Leiter International Performance Scale – Revised Gale H. Roid Lucy J. Miller Presented by, Eirini Lammi – Registered Psychologist Revised 1997 Developed on the basis of widely employed hierarchical models of intelligence and cognitive abilities The content of the scale was later compared to the combined Cattell-Horn-Carroll cognitive model CHC model has become popular as the basis of several published cognitive tests including SB5, WRAT, Woodcock-Johnson Allows for cross battery assessment The Leiter-R was designed to measure several cognitive factors from CHC theory: Fluid reasoning Visual-spatial ability Some aspects of crystallised general knowledge (visual matching, familiar objects) Short-term memory Long-term retrieval Processing speed

Unifying Model of Nonverbal Intelligence At the apex of the hierarchical model is a general intelligence or “g” factor. At the second level, a smaller number of broad factors have been identified ( 3 or 4 by Gustafsson, 8 by Carroll) including the fluid and crystallized factors of Horn and Cattell At the second level were eight ability domains similar to those identified by Horn and Cattell (1966), Gustafsson (1984), Woodcock (1990) and others. These include fluid reasoning (Gf), crystallized ability (Gc), General visualization (Gv), various memory and learning domains (Gsm), auditory ability (Ga), two factors of processing speed (Gs), and long-term retrieval (Glr).

Definition of General Intellectual Ability “Intelligence is a multifaceted array of cognitive abilities with a general (g) overarching ability, composed of at least 8 major dimensions.” (Leiter-R; Roid & Miller, 1997) “The general ability to perform complex nonverbal mental manipulations related to conceptualization, inductive reasoning, and visualisation” (Leiter-R; Roid & Miller, 1997, p. 103) The term nonverbal is somewhat imprecise because individuals responding to the tests probably use verbally guided thinking to solve the visually presented tasks. Thus, the examinee probably uses a number of unspoken verbal abilities on the items

The Assumption of Comparability “When we test students using a standardized device and compare them to a set of norms to gain an index of their relative standing, we assume that the students we test are similar to those on whom the test was standardized.” Tests are culturally loaded Nonverbal tests have a degree of cultural bias Tests require communication

Acculturation and Language Differences “The difficulty with norms in the case of culturally and linguistically diverse children rests with the issue of what constitutes ‘representative’. All too often, race and ethnicity are equated with culture. Culture is neither, thus measured performance is more likely to reflect varying levels of individual acculturation more so than variation in actual or ‘true’ cognitive ability.” Flanagan & Ortiz, 2001

Definitions of Nonverbal What is a nonverbal test? A test that reduces language loading in directions, items, or responses. When should examiners use nonverbal tests of intelligence? Conditions associated with language differences, hearing impairment and disabilities indicate use of nonverbal tests. The non vocal administration of the scale makes it highly useful in special education

Recognizing the Limitations of Nonverbal Assessment Tend to measure a narrow range of intellectual abilities (eg, visual processing) May result in constructs that are restricted and fail to capture the full range of intellectual functioning Often don’t have adequate norm sample representation.

Referral Flow Chart Multi method assessment approach Obtain information from several sources The assessment process is a multi stage process: planning collecting data evaluating results formulating hypotheses developing recommendations conducting reevaluations following up on child’s progress

Generating a Hypothesis Collecting relevant information and data Knowledge of a child’s language proficiency Distinguishing language differences from disorders Assess and evaluate opportunity for learning Assess and evaluate relevant cultural and linguistic factors Evaluate, revise and re-test hypotheses Determine test protocol Clinical history: Birth Development Past medical School Family Social

Choosing appropriate battery of tests What is the purpose of the testing? When was it published? What was the standardization group? How reliable is the assessment measure? Are data presented about the performance of diverse groups on the test? What validity measures are provided? Assessment measures should have adequate reliability and validity for the situations in which they are being used Assessment measures should have representative standardization groups Assessment measures should provide information needed to answer the referral question and make useful recommendations The tests you select should be related to the referral question

Introduction of the Leiter-R Mainly used for individuals with poor verbal communication skills, hearing impairment, traumatic brain injury, English as a second language. To be precise, the Leiter-R should be described as measuring nonvocal cognitive abilities although it is classified as a nonverbal assessment in common psychology terminology

