Summary - Psychological Testing and Assessment - Van der Molen - Chapter 5 & 7

Which theories about individual intelligence and performance tests are there? - Chapter 5

Topic 5A Theories of intelligence and factor analysis

Definitions of intelligence

Intelligence is one of the most researched topics in psychology.

Operational definitions define a concept in terms of the way the test is used. This way of defining is not useful for intelligence. Intelligence tests have been developed for measuring intelligence and not for defining it. In addition, operational definitions block the further development of the understanding of the nature of intelligence. A second problem is that the operational definition does not allow innovation. The old tests serve as a validity criterion for new tests. There must be a correlation between the old and new tests and in newly developed, innovative tests this correlation may weaken. According to the operational definition, they may therefore not be valid.

The real definition explains the true nature of the concept. To arrive at a true definition of intelligence, you have to ask experts. Several experts have come up with a definition. This list of definitions is Western and does not speak of cultural differences. The East thinks that benevolence, humility, freedom from conventional standards in assessment and doing what is right are essential for intelligence. In Africa, the social aspects of intelligence are emphasized. Despite all the differences, there are usually two things that everyone agrees on: intelligence is the capacity to learn from experiences and the capacity to adapt to the environment.

Another approach to understanding a construct is to study the popular meaning. Sternberg and colleagues did this and found out that the behaviors that are thought to belong to intelligence are quite similar to what experts think. The difference lies more in the order of importance.

When laymen think of intelligence, the following aspects are mentioned: practical problem-solving skills, verbal skills, and social competence. Experts mention the following aspects: verbal intelligence, problem-solving skills, and practical intelligence. This shows that public opinion on intelligence can be recognized in the intelligence tests.

The two parts that lay people and experts agree on are also not included in the tests. Firstly, because it is difficult to find items for these and secondly, because there has been little change in the design of the tests since the Binet time. Sternberg and colleagues have presented innovations that are based on their model of intelligence. Another interesting tool is the Everyday Problem Solving Inventory. In this test, participants should indicate how their response would be to everyday problems, such as forgetting your money when you go for lunch with a friend.

Factor analysis

Factor analysis is used to measure the validity of tests. There is confirmatory factor analysis and exploratory factor analysis. The first one checks whether test scores and variables fit in a predicted pattern and is therefore good for validation research. In the second, the relationships between many variables are summarized and it is examined whether they can be reduced to a lesser number of factors. It ensures that a simplified representation is given of a large, complex data set. Each collection of aptitude tests reflects a number of underlying basic skills.

A factor analysis starts with a correlation matrix, a table that shows correlations between different variables. Variables that have a high correlation with each other seem to explain a factor.

The smallest number of factors are determined by means of fast computers. Then a factor matrix can be made, which shows how heavily a variable weighs on a certain factor (values ​​from -1.00 to +1.00).

The factor loads are then 'rotated', the best fitting variable is stopped at the best matching factors. In 'rotation to positive manifold', as many negative factor loads as possible are removed per factor. In 'rotation to simple structure', factor loadings are classified in such a way that each variable has a significant load with as few factors as possible. The aim is to make the factor matrix as clear and analysable as possible.

The interpretation of the factors that come out has to be done by the researcher himself. The researcher looks at what the variables of the factor have in common and then gives that interpretation. There are also disadvantages for factor analysis. Firstly, a certain factor can only come out of the analysis if the tests and measurements also contain that factor. Second, the sample size is important: around n=300 a factor analysis is stable and reliable. Finally, factor analysis is always subject to subjective choices and theoretical prejudices of the researcher.

Galton and sensory keenness

Galton thought that intelligence was explained by keen sensory abilities. The theory, however, had a dead end with regard to psychometry: it could hardly be tested. Later in the 1980s, however, reaction time tests were developed that measured the speed of processing. Even though these tests suggested that speed-of-processing measures could be a useful addition to standardized intelligence tests, these tests as well as their data analysis were not sufficiently standardized. That is why this method has not been developed much further.

Spearman and the g factor

Spearman stated that intelligence was based on a general intelligence factor g and several specific factors s1, s2, s3, etc. He focused primarily on defining the nature of g which he called the energy or power that generally appears in every context. The specific factor s represents a specific group of neurons that were used in certain mental actions. He found that individual differences in g reflected three principles of cognition: apprehension of experience, eduction of relations, and eduction of correlations. The word eduction can be rephrased as figuring things out. Even though Spearman’s speculations about physiology are for the most part dismissed, his notion of a general factor of intelligence remains relevant until today.

Thurstone and primary mental skills

Thurstone used factor analysis and concluded that not a single intelligence factor, but a group of factors, could explain empirical results. That brought him to his 7 primary mental skills: verbal comprehension, word fluency, number, space, associative memory, perceptual speed, inductive reasoning.

Later he argued that probably a general g factor was at the head and Spearman also stated that there are group factors that explain intelligence. Vernon felt that there was a hierarchy with the g factor at the top, then two group factors verbally and practically and then the 7 skills of Thurstone divided over the two factors.

Cattell-Horn-Carroll (CHC) theory

According to this theory, intelligence consists of three hierarchical strata. Stratum III is the general intelligence factor g. Stratum II consists of 8 parts, each with a number of specific skills of stratum I which are in brackets:

1. fluent intelligence (5 specific skills): requires high thinking and reasoning and is used for new tasks that are not done automatically.
2. crystallized intelligence (10): the breadth and depth of the knowledge of the individual.
3. domain-specific knowledge (7): knowledge about specific domains that are not typical of culture.
4. visual-spatial abilities (11): presenting, retaining and transforming mental representations of images.
5. auditory processing (13): the accurate observation of auditory information and with it the capacity to analyze, understand and imitate patterns or groups of sounds.
6. memory (13): the ability to consolidate new information and store it in the long-term memory and recall it later through associations.
7. cognitive processing speed (7): the speed of performing automatic processes, especially when a lot of attention and concentration is needed.
8. reaction time (5): the ability to make quick decisions in response to simple stimuli.

