L&B6 - Intelligence Tests

Stanford- Binet Test

- Binet and Simon described the problems that they encountered in their research into the objective of using a test to determine a person’s intellectual capacities;

> They developed a test for mapping out different intellectual functions, which consisted of thirty items that increased in difficulty

intelligence quotient

- Stern developed a test, which involved determining the ratio between the mental age and the chronological age

Factor analysis

- Spearman assumed a small test-specific factor (s) and a general intelligence (g) that mainly explained a person's performance on intellectual tests

Das' PASS theory

- distinguishes four basic functions: planning, attention, simultaneous processing and successive processing

Cattell-Horn-Carroll theory (CHC theory)

- This is hierarchically structured and consists of three layers:

> Stratum III includes g

> Stratum II contains broad cognitive abilities

> stratum I distinguishes between a large number of specific cognitive abilities

Hebb

- referred to the biological capacity for solving problems as intelligence A and culturally determined skills as intelligence B

Cattell

- Cattell saw the ability to use inductive reasoning as fluid intelligence (Gf) and verbal ability as crystallized intelligence (Gc);

> This distinction is also central to the CHC model

> Crystallized intelligence remains relatively unaffected by disruptive factors, such as age or brain damage whilst fluid intelligence is much more sensitive to disruptive factors

Twin studies and intelligence

- Twin studies showed that the majority of intelligence in a Dutch population has a genetic basis, and that less than fifty percent of intelligence is determined by the environment

the measurement of intelligence

- Intelligence tests mostly contain verbal, spatial, numerical, and abstract items;

> Verbal tests are been used for measuring crystallized intelligence, while non-verbal tasks are used for measuring fluid intelligence

- A person’s final performance on an intelligence test is usually compared with the performance of a group of peers and is expressed in a standard score, the deviation IQ;

> The IQ reflects a participant's overall performance on all subtests in comparison with a group of peers

Wechsler Adult Intelligence Scale- IV (WAIS-IV-NL)

- The WAIS-IV-NL consists of fifteen subtests that cover different cognitive abilities, including Arithmetic, Working memory, Vocabulary, Information Processing Speeds, Power of Abstraction, Visual Spatial Problem Solving;

> Some of them are administered under time pressure

- The total administration time is approximately 70 minutes

- The raw scores on the subtests are converted to scaled scores on the basis of age-related norms;

> The scaled scores may subsequently be used to calculate the IQ

four factors that form the basis of the WAIS-IV

- Verbal Comprehension

- Perceptual Reasoning

- Working Memory

- Processing Speed

- A 5th factor was found that divided perceptual reasoning into a visual processing factor and a fluid intelligence factor; consistent with the CHC theory

Groninger Intelligence Test-2 (GIT-2)

- The ultimate version of this test consists of nine subtests;

-> Vocabulary, Doing Sums and Word Enumeration tend to measure crystallized intelligence while the six other subtests tend to measure fluid intelligence

- The GIT-2 has undergone the following changes where necessary:

> Instructions have been modified, the time limits for the non-speed subtests have been omitted and the implementation of the test materials has been revised

- The raw scores (between 0 and 20 for most of the subtests) were converted into C scores, which can be compared with peers;

> The summed C scores can be converted into IQ values

vocabulary subtest

- first subtest; In the Vocabulary subtest, one word with the same meaning as a stimulus word must repeatedly be chosen from a group of five words:

> Primarily measures verbal comprehension and verbal knowledge

puzzles subtest

- The Puzzles subtest measures spatial ability and consists of multiple-choice items in which an empty box must be mentally filled with figures

sailing instructions subtest

- In the Sailing Instructions subtest, the duration of traveling by boat with an increasing number of handicaps must be calculated:

> This subtest measures the ability to reason logically with numbers

sorting subtest

- The Sorting subtest measures the ability to reason logically with shapes:

