📘 Lecture 23 — Item Analysis
📖 Overview: This lecture expands on item analysis by introducing advanced concepts beyond basic difficulty and discrimination indices. It covers Item Response Theory and item-characteristic curves for graphical item evaluation, explains cross-validation for assessing test validity on new samples, and discusses qualitative analysis methods for gathering test-taker feedback. The lecture concludes with a review of core item analysis concepts.
🗂️ Topics Covered
This lecture covers Item Response Theory (IRT) and its alternative names, the construction and interpretation of Item-Characteristic Curves including positive slope, negative slope, and inverted-U patterns, the process and factors affecting Cross Validation, the areas explored in Qualitative Analysis of tests, and a Review of Item Analysis covering item difficulty and item discrimination indices.
📝 Lecture Summary
Item Response Theory
Item response theory (IRT) is an approach that considers the probability of answering each individual item in a test correctly or incorrectly. The information regarding each item is plotted graphically. This approach is also known as Item-Characteristic Curve Theory and Latent Trait Theory. The graph containing information about the items is called the item-characteristic curve. Decisions regarding the items of a test can be based upon this information.
🔑 Definition — Item Response Theory (IRT): An approach that takes into consideration the probability of answering, right or wrong, each individual item in a test, with information plotted graphically.
Item- Characteristic Curves
Item difficulty and item discrimination can be presented graphically using item-characteristic curves. These are graphs where the horizontal axis (X axis) represents the ability being tested, and the vertical axis (Y axis) represents the probability of correct responses or the proportion of examiners responding correctly to the item. According to Kaplan and Saccuzzo (2001), it is “a graph prepared as part of the process of item analysis. One graph is prepared for each item and shows the total test score on the X axis and the proportion of test takers passing the item on Y axis” (p. 637). The shape or slope of the graph or curve indicates whether the item is a good one or not, and how far it discriminates high scorers from low scorers.
A good item has a positive slope. The proportion of high scorers responding correctly is higher than the proportion of low scorers. More of the low scorers are not responding correctly. A steep slope indicates that the test discriminates between the two groups. Scores of a highly discriminating test will yield a very steep slope.
A negative slope means that the item is not a good one. The item does discriminate between the two groups, high scorers and low scorers, but in the opposite way. This graph indicates that more of the low scorers did the item correctly than the high scorers. The probability of answering the item correctly is higher for the low scorers rather than the high scorers. This, therefore, is a bad or poor item that needs to be removed or replaced.
Another type of poor item is one where the majority of neither the top scorers nor the low scorers do the item correctly. It is the middle, moderate, scorers who attain the maximum proportion of correct responses. This type of item is also a bad item.
💡 Why this matters: Item-characteristic curves provide a visual and immediate way to assess whether an item is functioning as intended, distinguishing between knowledgeable and less knowledgeable test-takers.
Cross Validation
The validity of a test may be determined from a sample that was used for item selection. However, to have a better estimate of the validity of the test, the entire test needs to be validated on different samples as well. This process is called cross validation. According to Cohen and Swerdlik (1999, p.246), “The term cross validation refers to a revalidation of a test on a sample of test takers other than the ones on whom test performance was originally found to be a valid predictor of some criterion.” The regression equation is used to predict performance in a sample of test takers who are different from the ones on whom the test was validated.
If validity is computed from the original sample, the validity index may be higher than expected from a new sample due to possible chance variations. It is expected that validity will shrink in the process of cross validation. “The amount of decrease in the strength of the relationship from the original sample to the sample with which the equation is used is known as shrinkage” (Kaplan & Saccuzzo, 2001).
Factors that may affect the amount of shrinkage include:
- The size of the original item pool
- Proportion of test items retained
- Sample size
A high validity coefficient can be expected if the original item pool was large while the proportion of retained items is small. The size of cross validation sample also affects shrinkage. Greater validity shrinkage may be expected if smaller samples are used.
🔑 Definition — Cross Validation: The revalidation of a test on a sample of test takers other than the ones on whom test performance was originally found to be a valid predictor of some criterion.
🔑 Definition — Shrinkage: The amount of decrease in the strength of the relationship from the original sample to the sample with which the equation is used.