Standardization and Psychometric Properties Normative sample stratified on basis of gender, race, socioeconomic status, etc. VR battery normed on 1710 children, adoelscents, adults AM battery normed on 763 children, adolescents, adults – due to its smaller size, it has proportionately larger deviations from the US population The VR norms were based on an average sample size of fewer than 100 individuals per age level. Fewer than 300 subjects total were included in the 7 year age span between 14 to 20 years, with an average of 41 subjects per age level. The AM Battery included fewer than 100 examinees at every age level across the entire age span, with samples ranging from a low of 42 examinees at four age levels (7, 8, 11, and 18 to 20 years) to a high of 86 children sampled at the 2 year old level. At individual age levels, the samples are under or overrepresented by as much as 6% to 8% for individual stratification variables. The AM sample, given its overall smaller size, has proportionately larger deviations from the U.S. population.

Technical Properties Technical properties are generally good, with some variability. Average internal consistency for the VR subtests across the age levels range from .75 to .90; average internal consistency for the AM subtests are generally lower and range from .67 to .87. Composite reliabilities VR Battery range from .88 to .93 (FSIQ .91 to .93); composite reliabilities for the AM Battery also tend to be lower than the VR composites and range from .75 to .93. Reliability estimates for the rating scales vary considerably depending on the particular scale. Average internal consistency estimates across the age levels range from .90 (Anxiety), to .97 (Attention).

Validity Studies Almost all of the subtests are Fair or Poor measures of g. Moderate to strong correlations between Leiter-R Brief IQ and FSIQ (but no means and SDs were reported). Mean FSIQs for various 6-20 year old samples are lower than anticipated. McGrew and Flanagan (1998) classified 8 of the 20 subtests as containing high levels of cultural content A number of small sample studies (17 to 33 subjects) were conducted (Examiner’s manual) which compared the Leiter-R Brief IQ and FSIQ with aspects of various memory scales and achievement measures. The manual reports moderate to strong correlations, but the means and SDs are not reported. This makes it impossible to determine whether the tests produce comparable scores. The authors suggest the subtests fit the underlying hierarchical g model used to design the instrument. Loadings on the g-factor, almost all of the subtests are classified as either Fair or Poor measures of g. McGrew and Flanagan (1998) examined the prominence of cultural content embedded in various intelligence tests. They classified 8 of the 20 subtests as containing high levels of cultural content, and 3 subtests as having moderate levels of cultural content.

Organisation of the Leiter-R 20 subtests complete the Leiter-R. Not all subtests need to be administered at a given age. Rather, the clinician should administer only those subtests needed for selected composites at a specific age or those that will generate clinical hypotheses related to a specific case. This table lists the subtests in recommended order of administration for the four age levels

Hierarchical Model of the Leiter-R 2-5 years The Leiter-R includes an increasingly complex set of subdomains, as the age of the child progresses. In the age range of 2 and 5 years, the major factors include Fluid Reasoning (Gf), Fundamental Visualisation (Gv), and Attention. At ages 4-5, a separate factor of Recognition Memory appears, due to the introduction of the Immediate Recognition subtest at age 4.

Hierarchical Model of the Leiter-R 6-10 years At the ages 6-10, the number of factors expand to 6, as the number of available subtests increases. The factors are reasoning (Gf), visual/spatial (Gv), attention, and facets of short-term memory (Gsm) and retrieval (Glr), labelled Recognition Memory, Associative Memory, and Memory Span.

Hierarchical Model of the Leiter-R 11-20 years In the ages 11-20, there are five factors due to the deletion of Recognition Memory because it’s too easy for the adolescents.

Full Scale IQ Scores Brief IQ: obtained from Brief IQ Screener consisting of 4 subtests. Full Scale IQ: obtained from administering 6 subtests. Mean = 100, SD = 15 Two IQ scores are available. One can be obtained from administration of a Brief IQ Screener consisting of four subtests; the second is the Full Scale IQ (FSIQ) and can be obtained from administration of six subtests. These subtests vary depending on age. These scores use a mean of 100 and SD of 15.

General Ability IQ Scores Two general ability scores, which are nonverbal intelligence quotients (IQ) are measured on the Leiter-R. The IQ scores are the sums of the scaled scores for the subtests that compose the IQ estimate. The IQ scores are intended as measures of “g”, or general nonverbal intelligence. There are two nonverbal IQ scores, a Full Scale IQ and a Brief IQ Screener.