> Eight cards must repeatedly be divided into two groups according to a principle that the subject must figure out for himself

figure discovery subtest

- The Figure Discovery subtest measures an aspect of perceptual intelligence and consists of incomplete drawings, for which the subject must state what the drawing suggests

doing sums subtest

- In the Doing Sums subtest, the subject must calculate as many simple sums as possible in one minute, to test numerical ability

rotating cards subtest

- The Rotating Cards subtest consists of multiple-choice items, in which a demonstrated manipulation with an L-shape must be mentally carried out with another figure:

> Measures spatial representation

matrices subtest

- In the Matrices subtest, the subject must figure out what the logical principle is that links two word pairs and then apply this principle to a third word pair

word enumeration subtest

- The Word Enumeration subtest measures the fluidity of association and verbal fluency:

> In the first part of this subtest, the subject must name as many animals as possible in one minute

> In the second part of the subtest, he must name as many professions as possible

Kaufman Adolescent and Adult Intelligence Test (KAIT)

- The KAIT can be administered in two ways:

> The so-called core battery consists of six subtests, three for measuring fluid intelligence and three for measuring crystallized intelligence; it can be administered in 60 minutes

- The expanded battery consists of the same six subtests plus four additional subtests involving a fluid and a crystallized test that can be used as a replacement for one of the subtests from the core battery

- In addition to separate scores for fluid and crystallized intelligence, a total IQ score can also be determined for the KAIT, as a summary score for the entire test;

> The authors’ recommendation is that the total IQ should not be interpreted when the crystallized and fluid scores significantly differ from each other

- Has been given a reasonably good COTAN rating, with the exception of the criterion validity, which was rated as ‘unsatisfactory’

symbol learning (KAIT)

- The subject must learn the meaning of a number of symbols and then decipher sentences that are composed of these symbols

logical reasoning (KAIT)

- The subject is offered everyday situations in which changes take place, about which the subject must answer questions

secret codes (KAIT)

- The subject must figure out which symbol matches which characteristic of the icon and, on this basis, must assign the correct symbols to a new icon

definitions (KAIT)

- The subject must indicate which word is meant on the basis of a number of letters in a word and a cryptic description

auditory comprehension (KAIT)

- The subject is offered a news item by means of a sound carrier and must subsequently answer questions about this news item

double meanings (KAIT)

- The subject is repeatedly offered two word pairs, on the basis of which he must identify a target word, associated with both pairs

Raven's Progressive Matrices

- A non-verbal test that measures visual problem solving and can consequently be used as a measure of fluid intelligence;

> The Raven's test may be regarded as an analogue reasoning task, which focuses on the capacity for abstraction and dedicated problem solving

- The administration time is between 40 and 60 minutes

- The raw scores can be converted to percentiles, on the basis of which the IQ can be estimated

three common forms of Raven's Progressive Matrices

- The Standard Progressive Matrices

- The Coloured Progressive Matrices

- The Advanced Progressive Matrices

-> Each version of the test is made up of a number of test booklets, which each contain a number of items that each consist of a pattern with a missing piece;

> The multiple-choice task involves selecting the piece with the pattern that is best suited to the rest of the pattern

advantages of Raven's test

- it does not require a verbal response and it is therefore suitable for use with clients whose mother tongue is not Dutch;

> Out of all intelligence tests, it is the one with the least cultural bias

> Performance on the test is closely linked to the subject's verbal reasoning ability and, for this reason, it may not be seen as a purely non-verbal test

The Standard Progressive Matrices (SPM)

- most widely-used version, consisting of five test booklets (A, B, C, D, E), with items that increase in complexity and level of abstraction;

> The first set (A) consists of a simple visual pattern that must be completed with a missing piece

> Set B consists of a 2 x 2 matrix, in which a missing box must be filled in

> The other sets (C, D, and E) consist of a 3 x 3 matrix, in which the complexity also increases

-> The raw scores can be converted to percentile values or to IQ estimates:

