Language Assessment for Remediation (1981)

David J Muller, Sian M. Munro, Christopher Code

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781351215282
• Edition: 1
• Topic: Education
• Subtopic: Education General

Published in 1981, this book describes and critically examines the standardised tests and modes of assessment available and most commonly used by speech therapists, psychologists and educationalists.

Tests and other assessment procedures are discussed and therapeutic strategies suggested. Thus, psycholinguistic approaches such as ITPA, the Reynell Developmental Language Scales and the Aston Index; linguistic techniques such as LARSP and phonological assessments are described, and adult disorders as well as childhood problems, are reviewed. There is also a brief consideration of the problem of assessing the language of those not speaking English as a first language.

The book serves as a core text for student speech therapists and also as a reference for those practicing or researching in speech therapy, special education and linguistic pathology.

1  LANGUAGE, INTELLIGENCE AND REMEDIATION

Chapter Outline

The role of language in intellectual functioning is discussed. It is argued that in order to remediate deficient intellectual performance, special emphasis needs to be given to language and this naturally involves the accurate assessment of linguistic behaviour. Three standardised intelligence tests are evaluated in the light of their success in diagnosing language deficiencies. The suggestion is put forward that although of value in identifying potential language difficulties, these tests are of limited use in making diagnoses and in suggesting remediation. It is concluded that intelligence tests need supplementing with other measures specifically designed to identify and remediate deficient language performance.

Introduction

It is almost impossible to discuss intellectual behaviour without referring in some way, either explicitly or implicitly, to the role of language. Yet, there has been little agreement concerning the exact nature of intelligence and Pyle (1979) begins an excellent text by noting that even after years of investigation and debate, there is still no consensus on what intelligence actually is. This has led to considerable uncertainty concerning the best way in which to assess intellectual activities in order to implement remediation.

Despite this uncertainty, it is beyond dispute that children and adults engage in activities which are termed intellectual, and that in doing so they use language a great deal of the time. In fact Viaud (1960) has argued that from an evolutionary perspective language has acted as the dividing line between specifically human ways of thought as compared to those ways which are common to man and animals.

Consequently, where it is intended to remediate intellectual behaviour, it is almost impossible to do so without recourse to language. At the simplest level, therapists are almost certain to work through the verbal medium, and it is likely that the treatment itself will be directed at various forms of linguistic activities and involve quite complex procedures. This means that in order to plan and implement effective remediation, careful consideration must be given initially to the role of language in intellectual behaviour.

Language and Intelligence

Discussion of the relationship between language and intellectual development, has often centred around the debate of whether language precedes and hence is causal in the development of thinking, or if it is simply a means of representing what is already known. The former view owes a great deal to the seminal Soviet work of Vygotsky (1934) which was not translated into English until 1962. This is in contrast to the Genevan school led and dominated by Piaget and his co-workers (see for example Piaget, 1968; Sinclair-de-Zwart, 1969). The practical value of this debate is unclear, but an extremely lucid review of the theoretical viewpoints and related research has been undertaken by Cromer (1974).

However, Bruner (1975) has put forward a solution to this problem which seems to have the greatest therapeutic potential. He has suggested that at times language is causal and ‘moulds’ an individual’s thinking and that at other times language acts as a ‘cloak’ encompassing and representing what has already been learned. Implicit in this argument is the view that the relationship between language and thought is determined situationally. Developmentally, it appears reasonable to assume that as children mature they learn to utilise language in ways which are most appropriate to their intellectual activities.

Consequently, the debate concerning whether language precedes thought or thought precedes language, is of limited value to those with an interest in remediation. It is clear that language interacts with and influences intellectual behaviour in two main ways. First, language is more often than not the medium through which an individual’s ideas or thoughts are made public and hence serves an important representative function. Secondly, in a variety of situations, language initiates and refines the thinking process and enables new ways of perceiving the world to emerge. In serving both these functions, language is without doubt a major influence in intellectual behaviour, and it is almost impossible and of little therapeutic value to conceive of language without thought or thought without language.

This suggests that in order to remediate intellectual functioning, it is important to consider these particular functions of language. In particular, careful attention must be paid to assessing language in order to diagnose the problem and hence plan remediation. A starting point is to examine how successfully standardised tests of intelligence measure language and to see whether they provide sufficient information to suggest reasons for any deficiencies.

The Stanford-Binet Intelligence Scale (1960)

The Stanford-Binet has undergone a number of revisions since its inception by Binet at the turn of this century. The version currently in use is the 1960 revision (Terman and Merrill) which is suitable for children aged from 2 to 14 years. There are also four adult scales. In the age range 2-5 years, children gain one month towards their mental ages for each subtest correctly completed, whereas from the age of 5 years they gain two months. Hence six subtests represent each half year in the first part of the Scale and a year in the second part.

This procedure of gaining credits in the form of months results in intelligence being stated in terms of a mental age. Hence, a child of 6 years who is able to complete as many subtests as might be expected of a 7 year old, would have a mental age of 7. This meant that in designing the Scale, the subtests had to be arranged in order of difficulty rather than by content and had to reflect what children might be expected to do at given chronological ages. Therefore, the mental age score reflects overall intellectual functioning rather than specific abilities and the IQ (Intelligence Quotient) score which can be attained from the mental age score is related to general performance.

