records the actual behav or person - electronic pagers - coding videotaped interactions - simple counts of specific behav

CHAPTER 3 Flashcards by N R

how to develop culturally useful measures

see if cut-off scores differ for groups i.e. “problematic” dependence differs
see if content is applicable

How well did you know this?

Not at all

Perfectly

what to do if no measure if validated for a culture

avoid using scores
use tests only for general hypotheses
explicitly acknowledge limitations to validity

How well did you know this?

Not at all

Perfectly

new vs standard pracitce

new practice: use measures validated for groups

standard practice: consult published norms in interpreting findings

standard clinical practice: use mult assessment methods to dec bias

How well did you know this?

Not at all

Perfectly

MMPI-2

565 items, T/F

7 validity scales to see if exaggerated, truthful, etc.

also scales i.e. dep, anx, schizo

scored by computer that interps patterns to resp

tells things like validity of profile, symptomatic patterns, interpersonal relations

How well did you know this?

Not at all

Perfectly

PAI

personality assessment inventory: used instead of MMPI-2

uses many of same scales but w fewer items and 4 POINT RATING

22 scales:
- 4 validity
- 11 clinical, interpersonal relations, therapy readiness

has critical items that need attention

How well did you know this?

Not at all

Perfectly

multidimensional perfectionism scale

looks at 3 types of perfectionism

OOP
SOP
SPP

How well did you know this?

Not at all

Perfectly

socially-prescribed perfectionism

belief that others make unrealistic demands of self

inc anx, distress, anger, dep

How well did you know this?

Not at all

Perfectly

self-oriented perfectionism

unrealistic demands of self, assoc w ED, anx, proscrastination

How well did you know this?

Not at all

Perfectly

other-oriented perfectionism

expectations of others, assoc w narcissism

parents of OOP parents inc risk of dev SOP or SPP

How well did you know this?

Not at all

Perfectly

NEO PI-R

designed to assess personality characteristics in five factor model

FFM used more in research than clinically

How well did you know this?

Not at all

Perfectly

types of self report checklists

SCL-90-R: 90 items w 9 subscales
- indicates distress past 2 wks
- may overpathologize takers

outcome questionnaire 45 (OQ-45): 45 items, 3 subscales, 5 mins
- i.e. symptom distress, interpersonal relations
- valid across many pops

california q-sort (CAQ): 100 personality descs
- compare characteristics to self

How well did you know this?

Not at all

Perfectly

3 judgment by others measure types

behavioural obs
interviews
life stories

How well did you know this?

Not at all

Perfectly

interviews

unstructured: rich info, questionable validity…preferred by interviewees

structured: inc valid, not see nuances

How well did you know this?

Not at all

Perfectly

life stories

rich info via analysis of writings i.e. soc med, letters, diary

How well did you know this?

Not at all

Perfectly

behavioural obs

records the actual behav or person

electronic pagers
coding videotaped interactions
simple counts of specific behav

How well did you know this?

Not at all

Perfectly

hare psychopathology test

is a revised PCL-R test….predicts recidivism, violence, responsiveness to therapy

20 item scale, 40 score max

clinican carries out review and gets info from ppl close to individ

1% meet criteria vs 15-25% of ppl in prison
- 10-15% in prison almost psychopathic

How well did you know this?

Not at all

Perfectly

psychopathy

superficial charm
grandiose
lack empathy

absence of fear in social and phys situations that scares others

projective tests

presents ambig stim, task, situation

wants to access unconscious mind as taker projects their fears, hopes, etc. onto test

sentence completion

i.e. my mother thinks i’m

TAT

thematic apperception test

show pics and ask for a story
- reveals themes, expectations, defenses
- ask before, now, future

rorscach ink blot test

10 cards presented in specific order

4th most used test, by 82% clinicians

scoring system is OUTDATED

now use R-PAS

R-PAS

shows high inter-rater reliability

mixed findings abt VALIDITY, prob bcs some studies don’t specify scoring methods

R-PAS is empirical
- use computer to score
- available cross-cultural norms
- difficult to fake resp

pros/cons projective tests

pros:
- good to break ice
- hard to fake
- skilled clinicians can get deep info

cons:
- time costly
- expensive
- sus and mistrust scientifically

objective test

personality test of qs, esp if uses computer answer sheet

T/F, Y/N, or numbered scale

what determines number of questions in objective tests

spearman brown formula principle of aggregation: sum of many measures is more stable than a single measure

3 methods of making objective tests

1. factor analytic 2. rational 3. empirical

rational method

items are chosen based on FACE VALIDITY that they relate to content - based on theory, gives S data 4 conditions for validity: 1. willingness to make acc report 2. items are valid indicators of construct 3. capability for self-assessment 4. items mean same thing to taker/writer

empirical method

gather many items and divide sample into groups...administer test, then compare answers - use cross validation

factor analytic method

generate a long list of objective items, administer to ppl, then analyze w a factor - find clusters of related traits i.e. put music into clusters like "pretentious" then give test to find correlated items has been used to find fundamental traits i.e. FFM

ways to evaluate research

1. significance testing 2. effect size 3. correlations 4. replication

significance testing

stat significance: result only occurs by chance less than 5% of the time p-level: probability of getting the same outcome from a test if variables were not manipulated

null hypothesis vs NHST

null hypothesis: there is no difference b/w groups NHST: null hypothesis significance testing - trad method of stat data analysis - determines chance of getting result if nothing manipulated problems with NHST: - 0.5 is abritrary - nonsignificant results can be interpreted as no result - only gives info on TYPE 1 ERROR

type 1 vs type 2 error

type 1: false positive type 2: false negative mistakenly believe variables have this effect on e/o

effect size

number reflecting degree to which a variable is related to another more meaningful than p lvl imp bcs sometimes small effect size may be stat sig but not meaningful in practical terms i.e. only affects certain membs of pop

correlations

b/w +/-1 0 = no relationship correlation coefficient: deg to which pos/neg 1 in linear fashion confidence intervals: estimate of range within the value of a stat probably lies

p hacking

manipulation of data analysis until is produces a stat sig result i.e. can't replicate precognition article

replicability

to make more dependable: - use large numb participants - share all data and methods - report studies that don't work - never regard 1 study as definitive proof

publication bias

tendency for research w strong results to be published

purposes of personality testing

help ppl, select employees, learn abt ppl i.e. gov arguments against it: - rude to be defined as scores - can discourage ppl from joining fields - fields may not grow if similar ppl join - tests can control ppl responses to criticsm: - we're always going to be judged - tests are more valid than informal judgments

deception and privacy

deception allowed, but rare in personality research privacy concerns w B data i.e. soc med

uses of psych research

should have good > harm honesty and open science: avoid plagiarism and fabrication