, children describe themselves in terms of personality traits (before global “good vs. bad”)
Adult personality questionnaires demand reading comprehension not reached until adolescence
Many behaviours asked for in personality questionnaires not relevant for toddlers, children or even adolescents
Measurement Problems, cont’d
Observer reports (e.g., parents, teachers) may be problematic (e.g., “halo effects”, systematic bias, unobservable behaviours)
Measures in children are less reliable
Measures for child → adolescent → adult differ (method variance/non-equivalence) making means non-comparable and underestimating stability
cf. the puppet interview (Measelle et al., 2005)
Structure
*Despite these problems, …
*Good evidence for 4 of the Big 5 personality traits in children:
*Neuroticism
*Extraversion
*Agreeableness
*Conscientiousness
*… but not (so much) for:
*Openness
*Honesty–Humility
Personality structure in children (Lamb et al., 2002)
Developmental change in children (Lamb et al., 2002)
Trait stability in children (Lamb et al., 2002)
From childhood to adulthood
The
Summary of Personality Structure, Change, and Stability
Personality Structure: Classifying Traits
Factor Analysis (FA):
A method to find hidden factors that explain relationships between many variables.
Helps in creating and testing ideas about personality.
Lexical Approach:
The idea that important personality traits are found in the words we use.
The Five-Factor Model (FFM) or "Big Five" includes Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness (OCEAN).
HEXACO Model:
An alternative model that adds Honesty-Humility to the Big Five. exploring how language reflects these traits can provide deeper insights into individual differences and psychological well-being.
The HEXACO model is an alternative personality model to the Big Five. It includes six dimensions: Honesty-Humility, Emotionality, Extraversion, Agreeableness, Conscientiousness, and Openness.
Each of these dimensions contributes to a comprehensive understanding of personality and offers valuable perspectives on how these traits can influence behavior and interpersonal relationships.
Developmental Change and Stability of Personality
Change vs. Stability:
Change looks at how personality changes on average over a lifetime.
Stability looks at how consistent individual differences are over time.
Types of Studies:
Cross-sectional studies: Compare different people at one time.
Pros: Quick and easy.
Cons: Affected by generational differences.
Longitudinal studies: Follow the same people over time.
Pros: Shows changes in the same individuals.
Cons: Slow, expensive, and affected by historical events.
Why Do We Change?:
Social Investment Hypothesis: Suggests that investing in social roles (like jobs or marriage) leads to personality changes.
Research studies mentioned:
Roberts, Walton, & Viechtbauer (2006):
What it is: A study that combined the results of 92 different studies (a meta-analysis).
What it found: Personality traits change most during young adulthood (ages 20 to 40), not just adolescence. All six trait domains show significant changes past the age of 30
Costa & McCrae (1992):
What it is: self-reports of adults, few weeks apart ð r » .80
What it found: self-reports of adults, 6 years apart ð r » .70
What it found: observer reports of adults, 6 years apart ð r » .70
What it found: self-reports of adults, 24 years apart ð r » .65
*Regardless of the trait which is considered
Robins et al. (2001):
What it is: 4 years in college ð r » .60
What it found: Followed college students over four years.
Roberts et al. (2001):
What it is: between age 18 and 26 ð r » .55
What it found: Looked at people between ages 18 and 26.
McCrae et al. (2002):
What it is: between age 12 and 16 ð r » .40 (.