The skills of stratum I are always expanded and renewed with research. In addition, this model is widely used because the broad and specific skills are empirically verifiable and they include meaningful and implications for the real world. Many psychometricians consider it having the strongest empirical foundation of any theory of intelligence. Others have proposed a larger list that also includes psychomotor, olfactory, and kinesthetic abilities.

Guilford and structure-of-intellect model

Guilford classified intellectual skills in dimensions of operations, contents and products. Operations are the type of mental actions that the test must contain. Contents refers to the nature of the materials or information that is offered. Products includes the different types of mental structures that the brain has to produce in order to arrive at a correct answer. He then had 5 x 5 x 6 = 150 factors of intelligence respectively.

It is a complicated model, but he has raised the suggestion that divergent production, the creation of several good possibilities for a single stimulus (as opposed to only constructing one correct answer to a stimulus situation), can also be part of intelligence.

Planning, attention and simultaneous and successive theory (PASS)

Luria adopted the theory ‘plans, attention, simultaneity and succession', a general theory about cognitive processing. His approach focuses on the mechanisms that process information and is therefore also called the information processing theory. According to him, analysis provides strong evidence for distinguishing between two basic forms of integrative activity of the cerebral cortex, allowing different aspects of the outside world to be reflected. The theory contains three functional units in the brain, whereby the processing of information takes place from the lower units to the higher units:

1. The first part includes subcortical areas like the brainstem, midbrain, and thalamus. It allows us to regulate attentive processes and selective attention and to resist distractions.
2. The second part contains the sensory parts of the cerebral cortex (parietal, temporal, and occipital lobes). Their functions are more lateralized and include simultaneous and successive processing.
3. The third part is in the frontal lobes and is responsible for planning and engine output.

According to this theory, processing starts with attention and planning is the final step.

1. Simultaneous processing (of information): this is characterized by performing different mental activities simultaneously. An example for which you need this is for example drawing a cube, you need spatial insight, motor skills etc.
2. Sequential processing (of information): is necessary for mental activities where a correct sequence of activities is needed. This is necessary, for example, for remembering a series of numbers.

Finding tasks that can test these two ways of processing in their pure way is the challenge for applying this theory to intelligence.

Information processing theories of intelligence

These theories propose models on how people represent and process information mentally. It is useful to make the analogy of a computer. The architectural system (hardware) comprises the biological requirements for information processing, such as coding speed. It is quite impenetrable to change by the environment.

The executive system (software) is easily influenced by the environment and it includes components related to the environment that control the problem-solving capacity and leading functional components. An example is metacognition, thinking about thinking, which is important for learning and intelligence.

Gardner and the theory of multiple intelligences

Gardner drew up a number of criteria for an autonomous intelligence:

• Potential isolation due to brain damage
• The existence of exceptional individuals in an area
• Identifiable core operations
• Distinctive developmental history
• Evolutionary plausibility
• Support from experimental psychology
• Support from psychometric studies
• Susceptibility to symbol encoding

Gardner thus established seven natural intelligences:

1. Linguistic
2. Logical-mathematical
3. Spatial
4. Musical
5. Bodily-kinesthetic
6. Interpersonal
7. Intrapersonal

Bodily-kinesthetic intelligence includes the skills used by athletes, dancers, mime artists, etc. Personal intelligences include the ability to access both one's own feelings (intrapersonal) and the ability to notice and distinguish someone else’s moods, temperament, motivation, and intentions (interpersonal). People with musical intelligence can easily learn to play an instrument and make their own compositions.

Savants provide an important insight into the existence of different kinds of intelligences. A savant is someone who is mentally inadequate, but has developed himself in a certain area, such as music or art. The theory is compelling because of its simplicity, but the empirical validity is difficult to investigate.

Sternberg and the triarchic theory of successful intelligence

Sternberg has developed a theory that is called triarchic, because it emphasizes three aspects of intelligence.

Componential / analytical intelligence

Internal mental mechanisms responsible for intelligent behavior.

• Metacomponents or executive processes: these control the activities of the other intelligence processes, such as planning.
• Performance components: the processes that are used to perform a task or solve problems, such as working memory or syllogistic reasoning.
• Knowledge acquisition components: processes that are used in learning, such as the use of certain vocabulary in the right context.

Experiential / creative intelligence

• Ability to deal with novelty
• Ability to automize information processes

Contextual / practical intelligence

• Adaptation to the 'real world' environment: if you possess the skills needed for the culture in which you live.
• Selection of an appropriate environment: the ability to leave the environment you are in and to select an environment that is better suited to your talents and needs.
• Shaping of the environment: the skill to form the existing environment in such a way that it is more consistent with your needs.

For the theory, Sternberg has made the STAT (Sternberg Triarchic Abilities Test), which also contains creative and practical questions.

Topic 5B Individual tests of intelligence and achievement

Intelligence testing is one of the major achievements of the field of psychology. The score of general intelligence is often not what researchers are interested in. If they intending to determine g, any test would do. Often, however, they are looking for specific intellectual functioning and therefore look at the results of subtests, so it is important that they use the test that best describes what they want to know. There are many intelligence tests and new ones are added every month. The following tests are likely to be used in 95% of the assessment cases.

Orientation

This chapter discusses the following intelligence tests:

• Wechsler Adult Intelligence Scale (WAIS-IV)
• Wechsler Intelligence Scale for Children (WISC-IV)
• Stanford-Binet: Fifth Edition (SB5)

• Detroit Tests of Learning Aptitude-4 (DTLA-4)
• Cognitive Assessment System-II (CAS-II)
• Kaufman Brief Intelligence Test-2 (KBIT-2)

Wechsler Scales of Intelligence

The first intelligence test by Wechsler, the Wechsler-Bellevue Intelligence Scale, was released in 1939. He made sure that the test did not have the shortcomings of already existing tests. The existing tests were not available for adults, were too focused on manipulative words, were too focused on speed rather than accuracy, and the notion of mental age was irrelevant.

He drew up the new formula for IQ:

IQ = obtained score / expected mean score for age

He did this, because according to him IQ remains almost constant (IQ constancy), but intellectual skills can change over time. In addition, Wechsler hoped to use the test for psychiatric diagnoses. For this he distinguished between verbal and performance intelligence.

