Psychological Testing  

Introduction

Review of the course: tests- WAIS-III, WMS-III, MMPI, PAI, TAT

Format: lecture + TA

Grading + requirements: Class participation, TA participation, Midterm take home, Take home Final

Contact:

Psychological assessment -> behavior / appearance / relatedness, affect, mood, thought process, speech, cognitive functioning,  psych Sxs, defensive organization, intelligence, memory, psychodynamic formulation, DSM Dx

Concept of Intelligence

Multidimensional construct

Biopsychosocial construct ->genetic-environmental-psychological

Intelligence – Cognitive functions – Ego functions – Defensive organization – personality / self

Temperament (brain / nervous system, genetic)  – Character (internal core, largely unconscious) – Personality (public persona , mainly pre / conscious) 

Weschler (1944) – “capacity to act purposefully, to think rationally, and to deal effectively with one’s environment” – global entity and an aggregate of specific abilities. (1975) – “Nonintellective factors, including abilities to perceive and respond to social, moral, aesthetic values” – personality traits

Global entity g + specific abilities

Intelligence-intellectual ability-information processing-working memory-cognitive functions-ego functions<-personality (motivation, impulsivity, perseverance, anxiety, cognition, thinking, affects, etc), Self

Working memory – person’s information processing capacity

Statistical Properties

Age-adjusted norms vs Reference-group norms (20-34) (general population)

I.Reliability – accuracy, consistency, stability 

measurement error – inversely related to reliability

“True score , error of measurement , confidence intervals

Test-retest stability range .70ies - .90ies, split-half reliability ranges .88 - .97, interscorer agreement for V S C >.91

Score differences - two different aspects:

1. Is difference real and not due to measurement error? - statistical significance of the score - (.15 and .05 levels of significance)

2. Is the difference clinically meaningful? Frequency in general population –how “abnormal” the score / difference is.

Score differences :

-between IQ scores and Index scores – Tables B.1 and B.2

-between single subtest score and an average of scores Table B.3

-between subtest scaled scores B.4

-intersubject scatter – (max-min)

-Digit Span Forward and Backwards: longest span Table B.6; Differences Forward/Backwards Table B.7

II. Validity – the degree to which the test measures what it claims to measure.

1. Content validity=content coverage (content sampling is adequate)+ content relevance content is related to actual abilities)

2. Criterion related validity – correlations with other measures (concurrent or predictive)

correlations with other measures:

WAIS-R: 

FSIQ=.93 (WAIS-III 2.9 lower); VIQ=.94 (WAIS-III 1.2 lower) PIQ=.86 (WAIS-III 4.8 lower)

Verbal subtests: .76 - .90; Performance subtests .50 - .77

WISC-III (sample of 16 year olds)

FSIQ=.88  VIQ=.88 PIQ=.78  (statistically significant)

Verbal subtests: .60 - .83; Performance subtests .31 - .80

Standard progressive Matrices

FSIQ=.64 VIQ=49, PIQ=79, VCI=.55, POI=.65; PSI=.25; Matrix reasoning=.81

Stanford-Binet Intelligence Scale (global SB-IV composite)

FSIQ=.88; VIQ=.78; PIQ=.89;

WIAT and WMS-III (see the technical manual)

3. Construct validity (convergent + discriminant validity) – the extent to which the test measures the theoretical construct /attribute of interest. Meaningfulness of the test score.

Intercorrelatins of WAIS-III subtests, scales and indexes (reflective of g factor)

(see copy of page 98 in TM) all are statistically significant

Factor analyses – exploratory + confirmatory -> 4 factors / indexes (5 factors include quantitative/numerical ability)

Convergent / divergent construct validity – selected external measures of:

-cognitive ability– WAIS-R, WISC-III, Standard Progressive Matrixes, MicroCog, DRS (Table 4.25 p.120)

-attention / concentration – WMS-R, Trail-Making Test, Halstead-Reitan Neuropsychological Battery, MicroCog (Table 4.26)

-memory – WMS-III, California Verbal Learning test, Rey-Osterrieth Complex Figures Test, MicroCog Memory Index

-language – Boston Naming Test, Multilingual Aphasia Examination, Category Naming Test

-Fine Motor Speed / Dexterity – Grooved Pegboard, MicroCog,

-spatial processing – MicroCog, Judgment of Line Orientation, Rey-O,

-executive functioning Wisconsin card Sorting Test (problerm solving strategies, cognitive flexibility, feedback use).

