Assessment L4 - PAI Flashcards
(30 cards)
What is the PAI?
Personality Assessment Inventory
- Self administered, OBJECTIVE inventory of adult personality (bc it actually tries to determine the reliability of clients’ response).
- provides info on critical clinical variables
- includes VALIDITY scales - are they performing consistently across q? are they being genuine?
- Orthogonal scales
What does the PAI consist of?
- 344 items - so many questions - very difficult for respondent to try and manipulate
- Comprising 22 non-overlapping full scales 4 Validity scales
- 11 Clinical scales
- 5 Treatment Consideration scales
- 2 Interpersonal scales
takes approx 50 mins to complete
- personality scales are equivalent to the big 5, but also allows us to be able to evaluate the likelihood that this person will benefit from treatment. The degree of suicidality, if their condition impinges on their function, and how responsive they will be to treatment.
Interpretive report will also allow for a DSM diagnosis - so we can use it to make diagnostic arguments
What is the goal of using PAI? what kind of information will it produce?
allows us to be able to evaluate the likelihood that this person will BENEFIT FROM TREATMENT. The degree of suicidality, if their condition impinges on their function, and how RESPONSIVE they will be to treatment.
Interpretive report will also allow for a DSM diagnosis - so we can use it to make diagnostic arguments
What are the applications of the PAI?
- PAI developed and standardised on a sample of adults ages 18 years+
- written at 4th grade reading level
- 344 items take approx 50 mins to complete
PAI administration and scoring?
- easy to administer and score
- can be administered by technicians who are trained in the admin of self report tools
What are the critical items in PAI?
- 27 critical items
- they are identified as indicators of potential CRISIS SITUATIONS and have very low endorsement in normal sample
- critical items facilitate follow up questioning! Just this assessment may not be enough - need further questions
How is the PAI interpreted?
- Interpretation should only be performed by professionals trained in the interpretation of psychological tests
- Interpretive guidelines are provided in the Professional Manual
- Interpretive software is also available that provides a comprehensive individualised report
Describe the development of the PAI
- based on construct validation framework emphasising RATIONAL as well as QUANTITATIVE framework.
- strong emphasis on the dev and selection of items
- non-overlapping nature of scales (ORTHOG)
10 STEPS:
1) Test users surveyed, literature reviewed,
and current diagnostic schemas used to
develop item content
2) Scale content areas defined
3) Scale and subscales selected
4) Development of initial 2,200 items
5) Reduction of item pool to 1,086 items
based on research team ratings
6) Review by external bias panel and expert judges reduced item pool to
776 items
7) Alpha version (776 items) analyzed
8) Item deletions and revisions to produce a 597-item beta version
9) Beta version tested and analyses conducted
10) Standardization version (344 items) produced
following further reliability and validity analyses
GOOD SCALE DEV!!! empitically determined whether these items distinguish between condition
Why is it good that the PAI is orthogonal?
Bc there are a lot of questionnaire type of measures which actually include a smaller number of items, but weight some items more on a scale…
Problematic bc it makes it non orthog - causes conflation between conditions bc the items are the same - better to have 1 item per scale.
–> not enough differentiation between items –> conflation of diagnosis
Reliability of PAI?
- internal consistency of alphas across 3 samples
- median alphas for full scales are 0.81, .82, .86 for the normative, college and clinical samples. respectively.
- pretty good!!!!!!
wais is 0.96 and 0.98. but diff construct.
- median test-retest was 0.83 across all samples
- mean absolute T score change tended to be 2 to 3 T score points for most full scales.
PAI Validity scales?
- Four Validity scales built into PAI
- Inconsistency and Infrequency assess
DEVIATION FROM CONSCIENTIOUS RESPONDING - Negative Impression and Positive Impression assess IMPRESSION MANAGEMENT.
—> assessing how reliable the client’s responses are
How good is PAI Validity?