> There is also an abridged administration (only using sets B, C and D), in which the scores can be converted to total scores

> The COTAN rating of the SPM is reasonably good, with the exception of the norms, which were rated ‘unsatisfactory’

The Coloured Progressive Matrices (CMP)

- is easier than the standard version, and is especially suitable for administration to children (5-12), to people in whom a low level of intelligence may be suspected and to patient groups in which there may be severe decline in intellectual functioning;

> The CPM consists of three test booklets (A, B and Ab) and makes use of coloured items, which increase in level of difficulty

> The CPM score results in a reasonably accurate estimate of the overall level of intellectual functioning, but unusual scores are difficult to interpret

> The CPM scores ‘sufficient’ to ‘good’ on all COTAN criteria

The Advanced Progressive Matrices (APM)

- consists of two test booklets with 12 and 36 items;

> Has been developed for use with people who have a high cognitive level

> The items in set I of the APM do not increase in difficulty:

> Set I can also be administered as an abridged version, for which a separate norm table is available

> Factor analytic research has identified four underlying components of the APM: processing speed, working memory, the ability to apply rules and spatial ability

> Dutch norms have not yet been published for the APM and the test has not been submitted to the COTAN for evaluation

Dutch Reading Test for Adults (NLV)

- Is derived from the American National Adult Reading Test (NART);

> Was originally developed for obtaining a reliable estimate of the level of premorbid intellectual functioningin patients with cerebral dysfunctions within a short time

- Involves reading fifty words out loud, for which the correct pronunciation is scored

- The raw score can then be converted to an IQ estimate;

> The basic assumption is that a person’s vocabulary, as a part of the crystallized intelligence, is resistant to cognitive decline

- Administration to a group of neuropsychological patients has shown that there is a high correlation between the NLV IQ, and the verbal IQ in WAIS;

> Higher correlation with level of education

- Patients who suffer from mild forms of dementia score approximately the same as non-demented elderly individuals; more severe forms resulted in lower IQ estimates

-> Indicates that an IQ estimate that is based on the NLV must nevertheless be made with some caution when it involves patients with severe cognitive disorders

> High IQs are underestimated due to the presence of a ceiling efect in the norm group scores

- The COTAN rating of the NLV is reasonably good, with the exception of the criterion validity, which was rated ‘unsatisfactory’;

> The norms are slightly outdated and some words are not as familiar to adolescents, which may result in an overestimate or underestimate of the IQ

> The NLV is not suitable for administration to non-native speakers and dialect speakers, or people with reading disabilities

- The NLV is a reliable and valid test for estimating the level of verbal intelligence, but it is obviously not suitable for determining the level of performance intelligence

The IQ- score as an 'absoulte number'

- The deviation IQ (with an average of 100 and a standard deviation of 15) is used in the Netherlands for selection purposes;

> Problem: the reliability/unreliability of the IQ test fails to be taken into account and the IQ is often reported as an ‘absolute number’, instead of being reported as a valude with a corresponding confidence interval

> The reporting of IQs is strongly discouraged

- In the Netherlands, a classification system would lead to a score of 70 being used as an upper limit for intellectual disability;

> Problem: most intelligence tests are barely able to differentiate between the IQ scores that fall below 70

- In an intelligence test, an IQ value should always be mentioned together with the corresponding confidence interval

-> Using different intelligence tests may also result in different IQ scores and, subsequently, different descriptors

- There will also be fluctuations in a person's IQ during the course of his life;

> As a rule, it can be said that a person’s IQ increases from adolescence to the 40th year of life, after which it slowly decreases before showing a large decrease in high age:

> This does not apply to two of the more ‘crystallized’ subtests in the GIT 2, Vocabulary and Doing Sums;

- This could also be due to a decrease of information processing speed instead of a decrease in fluid intelligence

Flynn effect

- the scores in an intelligence test in the average population increase by approximately five IQ points every ten years;

