PSYB30H3 Lecture Notes - Convergent Validity, Confirmation Bias, Discriminant Validity
5 views5 pages
Personality Assessment
The process of evaluating individual differences through
- tests
- interviews
- observations
involves the use of systematic, standardized procedures for observing, measuring, and recording behaviours and
personality features.
What makes a good personality test?
- Reliability across: will give you the same results each time you test the person
o Time
test-retest
o Items
Chronbach's alpha (greater than 0.7 but normal want greater than 0.8)
A similar to correlation – tells you how all the items of a test hang together
o Raters
Inter-rater reliability
- Validity: it measure what it says it is measuring
o Construct: test is actually closely aligned with the construct/concept you are trying to measure
o Face: does the test look like/ seem to measure what you want to
o Criterion: predictive validity – how reflective it is the test as a criterea for what you expect the person to do
Of special concern in test development:
Convergent validity: how well do your scores on one measure converge/are similar to the scores on
another measure that is used for the same purpose
Discriminant validity: if two measures are used for two different things then if the same person takes the
test then they should have different data on both… like there should not be a correlation between the two
Types of Personality Assessment
- Dream analysis
- Interviews
- Observation
- Objective Tests
o MMPI
o MBTI
o NEO - PI (r)
- Projective Tests
o Sentence completion
o Rorschach Inkblot Test
o Thematic Apperception Test
Dream Analysis
- Psychodynamic approach
- Freud: Dreams represented the unconscious desires of the dreamer
- "royal road" to the unconscious
- To reduce anxiety, ego disguises the true (latent) content of dreams using symbols (manifest content).
o Manifest content is analyzed to discover repressed feelings
Interviews
- Face-to-face meeting to get information about someone’s personality, current psychological state, personal history,
etc.
o Unstructured: conversation informal, topics discussed as they arise
o Structured: follow a pre-arranged plan, using series of planned questions
Can branch off and ask different questions based on the answers you get
o There are also semi-structured
Potential biases
- Confirmation Bias: making an inference at the outset & asking questions to confirm it
- Halo effect: making generalizations based on predominant features, which may be unrelated to personality
o Usually used for people that are really attractive
o Horn effect: the opposite – for people that have many negative qualities
- Social desirability: faking good, dissimulation (covering things up, not admitting things), faking bad
Observation
- assessing behaviour direct surveillance
- Selecting observers
o Professional personality assessors
Pro: well trained = better results
Con: little access
o People who actually know the target person
Often in better position to observe target's natural behaviours in a variety of contexts
May be biased because of relationship to target
Observations can be Naturalistic vs. Artificial Observation
- Naturalistic: observers record events that occur in the normal course of lives f the participants
o Pro: their natural reactions and behaviours; how it pans out in real life
o Con: not as controlled
- Artificial: occurs in artificial settings of situations
o Advantage: controlling conditions and eliciting relevant behaviour
o Con: may behave differently because they know that they are being watched and/or taped
Objective Personality Measures
- Involve standard set of questions or statements.
o "I am easily embarrassed: T/F"
- Responses compared to established standardized norms.
- From responses, develop a picture called a ‘personality profile’
- Possible response options:
o T/F or Y/N
o Dimensional scale: 1-5 scale etc
o Visual analog: graphic indicator but similar to dimensional scale above
o Adjective checklist: need to check of items from a list that answers the questions
Terminology
- Administration: giving someone the test
- Test, scale, measure, or instrument: the entire collection of questions intended to assess an aspect or aspects of
personality
- Item: an individual question of the test
Minnesota Multiphasic Personality Inventory-2ND Edition
- One of the most widely used personality tests
- Originally published in 1943
o to measure of psychopathology
- has been used for many predictive purposed, from personnel selection to marriage suitability
- Norms based on profiles of normal people and psychiatric patients
o Items were chosen based on ability to discriminate between clinical groups & controls