• Verbal> Performance (V>P): gave results for organic brain diseases such as psychosis and emotional disorders.
• Performance> Verbal (P>V): showed slight mental retardation in adolescents and adults.

Research has shown that there are many exceptions to this diagnostic rule, but it nevertheless showed that distinguishing between verbal understanding and perceptual reasoning (as we now know it) is useful in many areas, such as studying brain-behavioral relationships.

The later Wechsler tests were particularly successful because of the reliable content and formula of the first test, which meant that examiners, among other things, needed little training to switch to a new version.

The latest versions of Wechsler tests all contain the following common features:

• Thirteen to fifteen subtests that are used to measure intra-individual strengths and weaknesses rather than one global score.
• Empirically based scores and IQ scales. First it was subdivided into verbal IQ and performal IQ, but the latest versions have index scores in four areas:

1. Verbal comprehension
2. Perceptual reasoning
3. Working memory
4. Processing speed

• A common metric for IQ and index score, being an average IQ of 100 with standard deviation of 15. On a subtest this would be an average of 10 with standard deviation of about 3.
• Some of the same subtests for different test versions.

Wechsler subtests: description and analysis

Here the different subtests of the WISC-IV (children) and WAIS-IV (adults) are discussed.

• Information: this test tests factual knowledge about people, places and general phenomena. Most of these information questions are known when someone has grown up in a Western culture. This subtest also measures learning and memory skills, because one has to retrieve knowledge from memory. The information test is the best measure for general skills and relies heavily on the factor Verbal Understanding.
• Digit span: at digit span, participants get a series of digits which they then have to repeat (Digits Forward), have to repeat backwards (Digits Backwards) or put in the correct numerical order (Digits Sequencing). When the person is right, the next series that is a number longer, up to a maximum of 9 digits. Scores on this test can be influenced by anxiety or fatigue, which is often seen in psychiatric patients. The different types of test measure different skills, because repeating digits backwards requires more working memory and skills than forward, which is basically only repeating them.
• Vocabulary: in this test different words are mentioned after which the participant must tell what these words mean. They will then receive full scores if they mention all the characteristics of the word and some of the points if they have forgotten an important part. It is the second best measurement for general intelligence.
• Arithmetic: in this test calculation problems are verbally presented that the participants must answer. Some items can be tricky, because it must also be answered within a short time. Arithmetic charges the strongest on the factor memory.
• Compehension: this raises questions that do not have to do with factual knowledge, but where it is important to be able to explain some concept. An example is: 'What does the proverb mean: Better a bird in your hand, than ten in the air?' There is still a disagreement about whether this subtest has mainly to do with the social functioning of the participant and whether it therefore measures 'social intelligence'.
• Similarities: here the participant's ability is tested to separate important unimportant characteristics from objects, facts and ideas. For example, two objects are given (shirt and socks) after which they are asked where they are equal (common concept). The point is that the participant can distinguish the important equation from the unimportant equality (that they both start with the s).
• Letter-number sequencing: this gives the participant a series of letters and numbers, after which he first has to put the numbers and then the letters in the correct order. For example, the series R-3-B-5-Z-1-C should then be 1-3-5-BCRZ. This subtest, together with Digit Span and Arithmetic, measures the working memory.
• Picture completion: here images are shown in which the participants have to indicate which important part is missing. The test assumes that the participant has seen the object, so it can be bad for participants from a different culture.
• Picture concepts: this subtest is used in children. A row of pictures is shown and the child has to indicate which concept the pictures have in common. With each series the concept becomes more abstract and therefore more difficult to guess. This subtest is only in the WPPSI-IV and the WISC-IV.
• Block design: here participants must construct a two-dimensional image with three-dimensional blocks. It requires analysis of spatial relationship, visual-motor coordination and it is less susceptible to memory and previous experiences. In the more difficult series, bonus points can be earned when performing quickly.
• Matrix reasoning: here participants have to discover a pattern or relationship in a row of figures (simple) or in a 3x3 matrix (more difficult), where the last figure is missing, which has to be chosen from five choices. It mainly measures fluid intelligence.
• Object assembly: this is only a test for toddlers and preschoolers, where they get parts of a picture and have to put them together to get the big picture, like a mini-puzzle. This subtest only occurs in the WPPSI-III.
• Coding: this is a subtest where a number is tied to a certain symbol. After a few exercises, a series of numbers is given, after which the corresponding symbols must be given in that order. A number of trials must be done within two minutes. Encoding is unique, because it is the only Wechsler task that teaches a new task and tests it the same way.
• Symbol search: this is a subtest that measures processing speed, indicating whether a few symbols on the left are located between a larger group of symbols on the right.
• Cancellation: in this case, participants must cross out all animals on a large sheet of paper, standing between all non-animal objects. There are two trials: one randomized with the animals and one with a structured rows and columns of animals. In the WAIS-IV this test is more abstract and only two target stimuli are used.
• Visual puzzles: here six parts have to be chosen that form an example picture. It charges the strongest on the factor Perceptual Reasoning. This subtest only occurs in the WAIS-IV.
• Figure weights: this is part of the WAIS-IV that measures Perceptual Reasoning. There is a scale with figures that are missing on one side. The participant must choose from six options to keep the balance in balance. This subtest only occurs in the WAIS-IV.

Wechsler Adult Intelligence Scale (WAIS-IV)

The WAIS-IV is a significant revision on the WAIS-III. It has retained many old items, but has two new subtests, a simple structure and emphasizes index scores, allowing a sharper boundary to be drawn between discrete domains of cognitive functions. The WAIS-IV consists of 15 subtests, but only 10 core subtests are needed to calculate the IQ score and the index scores. The subtests are divided over the four index scores as follows, instead of only distinguishing between verbal and non-verbal:

• Verbal Comprehension Index

1. Similarities
2. Vocabulary
3. Information

• Perceptual Reasoning Index

1. Block design
2. Matrix Reasoning
3. Visual Puzzles

• Working memory Index

1. Digit Span
2. Arithmetic

• Processing speed Index

1. Symbol Search
2. Coding

The verbal comprehension index (VCI) is similar to the old verbal IQ concept. Perceptual Reasoning Index (PRI) is related to the former notion of Performance IQ. Verbal Understanding and Perceptual Reasoning are used today, because the concepts fit better in the factor analysis.

