Hadow (1924)

Notes on the text
Preliminary pages Preface, Contents, Membership, Analysis, Introduction
Chapter 1 History of development of psychological tests
Chapter 2 Summary of evidence
Chapter 3 Possible applications, conclusions and recommendations
Appendix I List of witnesses
Appendix II Recent experiments
Appendix III Use of tests in foreign countries
Appendix IV Standardisation and norms (Cyril Burt)
Appendix V Correlation in mental testing (Cyril Burt)
Appendix VI Grades in US schools (AE Twentyman)
Appendix VII Recent publications
Appendix VIII Examples of tests
Appendix IX General, special and group abilities
Index

Appendix IX
[pages 225 - 238]

The views of various psychologists who gave oral evidence before the Committee or sent memoranda on the following points:

(a) The factors involved in 'general' ability which must be included in testing for educable capacity.

(b) The need for testing for 'special' and 'group' abilities in determining educable capacity.

There was a difference of opinion between psychologists as to whether or not there was one central factor known as 'general' ability. Some said that there was such a factor, others said that there was not, whilst others again held that there were 'group' abilities. He personally had found evidence of the correlation between 'group' abilities. But the evidence was not sufficiently convincing to support the view that such correlation was governed by one central factor pervading all abilities. He was inclined to favour the view that there was such a thing as general ability with sub-factors. At any rate he was of opinion that the factors involved in 'general' ability could only be determined empirically, with the aid of the mathematical theory of correlation. Professor Spearman, working from the standpoint of the mathematical theory of correlation regarded his results as evidence of one central factor. Professor Thomson said that if one took a number of chance abilities and worked out the results one would get by chance the same hierarchy. Professor Thomson said, therefore, that the hierarchy was due to chance, but witness was not convinced that that view was right. He did not consider that such a divergence of views would affect the use of mental tests.

There was evidence in favour of the existence of both 'special' and 'group' abilities, and, therefore, both classes of abilities should be tested for in determining educable capacity.

Dr Burt stated that the educational capacity of a child at any period of his life might be assumed to depend on mental factors of two kinds: (i) inborn psychological abilities of a relatively elementary and general nature; and (ii) acquired capacities of a more complex and specific character, chiefly memories and habits, such as particular items of knowledge and particular forms of skill. The distinction between the inborn and acquired capacities was mainly theoretical.

Of inborn psychological capacities, 'general intelligence ' was by far the most important, as it was also by far the most easily tested. 'General intelligence' comprised a number of factors, but witness would regard it more as a single complex quality than as a group of independent elements. It was best measured by tasks requiring the voluntary maintenance of attention, quick and accurate learning (in the broader sense of the word, namely, adaptation to relatively novel conditions), and on the higher mental levels, reasoning. These should, perhaps, be regarded rather as modes of general ability, than as elements entering into general ability as component factors. Of other inborn elementary capacities, the next most important for educational purposes was probably that which might be described as 'long distance mechanical memory'. By 'long distance' memory witness meant a memory that could retain its powers over a period of weeks or months rather than of minutes; and by 'mechanical memory' he meant recollection of things in arbitrary connection, e.g. historical dates rather than the gist of a continuous passage of prose. The factors he had mentioned were not the only factors, but probably the most important. There was, for example, another factor of lesser importance, namely, imagery. These inborn abilities were inherited at birth and could not be inculcated. There was, therefore, a point of maximum attainment fixed by congenital limitations beyond which a child could not go, and to which many in fact did not reach.

Like most intellectual characteristics that had been subjected to statistical analysis, educational attainments appeared to depend upon capacities of two orders: (i) a more general capacity entering into all subjects alike - a capacity in which 'general intelligence' was the chief factor; (ii) special capacities limited to particular subjects or particular groups of subjects. Of these latter, the most important, or at least the most easily demonstrable in the ordinary elementary school, appeared to be: (a) linguistic or literary ability; (b) arithmetical ability; and (c) manual ability, to which might perhaps be added (d) artistic capacity; and (e) musical capacity. Dr Burt considered that it was possible, and would be useful, to develop a series of tests for measuring most of these capacities in schools.