Qualitative Analysis
Qualitative analysis of a test may also be conducted along with quantitative analysis. After test administration is over, the test takers may be asked questions about various aspects of the test. These questions can be asked and answered orally or in writing. Different formats can be adopted, such as interviews and discussions. The respondents’ responses can be of great help in improving the individual test items, test format, and the entire test itself. Cohen and Swerdlik (1999) have pinpointed areas that may be explored, including:
- Cultural sensitivity
- Face validity
- Test administrator
- Test environment
- Test fairness
- Test language
- Test length
- Test taker’s guessing
- Test taker’s integrity
- Test taker’s mental/physical state upon entry
- Test taker’s mental/physical state during the test
- Test taker’s overall impressions
- Test taker’s preferences
- Test taker’s preparation
Review of Item Analysis
Methods used for assessing and evaluating characteristics of test items and the test itself. Primarily two characteristics are measured: item difficulty and discriminability.
Item-Difficulty Index: The item difficulty index is either in the form of percentages or proportions of the total number of test takers who attempted an item correctly. Item difficulty is calculated separately for every item. It is denoted by a lowercase italicized ‘p’. A number attached as subscript to this ‘p’ indicates the item number whose difficulty level is described. For example, p₁ indicates item difficulty of item number 1, p₂ is the difficulty level of item number 2, and so on. On occasions, the term ‘facility index’ may be used rather than difficulty index, referring to the percentage of responses to correct choices. Both terms refer to the same procedure.
Item Discrimination: A test is supposed to discriminate between those who know and those who do not know; those who score high and those who score low; those who have acquired a skill and those who have not. A test will not be a good test if the people who are supposed to know the correct answer fail and those who are not supposed to know succeed. A good test differentiates between the high and low scorers. If some items are correctly answered by high scorers and some by low scorers then something is wrong with the test.
Item Discrimination Index: Every test has its discrimination power. To see if the test discriminates between high and low achievers, a certain percentage of the high and low achievers is taken. The discrepancy between their attempts to correct responses is calculated in terms of percentages. The item discrimination index is denoted by a lowercase italicized letter.
⭐ Key Takeaways
You must understand that Item Response Theory (IRT) focuses on the probability of answering each item correctly and uses item-characteristic curves for graphical analysis. The slope of an item-characteristic curve is critical: a positive slope indicates a good discriminating item, a negative slope indicates a poor item where low scorers outperform high scorers, and a curve peaking in the middle is also poor. Cross-validation is essential to estimate a test's true validity on a new sample, and the expected decrease in validity is called shrinkage, which is influenced by the original item pool size, proportion of retained items, and sample size. Finally, qualitative analysis complements quantitative data by gathering test-taker feedback on areas like cultural sensitivity, test fairness, and their mental state, helping improve the test overall.
🧠 Quick Revision Questions
- What are the three alternative names for Item Response Theory mentioned in this lecture?
- In an item-characteristic curve, what does a negative slope indicate about an item?
- Define cross-validation and explain why the validity of a test is expected to shrink during this process.
- Name three factors that affect the amount of shrinkage in cross-validation.
- List five specific areas that can be explored during the qualitative analysis of a test.
📘 Lecture 24 — Assessment of Intellectual and Cognitive Abilities
📖 Overview: This lecture introduces the complex concept of intelligence and its measurement. It explores foundational questions about what intelligence is, why we measure it, and whether tests are reliable and valid. The lecture then provides a comprehensive historical and theoretical overview, from early pioneers like Galton and Binet to modern multi-factor theories by Spearman, Thurstone, Gardner, Sternberg, and Goleman, establishing the essential context for understanding intelligence testing.
🗂️ Topics Covered
The lecture begins by posing fundamental questions about the nature, origin, and measurement of intelligence. It then defines intelligence and reviews its key biological and environmental influences. The majority of the lecture is dedicated to a chronological and theoretical survey of major intelligence theories: Spearman's g-factor, Thorndike's three divisions (social, abstract, concrete), Thurstone's seven primary mental abilities, Cattell's crystallized and fluid intelligence, Guilford's Structure of Intellect model, Gardner's multiple intelligences, Sternberg's triarchic theory, the concept of emotional intelligence, and Piaget's stage theory of cognitive development.
📝 Lecture Summary
What is intelligence?
