Vocational Testing in Relation to Professor Spearman’s Theories
- Author:
Earle.
(Reprinted by courtesy of the National Institute of Industrial Psychology. The use of psychological tests for various educational and vocational purposes has been growing rapidly during the last ten years and many persons are con- vinced of their value. Yet the nature of the mental functions which are measured by these tests is still disputed to an extent which threatens to delay their full and proper application. Quite recently 1 Professor Spearman has suggested that the procedure of many compilers and users of psychological tests is not founded upon any sound or consistent theory as to the nature of ability and that, in consequence, their conclusions are not sufficiently reliable.
Such criticism from the author of some of the most original and constructive contributions to modern theories of intelligence must be taken seriously. But, although no one who has thoroughly examined the processes of test construction and standardisation will deny that a special knowledge and technique is required for such work, wholesale scepticism of its value is difficult to justify. When they have been carefully constructed and used, psychological tests have proved of great service in the organisation of school activities, in the important work of vocational guidance, and in other ways. It will be interesting, therefore, to examine current practice in mental testing in the light of Professor Spearman’s present theories as to the nature of ability.
1 Spearman, C., The Abilities of Man (London : Macmillan & Co. 1927), Appendix, p. xviii THE TWO-FACTOR THEORY.
For those who are unacquainted with Professor Spearman’s writings some explanation is necessary. The Two-Factor Theory, in brief, is that all human activities which are mental in origin depend for their successful operation upon at least two factors?one a ‘ general ‘ factor common to them all, the other a ‘ specific ‘ factor peculiar to each activity, In other words, the general factor (called ‘ g ‘) is supposed to operate to a greater or less extent in every mental activity, whereas the specific factors vary from one task to another. The specific factors are regarded as almost entirely uncorrected, that is, there is no factor common to them all, and the possibility of one or more factors common to small groups of abilities is regarded as remote. If there were bonds of connection be- tween abilities other than that due to the general factor ‘ g,’ we should be able to use our knowledge of a person’s proficiency in one activity to predict his prob- able success in another activity belonging to the same group. Professor Spearman, however, denies the existence of such so-called ‘ group ‘ factors, at all events over such a wide range as would be necessary to justify the doctrine of ‘ mental faculties.’ 1 For him there are no ‘ faculties ‘ such as ‘ the power of attention,’ ‘ inventiveness,’ the ‘ capacity for abstract thought,’ and the like. There is only the single general ability ‘ g ‘ and many unrelated specific abilities’s.’ If ‘ over- lap ‘ of specific factors occurs, as between the ability to perform one task and that required for success in another, it is usually, he says, because the tasks very closely fesemble one another, so that the specific abilities are substantially the same. Nevertheless Professor Spearman allows that there may be ‘ group ‘ factors over a somewhat narrow range, such, for instance, as in operations which require the perception of shape and form, (in geometry, for example), or in tasks which involve constructive mechanical ability.2
Persons may be regarded as differing greatly in their endowment in regard both to ‘ g ‘ and to ‘ s.’ Accordingly, in order to predict the success of any person in a given activity, it is necessary to know (i) the relative values of the ‘ g ‘ and the ‘ s ‘ which together make up the ability required for the task; (ii) the amount of ‘ g ‘ the person is ordinarily capable of employing; (iii) the extent to which a specific ability for the task in question supple- ments or becomes a substitute for the success which is otherwise due to ‘ g.’ This analysis becomes more intelligible if, with due regard to Professor Spearman’s views,3 we take ‘ g ‘ to be roughly equivalent to the ‘ general in- telligence ‘ which is universally presupposed in successful work. It is possible ?and, in practice, it often occurs?that a person with a high endowment of in- telligence, who lacks a high degree of the specific ability needed for a particular task, will succeed in the task equally with one who possesses only low intelligence but greater specific ability. (There are also, of course, tasks in which high specific ability is always essential and in which high intelligence is hot necessarily an aid. Musical ability is probably if this kind.) Usually, however, unless we are able to separate the components of the activity in such a manner as to enable us to say how much of the person’s success is due to intelligence (or to ‘ g ‘) and how much 1 Op. cit., p. 40. 1 Op. cit., chap. xiii.
3 Professor Spearman does not identify these two, but many people mean by ‘ intelligence ‘ what ‘ g ‘ may very likely prove to be. is due to other abilities, we shall not be in a position to predict his probable success in other tasks. A great deal of Professor Spearman’s research has been directed towards perfecting methods by which values for ‘ g ‘ and for ‘ s ‘ for each person and for each task may be determined. At the same time it must be borne in mind that this analysis applies only to the cognitive aspects of mental processes. Temperament and character, which are vocationally of great im- portance, are not here under consideration.
RELATION TO CURRENT PRACTICE.