'General ability' is a somewhat indefinite expression. Its meaning will necessarily vary according to the criterion with reference to which ability is determined. The acid test of life may lead to a verdict very different from that based upon school or academic success, and in precisely the same way school success may not bear out the verdict based upon the results of our special tests for 'general ability'. Personally I think I succeed in keeping matters clear in my own mind by distinguishing between 'tendency' and 'capacity' in the natural endowment of the human being. In discussing the subject of mental testing, it is certain that some such distinction must be drawn and adhered to. As a general rule mental tests are tests of 'capacity' and do not profess to be more. Indirectly and incidentally they may throw valuable light on 'tendency', but they are not specifically adapted for this purpose, and it would seem well nigh impossible to adapt them, at least on any basis of mental testing yet proposed. In what follows it must therefore be understood that 'general ability' and 'educable capacity', though really including factors of the 'tendency' order - especially the former - are, for the purpose of the present discussion, being taken more narrowly, and in the strict sense of 'capacity' as opposed to 'tendency'.

The factors to be included fell into three groups:

(I) The processes basic to, and constitutive of, intellect, viz. motor ability, attention, memory, imagination, analytic comparison, generalisation, inference.

(II) The functioning of intellect:

(1) In the command of language.
(2) In the application of past experience to theoretical and practical problems.
(3) In the use of standards of judgement and in the quality of judgements.

The Binet-Simon tests in their original and adapted forms, and 'alpha' test of the American army, the college entrance tests, the Hamburg secondary school selection tests, in so far as they were not information tests, fell under groups I and II. The educational bearing of the psychological doctrines of instinct and interest was not adequately recognised in existing tests, but was of first-rate importance in interpreting results (Group III).

To a great extent interest and ability went together, and it was almost impossible to test one without the other. A child was born with certain range of instinctive ability varying in degree in different directions, such ability determining interest, but the child had also spontaneous activity and from this as well as from his instincts developed the interests which made up his individuality. Unless ability was tested along the lines of interest the child had not a fair chance of making a good score.

The whole question of 'special' and 'group' abilities was still in the melting pot. The validity of inferences based on correlation coefficients as to group abilities or a general common factor was questionable. The conditions upon which excellence of performance in a particular set of tests depended were insufficiently known. Theories as to group abilities, levels, transfer of training, application of a common method, appeal to a common interest, needed further elucidation. In the present state of ignorance a varied range of tests was essential.

A definite programme of tests extending over a considerable period of time and carried out on an extensive scale was necessary for solution of this and other problems.

Investigation had so far been based upon two assumptions:

(a) that intelligence was something in the nature of mind energy (a central factor), which entered, in varying degrees, into all controlled behaviour, from simple repeated movement through sensory discrimination to the recognition of analogies and opposites, from the most elementary function (single factors isolated as far as possible) to the most complex function (group factors working in combination).

(b) That a hierarchical system of correlations was evidence of the existence of a central factor.

On the second point the question whether a hierarchy of correlations could be artificially produced where there was no central factor at work, and whether the mere discovery of a hierarchy was, therefore, conclusive evidence of the existence of a central factor, were under dispute and the witness was awaiting the outcome.

At this point one might perhaps usefully draw a distinction between intellect and intelligence. The biologist distinguished tacitly between them when he spoke of an order of intelligence in the animal world. A dog might be very intelligent, but never intellectual. A workman was often very intelligent though even under the most favourable environment he would not have been intellectual. Conversely, highly intellectual men were sometimes less intelligent than they were intellectual. Gaucherie and absent-mindedness, and even childishness, were features commonly expressed in the caricature of the professor. This might, of course, be expressed in another way. Mind never worked in a vacuum. It might create a world of its own and determine behaviour with reference to that world. Within this intellectually created world, the professor was highly intelligent, but he had in the process of building his world lost touch with that of ordinary men, in which world therefore he often behaved foolishly.