The lecture begins by stating that intelligence cannot be answered with a single word or sentence. It defines intelligence as "the capacity to understand the world, think rationally, and use resources effectively when faced with challenges" (Feldman, 2002). Intelligence refers to the ability to adapt, reason, solve problems, and think abstractly; it also includes learning from experience. The lecture emphasizes that intellectual ability is based on a constant interaction between environmental factors and inherited potentials, with modern psychology viewing both heredity and environment as influential.
Theories of intelligence
One of the earliest contributions was made by Sir Francis Galton, the "founder of differential psychology." In his work "Hereditary Genius" (1869), he proposed that gifted individuals tend to come from families with other gifted individuals. He attempted to measure intelligence quantitatively to determine the role of heredity and also investigated the relationship between intelligence and head size, though this found no empirical support. Cattell, an American psychologist, focused more on mental processes and first used the term “mental test” for devices measuring intelligence, developing tasks for reaction time, word association, and weight discrimination. Alfred Binet developed the first proper formal intelligence test in 1905, which will be discussed in detail in later sessions.
Spearman’s g-factor theory
Charles Spearman proposed one of the earliest theories, suggesting intelligence is a single, general factor. Using factor analysis, he observed that people scoring high on one mental test also tend to score high on others. He proposed two factors: the “g” factor (general intelligence) and the “s” factor (specific intelligence). The g-factor accounts for the general ability common in all people, while the s-factor accounts for specific abilities that differ among individuals.
🔑 Definition — g-factor: The general intelligence factor, a single, overarching mental ability believed to underlie performance on all cognitive tasks. 🔑 Definition — s-factor: Specific intelligence, referring to unique abilities that vary across different individuals and mental tests.
Thorndike’s Social Intelligence
Edward Thorndike criticized Spearman’s g-factor approach, arguing that intelligence consists of multiple factors expressed in human actions. He proposed three main divisions of intelligence:
- Social intelligence: Enables one to understand and manage relationships.
- Abstract intelligence: Enables one to understand and manage ideas such as algebra, mathematics, or abstract concepts.
- Concrete intelligence: Enables us to manage concrete and mechanical concepts and ideas, e.g., accounting, economics, and architecture.
Thurstone’s approach: Primary Mental Abilities
Louis L. Thurstone (1938) argued that intelligence is not a general factor but is composed of small, independent factors or elements called “primary mental abilities.” With his wife, he prepared a set of 56 tests, administered to 240 college students. Factor analysis yielded seven primary mental abilities:
- Verbal comprehension: Ability to understand and define words.
- Word fluency: Speed of thinking of verbal material, such as rhyming or naming words in a category.
- Spatial visualization: Ability to recognize and manipulate objects in three dimensions, e.g., drafting and blueprint reading.
- Perceptual speed: Quick ability to perceive visual details and differentiate similarities and differences between designs.
- Reasoning/inductive reasoning: A logical ability to derive general ideas from specific information.
- Numbers/arithmetic ability: Capability to work easily with numbers, such as performing simple arithmetic tasks quickly.
- Memory: Capacity to remember and retain material like words and letters, and the ability to recall and associate different words.
Crystallized and Fluid Intelligence: R.B Cattell
R.B. Cattell proposed that intelligence consists of two types:
- Crystallized intelligence: The capability of using information learned through experience. It is affected by education and culture and includes accumulated knowledge, skills, and techniques applied in problem-solving. This type increases with age.
- Fluid intelligence: Largely influenced by biological factors, it is the capability for information-processing and solving problems that depend on neurological development, such as reasoning and memory. This type declines with age.
Guilford’s theory of the Structure of Intellect (SOI)
J.P. Guilford created a model where intelligence is the result of the interaction of operations, contents, and products. He proposed 150 such abilities. The components are:
- Operations: Potential of different ways of thinking, including evaluation, convergent thinking, divergent thinking, memory retention, memory recording, and cognition.
- Contents: Potential of what we think about, including visual, symbolic, semantic, and behavioral content.
- Products: The results obtained by applying certain operations to certain contents, including units, classes, relations, systems, transformations, and implications.