In order to give an indication of the range of abilities assessed by the Scales, Sattler (1974) has classified the subtests according to their content. He has suggested that seven areas are assessed as follows: language; memory; conceptual thinking; reasoning (verbal and nonverbal); numerical reasoning; visual motor skills; and social intelligence. Furthermore, factor analysing the subtests suggests that there is a heavy loading on verbal production and verbal reasoning. This is supported by an analysis of the number of subtests reflecting verbal skills. Approximately one third of the subtests can be classified as assessing language ability or verbal reasoning in some way.

From a therapeutic perspective, the assessment provided by the Stanford-Binet tends to be more descriptive than diagnostic. Many of the subtests involving language, assess knowledge of vocabulary including identifying parts of the body, recognising pictures and defining the meaning of words. Other subtests assess verbal abilities such as eliciting opposites or explaining differences between various items. There is little emphasis given to an objective assessment of verbal production and more stress is placed on the skills of comprehension.

In conclusion, although the Stanford-Binet enables language performance to be described in such a way that general problems can be noted, it has limited therapeutic implications. From the subtests alone, it is difficult to explain why children may be having difficulties and hence the Scales are of limited diagnostic value. But perhaps the greatest difficulty in using the Stanford-Binet to assess language is that the subtests are arbitrarily distributed. Whereas at the age of 2½ years at least four subtests assess language, at the age of 4½ years it can be argued that there is only one sub test relating to language. This makes it almost impossible to assess fully language at all the age ranges. These two factors make it difficult to use the Stanford-Binet as a guide for remediating deficient language performance.

The Wechsler Scales

David Wechsler has been responsible for producing a series of tests designed to assess intelligence. There are three tests readily available; the Wechsler Pre-school and Primary Scale of Intelligence (WPPSI) published in 1967 for use with children aged 4 to 6½ years, the Wechsler Intelligence Scale for Children (WISC) revised in 1974 (WISC-R) and for use with children aged 6 to 16 years 11 months, and the Wechsler Adult Intelligence Scale (WAIS) published in 1955.

All three tests have a similar format, being divided into a verbal scale and a performance scale. Each scale has up to six subtests. This format makes it possible to calculate separate verbal and performance IQs as well as providing an overall measure of intelligence. The concept of mental age used in the Stanford-Binet has been replaced by the use of a deviation intelligence quotient. This is obtained by comparing each subject’s test performance with the scores earned by others in their own age group, rather than with the complete age range for which the test is designed. Wechsler (1974) has argued that this approach produces more meaningful results.

The verbal scale is designed to measure those abilities which are based on previously learned verbal material and are to a large extent dependent upon experience. It consists of the following subtests: information (general knowledge); comprehension; arithmetic; similarities; vocabulary; and short term memory of digits. The performance scale is intended to measure immediate problem solving ability and relies less on the recall of past experience and more on practical expertise in novel situations. It consists of the following sub tests: picture completion; picture arrangement; block design; object assembly; and coding or mazes. Most of these are timed and bonus marks awarded for quick performance.

At a therapeutic level it would appear that the Wechsler Scales are particularly useful in identifying specific language deficiencies. Using the scales it is possible to compare any individual’s verbal score with their performance score and their overall general intelligence. Furthermore, discrete subtest scores can be compared and profiles of their overall abilities built up. In the case of dyslexic children for example, it has been shown by Thomson and Grant (1979) that they score lower than controls on Information, Digit Span, Arithmetic and Coding, but higher on Object Assembly and Picture Completion.

However, the diagnostic value of the Wechsler Scales is not as clearly evident as might be expected. The main difficulty in using them to assess language, is that it is questionable whether the subtests of the verbal scale are comprehensive. Most of them appear to measure only receptive language skills without testing for any understanding of grammatical forms. A second and related problem is the question of whether or not all the verbal subtests actually tap language ability. It is questionable whether remembering a list of digits or doing mental arithmetic are really tests of verbal ability. Similarly, testing for general knowledge is not really assessing language ability.

Therefore, although the Wechsler Scales measure verbal and performance abilities separately it can be argued that the subtests used to assess verbal ability are of little diagnostic value. There are similarities with the Stanford-Binet in that, although both include assessments of language ability, neither are sufficiently comprehensive or specific to give a clear indication of the therapeutic procedures which can be adopted.

The British Ability Scales

Elliott, Murray and Pearson (1978) designed the British Ability Scales to assess a wide range of capabilities for the age range 2½ to 17 years. The Scales are unique in that they have been standardised on a British sample using criterion referencing which allows each individual’s performance to be assessed in relationship to a particular ability, rather than with a sample of the population. This approach is based on a statistical procedure called Rasch scaling.

There are twenty-four separate subtests or scales altogether, including some which assess traditional school-based abilities. It is possible to calculate IQ scores by administering only four of these scales, separate combinations having been specified for three broad age ranges. The scales are grouped according to the underlying processes being assessed as follows: visual-motor speed; reasoning; spatial imagery; perceptual matching; short-term memory; and retrieval to facilitate the application of knowledge.