If this view were true, it seemed to follow -

(a) That intelligence was not a summation of parts, or a multi-dimensional entity capable of resolution in certain directions.

(b) That intelligence was correlative to the universe in which it worked. As the universe developed, intelligence developed. They conditioned each other, and intelligence was revealed by its universe of operation. In other words one could not measure a potential intelligence, though one might, in some sense, gauge an active intelligence by a comparative estimate of its efficiency within its sphere of work. Intelligence did not exist in vacuo. Professor Green did not agree with the expert psychologists who said that the intelligence of the individual did not grow after about the age of 15 or 16. He considered that as a man's environment expanded so his intelligence grew. He would accept the statement that a man of 40 would do no better than a boy of 16 in certain psychological tests (e.g. in the detection of absurdities), but he would not accept such tests as a measure of intelligence. The man of 40 required a different test of intelligence from the boy of 16. If we are to give both the same test, it must obviously be a test within the universe of the 16 year old. That meant a serious handicap for the older man. It was perhaps comparable to a test devised to compare the intelligence of a child and a dog. To make comparison possible the test must have reference to a world in which both child and dog could function. Might not the dog prove superior in such a case?

(c) That tests of intelligence must take into chief consideration the world in which the persons concerned had been living, and the possibilities of their rising to the demands of a larger world.

A in a universe a, had efficiency .75.
B in a universe b, had efficiency .5.

With regard to the need for testing 'special' and 'group' abilities. Professor Green said that if this meant the various types of universe in which different 'intelligence' might operate most efficiently there was probably real need for enquiry in this direction. Psychological tests were probably used very slightly anywhere except in America.

There were some practical difficulties in assessing intelligence (arising from its mode of working).

(a) Intelligence was often curiously shy. Even in the same person its operative force could not be steadily guaranteed. This was shown by the relatively low reliability coefficients often obtained in various kinds of problem work.

(b) Intelligence was apt to show itself in restricted fields, or to operate with varying efficiency in different fields, e.g. a class of fifty boys (Standard VA) working arithmetical problems and Burt's reasoning tests gave a correlation as low as .34.

There were many schoolboys (perhaps more schoolgirls) who were prevented from matriculating by the requirements in mathematics.

These differences were probably due to fundamental differences (e.g. simple quantitative relations) in sensitivity. Restriction of this kind, however, meant an intelligence of restricted range, and a test in a wrong field would give a very unreliable result for the individual concerned.

(c) Even though intelligence was defined as the integrative function of mind the integrative process really included analysis. It could only put together what it has previously taken to pieces, but the taking to pieces (when it was conscious), must have been done upon a plan which itself was the outcome of an integral view, and when it was not conscious, as in childhood, the analysis followed more or less closely, a plan determined by our racial history. The difference between a penetrating and a superficial intelligence was due to differences in analytical power. So we were brought back again to the capacity to see things as wholes as being a fundamental characteristic of intelligence.

Dr Lewis said that all tests of general ability were really tests of special abilities. Although there is a 'common factor' involved in general ability, no tests yet devised measure this factor directly. The most important factors involved in general ability are voluntary attention, reasoning, the ability to choose the essentials in any situation, and the power to apply previous knowledge in solving new problems.

One method of approaching this problem was to find what mental tests differentiated most definitely the normal from the subnormal person. Records with the following Binet tests showed the greatest disparity between these two classes of persons:

Comprehension of difficult questions.
Reconstruction of dissected sentences.
Detection of absurdities.
Definition in terms superior to use.
Definition of abstract terms.
Tests of immediate memory.
Counting backwards.

The present system of educating children in large classes made it difficult to recognise special abilities, and still more difficult to foster them. Specially devised psychological tests could give the teacher valuable information in this respect.

In the past, the failure to analyse carefully the child's mental processes had led us to attribute to general intelligence much that was really a manifestation of special abilities.

In the study of mentally defective children specific disabilities seemed to be of greater significance than the general mental retardation.