4. Special Group Studies [discuss in the interpretation section]

-Neurological disorders – Alzheimer’s, Huntington’s disease, Prkinson’s disease, traumatic brain injury, MS, temporal lobe epilepsy

-Alcohol-Related Disorders – dependence, Korsakoff’s syndrome

-Neuropsychiatric Disorders (schizophrenia)

-Psycheducational and Developmental Disorders – MR, LD math / reading

-Deaf and Hearing Impaired

Subtests and Indexes

[see the table]

Scoring

Evaluate each response item -> raw score(s)

Tally  up raw scores à total raw score for each subtest

Transcribe raw scores ---à scaled scores (age-adjusted)

Tally up selected scaled scores  --à means, indexes and IQ scores

Calculate strengths and weaknesses

Generate graphs

Discrepancy Analysis

Interpretation

Empirical, cognitive, clinical

Empirical –>

loadings on General Factor

Specificity of subtests (unique vs. shared variance)

FSIQ – general level of intellectual / global cognitive functioning

VIQ – acquired knowledge, verbal reasoning, attention to / processing of / verbal materials

VCI – “pure” verbal acquired knowledge and processing, verbal conceptualization and expression (V, I, S)

WMI – information processing capacity, number ability and sequential processing  (A, DS, LN)

PIQ – fluid reasoning, spatial processing, attentiveness to detail, visual-motor integration / processing.

POI – nonverbal, fluid reasoning, attentiveness to detail, visual-motor integration (PC,BD, MR)

PSI –  speed of visual information processing (DS-Coding, SS)

I. Full Scale IQ

IQ, confidence intervals, percentiles

    < 69              Extremely Low

70 – 79             Borderline

80 – 89             Low Average

90 – 109            Average

110 – 119          High Average

120 –129           Superior

130 +                Very Superior

FSIQ -> VIQ + PIQ

VIQ = VCI [V+S+I] + WMI [A+DS+LN] + C

PIQ = POI [PC+BD+MR]  + PSI [Cd+SS] + (OA)

II. Statistical significance of differences of IQ and Indexes [‘’real” vs. chance error]

In NP mean VIQ – PIQ = 8.6; VCI-POI = 9.7

VIQ – PIQ min 9 points [.05] or 12 points [.01] 

VCI – POI min 10 [.05] or 13 [.01]

If < 9/10 -> verbal and nonverbal skills are fairly evenly developed.

III. Frequency (abnormality=the extreme 15% of population [85 percentile]) in normal population NP of differences of IQ and Indexes Table B.2 ASM

VIQ – PIQ min 17

VCI-POI min = 19

IV. Is VIQ – PIQ Interpretable? (Global ability responsible for scaled scores on subtests?)

1.         VCI-WMI min 10 points [.05]; 13 points [.01] if >= 10/13 -> VIQ not a unitary construct

POI-PSI min 13 [.05]; min 17 [.01] if >= 13/17 ->  PIQ not a unitary construct

2.         Subtest scatter -> min-max scores within VIQ and PIQ (separately) < 8? If  >= 8 do not interpret (not a unitary construct)

V. Is VCI – POI interpretable?

Subtest scatter -> min-max scores within VCI < 5?; within POI < 6? If >= 5/6 do not interpret (not a unitary construct)

VI. Are WMI and PSI interpretable?

Subtest scatter -> min-max scores within WMI < 6?; 4 within PSI? If >= 6/4 not a unitary construct.

VIIes. Interpret Global IQ and Indexes

Horn / Cattell Fluid-Crystallized Intelligence – Gf; Gc; Gsm (short memory), Gv (visual scanning)

Performance > Verbal -> Billingualism; Learning Disability

1. Verbal > Performance -> Depression; Multiple Sclerosis; Alcoholism, Alzheimer’s

2. Working Memory -> (Freedom From Distractibility) attention, concentration, anxiety, sequencing ability, sequential processing,                                                                        number ability, planning, STM, executive functions, hyperactivity

3.  Processing Speed (DS + SS)  ->  cognitive + visual psychomotor speed, motivation, perfectionism, anxiety, impulsivity, compulsiveness, reflectiveness, planning ability (errors), visual memory

VIII.  RR 4.15. Uniqueness of each individual (ipsative scores) – scatter around the average (mean)

If VIQ – PIQ >=|17| - use two separate means (all Verbal tests administered and all Performance tests administered)

If VIQ – PIQ <= |17| use the Total mean of all subtests administered

Check significance of the difference from the mean(s) in the ASM Table. B.3.

Determine percentile score for each subtest

Determine Strengths (S) and Weaknesses (W)

IX. RR 4.16 (p.148) Subtests Scatter -  Abilities shared with other subtests -

Cross-refer strengths and weaknesses from VIII against cognitive functions / abilities

Compare scores on all subtests for any S and W (just higher or lower than the mean is sufficient)

Strong vs weak abilities - for 2 subtests -> 2/2 above/below  mean; for 3-4 subtests min 2 above/below mean, no more than 1 equal mean; for 5+ subtests – min 4 above/below mean 1 equal / above / below mean

X.   Subtest Scatter – Individual subtests - if no shared abilities have been identified à interpret individual subtests

[consider test specificity and loading on General Ability (g)]

            Ample specificity [unique variance]: DS [.50]; MR [.39]; Cd [.38]; PC [.35]; A [.30]

            Adequate specificity: PA [.31]; BD [.27]; I [.23]; C [.20]; S [.20]; V [.19]

Clinical Interpretation

Multi-level observation and interpretation

---> test-taking behavior

--àintellectual / cognitive functions 

--à symptoms [of DSM –IV Axis I disorders] ( see Table)

---à defensive organization / personality style (DSM spectrum)

(See the Table)

Two main approaches:

1. Statistical/ actuarial data

2. Symptoms (DSM-IV) appear and/or interfere with test responses

No 1:1 correspondence between test results and clinical categories - what is the most plausible and elegant explanation of the results and the scatter. WAIS is a 1st step in the hypothesis generation process, other more specialized tests are always needed.