A number of these processes are considered to have a verbal component of varying degrees. Three out of four of the scales assessing reasoning rely on language: those of formal operational thinking and social reasoning are presented verbally, and the scale assessing knowledge of similarities is based on verbal items. Two of the scales assessing perceptual matching include copying or matching letter-like designs and the third involves finding objects after receiving verbal descriptions. For over half of the scales assessing short term memory the mode of response is verbal.

However, the most detailed assessments of language ability are grouped under the process of retrieving and applying knowledge, where five out of seven scales have an obvious verbal component. A test of naming vocabulary designed for children aged 2½ to 8 years assesses the ability to name objects presented pictorially. There is a scale to test verbal comprehension for the same age range which involves carrying out various actions in response to verbal commands. There is a test of verbal fluency designed for children aged 4 to 17 years to assess the creative use of language. Finally, there are two tests of reading and defining words, for children aged 5 to 14 and 5 to 17 years respectively.

This wide range of scales measuring language is far more comprehensive than the assessments included in either the Stanford-Binet or Wechsler scales. Using the British Ability Scales it is possible to assess knowledge of vocabulary, comprehension skills, production skills (creative and non-creative), reading levels and the ability to reason verbally. Furthermore, it is possible to select scalesf which measure specific language abilities as required.

However, it can be seen that like the tests previously mentioned, the British Ability Scales are only indicative of a problem and are not really diagnostic in the sense of highlighting specific disabilities and suggesting therapy. They appear to describe quite accurately what children are capable of, given certain tasks, but they do not fully explain the reasons behind any individual child’s performance. Knowing that a child has a limited vocabulary or has difficulties in reading does not in any way guide the therapist towards possible therapeutic techniques, despite being able to identify the problem. In conclusion, it does not seem that the British Ability Scales are of any more practical value in guiding remediation that either the Stanford-Binet or Wechsler scales, although such a conclusion may be rather premature at this stage and should consequently be viewed with caution.

Conclusion

Although definitions of intelligence rarely include specific reference to language it is implicit in most viewpoints that language plays a major role in intellectual behaviour. This is reflected in the range of abilities assessed in traditional intelligence tests, all of which include items specifically designed to assess language. Thus by using intelligence tests it is possible to get a clear indication of how well an individual is able to perform on tasks requiring the use of language. In this sense they can be viewed as screening tests which enable attention to be focused on individuals identified as having distinct language problems.

However, rather than simply describing and identifying a problem it is of value in guiding remediation to employ a diagnostic approach towards assessment. There are at least three stages in this process. First, the problem must be identified and it has to be shown that it can be attributed to the relationship between language and intellectual performance. This can be seen as the screening stage where an attempt is made to identify individuals at risk. Intelligence tests are able to fulfil this function with a considerable degree of reliability.

The second stage is to attempt to describe any deficient language performance as accurately as possible in order that the specific problem can be identified. Intelligence tests meet with varying degrees of success at this stage. The most comprehensive test discussed in this chapter is the British Ability Scales, from which specific subtests can be selected as required. However, there are considerable limitations in the descriptive capacity of the British Ability Scales. By themselves the Scales do not really provide sufficiently precise descriptions of the language behaviour under investigation nor is it possible to generalise from them to include areas not specifically tested. It may simply be, for example, that an individual’s poor performance on comprehension is purely test specific and may not be repeated on a different test. Taking the second stage overall, it can be concluded that although intelligence tests meet with some success in providing descriptions of the behaviour in question, considerable limitations are more evident.

The third stage is to try and make a diagnosis which in turn suggests either one or more therapeutic strategies which might be implemented and evaluated. In other words the tests used need to generate possible reasons for poor performance. Circular arguments such as the notion that a child has poor language ability because of low intelligence, give no indication of possible approaches towards remediation. It is at this stage that intelligence tests are of less value. Even the British Ability Scales give very little information to assist the therapist in implementing treatment. Furthermore, intelligence tests are of even less value at this stage for use with adults suffering from acquired language disorders, due to the difficulty of estimating intelligence prior to the onset of their problem.

This distinction between description and diagnosis will become an important theme in this book. Remediation depends not only on recognising and describing the problem, but more importantly on being able to make a prognosis and providing some indication of the therapeutic strategies which can be implemented and evaluated. The following chapters will examine this in more detail, concentrating on those approaches to assessing language which have more direct implications for therapy than would be gained from using intelligence tests.

2 EVALUATION OF PHONOLOGICAL AND GRAMMATICAL ABILITIES IN CHILDREN

Chapter Outline

Six of the most frequently encountered tests and techniques for assessing phonological and grammatical developme...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Preface
• 1. Language, Intelligence and Remediation
• 2. Evaluation of Phonological and Grammatical Abilities in Children
• 3. Developmental Psycholinguistic Approaches
• 4. Childhood Bilingualism
• 5. Assessing Adult Aphasia for Remediation
• 6. Applying Behaviouristic Techniques
• 7. Naturalistic Considerations
• References
• Name Index
• Subject Index