Mr Muscio suggested that educable capacity might not be a unitary thing, but that a person might be educable in one direction and not in another. Sufficient investigation had been done to justify persistent research into this suggestion. Probably, however, educable capacity of any sort demanded the presence of the following three factors, involved in 'general' ability.

(a) The capacity to perceive or apprehend relations.
(b) The capacity to adjust thinking and action to new conditions; and
(c) The capacity to combine particular experiences into significant wholes.

Dr Nunn stated that in testing a person the aim was at bottom, always the same; namely, to infer from the way in which he dealt with a given situation the way in which he was likely to deal with others. The inference was safer the more closely the subsequent situations resembled the first, but it was never free from an element of risk. A boy might jump a certain height today, but it was not certain that he would be able to reproduce the feat tomorrow. In general, however, the situations contemplated in the inference were of necessity different from the test situation, and might differ from it very widely - as when from the quality of a young man's Greek verses his probable capacity as an administrator was inferred. In such cases the inference was valid only if it were known beforehand that the capacities brought into play in the test situation and in the subsequent situations were, if not identical, at least definitely correlated. In other words, the procedure must be based upon a well-founded theory of abilities.

At present there was no such theory universally accepted by psychologists. It might be taken as agreed that situations could be classified into groups, such that inference was always possible from observed ability to deal with one situation to probable ability to deal with another of the same group. For instance, a boy who was good at tennis was very likely to be good at racquets, fives and badminton; and one who learnt Latin easily would probably learn certain other languages easily. The question in dispute was whether ability in one group was a valid basis for inferring ability in a group which appeared to have nothing in common with the former. To that question Professor Thorndike has replied 'No', Professor Spearman, 'Yes'.

It was somewhat widely believed that Professor Godfrey Thomson who had dealt with the subject very acutely, had recently proved Dr Spearman's answer to be untenable. This was a mistake. Spearman had expanded his 'Yes' into the statement that a single 'central intellective factor' entered into the constitution of all varieties of intellectual performance; it was against this statement, and, in particular, against the chief mathematical argument used to support it, that Thomson had directed his attacks. It must not, however, be assumed that these attacks had been successful. From personal discussion. Professor Nunn had gathered that Professor Thomson did not dispute the position in the form in which witness had expressed it; namely, that a person's powers to deal with situations of widely differing character were, in so far as they involved intellect, connected systematically in such a way that it was always possible to make an inference from his ability in dealing with one type to his probable ability in dealing with another type. And that Professor Nunn took to be the essence of Professor Spearman's contention. He himself agreed with this contention.

The controversy had far more than mere academic interest; for unless Spearman's view, as Professor Nunn had put it, was sound, inferences based upon the results of tests - whether the new psychological tests or the older competitive examinations - must often be sadly precarious.

He agreed that there were sufficient instances of people being abnormally gifted in one activity and far below normal in all others to make the subject an important one for discussion. At present there was no explanation; a partial explanation might be that the one subject had sapped all the individual's interest, leaving none for other subjects. There was also evidence of people whose interest took one direction at first and another direction later. With regard to great musicians witness considered that ability in music must also presuppose general ability.

He would define precocity in children as the premature emergence of certain faculties as compared with the average time of emergence in normal children. He was not prepared to say that a child could be precocious in one form of ability and of less than average ability in other forms; it was probably more true to say that precocity meant more general capacity at a certain age.

Note. It is to be understood that the opinions expressed by Mr Richardson are his private views, and must not be taken to represent the opinion or the policy of the Board of Education.

Mr Richardson said that while it should not be forgotten that in all mental processes the mind worked as a whole, and therefore any isolation of special factors was largely arbitrary and abstract, the various aspects of intelligence might be conveniently classified so far as testing for educable capacity was concerned, as follows:

(1) Readiness and ability in applying knowledge.
(2) Discrimination of essentials.
(3) Associative processes: (a) richness and maturity; (b) logical integrity.
(4) Power to control and concentrate attention.
(5) Power of comprehension.
(6) Ability to hold in mind the conditions of a problem.
(7) Practical judgement and ingenuity.
(8) Steadiness of purpose.
(9) Power of forming abstract ideas.
(10) Power of generalisation.
(11) Critical ability, including self-criticism.
(12) Ability to manipulate imagery.

