Controversy In Personality Research Lecture 12

Front 1

What are some of the challenges to personality research?

Back 1

- unsound assumptions? (lexical hypothesis)

-usefulness of personality to predictions in psych.

- Issues surrounding questionnaire measures

- Personality could be an illusory construct.

Front 2

What is the ‘Forer Effect’ ?

Forer (1949)

Back 2

- An experiment that casts doubt on personality questionnaire validity.

- A vague/generalised personality profile. (Most participants agreed that the profile described them).

- The 'better then average effect' on all things positive.

Front 3

How does the Forer Effect work?

Back 3

-Descriptions are sufficiently generalised and amenable to multiple interpretations (‘one size fits all’)

- If the personality profile description is positive, people agree if is accurate of them.

-They also think it's more accurate if they believe in the authority of the source, ie parent over stranger.

Front 4

What does the Forer Effect teach us?

Back 4

-Shows that if personality tests were bogus the average test taker would not know

- Face validity is a poor guide to actual validity

Front 5

What is the ‘conceptual similarity’ critique?

Back 5

A critique of the validity of personality questionnaires…

Front 6

‘conceptual similarity’ =

Back 6

The problem of “how to classify people” becomes mistaken with “how people classify” If so, then personality traits (by factor analysis) simply reflect judgements of conceptual similarity

So ample, large and bulky ‘go together’ not because they similarly describe a target but because they similarly describe a concept

Front 7

Supportive data for the 'Conceptual Similarity' Critique...

Back 7

Summer camp experiment.

Sweder, 1975

Front 8

What were the key findings by Sweder, 1975....

Back 8

That conceptual similarity correlated the same as trait frequency counts. Suggesting that we really just profile peoples personalities based on general concepts- large, bulky etc.

Front 9

Problems with Sweders ‘conceptual similarity’ critique study...

Problem 1

Back 9

Shweder’s study had a methodological confound

Frequency counts versus

Likert scale. This is called common method variance

Front 10

What happened when Romer & Revelle, 1984 replicated the study and had both groups you frequency likert scales?

Back 10

Personality traits correlated more strongly with actual behaviour as opposed to conceptual similarities.

Suggesting that we can have trait based personality profiles and don't necessarily just use vague concepts.

Front 11

Problems with Sweders ‘conceptual similarity’ critique study...Problem 2

Back 11

- If trait measures did only reflect conceptual similarity, then how come… A judge gives different trait ratings for different people?

Front 12

What 's a possible study that Kenrick& Funder, 1988 suggest to support Romer & Revelle's fingings?

Back 12

Have multiple raters rate targets on personality-irrelevant descriptors (e.g., edible, comestible, digestible, etc)

Front 13

However, we did learn that....

Back 13

When we rate ourselves we have to decide between conceptual similarities such as talkative and sociable.

However as familiarity decreases (e.g., stranger vs. peer ratings), trait ratings are more influenced by conceptual similarity and less by the target’s characteristics.

So, conceptual similarities are not totally irrelevant to trait personality.

Front 14

Who sparked the Person- Situation debate?

Back 14

Walter Mischel in 1968

Front 15

What were Walter Mischels two key claims?

Back 15

Two key claims:

1. Personality weakly predicts behaviour (r ~ .30).

2. Behaviour varies considerably over situations.

Front 16

What did Walter Mischel conclude?

Back 16

The concept of a ‘personality trait’ is not useful for psychology. Behaviour is driven by situations, not persons (i.e., personality)

Front 17

Not every one agreed with W. Mischel...Why?

Back 17

Showing that traits are imperfect predictors does not mean they are not useful, or that situations are more useful.

Effects of situations on behaviour are, on average, no stronger (Funder & Ozer, 1983; Sherman et al., 2015). Average effect sizes across all of personality and social psychology equate to r ~ .21 (Bond et al., 2003) So, is a correlation of .30 “weak”?

Front 18

For example....

Back 18

Effects of situations on behaviour are, on average, no stronger!

In fact, average effect sizes across all of personality and social psychology equate to r ~ .21.

So, is a correlation of .30 for personality predicting behaviour really that weak??

Front 19

Claim 2 is true: Behaviour does vary across situations...

What did Will Fleeson's study find?