> The result of the Flynn effect is that the normative data for an intelligence test quickly become outdated

intelligence and level of education

- IQ is a good predictor of scholastic success and a measure of learning aptitude and most tests rely on factual, scholastic knowledge and the ability to solve problems

- score gained on an intelligence test is the best predictor of the relationship between intelligence scores and academic success;

> The highest correlation between intelligence and school performance is found in tests in which the emphasis is placed on reasoning and not on learning

- In the case of working situations, the performance on an intelligence test mostly appears to be linked to success;

> It may make more sense to test specific skills that are useful in the profession, instead of a general intelligence

- Intelligence tests are useful in the school and work space to establish the possibility of an intellectual disability that could lead to problems in that environment;

> A too high IQ score may cause a problem itself in certain professions, if the person perceives the work as uninteresting and not challenging enough

Neuropsychological disorders and intelligence tests

- Administration of an intelligence test can be useful in the early stages of dementia and related disorders, to determine the pace of the decline;

> At later stages of the disorder, intelligence tests can often not be administered due to the complexity of the instructions and the tests themselves:

> The Raven’s CPM can be used for obtaining IQ estimates for this group of patients, because it has a low level of difficulty

- Intelligence tests’ subscales are also useful to examine a specific cognitive function, for example when there is suspected relationship between a brain dysfunction and behaviour problems;

> The specific disorder may go undetected when the focus is on the entire IQ

three levels used for intelligence tests with neuropsychological questions

- The level of the test performance itself (the current level);

> This test performance is related to the patient’s background for the purpose of establishing whether there is a decline in the intellectual functioning

> When there is evidence of specific failure or disorders in certain subtests, it is difficult to provide a valid interpretation at this level

> The performance at this level of an intelligence test is used for interpreting results from other neuropsychological tests

- The second level involves the inventory of disorders; it is established if the performance has worsened or not;

> The test is no longer used for its original purpose, namely determining the IQ, but it is used as a neuropsychological test battery

> Use can be made of the fact that certain subtests are insensitive to decline (‘hold’ tests) while other subtests are sensitive to it (‘don’t hold’ tests)

- The third level relates to statements about abnormal performances on IQ tests, which is often made in terms of the localization of brain damage;

> Making statements about individuals at this third level is undesirable due to of the limited validity of these relationships

subtest within WAIS-IV-NL useful for diagnosing neuropsychological disorders

- The Processing Speed factor is sensitive to the presence of disorders after traumatic brain injury

- The Working Memory factor is a good index for assessing disorders in the working memory

subtest within GIT-2 useful for diagnosing neuropsychological disorders

- Word Enumeration subtext is extremely useful in neuropsychological examination;

> This measure of semantically assisted word fluency can be seen as a measure of the semantic memory and the capacity for verbal expression

subtest within KAIT useful for diagnosing neuropsychological disorders

- The fluid intelligence score is closely associated with neuropsychological tests that make use of executive functions such as planning or updating of the working memory

profile analyses

- Use is often made of profile analyses and discrepancy scores within clinical practice;

> The WAIS-IV-NL may examine the differences between the index scores:

> It can also be used to examine the differences between the individual performances on subtests, taking into account measurement errors and the (un)reliability of difference scores

> The manual also contains data for determining the statistical significance between the index scores and the individual subtests

- A profile analysis, such as a strengths and weaknesses analysis, is often made on the basis of these statistical differences;

> In the case of the GIT, it has been shown that profiles are dependent on IQ scores:

> Individuals with high IQs generally score relatively high on Sorting and low on Do Sums

> People with low IQs do relatively well on Vocabulary but do poorly on Rotating Cards

-> Only a few participants will perform equally well on all of the components of an intelligence test

> The use of statistical-based intelligence profiles in clinical practice without preconceived hypotheses or clear indications of cerebral dysfunction, has no empirical foundation and is strongly discouraged