Multiple Intelligences: Howard Gardner’s Approach
Howard Gardner (1985) maintained that intelligence does not consist of a single factor but consists of eight independent intelligences, possessed by all individuals in varying degrees:
- Linguistic
- Logical-mathematical
- Spatial intelligence
- Musical intelligence
- Bodily-kinesthetic
- Interpersonal intelligence
- Intrapersonal intelligence
- Naturalistic intelligence
Sternberg’s Triarchic Theory
Robert Sternberg’s triarchic theory (1980s) posits that intelligence consists of three main components:
- Analytic intelligence
- Creative intelligence
- Practical intelligence
Sternberg emphasized that practical intelligence is related to overall success in living, unlike traditional tests which measure academic success. Practical intelligence is learned through observation of others' behavior, while academic success comes from reading and listening.
💡 Why this matters: Sternberg’s theory highlights a critical gap: traditional IQ tests may not predict career or life success, which requires practical, street-smart intelligence.
Emotional Intelligence
Daniel Goleman (1995) proposed that emotional intelligence is the ability to go along with others. It involves the accurate assessment, evaluation, expression, and regulation of emotions. Emotional intelligence involves the realization and regulation of personal emotions as well as empathy and understanding of others' emotions. Key aspects include social skills, self-awareness, and empathy.
Piaget’s View of Intelligence
Jean Piaget defined intellectual development in terms of qualitative changes in thinking apparent in children of particular ages. His theory emphasized universal patterns of intellectual development and how children acquire knowledge. He proposed four universal and invariant stages of cognitive development:
- Sensorimotor
- Preoperational
- Concrete operational
- Formal operational
⭐ Key Takeaways
The most critical concepts from this lecture are the shift from viewing intelligence as a single, general ability (Spearman's g-factor) to complex, multi-faceted models (Thurstone, Guilford, Gardner, Sternberg). Students must understand the distinction between crystallized and fluid intelligence (Cattell), as this explains age-related changes in ability. Sternberg's introduction of practical intelligence and Goleman's emotional intelligence are crucial for understanding that traditional IQ tests do not capture all dimensions of human capability. Finally, the foundational work of Galton, Binet, and Cattell established the historical trajectory of intelligence testing, leading to these modern theoretical frameworks.
🧠 Quick Revision Questions
- What is the fundamental difference between Spearman's g-factor theory and Thurstone's theory of primary mental abilities?
- Describe the two types of intelligence proposed by R.B. Cattell and explain how they change across a person's lifespan.
- According to Sternberg's triarchic theory, what are the three components of intelligence, and why is "practical intelligence" considered important for career success?
- What are the eight independent intelligences proposed by Howard Gardner, and what is a key characteristic of how they are distributed among individuals?
- How does Daniel Goleman's concept of "emotional intelligence" differ from the traditional understanding of intellectual ability measured by standard IQ tests?
📘 Lecture 25 — Measurement of Intelligence
📖 Overview: This lecture provides a comprehensive historical and practical overview of intelligence testing, beginning with Alfred Binet’s foundational work in France. It traces the evolution of the Binet Scale through multiple revisions to the Stanford-Binet and then details the development and structure of the equally influential Wechsler Scales. Understanding this evolution is critical for appreciating how modern intelligence tests are constructed, standardized, and interpreted.
🗂️ Topics Covered
This lecture covers the historical evolution of intelligence testing starting with the Binet and Simon scale from 1905 through its major revisions in 1908, 1916 (Stanford-Binet), 1937, 1960, 1972, and 1986 (4th Edition). It explains the core concepts of Mental Age and Intelligence Quotient (IQ), including the shift from ratio IQ to deviation IQ. The lecture then transitions to the Wechsler Scales, detailing the three main tests (WAIS-III, WISC-III, WPPSI-R), their verbal and performance subtests, and psychometric properties. Finally, it provides a standard classification of IQ scores.
📝 Lecture Summary
The Binet Scale
The first formal measure of intelligence was developed in 1905 in France by Alfred Binet and Theodore Simon. Its primary purpose was to help the education ministry identify “dull” students in the Paris school system so they could receive remedial training. The core belief was that intelligence could be measured by a child’s performance on age-specific tasks. If a child could perform tasks typical for their age, they were considered of average intelligence. If they could only perform tasks for younger children, they were below average; if they could perform tasks for older children, they were above average.