Back 19

He found daily fluctuations in personality ‘states’

The is huge variablity across time, however ironically there is also stability.

Even though states vary from day to day, average levels of behaviour over weeks are quite stable.

Front 20

What did Will Fleeson's 'mega-analysis' find?

Back 20

I actually don't know! - Sophie can we please discuss??

Front 21

What did Sherman et al 2015 find?

4 key findings....

Back 21

- State expressions (behavior/experience) vary widely between and within participants

- Situations experienced/encountered vary widely between and within participants

- Personality traits (modestly) predicted situational experience (situation selection)

- Personality and situations independently predict state expressions S

Front 22

What are the 2 main lessons learnt?

Back 22

1. Global personality traits are usefulfor predicting state expressions (i.e.,cognitions, emotions, and behaviors) across many situations

2. Single state expression by a given individual in a specific situationis substantially dependent on the characteristics of his or her situation

Front 23

What are Person by situation perspectives known as?

Back 23

Interactionism

Front 24

What is the situational strength hypothesis?

Back 24

Personality will predict behaviour in weak situations... however

Personality will be less predictive in “strong situations”,

Front 25

What are strong situations characterized by?

Back 25

1. Clear behavioural expectations

2. Incentives for compliance (or threats for non-compliance)

3. Individual ability to meet the demands of the situation

(Traffic light analogy)

Front 26

However some strong situations such as Milgram’s Prison experiment (1962) showed what?

Back 26

That strong situation do not completely drive behaviour: 65% delivered the maximum shock level, but 35% did not.

Therefore -Personality partly influenced obedience…

Front 27

And what about Begue et al's 2004 TV game show which was a replication of Milgram prison experiment?

Back 27

Same thing!

Even this ‘strong situation’ did not overwhelm personality effects.

Larger shocks predicted by Agreeableness, r = .26, p = .0239 Conscientiousness, r = .34, p = .006

Front 28

But could strong situations also “activate” relevant personality effects?

Back 28

Personality traits are probabilistic descriptions of regularities in behaviour and experience arising in response to broad classes of stimuli and situations (DeYoung, 2015) e.g., neuroticism predicts physical signs of nervousness more strongly in strong, trait-relevant, situations designed to elicit anxiety (Hirschmüller, et al., 2015).

Front 29

Strong situations or strong traits?

Back 29

Some strong situations activate as opposed to repress personality traits. eg: Neuroticism.

Front 30

Or put another way....

Back 30

People high on neuroticism will be effected by the situation while people low will not. This suggests that it is a strong trait, not the situation that predicts behaviour.

Front 31

What did the review by Cooper and Withey 2009 show?

Do any of these examples really test the situational strength hypothesis?

Back 31

No. Authors tend to simply state that the data seem to capture Mischel’s idea… Essentially no studies attempt to measure situational strength dimensions, i.e., participants’ perceptions of

1. Behavioural expectations

2. Incentives for compliance

3. Ability to meet the demands

Front 32

Rethinking Situational Strength

Conclusions =

Back 32

-Direct tests of the theory are needed

-The theory may need to be refined (e.g., to accommodate ‘trait activation’ effects)

- Perhaps best framed in terms of ‘joint effects’ of person and situation?

Front 33

What are some of the personality measurement issues?

Back 33

Trait measures appear to assume that people can accurately recall their typical behaviours and experiences. Eg, Do you often act on the spur of the moment?

However, trait measures correlate only moderately with momentary measures (rs ~ .45) Connor and Barrett 2012

Front 34

What are the conceptualisation issues?

Back 34

- How we frame our own Identity and reputation (Hogan, 2008)

- Personal narratives as myth (McAdams,1994)

How do we know we are how we believe we are?

Front 35

Cont:

-Personality traits assess:The ‘believingself’Are a relatively biasedrecordof actual experience

- Reflect semantic memory –how you conceptualise andunderstand your experience

Back 35

This reliance of self evaluation means that the data is floored.

Front 36

The Lexical Hypothesis

Allport & Odbert, 1936…(words from dictionary)

Catell, 1943… (Factor analysis)

Back 36

An implicit assumption: Aspects of personality that are most important for describing ones self and others will eventually become coded into language Primarily focussed on adjectives: warm, friendly, aloof, obedient, tense, calculating, secretive, brash… Later formalised as an explicit hypothesis

Front 37

What are some criticisms of the Lexical Hypothesis?