- 1,000 computer-generated random response protocols were compared against profiles from the three subsamples and there was marked separation
- Fake-good and fake-bad studies also conducted on a sample of 90 adults
- Comparison of fake-good and fake-bad profiles to those from the normal and clinical samples revealed clear separation
- Response set studies also conducted using 1,000 computer-generated profiles for each response set type
- These simulation studies suggest that the Validity scales would capture a large majority of invalid protocols
What is the process of administering a PAI
- Administration
- Look at omissions- 17/344 = CAN’T SCORE
- Consistency of endorsement = if unnaceptable, stop
- accuracy of item endorsement = if underreporting or overreporting psychopathology, stop, or proceed with caution
- interpret configurations for self-concept, interpersonal style, and perception of environment
- interpret clinical scales
- examine critical items
- interpret treatment planning configuration.
What are PAI-SP clinical reports?
- PAI-SP software generates PAI Clinical Reports
- Clinical Report includes PAI full-scale and subscale profiles
GIVES SKYLINE REPORT!!! - t scores - mean = 50, SD= 10
- Includes all 22 non-overlapping scales
- 4 Validity scales: ICN, INF, NIM, and PIM
- 11 Clinical scales: SOM, ANX, ARD, DEP, MAN, PAR, SCZ, BOR, ANT, ALC, and DRG
- 5 Treatment Consideration scales: AGG, SUI, STR, NON, and RXR
- 2 Interpersonal scales: DOM and WRM
What scales does the PAI have
- Includes all 22 non-overlapping scales
- 4 Validity scales: ICN, INF, NIM, and PIM
- 11 Clinical scales: SOM, ANX, ARD, DEP, MAN, PAR, SCZ, BOR, ANT, ALC, and DRG
- 5 Treatment Consideration scales: AGG, SUI, STR, NON, and RXR
- 2 Interpersonal scales: DOM and WRM
What is the ‘skyline’?
This is 2SD’s above the clinical population.
If someone goes above it, it is a sign of malingering, lack of self awareness, over estimation or a cry for help.
What does a low score on a subtest mean?
Simply that they performed badly on one subtest.
Can’t conclude anything from the fact that they didn’t do well from just one subtest.
What subtest scores should we pay attention to, to indicate treatment responsivity.
Low scores on drug and alcohol, or preparedness to treat.
What does the PAI Clinical report subscale profile include?
Includes nine Clinical scales and one Treatment Consideration scale with their respective subscales
- Somatic Complaints: SOM-C, SOM-S, and SOM-H
- Anxiety: ANX-C, ANX-A, and ANX-P
- Anxiety-RelatedDisorders: ARD-O, ARD-P, and ARD-T
- Depression: DEP-C, DEP-A, and DEP-P Mania: MAN-A, MAN-G, and MAN-I Paranoia: PAR-H, PAR-P, and PAR-R -
- Schizophrenia: SCZ-P, SCZ-S, and SCZ-T
- Borderline Features: BOR-A, BOR-I, BOR-N, and BOR-S
- Antisocial Features: ANT-A, ANT-E, and ANT-S
- Aggression: AGG-A, AGG-V, and AGG-P
What’s the difference between looking at the full scale profile and the subscale profile?
full scale has 22 scales, subscale has 10.
subscale gives you subscores, which can give you overall score for the super group - this allows you to be able to unpack the nature of the response - eg. when some are high and low within a group.
full scale gives you overview of all the supergroups. but no subscales.
What is the inconsistency scale?
- validity scale
- 10 item pairs, each with related content - same questions asked in diff ways
- measures RESPONDENT CONSISTENCY
- If a person responds inconsistently, it undermines you faith in reliability.
- If they’re reliably responding on these, we infer theyre reliably responding on others.
What scores on the inconsistency scale suggests inconsistency?
- below 64T suggests consistent
- 65-72T = some inconsistency
- 73T+ = NOT CONSISTENT!!!! call invalid. this is 2 SD above mean.
What is the infrequency scale?
- 8 items that are very low endorsed.
- 4 very true and 4 very false.
- measures RANDOM responding, INDIFFERENCE, CARELESSNESS, CONFUSING AND READING DIFFICULTIES
What scores on infrequency scale indicate a problem?
59T or less = approriate attention and comprehension
60-74T = some atypical responding
75T+ = call INVALID