The Development of Binet and Simon Scale
The test developers first created a number of tasks. These were then presented to groups of students who had been labeled as ‘dull’ or ‘bright’ by their teachers. The tasks that were successfully completed by the ‘bright’ students were retained, as these were considered indicative of intelligence. The key insight was that tasks were age-related; a child who could perform tasks meant for a higher age group was considered above average.
Binet’s scale gained rapid popularity, leading to translations and adaptations worldwide, particularly in the United States. A training school in New Jersey used it as early as 1908, and a modified version published in 1912 extended the age range down to three months. The most significant developments, however, took place at Stanford University.
The 1905 Scale
The original scale consisted of 30 items arranged in increasing order of difficulty. 🔑 Definition — Norms: Performance standards derived from a sample. The norms for this first scale were obtained from a very small sample of only 50 children who were reported to be ‘normal’ based on their average school performance. Although the normative and validity data were insufficient, this scale was a major milestone in psychological measurement.
The 1908 Scale
The 1908 revision was an age scale that retained the principle of age differentiation. Its most important innovation was the introduction of the concept of mental age. 🔑 Definition — Mental Age (MA): The average age of children who achieve a particular score on a test; it represents the typical intelligence level found for people at a given chronological age. A test taker’s performance was compared with the average performance of others of the same chronological age. The standardization sample for this revision included 203 individuals.
The 1916 Revision: The Stanford-Binet Intelligence Scale
In 1916, American psychologist Lewis Terman produced the first Stanford revision, creating the Stanford-Binet Intelligence Scale. This was standardized on an American sample for ages 3 to 14, plus “average and superior adults.” The sample, however, was criticized for containing only white, native Californian children.
This scale had three significant characteristics:
- It was the first American test to use the concept of Intelligence Quotient (IQ).
- It introduced the concept of an ‘alternate item’, an item that could replace a regular item if needed.
- It provided detailed, organized instructions for administration and scoring, a hallmark of standardized testing.
The 1937 Revision
This revision, by Terman and his colleague Merrill, contained new tasks for preschool and adult levels. A major improvement was the creation of two equivalent forms, ‘L’ and ‘M’ (the initials of the authors’ first names). The standardization sample was much larger, with 3,184 individuals from 11 U.S. states. However, it was still criticized for a lack of representativeness, as the subjects were all white and mostly from urban areas.
The Concept of Mental Age
Children taking the Binet-Simon test were assigned a score corresponding to their age group. This score was their “mental age” . 🔑 Definition — Mental Age (reiterated): The average age of children who secure the same score on a test. Mental age can differ from chronological age, reflecting whether a child is performing above, at, or below the level of their peers.
The Concept of Intelligence Quotient or IQ
To address problems with using mental age alone, the Intelligence Quotient was developed, which gives consideration to both mental and chronological age.
📐 Formula: IQ score = MA / CA x 100 → Plain-English meaning: This formula calculates the ratio of a person’s mental age (MA) to their chronological age (CA). Multiplying by 100 removes decimal points. 📌 Example: A 10-year-old child (CA = 10) who has a mental age of 12 (MA = 12) would have an IQ of (12/10) x 100 = 120. If their MA was exactly 10, their IQ would be 100. 💡 Why this matters: This formula provided a single, standardized number to represent intelligence, allowing for easy comparison. An IQ of 100 is the average, scores below 100 indicate below-average performance for one’s age, and scores above 100 indicate above-average performance.
The 1960 Revision
A major change in the 1960 edition was the creation of a single form, ‘L-M’ , containing the best items from the two previous forms. More importantly, it shifted from using ratio IQ to deviation IQ. 🔑 Definition — Deviation IQ: A standard score where a person’s performance is compared to the performance of others of the same age in the standardization sample. The mean is set to 100 and the standard deviation to a constant value (16 for this test). Using deviation IQ allowed for a precise estimation of a test taker’s relative position (e.g., percentile rank) within their age group, solving a key limitation of ratio IQ.
The 1972 Revision
This revision introduced an improved normative sample of 2,100 subjects, which finally included nonwhites. Despite this improvement, the sample was still criticized for not having enough non-white participants.