Back 37

- Adjectives are too simple…

- Adjectives are ambiguous… “Laypersonssometimes use the term ‘‘aggressive’’ to mean assertiveness oropportunity-seizing and sometimes to mean the expression of hostility.” (Block,1995)…

Front 38

What are some defences of the Lexical Hypothesis?

Back 38

-Personalities are complex, therefore we need complex and varies descriptors

- Theway ordinary individuals perceive themselves and others is, in fact, central tomost definitions ofpersonality

- Layman language is a problem of all questionnaire measures, not just personality.

Front 39

What is the most widely used personality inventory in history?

Back 39

Meyers-Briggs Type Indicator (MBTI)

which classifies people as types.

Front 40

What are Carl Jungs four basic 'psychic functions?'

Back 40

Perceiving

1: Sensation

2: (v's) Intuitions

Judging

3: Thinking

4: Feeling

Front 41

Who was the MBTI developed by and whose ideology was it based on?

Back 41

Developed by a mother-and-daughter (Katherine Cooks Briggs and Isabel Briggs Myers) with an interest in Carl Jung's philosophy.

Front 42

What are some of the criticisms of the MBTI

Back 42

1: Reliability (fewer then 50% get same type twice).

2: Predictive Validity (same as random population found in 'type' occupations)

3: Impartiality (very little academic research on the MB. Owned by a company $$)

Front 43

We no longer think in terms of types, but rather traits....why?

Back 43

Taxometrics:

This is unclear to me- let's discuss!

Front 44

What are some of the lesson we're learnt from Personailty Psychology

Back 44

1. Trait ratings are not free from bias (e.g., recall errors, implicit judgments regarding conceptual similarity)

2. Behavior is a complex function of both traits and situations

3. Our assumptions are often implicit and need to be carefully explicated and directly evaluated (e.g., the lexical hypothesis, situational strength) 4. Invalid measures may successfully masquerade as valid ones (e.g., MTBI, Forer effect)

Front 45

What is a current challenge of personality psychology?

Back 45

The “replication crisis”

25% replication rate forpersonality-social psychology (vs. 50% in cognitive psychology) (Nozek etal., 2015)

Front 46

What are some of the reasons for failed replication?

Back 46

-Low sample sizes (statistical power)

-Biases in methods and analyses

Front 47

Why is small sample size a problem?

Back 47

Small sample sizes are a problem from theperspective of statistical power

-Statistical Power = thelikelihood of correctly rejecting the null hypothesis

- Low power:Less likely to correctly reject the nullhypothesis (false negative)More likely to incorrectlyreject the null hypothesis (false positive)

-Publication bias: Journals are more likely to publishsignificant findings, so probably many false positives (vs. falsenegatives) in the literature

Front 48

What are some of the ambiguities in research that open thedoor to bias?

Back 48

-When to stop data collection?

-What to do with an outlier?

-What do do with conditions/measures that“didn’t work” ?

Theanswer to these will often be “let’ssee what effect it has on the results…”

Front 49

Ed Vul and colleagues (2009):Published correlations betweenpersonality traits and brain activity often approach singularity....

Back 49

However methodological factors such as imperfectreliability gives us reasons to doubt such strong associations!

Front 50

Biased methods and analyses seemto be a major factor is nonreplicabiliy...for example?

Back 50

Researchers who reported using‘non-independent’ analyses (more biased) tended to report the strongestfindings reviewed

Front 51

How has the personality psychology community responded?

Back 51

Become more self-critical..

-Special journal sections for replicationstudies -Guidelines for minimising‘researcher degrees of freedom’

-Methodological ‘wiggle room’ that createsbias toward significant effects

- Publication of null results

-Evaluation of journals and universitiesin terms of replications

Front 52

Problem:

Unreliable findings and questionablepractices undermines psychological science.

In recent times the problem has beenacutely felt in personality and social psychology

Back 52

Solution:

Less emphasis on p <.05

More emphasis on

-statistical power

-replication, and

-full disclosure of research practices.