The Working Memory Index (WMI) consists of subtests that require attention and direct memory. Relatively low scores can therefore indicate concentration problems. The Processing Rate Index (PSI) consists of subtests that require rapid processing of visual information and is sensitive to many conditions. The WAIS-IV is stricter in its criteria than the previous three WAIS tests. In addition, this successor is useful for both the healthy population and the population with mental problems.

• Standardization: the standardization of the WAIS-IV was done in 13 different age groups from 16 years old with a total sample of 2200 adults. The sample was stratified on the variables gender, ethnicity, education and region. It was done with cooperative, healthy, English-speaking people without brain damage.
• Reliability: the WAIS-IV has good reliability in terms of split-half reliability for the index scores and the full IQ scores. This means that the WAIS-IV is also a good test for a special population in which only one index score needs to be examined.
• Validity: the WAIS-IV has a good criterion validity, because it has high correlations with existing intelligence tests. It also has good convergent and discriminant validity. The validity is also well proven by the strong overlap with the previous three editions of the Wechsler intelligence tests

Wechsler Intelligence Scale for Children-IV (WISC-IV)

Several errors were found in the first version: in the standardization sample there was only one culture group (whites), ambiguities about scoring, inappropriate items for children and no use was made of women or African Americans in the pictorial content of items. This has all been adjusted in the WISC-IV.

The WISC-IV has 10 core subtests and 5 additional subtests. The core subtests are block design, similarities, digit span, picture concepts, coding, vocabulary, letter-number sequencing, matrix reasoning, comprehension and symbol search. The additional subtests are picture completion, cancellation, information, arithmetic and word reasoning.

The additional subtests are not used to calculate the general IQ scores, but may be useful firstly because of the important diagnostic information they provide. Secondly, they can serve as an alternative to a core subtest, when a child is not able to perform such a test, for example due to motor problems.

The standardization of the WISC-IV was good, because the agreement between the standardized sample and the data from the population register was almost perfect. This is the case, because 5.7% of the sample consisted of children who had certain characteristics, such as giftedness, learning disabilities, autism or motor problems.

The reliability of the WISC-IV is good and comparable with previous versions. The coefficients of the entire IQ score are around .90, but for each subtest the coefficients vary from .79 to .90.

The validity of the WISC-IV is also good. The correlations between the WISC-IV and the WISC-III with subtests rank from high in the .70 range to low in the .80 range. The correlation for the overall IQ score is higher, namely .89. The convergent validity and discriminant validity of the WISC-IV was also high.

By means of factor analysis, different indexes can also be ordered for the WISC-IV, also with an average of 100 and standard deviation of 15:

• Verbal Comprehension Index

1. Similarities
2. Vocabulary
3. Comprehension

• Perceptual Reasoning Index

1. Block design
2. Picture Concepts
3. Matrix Reasoning

• Working memory

1. Digit Span
2. Letter-Number Sequencing

• Processing speed

1. Coding
2. Symbol Search

Stanford-Binet Intelligence Scale: Fifth Edition (SB5)

The versions for the fourth version of the Stanford-Binet Test could not be used for an analysis of the subtests. Only an overall IQ score was given.

The SB5 has five factors of intelligence that are divided over two domains, resulting in 10 subtests:

Before the test is taken, a routing procedure is carried out to estimate the general cognitive skills of the participant. It ensures that the number of items can be reduced during the test without introducing measurement errors. This is possible because the test is based on the item response theory and difficulty levels and other parameters are already known since the development phase.

The SB5 can be used for very young children up to gifted adults. The non-verbal subtests can be used well for people who do not have a good command of the language, such as immigrants or the deaf. An important characteristic of the SB5 is that there are items that can distinguish giftedness and mental retardation from the 'normal' IQ scores. In addition, we also looked at whether the test is fair when it comes to cultural, gender, and religious differences. This is the first test that looks at religion. Hearing-impaired people and people who speak poor English or have communication problems can also make this test. The factor on working memory contains both verbal and non-verbal memory which can offer a solution in research on children with ADHD.

The standardization was done at 4800 persons based on the American population register. The validity of the SB5 is good with correlations above .70. The robust correlations with the SB4 show that the SB5 is valid, but also more usable than the SB4. Because the SB5 is so useful for both extremes of the cognitive spectrum, it will be a good tool to test individual intelligence.

Detroit Tests of Learning Aptitude-4 (DTLA-4)

The DTLA-4 is a test used for children between 6 and 17 years of age. With the test, 16 compositions are calculated, including general intelligence, optimal level, and 14 skills. It contains 10 subtests that are very similar to the Binet-Wechsler tests. The general mental skill composition is calculated using the standardized scores of the 10 subtests.

The optimal level composite is based on the four highest standard scores and shows how well the participant performs under optimal conditions. For the other 14 composite scores, combinations of certain subtests are used that together measure one of the following attributes:

• Linguistic (verbal vs. non-verbal)
• Attention (attention-enhanced vs. attention-reduced)
• Motoric (motor-enhanced vs. motor-reduced)
• Smooth intelligence vs . crystallized intelligence (Horn & Cattell)
• Simultaneous vs. successive (Das & Naglieri)
• Associative vs. cognitive (Jensen)
• Verbal vs. Performal (Wechsler)

They are used to give contrasting composites with significant differences that can result in a diagnostic explanation. For example, someone who scores well on attention-reduced tasks, but very badly on attention-enhanced tasks, probably has attention problems.

A problem with this test is that there is not enough empirical evidence for the composites. Another problem is that there are more composites than subtests, so that the composites will have high intercorrelations. So it is a good test for general intelligence but not for the composites. Reliability and validity are okay.