The 1986 Version: The Stanford Binet (4th Edition)
This version successfully overcame many previous criticisms. It used a large standardization sample of 5,000 subjects from 47 states, stratified by geographic region, community size, ethnic group, age, and gender. It covered four content areas:
- Verbal reasoning (Subtests: Vocabulary, Comprehension, Absurdities, Verbal relations)
- Abstract/visual reasoning (Subtests: Pattern analysis, Copying, Matrices, Paper folding and cutting)
- Quantitative reasoning (Subtests: Quantitative, Number series, Equation building)
- Short-term memory (Subtests: Bead memory, Memory of sentences, Memory of digits, Memory of objects)
Administration of Stanford-Binet Test
The test uses an individual-oral administration format. The examiner establishes a basal age (the lowest level where two consecutive items are passed) and a ceiling (the point where at least three out of four items are missed). 🔑 Definition — Basal age: The lowest level in the test where the examinee passes two consecutive items of approximately equal difficulty. 🔑 Definition — Ceiling: The point in the test at which the examinee fails at least three out of four items. Scores from all 15 subtests are converted into standard age scores (mean = 50, SD = 8), which are then grouped into the four area-content scores (mean = 100, SD = 16). A composite score is also calculated.
Some Sample Items from Early Versions of Simon-Binet Scale
The lecture provides examples of tasks for different ages, showing the increasing complexity:
- Three years: Shows nose, eyes, mouth; repeats two digits.
- Six years: Distinguishes morning from afternoon; copies a shape; counts 13 pennies.
- Eight years: Counts from 20 to 0; indicates omissions in pictures; repeats five digits.
- Fifteen years: Repeats seven digits; gives three rhymes; solves a problem from several facts.
The Wechsler Scales
The Wechsler Scales, developed by psychologist David Wechsler, are perhaps the most commonly used intelligence tests today. The three main scales are:
- WAIS-III (Wechsler Adult Intelligence Scale, 3rd ed.): for ages 16 to 89.
- WISC-III (Wechsler Intelligence Scale for Children, 3rd ed.): for ages 6 to 16.
- WPPSI-R (Wechsler Preschool and Primary Scale of Intelligence-Revised): for ages 3 to 7 years, 3 months.
The first Wechsler scale, the W-B I (Wechsler-Bellevue), was published in 1939. It was a point scale, giving credit for every correct response, unlike the Binet’s age scale. The WAIS was developed in 1955, revised as the WAIS-R in 1981, and the latest version, WAIS-III, in 1997.
The Wechsler scales are divided into two categories: Verbal and Performance scales.
- Verbal Scale subtests include: Vocabulary, Similarities, Arithmetic, Digit span, Information, Comprehension, and Letter-number sequencing.
- Performance Scale subtests include: Picture completion, Digit symbol-coding, Block design, Matrix reasoning, Picture arrangement, Symbol search, and Object assembly.
Psychometric Properties of WAIS-III
The standardization sample for the WAIS-III consisted of 2,450 adult subjects in 13 age groups (from 16-17 to 85-89). Stratification was based on gender, race, education, and geographic region, taken from the 1995 U.S. census.
The Meaning of IQ Test Scores
The lecture provides a standard classification of IQ scores:
- < 70: Retarded
- 85: Borderline
- 100: Average
- Above 115: Superior
- Above 140: Gifted
⭐ Key Takeaways
- The Binet Scale introduced the foundational concepts of Mental Age and Age Differentiation, but its ratio IQ was later superseded by deviation IQ in the Stanford-Binet, which allows for direct comparison of an individual’s performance to their age peers.
- The evolution of the Binet/Stanford-Binet Scale is a textbook example of test development, showing a progressive refinement in standardization, sample size, and representativeness (from 50 white children to 5,000 stratified subjects).
- The Wechsler Scales revolutionized the field by moving from an age scale to a point scale and introducing distinct Verbal and Performance subscales, providing a more detailed profile of cognitive strengths and weaknesses.
- Both major test families rely on individual administration by a trained examiner, establishing a basal age and a ceiling to ensure the test covers the appropriate difficulty range for each test taker.
- The meaning of IQ scores is only interpretable relative to a standardized distribution (mean = 100, SD = 16 for many tests), with labels like “Average,” “Superior,” and “Retarded” defined by specific score ranges from the normative sample.
🧠 Quick Revision Questions
- What was the primary purpose of Alfred Binet and Theodore Simon’s original 1905 intelligence scale?