Composite Scores VR Battery produces 5 composite scores: Mean = 100, SD = 15 Brief IQ and FSIQ The VR Battery produces five composites; The AM Battery produces six composites, which include a Brief IQ Screener (ages 2 to 20)

Visualization and Reasoning Record Forms Visualization and Reasoning

Composite Scores Attention and Memory Battery composite scores: Mean = 100, SD = 15 The VR Battery produces five composites; The AM Battery produces six composites, which include a Brief IQ Screener (ages 2 to 20)

Record Forms Attention and Memory Battery

Subtest Scores 20 subtests First 10 subtests make up Visualization and Reasoning (VR) Battery Second 10 subtests make up Attention and Memory (AM) Battery Subtests Mean = 10, SD = 3 There are 20 subtests that together measure nonverbal intelligence and abilities, with a mean of 10 and SD of 3. The first 10 subtests form the Visualization and Reasoning Battery, measure traditional intelligence constructs such as fluid reasoning, visual-spatial abilities and problem solving. The second 10 subtests, the Attention and Memory Battery, measure attention, memory and learning processes.

Subtests in the Leiter-R Visualization and Reasoning (VR) Battery Attention and Memory (AM) Battery 1. Figure Ground 6. Repeated Patterns 2. Design Analogies 7. Picture Context 3. Form Completion 8. Classification 4. Matching 9. Paper Folding 5. Sequential Order 10. Figure Rotation 20 subtests complete the Leiter-R. Not all subtests need to be administered at a given age. Rather, the clinician should administer only those subtests needed for selected composites at a specific age or those that will generate clinical hypotheses related to a specific case. This table lists the subtests in recommended order of administration for the four age levels 11. Associated Pairs 16. Visual Coding 12. Immediate Recognition 17. Spatial Memory 13. Forward Memory 18. Delayed Pairs 14. Attention Sustained 19. Delayed Recognition 15. Reverse Memory 20. Attention Divided

Four Optional Social-Emotional Rating Scales All rating scales have the following composite scores: Composite: Cognitive/Social Composite: Emotions/Regulation Each of the behaviours is described in positive terminology eg “pays attention to details within tasks”. Therefore, a high score indicates more frequent positive or constructive behaviours. For the subscales, the scales scores range from 1 to 10, with 10 being average, or, of no clinical concern.

Examiner Rating Scale Each of the behaviours is described in positive terminology eg “pays attention to details within tasks”. Therefore, a high score indicates more frequent positive or constructive behaviours. Subscales: Attention, Organisation/Impulse Control, Activity Level, Sociability Composite: Cognitive/Social Subscales: Energy and Feelings, Mood and Regulation, Anxiety, Sensory Reactivity Composite: Emotions/Regulation

6 Steps in Interpretation Interpret the Full Scale IQ and Composite Scores 2. Determine the statistical significance of Composite Score differences 3. Determine whether the differences between Composite Scores are abnormally large Describe the performance at a composite level (i.e., FSIQ, Fundamental Visualization). Use qualitative description to interpret scores. Most examiners will be mostly interested in obtaining an IQ and may only administer the VR Battery. FSIQ includes aspects of reasoning, visualization, and spatial abilities, and is typically the best representation of the examinee’s overall or global intelligence, it is the first score to be examined. When there is significant variability in an examinee’s subtest/and or scale scores, the FSIQ may not serve as a good representation of global ability.

6 Steps in Interpretation 4. Interpret the significant strengths and weaknesses of the Profile 5. Subtest interpretation 6. Generate hypotheses about fluctuation in the profile 4. Begin by looking at the theoretical underpinnings of the test. If the global scales or composites exhibit statistically significant and abnormally large difference, interpretation proceeds along theoretical lines (e.g., Fluid Reasoning greater than Spatial Visualization or Associative Memory greater than Attention).

Case Study Psychological Evaluation Name: John Smith Age: 3 years, 10 months Language: English, Spanish Reason for Referral: John was referred by his parents because of speech and language difficulties, as well as assistance with educational planning.

Leiter International Performance Scale – Revised Gale H. Roid Lucy J. Miller Eirini Lammi lammi@acer.edu.au / (02) 8338 68040