Mr Richardson thought it unlikely that special abilities existed in the same sense as general ability. It was becoming sufficiently evident that in all forms of 'intelligent' mental activity there was present a certain factor to which the name 'general ability' might fairly be given. It then seemed probable that special ability or (better) 'aptitude', was a combination of general ability and special interest, and, in some cases, special temperamental and physical characteristics.

Hence in determining educable capacity the prime necessity was to determine the level of general ability. In children of superior intelligence this would usually be sufficient, for as a result of the tests they would be known to be easily educable, and any special aptitude would, without doubt, show itself in the natural course of their development. But in the case of children of a low degree of intelligence, it would be extremely useful to discover their special aptitude (if any), for it would then be possible to make the most of their limited educable capacity.

Briefly then, supposing all children to be tested for general ability, it was desirable in the case of those testing below a certain level, to make further suitable tests in order to discover their special aptitudes.

He was not personally a believer in the existence of a 'faculty' called general ability. He considered that the statistical work of those who affirmed this theory was of doubtful validity. He had, of course, no objection to the popular use of the term to indicate the average ability which a man showed.

His own opinion was that ability in man was a very much more complex affair than this, and he was convinced that differences of type occurred which were not explicable by a general factor plus a unique or specific factor. It was not necessary to postulate the existence of a general factor, but educational units should be considered. The fact that a man was better in any particular activity was due to the pruning away of unnecessary mental factors.

Note. It is to be understood that the opinions expressed by Mr Watts are his private views, and must not be taken to represent the opinion or the policy of the Board of Education.

The ability that became and remained 'general' was probably rare. 'General ability', as the psychologist knew it, was the form of intelligence which was of the nature of intellect as distinct from instinct or intuition - it involved the capacity to deal with 'general' ideas. Tests like the USA army Alpha tests would discover general intellectual ability. The young child had general intelligence rather than 'general ability' - if one might be allowed to differentiate; it was general ability undeveloped, still at the concrete level, which he displayed. The only method of testing general intelligence at this stage would seem to be to take samples of ability in a very large number of diverse performances. This was impossible practically. A sampling would usually introduce the element of unfairness.

The 'special' abilities which needed investigation were:

(1) Musical ability Witness recommended the use of the Seashore Gramophone Test Records for this purpose. (Ability to discriminate differences of pitch and intensity, sense of consonance and tonal memory were tested).

(2) Arithmetical capacity The ordinary examinations usually did this, but they were occasionally apt not to produce thoroughly reliable results.

(3) Language capacity This was usually tested via composition. Added information could be secured from children's gradings of a number of compositions of obviously varying merit.

(4) Drawing capacity and mechanical dexterity Witness knew of no reliable tests.

Arising from the above and energised by particular 'instincts' and 'interests' are many of the aptitudes making for success in industry, trade and commerce. At present these are untestable.

Mr Winch said that the meaning and existence of general ability was much disputed among psychologists. The older conception of faculties was largely given up. Personally he would include a test in rote memory, substance memory, imagination and reasoning as necessary to estimate so-called general ability, with possibly a perception test. There was a difference between 'rote memory' and 'substance memory': the one entailed the memorising and reproduction of words and the other the reproduction of ideas underlying the passage to be memorised. Of the two 'substance memory' was usually the better done, as it was the easier. The ability for rote memory and for substance memory steadily rose up to the end of the elementary school age.

Whether abilities did fall into 'groups' was again a matter of dispute among psychologists. Using the term 'general ability' loosely as a sum or average of several factors, witness thought that tests should be used in all cases, where possible, as above indicated. Special tests would be required for draughtsmanship and constructive ability of a mechanical kind in addition to those of linguistic constructiveness involved in the general ability tests.