- How does a ratio IQ differ from a deviation IQ in how it is calculated and what it tells us about a person?
- What is a basal age and a ceiling in the administration of the Stanford-Binet test, and why are they important?
- Name the three main Wechsler scales currently in use and the specific age range for which each is intended.
- According to the standard classification provided, what is the IQ score range that is considered “Superior”?
📘 Lecture 26 — The Kaufman Scales: Intelligence Tests
📖 Overview: This lecture examines several major intelligence test batteries developed by Alan and Nadeem Kaufman, including the K-ABC, K-BIT, and KAIT, alongside the Differential Ability Scales (DAS). It also addresses critical issues of cultural bias in intelligence testing and introduces alternative formulations of intelligence such as moral, social, and emotional intelligence that expand beyond traditional IQ measurement.
🗂️ Topics Covered
The lecture covers the Kaufman Assessment Battery for Children (K-ABC) with its five global scales and 16 subtests, the Kaufman Brief Intelligence Test (K-BIT) as a screening instrument, and the Kaufman Adolescent and Adult Intelligence Scale (KAIT) featuring crystallized and fluid scales. It then presents the Differential Ability Scales (DAS) with its core subtests, diagnostic subtests, and achievement tests across different age groups. The discussion shifts to cultural biases in intelligence testing with culture-fair tests like Raven's Progressive Matrices, followed by significant questions for IQ test use and alternative formulations including moral, social, and emotional intelligence.
📝 Lecture Summary
The Kaufman Scales: Intelligence Tests
The husband and wife duo, Alan and Nadeem Kaufman, have contributed significantly to intelligence testing by developing the K-ABC (1983), K-BIT (1990), and KAIT (1993). Each test serves distinct populations and purposes within intellectual assessment.
K-ABC: Kaufman Assessment Battery for Children
The Kaufman Assessment Battery for Children includes five global scales. Sequential processing involves subtests like hand movement, number recall, and word order. Simultaneous processing includes magic window, face recognition, Gestalt closure, triangles, matrix analogies, spatial memory, and photo series. Mental processing Composite combines sequential and simultaneous processing scales. The Achievement scale includes expressive vocabulary, faces and places, arithmetic, riddles, reading/decoding, and reading/understanding. The Nonverbal scale includes face recognition, hand movements, triangles, matrix analogies, spatial memory, and photo series.
💡 Why this matters: This battery focuses on the information processing approach, examining how individuals process information rather than just what they know.
The scales consist of 16 subtests other than the nonverbal scale. A national sample of 2000 American children aged 2.5 to 12.5 years was used for standardization. The sample considered age, gender, geographic region, parental education, community size, educational placement, and race. Gifted and talented, mentally retarded, and learning disabled individuals were included proportionally to their representation in the general public.
K-BIT: Kaufman Brief Intelligence Test
The Kaufman Brief Intelligence Test, meant for ages 4 to 90 years, is a quick screening instrument. It is an individual test for assessing intellectual functioning through individual administration. The K-BIT yields three scores: verbal, non-verbal, and composite. The verbal subtest includes 45 Expressive Vocabulary items and 37 Definitions. The non-verbal subtest contains 48 matrices. This test used nearly 20% of the KAIT standardization sample.
🔑 Definition — K-BIT: A brief, individually administered intelligence test providing verbal, non-verbal, and composite scores for ages 4 to 90 years
KAIT: Kaufman Adolescent and Adult Intelligence Scale
The KAIT was developed for measuring intelligence of subjects aged 11 to 85 plus. It consists of a Crystallized scale measuring concepts learned from schooling and acculturation, and a Fluid scale measuring the ability to solve new problems. Three subtests are used in each scale in the Core Battery. An Expanded Battery is available for subjects suspected of having neurological damage, where any of four specified subtests are added. A brief mental status test assessing attention and orientation is included for cognitively impaired test takers who cannot complete the whole battery.
Differential Ability Scales
The Differential Ability Scales (DAS) were developed by C. D. Elliot (1990) in Great Britain as a revised form of the British Ability Scales. There are three major components containing 20 subtests in all.
