Textbook Notes (280,000)
CA (170,000)
UTSC (20,000)
Psychology (10,000)
PSYB30H3 (500)
Chapter 4-5

PSYB30H3 Chapter Notes - Chapter 4-5: Criterion Validity, Mirror Test, Personality Test


Department
Psychology
Course Code
PSYB30H3
Professor
Lisa Fiksenbaum
Chapter
4-5

This preview shows pages 1-3. to view the full 10 pages of the document.
Chapter 4 Personality Assessment
- Personality Assessment is the measurement of the individual characteristics
of a person. Most commonly used are personality tests.
o In clinical settings used to develop modules of disorders, intervention
programs, treatments and diagnosis. 
What Makes a Good personality test?
- Developers of personality test must demonstrate that the test is valid and
reliable and specify the conditions, populations, and cultures the test applies
to
- Developers must also provide theoretical background and research evidence
to show the test is related to certain outcomes.
- Results should not be due to biased responding this evidence should be
published in scientific journals
- Legitimate personality tests have
o Reliability
o Generalizability
o Validity
o Supported by research evidence that is published to review
- Good tests allow replication
- Allows us to make inferences about behavior
Test Reliability
- Reliability is an estimate of how consistent a test is: good test gives
consistent results overtime
- Consistency across time, items and raters.
- Temporal consistency reliability: have respondents take the test a second
time to se e if their scores are similar.
- When demonstrating test- retest reliability be careful that participants are
not remembering what they previously said- no practice effects
o Second test should be later in time to avoid these
- Internal Consistency reliability: if different items of the test give similar
results
- Parallel forms of reliability: 2 versions of a test that we compare and check to
see that the scores on the tests were similar
- Split have reliability: spilt a test in half and see if a test takers scores on one
half correlated with scores on the other half
- Cronbachs Alpha: TODAY taking the correlation between the scores of 2 half
tests and calculating the average correlation of all possible haves of the rest
o Estimates the generalizability of the score from on set of items to
another (alpha of .07-.08, sometimes higher like in IQ when testing
individuals
- Interrater reliability: measures are reliable across multiple raters
- Often calculate the average correlations among the scores of all raters or the
percentage of agreement- if consistent, correlations are high
Test Validity
find more resources at oneclass.com
find more resources at oneclass.com

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

- Validity: extent to which a test measures what its supposed to measure.
Many types
- Every test aims to measure a construct- derives from a theory
- Construct Validity: measure the theoretical concept the test was deigned to
measure
- A test has face validity when it appears to measure the construct of interest
o Useful under 2 conditions
Important for personnel testing, or other situations where the
cooperation and motivation of the test taker can affect the
results of a test. When test has face validity respondents view
test as fair, relevant- take test seriously
Researchers disguise contents of test to lessen possibility of
fake/bias answers.
- Criterion Validity: determines how good a test is, compares the results of the
test to an external standard
- Need for Cognition Scale: measures individual differences in the need for
enjoyment of effortful thinking
o To establish criterion validity, compared 2 groups of participants
assumed to differ in their preferences for thinking differences
existed, demonstrating criterion validity
- Convergent validity: test is similar to other tests of the same, related
constructs.
- Discriminant validity: our test looks at a different concept
- To establish construct validity demonstrate what a test measures and what
is doesnt measure.
- Barnum Effect: people rapidly believe things that are relevant to all-
experiment where prof gave same feed back (in regards to personality) to all
students, after which they rated the feed back and all but one said he was
correct
Test Generalizability
- Generalizability: establishes that boundaries or limitations of a test
- I.e.: mental illness test cannot lest fitness
- Challenge is to ensure that test is generalizable for populations other that
college students
- USING AN UNRELIABLE TEST WILL MAKE IT HARD TO FIND A TRUE
EFFECT; USING AN INVALID TEST WILL RENDER THE RESULTS USELESS
NEO-PI-R Test
- Page 85- review and look at charts to understand
- To check reliability Costa and McCrae report Cronbachs alpha for each of
the factors.
- To check test-retest reliability college students too the test at 2 types, 3
months apart
o Correlations were high indicating reliability
- Construct validity; does the test measure 5 factors of personality?
o Extraversion, openness, agreeableness, conscientiousness
find more resources at oneclass.com
find more resources at oneclass.com

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

-
- Costa hypothesized that individuals in psychotherapy would score high and
that drug abusers would score low on agreeableness and conscientiousness-
they do.
- High correlations on Neo score and their friends ratings of their personality
demonstrates criterion Validity
- Demonstrates convergent and discriminant validity by correlating scores
with other personality tests
- Cautioned that if more than 40 questions unanswered, scores may not be
valid
- Generalizable to populations, adults and elders (not under 18), clinical
settings and in other cultures.
PERSONALITY TESTS
Types and formats of personality tests
Self reports: respondent answers questions
- May use dichotomous two choice scale (yes/no), likert type (rating
agreement), degree, similarity, frequency, scale point, checklists, visual
analog scales: measuring intrinsic and extrinsic motivation
- Adjective checklist: based on the # and items checked
- Forced choice format: present respondents with a limited number of
choices (locus of Control Scale and Machiavellianism Scale) Machiavelli
believed people were too trusting- research conducted on this scale suggest
that people who are high in the scale are nor more hostile, vicious or
vindictive than low machs- they have a cool detachment, emotional issues
and experienced embarrassing situations. Forced choice controls for the
effects of social desirability
Performance based: unstructured format in which participants respond to
ambiguous stimuli- associate own meaning, interpretations etc. that serve as part of
answers
- Rorschach inkblot rest, thematic appreciation test and often used in clinical
and forensic setting controversial
- 5 categories of projective techniques
1) Association techniques- word association tests, like above inkblot test
2) Construction techniques- draw a person test (TAT test)
3) Completion techniques
4) Arrangement or selection of stimuli test
5) Expression techniques
- Used to measure motives- need for affiliation, power and mostly achievement
- Goodenough Harris Draw a Person (DAP) test may be interpreted according
to the presence or absence of certain elements (size of eyes indicate
paranoia)
Response Sets
- self-reports tests can give us access to an amazing amount of information
about various aspects of peoples thoughts and experiences. easy to
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version