Cognitive Assessment System-II (CAS-II)

The CAS-II is based on the PASS theory of Das & Naglieri discussed earlier. The test is for children from 5 to 17 years. The test consists of 12 subtests and takes 60 minutes. There is a smaller version of 8 subtests but it is not recommended to use them. For each scale certain subtests are available:

Planning

• Matched numbers: the child must underline the two identical numbers in a number of rows of 6 numbers.
• Planned codes: a code is learned and then this is tested by having to enter the missing codes.
• Planned connections: lines must be drawn between points in order of alphabet and number (1-A-2-B-3-C, etc.).

Attention

• Expressive attention: this is a kind of Stroop task, in which the color of the words should be mentioned.
• Number detection: the child must underline specific digits, but only if they are printed in bold font.
• Responsive attention: first, the child underlines letter combinations that are physically the same (TT, not Tt), and then combinations that are the same in name (Bb, not Ba).

Simultaneous processing

• Non-verbal matrices: a logical shape must be found in a 3x3 matrix so that the empty box at the bottom right can be entered.
• Verbal spatial relations: six pictures with figures are shown and the child has to find the relationship in response to a specific question (e.g.: point out the square to the right of the circle).
• Figure memory: a drawing is shown and then the child has to recognize the drawing in a larger and more complex picture.

Successive processing

• Word series: a series of two to nine words must be repeated.
• Sentence Repetition: sentences must be read aloud and then repeated, when they can no longer be seen.
• Sentences questions or speech rates (8 to 17 years): questions are asked about sentences that need to be answered.

The scores of children with ADHD exceeded the scores that were expected based on already known information about the scores of children with ADHD.

The difference between black and white children is minimal when checking for background. In addition, high correlations between scores and school performance were found. The test is therefore useful for special education. It is a promising test that deserves to be used more often. The reliability and validity are very good.

Kaufman Brief Intelligence (KBIT-2)

There are two major drawbacks to the aforementioned tests:

1. The tests take a lot of time, at least one hour.
2. The person who takes the test must be well trained to be able to administer the tests properly.

Kaufman then developed an easy and short test to solve these problems. The test consists of a verbal scale containing two types of items: verbal knowledge and riddles and a non-verbal scale consisting of matrix items (2x2 and 3x3 figurative analogies).

This test is useful for people between 4-90 years and can be taken in about 20 minutes. Another M = 100 and SD = 15 is used. Kaufman makes it clear that this test can not replace the traditional tests, it is more of an instrument for screening, for example for the following purposes:

• Making a quick estimate of intelligence where accurate scores are not essential.
• Estimation of verbal versus non-verbal intelligence.
• Re-evaluating the intellectual status of previously tested people.
• Screening students who may benefit from placement in gifted programs.
• Screening students with a high risk who may need further assessment.
• To obtain a quick estimate of intelligence when treating adults in insitutions

Individual tests of achievement

Where the intelligence tests measure the broadness of a person's mental skills, achievement tests measure what a person has learned in school or a course. The focus here is on individual tests, so they are conducted 1 on 1.

Kaufman Test of Educational Achievement-II (KTEA-II)

The KTEA-II is used for children from 4.5 years to 25 years. There is a short version of 3 subtests that can be used for people who are even older than 90. For the mapping of learning problems only the full version is recommended. It consists of eight subtests divided into four areas:

Reading

• Letter and word recognition
• Reading understanding

Mathematics

• Math Concepts and applications
• Math computation

Written language

• Written expression
• Spelling

Oral language

• Listening comprehension
• Spoken expression

Three composite scores can be calculated: reading, math and written language. In addition, a total score can also be calculated. For diagnostic purposes there are also more subtests for reading skills. For older children this test takes 80 minutes, for younger children 30 minutes.

Nature and assessment of learning disabilities

A learning disability can be described as a serious difference between the general intelligence and specific performances in one of the following areas: spoken expression, listening comprehension, written expression, basic level reading, reading comprehension, mathematical calculation and mathematical reasoning. A serious difference was described as one or more standard deviations between a general intelligence and a specific performance. However, this definition was too strict, so that children who did have a learning disability did not come through selection procedures but did need more help.

The new definition of learning disability described it as intrinsic to the individual, a dysfunction of the central nervous system being the cause, and that it can reach into adulthood. With this definition a better separation was made between skill and performance.

Since 2004 there has been a new approach to identifying learning problems. No more attention is paid only to the discrepancy model, but also to the response to an intervention (RTI). Scientifically developed interventions are being used more frequently. This method is therefore not only about identifying, but also about increasing the capacity of the school system to be able to respond to the different needs of the pupils with disabilities.

The approach consists of several steps and starts at the beginning of the year with a general screening of all children to see who is at risk. The children who stand out will receive an intervention for 8 weeks, and if they do not catch up, another intervention will be used again for 8 weeks. If it still does not work, it will be examined on an individual basis where the problems come from and the child may be placed in special education.

Ultimately, there are five characteristics to indicate a learning disability:

1. A learning disability is a discrepancy within a person in terms of cognitive functioning. But it is not just a difference between a general IQ score and performance test scores.
2. The learning disability should not be determined by other limiting conditions, such as retardation or visual impairment.
3. A learning disability is heterogeneous, so there are many different variants of it.
4. A learning disability often arises early in childhood and can exist into adulthood. This is investigated by means of longitudinal studies. It is a developmental phenomenon.
5. People with learning disabilities often experience social and emotional problems that are often just as serious as deficits in academic performance. These problems can also exist in adulthood.

The cause sometimes lies, as with dyslexia, in the brain. In addition, the left hemisphere is limited. Restriction of the right hemisphere involves problems in non-verbal tasks. Often, however, the cause of a learning disability is not known.

It is clear, however, that learning disabilities have a major impact on the lives of the children. That is why individual performance tests are best for measuring this. These tests are taken individually and measure academic skills. This is the point where there are problems. It is often about identifying specific problem areas and this should be taken into account when choosing and administering a test.

How are 'special populations' tested? - Chapter 7

Topic 7A: Infant and preschool assessment

Assessment of the infant capacities

The baby and toddler period (preschool period) lasts from birth to about 6 years old. In this, children develop basic reflexes and motor skills. However, some children lag behind and therefore, parents want to know how far the child is behind and whether their cognitive emotional development is going well. On the other hand, there are also children whose development is far ahead of others. Assessments for infants and toddlers help to investigate this.