It is noteworthy that the tests which according to Professor Spearman’s results correlate most highly with ‘ g ‘ (and which, therefore, may be expected to be most useful in evaluating the ‘ g ‘ of the individual) are just those which psychologists and others have usually regarded as the best tests of ‘ intelligence.’
Such tasks as completing sentences from which words have been omitted, saying whether pairs of words are alike or opposite in meaning, solving simple problems in reasoning, and so forth, invariably appear in the best standard tests of in- telligence. They are also the tests which correlate most highly with Professor Spearman’s ‘ g.’ This means that in such tasks the specific factors, which may arise fortuitously from schooling or from environment, as well as from innate ability for the task itself, are comparatively so small that their influence is negligible. If the influence is not negligible it must be eliminated by some means or other. The empirical method has been to lengthen the tests by the inclusion of large numbers of items (or separate tasks) of such a kind that for all practical purposes, in comparing individuals, the special abilities neutralise one another. Professor Thorndike 1 has recently shown that by a very careful selection of tasks the need for lengthening tests is considerably reduced. Professor Spear- man 2 proposes to save our time by giving us other methods of eliminating the influence of the specific factors. He suggests that we should calculate the value of ‘ g ‘ more directly, and he seems to imply that by suitable methods any task may be used to give the value of a person’s ‘ g.’
It is, however, in regard to the use of vocational tests that Professor Spear- man has been most critical. Yet there are no really vital differences between the procedure which his theory sanctions and that which is in current practice, at any rate in Great Britain and by the National Institute of Industrial Psychology. This is without doubt partly due to the influence which Professor Spearman has himself exerted upon the theory and practice of mental testing. Nevertheless, since vocational psychology has had to serve a number of practical needs, it has been developing along lines of its own.
VOCATIONAL SELECTION.
In vocational selection there has been a strong tendency to follow up the psychological analysis of each factory process by arranging that the test problems shall resemble the actual process as much as possible. Sometimes the complete task is divided into elements and the tests are made analogous to these; sometimes the test is a complex one introducing all the various components of the task. Generally, the validity of this procedure is established by the agreement between a person’s success in the test and his skill in the actual work.
1 The Measurement of Intelligence. New York : Teacher’s College, Columbia University, 1926. ? Op. cit.
Now in many factory operations high general intelligence does not seem to be necessary; in most cases specific abilities, to some extent encouraged by train- ing in the work itself, are important factors in success. Many of the tasks require skill in the manipulation of limbs, tools and materials; the mental aspects of these processes (in the sense of activity at the highest cortical levels) are not conspicuous. Consequently it seems probable that the ‘ g ‘ factor tends to be reduced in importance, while, conversely, the specific factors become highly significant.
The question then arises whether abilities depending mainly upon uncor- rected specific factors rather than upon the ‘ g ‘ factor can be successfully pre- dicted by any other process than trial at the work itself; for if other tasks be used for purposes of testing, they necessarily depend upon a different set of specific abilities and any common element due to ‘ g ‘ is small. Professor Spearman, however, allows that ‘overlap’ will generally be found when activities sufficiently resemble one another. In the case of tests for selecting factory operatives there is very frequently considerable resemblance between the tests and the actual tasks. Many of the tests which have been most successfully used in the Institute’s investigations have been devised with this object in view. It should, however, be remembered that a superficial resemblance in form or in action is not nec- essarily a sign of psychological resemblance, and sometimes tests which seem exactly like the actual task are not the most successful. Reasons may be found in the underlying psychological processes. The fact that high degrees of re- semblance, which are not due to the influence of ‘ intelligence ‘ or the universal ‘ g ‘ factor, are shown by coefficients of correlation to exist between tests and tasks seems to point to the existence of an ‘ overlap ‘ arising from the similarity of the pschological processes taking place. Moreover, in standardising the tests there has been obtained abundant support for the belief that many specific abilities arise from innate suitability for the task, and the use of the tests for the selection of new operatives has been to that extent justified.
VOCATIONAL GUIDANCE.
In vocational guidance the purpose of special tests is to confirm what is known of a child’s attainment and progress in school subjects and to disclose and measure abilities which may have had no opportunity for expression. The problem is a much wider one than in vocational selection, since many possible occupations have to be considered and their requirements analysed before a decision can be made. It is, therefore, important to know to what extent a com- paratively short psychological examination can cover all the most likely occupa- tions.
Hitherto, the practice has been to examine each child first of all for general intelligence, and then for special abilities such as are required in the occupations under consideration. In many cases it is not necessary to consider numerous alternatives. The economic position of the child and his parents, his previous education and training, his general temperament and character all tend to limit the range of occupations from which a choice is to be made. It has therefore been comparatively easy to arrange tasks which will throw light upon a child’s suitability for one or other of these occupations. Tests so constructed often closely resemble the duties of the occupation. Thus, for example, in tests of clerical ability we employ tasks such as filing, indexing, number-checking, quick calculating and the like. Similarly, for children of fourteen who seek employ- ment on leaving the elementary schools, tests which provide information con- cerning the children’s ability to deal with mechanisms and with relations of form, or to use their hands dexterously and rapidly, have been found useful.
