Lecture 5 Flashcards
What high-stakes situations may people distort their responses?
o Employment selection
o Internet dating (e.g. RSVP)
o Selection for education
Is response distortion conscious or unconscious?
Both. People don’t always have an accurate representation of themselves (false sense of who they are)
What are the different kinds of response distortions?
♣ Self-deceptive enhancement (unconscious)
♣ Self-deceptive denial (unconscious)
♣ Impression management (conscious)
What are the types of biases in self-deceptive thinking?
- Egoistic bias (“Superman”)
- Moralistic bias (“Angel”)
What is egoistic bias?
o Exaggerate one’s status (social, intellectual, physical, etc.)
o Pro-self
o Value = agency
o Strong, competent
What is moralistic bias?
o Deny socially deviant impulses and claim sanctimonious, saint-like attributes
o Pro-social
o Value = communion
o Good, kind
What is egoistic bias linked with?
Linked with:
Self-deceptive enhancement
What is moralistic bias linked with?
Linked with:
Self-deceptive denial
What are the two types of faking?
- Faking good
- Faking bad
What is faking good?
Trying to get higher test scores
What is faking bad?
Trying to get lower scores
What situations may people try to fake good?
o Employment selection
o Education selection
o Dating/inter-personal evaluations
What situations may people try to fake bad?
o Legal context
♣ Obtain benefits
♣ Diminished responsibility
o Educational context
♣ Special treatment
o Military
♣ Discharge, special duties (especially conscription)
♣ WW2: “Black Psy-Ops” – dropped leaflets in enemies’ languages giving instructions on malingering to soldiers to get out of military serviceCan intelligence tests be faked?
Intelligence tests can’t be faked “good” but they can be faked “bad”
According to Mittenberg et al. (2002), what is faking suspected in?
♣ 30% of personal injury cases
♣ 30% of disability claims
♣ 19% of criminal evaluations
♣ 8% of clinical/medical cases
What are the two types of methods to deal with response distortion?
- Detecting fakers
- Reducing faking
What are the 4 methods for detecting fakers?
- “Lie” scales (a.k.a. Social desirability scales)
- Response time rubrics
- Over-claiming technique
- Bayesian truth serum
What are the 5 methods for reducing faking?
- Forced-choice formats
- Verifiable statements (especially biographical data)
- Other reports (e.g. self-assessments, referee assessments)
- Warnings
- Implicit measurement techniques
What is the basic idea/assumption of lie scales?
Everyone does, or has done at least once, these things (test items) so then, if you don’t admit to them, you are lying
What are the 4 lie scales?
o Marlow-Crowne Social Desirability Scale
o Minnesota Multiphase Personality Inventory (MMPI) Lie Scale
o Eysenck’s Lie Scale
o Paulhus Balanced Index of Desirable Responding (BID-R)
What is the issue with the basic assumption of lie scales?
Some people are just really nice and never do anything bad/wrong (i.e. have never done any of the test items)
What is a problem with the lie scales?
o Lie scales relate to substantive personality traits
♣ Possible interpretation: Lie scales do not measure lying but may measure actual aspects of personality
• Non-trivial correlations with
personality traits
♣ High scorers = nice people (not liars)
• Higher impression management and
self-deceptive enhancement
scores = higher conscientiousness,
higher agreeableness
o Statistically large correlations
• Kicking people out based on their scores actually means excluding the nicest people
What are the 3 other ways to detect fakers?
- Response time rubrics
- Over-claiming technique (Paulhus)
- Bayesian truth serum
What are response time rubrics?
o Participants response times measured because, theoretically, people take longer to respond because they are lying
o People pause to think about whether to fake an item (longer response time = faking?)
♣ Response time changes based on strategy – accuracy or lying?
o BUT people who decide to fake everything (rather than exaggerate from their own “real” personality) are quicker
What is the over-claiming technique?
o Ask people to rate their familiarity with a number of concepts
o Compares familiarity of “real” terms with “foils” to determine if people are “overclaiming”
What is the issue with the over-claiming technique?
People who know a little bit of the concept can be tricked because foils are similar to real terms they are vaguely aware of but are not necessarily lying
What is the assumption of the over-claiming technique?
Assumption that people who lie on this will lie on everything
♣ Ignores situational factors (e.g. really honest people may be desperate for the job so lie on application personality test)
What is the Bayesian truth serum?
For each (dichotomous) item, test-takers estimate the proportion of people who would answer the same way
What is the issue with the Bayesian truth serum
o False Consensus Effect (from Social Psych): People OVER-ESTIMATE the number of others who share their beliefs
What are the 5 types of warnings to reduce faking?
♣ Detection warnings
♣ Consequences warnings
♣ Reasoning warnings (convince test-takers that answering honestly is in their best interests)
♣ Educational warnings (convince test-takers of the validity of the tests and selection procedures)
♣ Moral warnings (appeal to the conscience of the test-taker)
What is ‘forced choice’ method to reduce faking?
o Force test-takers to choose from equally desirable alternatives
♣ E.g. Which is more like you: “I am kind to others” OR “I work hard”
What is the issue with the ‘forced choice’ method to reduce faking?
Scores on forced-choice measures have ipsative properties
♣ Relative profile for everyone but not the exact level
• Can’t be compared across people (only accurate for the person tested)
What is the ‘verifiable statements’ method to reduce faking?
o Asking people questions where their answers are “checkable”
♣ People are less likely to lie when information can be easily verified
What are the ‘implicit measurement techniques’ method to reduce faking?
Measures unconscious association between oneself and words representing a personality trait (e.g. Conscientiousness)
What are the paradigms in faking research?
- Group comparison
- Instructed faking
- Incentive manipulation
What is the ‘group comparison’ paradigm in faking research?
o Compare job applicants to others (job incumbents/student samples/community samples)
♣ High stakes vs. low stakes
o Measure lower limit of faking: not all job applicants will fake
o There may be real group differences (e.g. psychology students are not like normal people…)
What is the ‘instructed faking’ paradigm in faking research?
o Compare scores under “answer honestly” vs. “maximise your score” instructions
o Many different instructions for “maximise your score” – can lead to different findings
o “Answer honestly” instructions account for impression management but may still have self-deceptive biases
What is the ‘incentive manipulation’ paradigm in faking research?
o Compare scores under no stakes conditions with scores obtained when there is an incentive to do well
♣ E.g. The top 10% of scorers in the experiment will receive $10
What is the problem with the ‘incentive manipulation’ paradigm in faking research?
o Cannot create a high-stakes situation
♣ Hard to mimic real-life incentives like getting a job
How much does faking change scores?
- Can have potentially big effects
How should personality tests be used for selection to deal with faking?
- Selection ratio
o Personality tests should only be
used for ‘select-out’, not
‘select-in’, processes
What personality traits do people fake the most?
- Conscientiousness
- Neuroticism