The 'Neonatal Behavorial Assessment Scale' (NBAS) is used to investigate the behaviors of newly born children. 28 behaviors are scored on a 9-point scale. Examples are reaction to light, cuddiness and orientation to inanimate visual stimuli. 18 reflexes are also controlled, such as the Babinski reflex, sucking reflex, grasp reflex or the rooting reflex. Finally, the qualities of responsiveness of the weak children are tested with, among other things, the quality of alertness, general irritability, and emotional response.

The NBAS is used as feedback for the parents, so it does not have a real scoring table. The developers want to emphasize that a close parent-child relationship is most important. There are researchers who have issued a scoring system for the NBAS with 7 clusters: habituation, orientation, motor performance, arousal/lability, regulation, autonomic stability and reflexes. The reliability of this scoring system is low, so one has to be careful with interpretation. The development of babies can change quickly and strongly.

The Bayley-III is available for children from 1 to 42 months. The test consists of 5 domains with representative scales, namely:

• Cognitive scale: 91 items, including sensory accuracy, perceptive skills, attention, puzzle solving, matching of colors, counting and investigative and manipulating behavior. This scale has no subscales.
• Language scale: 48 items, includes receptive and expressive communication, such as recognition of sounds, non-verbal expression, identification of action pictures, naming objects, answering questions and following simple instructions. This scale has three subscores: expressive communication, receptive communication and language control score.
• Motor scale: 138 items, includes gross and fine motor skills, such as manipulation of objects, functional use of the hands and motor planning. This scale has three subscores: gross motor skills, fine motor skills and a general motor score.
• Social-emotional scale: 35 items, includes interactive and purposeful use of emotions, expressing feelings and connecting ideas and emotions with each other. This scale has no separate subtests.
• Adaptive behavior scale: parents fill in how the child scores on components such as communication, health and safety, self-care. This scale has a subscore for each component.

The Bayley-III does not follow an overall score, because this can be misleading due to the many skills that are tested. The test is precisely meant to use scale scores for diagnosis and assessment. The reliability of the test is only reasonable, because it is nevertheless a test for young children, which develop quickly, which makes later test scores different (especially for fine motor skills). The validity is good with correlations between .72 and .79.

The Devereux Early Childhood Assessment Clinical Form (DECA-C) is made as an assessment for children from 2 years to about 6 years with social or emotional problems. The test focuses on the protective factors of (the environment of) the child who can partly absorb the problems. The test also contains problem scales.

The test is based on the 'resilience theory' which states that protective factors exist on three levels: environmental, family and within-child. The protective factors can be formed into a Total Protective Factor Score by means of three scales:

• Initiative: investigates the child's ability to think independently and behave according to his needs. Items resemble statements like 'the child gets things by himself'.
• Self-control: examines the capacities of the child to experience and express his different emotions in a socially desirable way. Items resemble statements like "the child can control his temperament".
• Attachment: examines the formation of strong and long-lasting relationships with parents, family and teachers. Items resemble statements like "the child accepts the comfort of adults when he is upset."

The problem factors can be formed into a Total Problem Factor Score by means of four scales:

• Attention problems: examines the child's ability to focus on a task and not be distracted by the environment. Items resemble "the child loses concentration quickly."
• Aggression: measures aggressive or destructive behaviors towards other people or things. Items resemble "the child destroys personal property of others".
• Withdrawal / depression: investigates self-reflection and emotional / social withdrawal. Items resemble 'the child seems to live in his or her own world'.
• Problems with emotional control: measure the problems with controlling negative emotions related to purposeful behavior. Items resemble "The child is upset when things do not go his or her way."

Several studies showed that the reliability of the DECA-C is good. The test also appears to be valid and useful.

Of course, while researching young children one must take into account the troublesome fact that they are still in full development, and thus often do not listen to instructions. This makes the investigation very difficult.

Assessment of preschool intelligence

When examining the intelligence of young children, it must be taken into account that low scores do not directly mean that the child has low cognitive skills. It can also be due to the concentration of the child, or to, for example, environmental factors.

The following three tests are intended to measure intelligence in children.

DAS-II

The Differential Ability Scales-II (DAS-II) can be used for three age categories, but here we focus on preschool children aged 3 to 6 years. The DAS-II consists of ten core subtests, used for measuring cognitive skills, and 10 diagnostic subtests, used for additional information about readiness for school and information processing.

The core subtests correspond with the g factor of intelligence and contain three clusters for verbal power, non-verbal reasoning power and spatial power. From this, a General Conceptual Ability (GCA) and a Special Nonverbal Composite (SNC) can be calculated. The GCA is seen as a precursor of the IQ.

The diagnostic subtests of the DAS-II are only used for clinical analyzes and are less influenced by the g factor. They provide useful information for discovering learning difficulties and the readiness for school, and can therefore also be used to replace core subtests.

The DAS-II is a reliable measuring instrument due to the high reliability coefficients, but as is often the case with research on young children, test-retest reliability studies over several weeks are less high, due to the development of the children. The validity is good, given the high correlations with comparable tests.

WPPSI-IV

The Wechsler Preschool and Primary Scale or Intelligence-IV (WPPSI-IV) is used for children aged 2 to 7 years, but here only the test for older preschool children is discussed: from 4 to 7 years. An advantage of this test is that child-friendly and playful stimulus material is offered.

There are five primary index scales, each with two subtests that summarize the cognitive skills of children. The complete WPPSI-IV includes 13 subtests, but only 6 (the italic) are required to calculate an IQ score.

• Verbal comprehension, subtests: information and similarities.
• Visual spatial, subtests: block design and object assembly.
• Fluid reasoning, subtests: matrix reasoning and picture concepts.
• Working memory: subtests: Picture memory and zoo locations.
• Processing speed: bug search and cancellation.

In addition, there are four subordinate index scales: vocabulary acquisition, nonverbal, general ability and cognitive proficiency. These index scales can be useful in special circumstances, such as the assessment of deaf children or children with slow processing.