MENTAL WELI-AUE 61
How far do these practical methods receive support from Professor Spearman’s general point of view ? At first sight, it would appear that the theory of one single universal factor combined with innumerable other uncorrelated specific factors would, if correct, prove a bar to successful vocational guidance. The theory seems to imply that there must be a separate test for every specific ability which it is necessary to consider. If so, the process of examination would be so long and fatiguing both to the child and to the examiner that as a practical method it would be useless.
But the fact that most mental activities are probably correlated through the operation of the ‘ g ‘ factor helps considerably. We can get a satisfactory meas- ure of a person’s 1 g ‘ either by means of suitable tests of 1 general intelligence,’ or by tests specially constructed for that purpose. We can also determine more or less successful^ the extent to which intelligence (or alternatively ‘ g ‘) is an important factor in success in each different occupation. We can then estimate the extent to which a person possessing a given quantity of intelligence is likely to succeed in the occupation in question. In so doing we have to consider how far abilities of a specialised character are also needed, and how they arise. In many occupations they may be, and often are, the result of the operation of 1 g ‘ in a specialised environment. In other words, they are largely developed by train- ing, but the necessary general ability must be possessed if this development is to occur. In such occupations therefore, other things (such as temperament) being equal, success may be predicted if the requisite amount of ‘ g ‘ is known to be present in the individual.
On the other hand, in some occupations specific abilities of innate origin appear to be much more important than general intelligence or ‘ g.’ For example, the ability to maintain a series of operations in proper sequence, attending now to one, now to another, is often required. According to Professor Spearman there is no general ability to distribute attention which may be applied universally to different kinds of tasks; on the contrary, in his opinion, which is in accordance with the researches of Dr E. Neil McQueen, the distribution of attention is in every case specific to the particular task. In such cases, then, it would seem impossible to predict the success of a person in a given task on the basis of his performance in another task.
But the experimental work in the distribution of attention referred to above was concerned with processes which were extremely unlike one another. In situations which are similar, as the writer hopes soon to show elsewhere, there does appear to be a connection sufficiently close to enable prediction to be confidently made. If this should be the case generally, then the uses of vocational tests can be widely extended. Already we have tests which appear to be satisfactory under these conditions. The so-called tests of mechanical ability which require a child to put together correctly a number of common objects provide a good example. Although the ability to succeed in a test of this kind does depend in part upon intelligence, it also depends upon a specific ability (whatever its origin) for the task. In occupations such as those of the motor mechanic, the fitter, and the carpenter, activities similar to those required in the test are found, and it is significant that higher correlations occur between proficiency in the work of the carpenter and of the mechanic and success in this test, than between proficiency and intelligence measured in other ways.
VOCATIONAL GUIDANCE IN THE FUTURE.
The use of psychological tests in vocational guidance is, of course, of com- paratively recent origin, and much of the work has necessarily been of an empirical nature. Consequently there is enormous scope for progress and advancement in a field of inquiry which is full of difficult and involved problems. No one can claim that he has reached a final solution of any of them. Professor Spearman’s investigations and explanations of his results have been most helpful. Empirical inquiries have in general supported his principal conclusions. In searching for measures of abilites of different kinds investigators have sometimes found the influence of ‘ intelligence ‘ or of ‘ g ‘ so dominant that, in practice, it seems possible to determine the value of one by means of another. Thus, for example, in an early experiment by the Industrial Fatigue Research Board and the National Institute of Industrial Psychology tests for ‘ creative imagination ‘ gave such high correlations with tests of intelligence that it was reasonable to assume that the same ability was the determining factor in both. But there are also many instances in which specific abilities are quite as important as any general ability.
There is much work yet to be done in the examination and measurement of special abilities, and in the analysis of occupations in order that the nature of the abilities required may be more fully known. It is quite possible that the over- lap between different tasks may be found to exist more frequently than one would expect from the data given by Professor Spearman, for his investigations have necessarily been confined to a somewhat limited field. If this should be the case, testing for vocational guidance will be much easier and the conclusions more definite. Having determined the amount of a person’s ‘ g ‘ and having brought together the occupations which it will be most profitable to consider, one will begin the measurement of the specific abilities needed in them by a number of specially selected tests. Such procedure differs little in principle from that of the present day, but it will obviously improve greatly in value as our technical knowledge and skill increase.
Disclaimer
The historical material in this project falls into one of three categories for clearances and permissions:
Material currently under copyright, made available with a Creative Commons license chosen by the publisher.
Material that is in the public domain
Material identified by the Welcome Trust as an Orphan Work, made available with a Creative Commons Attribution-NonCommercial 4.0 International License.
While we are in the process of adding metadata to the articles, please check the article at its original source for specific copyrights.