Clinician needs to know what he/she needs to find (diagnostic symptoms) to infer that a clinical category may be present.

Quantitative vs qualitative data / interpretation

FSIQ – general level of intellectual / global cognitive functioning. GLOBAL EGO STRENGHT / LEVEL OF FUNCTIONING

VIQ – acquired knowledge, verbal reasoning, attention to / processing of / verbal materials [ VCI + WMI + C]

VCI – “pure” verbal acquired knowledge and processing, verbal conceptualization and expression (V, I, S)

WMI – information processing capacity, number ability and sequential processing  (A, DS, LN)

PIQ – fluid reasoning, spatial processing, attentiveness to detail, visual-motor integration / processing [POI + PSI + (PA) + {(OA)}]

POI – nonverbal, fluid reasoning, attentiveness to detail, visual-motor integration (PC,BD, MR)

PSI –  speed of visual information processing (DS-Coding, SS)

Neurological Disorders

Site + severity

Attention (focus, shift, sustain) +  memory + selected cognitive deficits [apraxia, aphasia, agnosia, executive (planning, sequencing, organizing, abstracting) functions

Indexes in the 70 – 95 range

PSI < WMI / POI < PIQ < VIQ

Deficits contrast with premorbid or expected level of functioning

Alzheimer’s Disease

Progressive, chronic, gradual loss of cognitive functions

Cortical atrophy of temporal-parietal and frontal regions of the brain + hippocampus and amygdala

Impairment in declarative memory, inefficient encoding and storage without prominent retrieval deficits

PSI < PIQ < VIQ

Huntington’s Disease

VCI relatively intact, PSI / WMI < PIQ < VIQ

Parkinson’s Disease

PSI < POI < WMI < VIQ intact

Psychotic Disorders

Schizophrenia (neuropsychiatric disorder) = gross neurological damage

Positive (hallucinations, delusions, thought disorder) and negative (alogia, anhedonia, thought disorder) symptoms

Attention, abstract reasoning, word fluency, sequential memory, verbal memory

Indexes in the 80 –90 range, PSI WMI more impaired

Developmental and Psychoeducational Disorders

Mental Retardation

2 – 3 SD below average Mild - IQ = 50 - 70; Moderate –  IQ = 35 – 55, Flat / low scatter

Arithmetic, Symbol Search, Vocabulary, Coding are the most impaired

ADHD

Inattention OR hyperactivity / impulsivity

A deficit in response inhibition

Attention, concentration, freedom from distractibility are impaired

“SCAD” profile [SS, Coding, (PSI),+ A, DS}

FSI Average

VIQ – PIQ nonsignificant

WMI < VCI up to 1 SD

PSI < POI up to 1 SD

Higher of VCI or POI vs lower of WMI or PSI – the difference of 1SD for 61.3 % vs 30% in NP, [2SD for 16.1 % vs 3.5%]

Most problems with: S-Coding DS, SS, LN-Sequencing.

Learning Disabilities

Achievement < IQ

Spatial > conceptual >acquired knowledge > sequencing (Bannatyne’s model)

“ACID” profile – Arithmetic, Coding, Information, Digit Span

WMI and PSI are most impaired

FSIQ is Average

WMI < VCI up to 1 SD

PSI < POI up to 1 SD

__________________________________________________________________________

Diagnostic spectrum (prompted by Mayman et al. reading):

An important caution about use of the WAIS to get at personality structure: may tell you more about the kind of character / disturbance than about the level (see Alevels of disturbance,@ at end of this note set).  E.g., may not be able to tell symptomatically obsessional person from essentially non-troublesome, obsessional character style.

Reason for including other tests, especially Rorschach: they are more helpful in fixing level of disturbance. 

But Rorschach has limits in this regard, too!  Can tell you about capacity for regression; may not tell fully about how overtly functional / dysfunctional someone is. 

Psychosis occurs in a variety of character settings, so can=t assume that all elements of psychotic profile result from the psychosis.

Note the Gartner et al. reading: a sign of borderline personality disorder is relatively clean WAIS, with thought-disordered Rorschach.  On less pathological end of spectrum, clean on all the tests.  In psychosis, distortions even in the WAIS.