Early SB5

The Stanford-Binet Intelligence Scales for Early Childhood (Early SB5) combines the SB5's sub-tests with new Test Observation Checklists (TOC) and parent reports. These are used to get an idea of ​​the child's behavior during the test. A certain behavior, such as not answering a question, can have multiple meanings in children. The child could actually not know the answer, but the child could also be bored, afraid or distracted.

The TOC consists of a part that looks at the characteristics of the child and a part that looks at specific behaviors of the child. Characteristics of the child include: motor skills, activity levels, attention / distractibility, impulsivity, language.

Specific behaviors include: consistency in performance, mood, frustration tolerance / aggressiveness, change in mental set, motivation, fear of failure, degree of cooperativeness or refusal, anxiety, need for redirection, parental behaviors, representativeness of test behaviors.

Practical utility of infant and pre-school assessment

Test scores obtained in the first two years of life often seem to have minimal predictive validity. The important role that these tests nevertheless play will be discussed after a review of predictive studies.

Predictive validity

It is often found that there is a positive correlation between the preschool results and the results of children when they are at school. However, this correlation is not very strong. The correlation only becomes stronger when the children are 19 months old. Only then do the results have a strong, significant and meaningful correlation with the later IQ. The following rule applies: the older the child is during testing, the stronger the relationship with the later IQ. Around the 8^th year of life IQ remains fairly stable. The results show that tests with infants have a low predictive validity and with preschool children or toddlers an average.

Practical utility of infant scales

The main reason for tests in infants is screening for developmental disorders, because it requires early intervention and thus ensures better outcomes later in life. Exceptionally low scores on the Bayley test, especially on the mental scale, can predict mental disability. Furthermore, for tested 'risky' infants the predictive validity is higher than for normal children. It is therefore useful to test infants because it can predict disability later in life.

Fagan Test of Infant Intelligence (FTII)

Lewis found that traditional tests overlooked early information processing processes that can predict cognitive functioning. For example, in his test visual habituation for new stimuli is examined in infants aged 3 months. The correlations with results at later ages were average, which means that cognitive functioning can be predicted early on.

Fagan developed his FTII with inspiration from this test. A face is shown to a baby, and later a similar face is shown, or the same face, but then the face looks differently. It is measured how long the baby looks at the new face. The reliability is high, due to a large inter-assessment agreement. The predictive validity of the FTII is strongly doubted because low correlations are found with the results for cognitive functioning in later life. That is why FTII is seen as a screening instrument for developmental delays rather than as a predictor of general intelligence.

Screening for school readiness

To define what school readiness is, five models are used:

• Maturationist model: according to this model, age is the best way to determine whether a child is ready to go to school, because it is the best indicator for the development of the child.
• Environmental model: here the readiness is based on what skills the child has in social experiences, so parents have to decide on this.
• Constructivist model: here a child is ready for school if it can not only interact well with teachers, but also with other knowledgeable peers and adults, so several people have to co-decide.
• Cumulative-skills model: here the readiness is determined by looking at how many skills the child already possesses that are needed to learn fundamental subjects, such as math and reading.
• Ecological model: here the readiness is seen not only within the child itself, but as an interaction between the child's developments and the child's environments. From this follows a complex and qualitative evaluation.

Screening tests must be quick and simple, but they may have certain errors. There may be false-positive children who fail the test, but who do not really lack anything. And there may be false-negative children who do pass the test, but actually have a developmental deficit. Glascoe and Shapiro (2005) gave five pitfalls in screening tests:

• Wait until the problem is observable.
• Ignore screening results.
• Rely on informal methods.
• Use of inappropriate tests.
• Assume that services do not exist or are limited.

A good screening test should meet the following criteria:

• The primary purpose is screening, so do not investigate, diagnose or predict.
• Screening is provided in areas such as motor, language, cognitive, social and emotional functioning.
• Test-retest reliability must be at least .70.
• Concurrent validity must be at least .70.
• Sensitivity and specificity for positives and negatives should be at least .70.
• The test should last less than 30 minutes.
• The test must be able to screen children from different cultures.
• A minimum of expertise is needed to take the test.

DIAL-4

The Developmental Indicators for the Assessment of Learning-4 (DIAL-4) is a fast and efficient screening test for ages 2 to 5 years. It tests five areas. Three of them are important development domains: motor, concepts and language. The other two (self-help and social-emotional) are filled in by the parents and teacher of kindergarten.

The scoring leaves room for subjective interpretations, which reduces the reliability of the test. Using norm scores, it can be said whether a child scores average in an area or if he has a 'potential delay’. The reliability is reasonably good. The content validity and criterion validity were also high. The construct validity was tested to see if the three areas matched the items and that was true.

Existing scepticism is mainly about the practical use of the DIAL-4. Sensitivity figures ranged from .73 to .82, which is high, but that still means that 18 to 27 percent of potentially delayed children are not recognized as such. Specificity numbers ran from .82 to .86, which means that 14 to 18 percent of normal children are screened as potentially delayed.

Denver-II

This test consists of 125 items divided into four areas: personal-social, fine motor adjustments, language skills and rough motor skills. The test does not produce a developmental quotient, but a score that is described as normal, questionable or abnormal. The reliability is very high with numbers above the .90. However, the predictive validity of the Denver-II is questionable, because the sensitivity is high, but the specificity is not high; many normal children were labeled as doubtful or abnormal.

HOME

The Home Observation for Measurement of the Environment (HOME) is used to observe the physical and social environment of the children while they are at home. The HOME followed after the SES. It was stated as a hypothesis that children of a higher social class enjoy a richer and warmer environment. This was indirectly measured with work and education from the parents.

However, the HOME allows one to directly investigate whether certain crucial interactions and experiences are present. It measures the quality and quantity of stimulation and support for cognitive, social and emotional developments that are available at home.

The researcher scores dichotomically whether certain items are present or not. These are subdivided into six subscales:

• Emotional and verbal responsivity of the parent
• Acceptance of the child’s behavior
• Organization of the environment
• Provision of appropriate play materials
• Parent involvement with child
• Variety of stimulation

The reliability of the HOME is reasonable. Some subscales have few items, so for these the reliability can not be very high. The interobserver agreement is high, as are the internal consistencies. The validity of the HOME is also average. The correlations with the SES should be significant, but not fully related, and studies showed that this is indeed the case.