Core Subtests: Block Building, Verbal Comprehension, Picture Similarities, Naming Vocabulary, Early Number Concepts, Copying, Pattern Construction, Recall of Designs, Word Definitions, Matrices, Similarities, and Sequential and Quantitative Reasoning.
Diagnostic Subtests: Matching Letter-Like Forms, Recall of Digits, Recall of Objects, Recognition of Pictures, and Speed of Information Processing.
Achievement Test: Basic Number Skills, Spelling, and Word Reading.
Four core subtests are used with preschoolers aged 2 years 6 months to 3 years 5 months. Six core subtests are for preschoolers aged 3 years 6 months to 5 years 11 months. Six subtests are for school level, ages 6 years to 17 years 11 months.
The standardization sample included 3475 persons aged 2 years 6 months to 17 years 11 months, representing non-institutionalized English-proficient individuals in the U.S. Age, gender, race/ethnic origin, parental education, and geographic region were considered for stratification.
Cultural Biases And Intelligence Tests
Tests used to assess intelligence have been frequently criticized for being biased against particular groups. Culture-fair IQ tests are developed to overcome this problem, such as Raven's Progressive Matrices, which do not discriminate against any minority or cultural group.
Some Significant Questions Pertaining To the Use of IQ Tests
Key questions to consider include: Is the test a validity test? Is it reliable? Was it standardized? Is it being used with people similar to those in the standardization sample? Are the differences between the test taker and standardization sample drastic and serious? Have consequences been anticipated against expected benefits? Can cultural background affect results? Can ethnic origin affect test results? Can administration be problematic due to personal, environmental, or physical reasons?
Alternative Formulations
Alternative formulations of intelligence include Moral intelligence, Social intelligence, and Emotional intelligence.
Moral Intelligence
Given by Coles (1997) and Hass (1998), moral intelligence is the ability to differentiate between right and wrong. More comprehensively, it is the capacity to make right decisions that are beneficial not only for oneself but for others as well.
🔑 Definition — Moral Intelligence: The ability to differentiate right from wrong and make decisions beneficial to oneself and others
Social Intelligence
Given by Hough (2001) and Riggio, Murphy, & Pirozzolo (2002), social intelligence is manifested as SQ. It is the ability to understand and deal with people, exhibited by salesmen, politicians, teachers, clinicians, and religious leaders. It also involves understanding and dealing with oneself by identifying one's thoughts, feelings, attitudes, and behaviors.
🔑 Definition — Social Intelligence (SQ): The ability to understand and deal with others and oneself through identifying thoughts, feelings, attitudes, and behaviors
Emotional Intelligence (EI)
Emotional intelligence is a type of social intelligence involving the ability to cope with one's own and others' emotions, differentiate between them, and use information for guiding thoughts and actions. It is indicated by a person's EQ and includes these aspects:
- Self-awareness
- Managing emotions
- Empathy
- Handling relationships
🔑 Definition — Emotional Intelligence (EI): The ability to cope with one's own and others' emotions, differentiate between them, and use emotional information to guide thoughts and actions
⭐ Key Takeaways
Students must remember that the Kaufman scales (K-ABC, K-BIT, KAIT) target different age ranges with distinct purposes—K-ABC focuses on information processing in children, K-BIT provides quick screening across a wide age range, and KAIT measures crystallized versus fluid intelligence in adolescents and adults. The Differential Ability Scales (DAS) contain 20 subtests organized into core, diagnostic, and achievement components with age-specific administration patterns. Cultural bias remains a critical concern in intelligence testing, addressed through culture-fair tests like Raven's Progressive Matrices, and test users must evaluate validity, reliability, standardization, and cultural appropriateness before administration. Finally, intelligence extends beyond traditional IQ to include moral, social, and emotional intelligence, each with distinct definitions and applications in understanding human capability.
🧠 Quick Revision Questions
- What are the five global scales of the K-ABC, and what does the Mental Processing Composite combine?
- How does the KAIT distinguish between crystallized and fluid intelligence, and what is the purpose of the Expanded Battery?
- How many core subtests are used for preschoolers aged 2 years 6 months to 3 years 5 months on the DAS compared to school-aged children?
- What are three critical questions that should be asked when determining if an IQ test is appropriate for a particular test taker?
- What specific aspects comprise emotional intelligence (EI), and how does it differ from social intelligence (SQ)?