The HOME is not only promising in research, but also as a practical instrument for interventions.

Topic 7B Testing persons with disabilities

Non-language tests

These are tests that have as little written or spoken language between examiner and examinee as possible. The Leiter-R test is a non-verbal test for children aged 2 to 21 who do not need a spoken language and can therefore be used for children who do not speak the language, children with autism, brain damage and speech or hearing problems. Children must match small cards with figures with an example. The Leiter-R test has 20 subtests divided over two test batteries: Visualization & Reasoning and Memory & Attention.

It is a reliable test with numbers above .90. Because the Leiter-R is also only non-verbal, it can be used for different target groups. The empirical studies support the use of the test, because it proved useful for medically weak children, low-functioning children with autism and children with a language deficiency. The validity was also high, namely r= .80 (correlation with another measure of nonverbal intelligence).

Many children enjoy drawing human figures and can do so routinely, so Florence Goodenough (1926) invented the Draw-A-Man test reviewed by Harris in 1963. However, this was not entirely non-verbal, because the instructions were given orally. The goal was to measure intelligent development, so the more detailed the drawing was, the better the score.

Naglieri used this technique and developed thein 1988 Draw A Person test (DAP), adding a quantitative scoring system. The DAP is praised for its clear scoring, strong reliability and good standardization, but the validity is less good. It does not effectively identify children with learning disabilities and developmental defects.

The Hiskey-Nebraska Test of Learning Aptitude (H-NTLA) is a non-linguistic test and is used for children aged 3 to 17 years. It consists of 12 subtests, from which a Deviant Learning Quotient (LQ) is calculated. The reliability is high, but becomes less after a number of years. The validity is reasonable, given the correlations with other performance tests.

The Test of Nonverbal Intelligence-4 (TONI-4) is a non-linguistic test for people aged 6 to 90 years. It consists of 60 abstract figures in which the participants have to discover relationships between a number of figures. The test has three scoring methods: age comparisons, percentile grades and TONI-4 quotients (average 100, sd 15). The reliability is good (> .90). Furthermore, the TONI-4 is recommended as an intelligence test for people with poor language skills.

Nonreading and motor-reduced tests

The Peabody Picture Vocabulary Test-IV (PPVT-4) is useful for people who can not verbally express themselves and for people who have weakened motor conditions. The test has two parallel versions, four of which have practice plates and 228 testing plates that contain pictures. The tester pronounces a stimulus word and the person must designate the picture that best suits it.

The reliability is high, but is still disappointing among children of foreign origin. The validity of the PPVT-4 is good, but only as a measuring instrument for vocabulary and not as a measuring instrument for general intelligence.

Testing persons with visual impairments

People who are legally blind, who see less than 10% in the right eye, need special tests to be properly examined. The Wechsler test that has been converted for this purpose is the Haptic Intelligence Scale for the Adult Blind (HISAB). It consists of six subtests of Digit Symbols, Block Design, Object Assembly and Picture Completion. The remaining two are Arithmetic sums and Matrix Reasoning using a pattern board.

Another instrument is the Blind Learning Aptitude Test (BLAT), for children aged 6 to 16 years. This test consists of six subtests that work with a kind of braille, also consisting of dots and lines. .

The Intelligence Test for Visually Impaired Children (ITVIC) replaces the nonverbal subtests of IQ tests requiring intact vision with tests that require only sensation, so-called haptic subtests. The verbal subtests remain as usual. This is for children from 6 to 15 years.

Testing people who are deaf or hard of hearing

A test in sign language can not always be the solution, because not everyone has the same sign language. Using someone who translates into sign language can be bad for validity because the translator can change the content unconsciously. That is why it is best for the examiner to be proficient in standard sign language, so that the test remains standardized. The Wechsler tests are best used for deaf or hearing-impaired people.

Assessment of adaptive behaviour in intellectual disability

It is difficult to investigate intellectually disabled people. Someone is intellectually disabled if his IQ is below 75. Intellectual disability is indicated on a continuum of mild, medium, severe and very serious. In addition, the person must also have difficulty with adjustments of behavior in conceptual, social and practical skills.

These skills are examined with different measuring scales. The Scale of Independent Behavior-Revisted (SIB-R) consists of 259 items divided over 14 subscales divided into four clusters: motor skills, social and communication skills, personal life skills and social skills. The scores from these four clusters give an Independence Score. In addition, the SIB-R indicates whether there are problem behaviors and how serious they are.

The SIB-R thus gives a good picture of how the person functions in real life at home, at school and in public places. In addition, standardization, reliability and validity are also sufficient.

The Inventory for Client and Agency Planning (ICAP) is used for children and adults who have become dependent from birth or later in life. The focus is on determining which kind of care is needed. The test is completed by a care provider and there is also room to indicate problem behavior and maladaptive behavior. The advantage of this instrument is that there is a Service Score that indicates how high the urgency is for attention, supervision and training for the person. The lower the score, the higher the urgency.

Lastly, there is the Vineland Adaptive Behavior Scale (VABS-II) which provides an evaluation in the following areas: communication, daily living skills, socialization and motor skills. The predictive validity of the scale is good.

Assessment of autism spectrum disorders

Autism is not a single disorder, but multiple disorders in a spectrum. People with autism differ from each other, but all have difficulties with social skills, communication skills and flexible behavior. In addition, people with autism often lack empathy.

A measuring instrument for autism is the Modified Checklist for Autism in Toddlers (M-CHAT) for 16 to 30 months old children. It is a checklist of 23 items that indicates that the child may be autistic. The M-CHAT, however, has a high false-positive ratio, but that can be acceptable for the number of children who are well diagnosed. It is very important that children with autism are detected early to help them at the earliest possible age.

Another frequently used instrument is the Baby and Infant Screen for Children with Autism Traits, BISCUIT-1. It consists of 71 items on a 3-point scale for toddlers aged 17 to 37 months. The BISCUIT-1 is also a good measuring instrument for discovering autism.