The Psychological Assessment of Candidates for Aircrew in the R.N.Z.A.F.
As you are aware, psychological methods were called on and widespread use made of them in the selection and placement of personnel in the Armed Forces during the recent war years. By the end of the war the psychological assessment of occupational suitability had virtually become an integral part of the induction of the recruit into service life.
The purpose of this paper is to outline to you some of the results of one of the many projects in this field: the psychological assessment used as part of the process of selection of aircrew in the R.N.Z.A.F. As a prelude, I shall describe briefly the composition and aims of the unit that undertook this work.
Composition and Aims of the Medical Investigation Unit.
In the early stages of the war psychological research played no part in R.N.Z.A.F. aircrew selection. Later, the desirability for it became apparent and early in 1942 a unit known as the Medical Investigation Unit was formed. The purpose of this Unit was twofold: first, the psychological suitability of aircrew volunteers was to be assessed and reported on to the Aircrew Selection Committee and, secondly, the Unit was to become a centre for the instruction of aircrew candidates (and to a certain extent qualified aircrew) in the physiological aspects of aviation. This physiological work comprised lectures in aviation medicine; practical demonstrations in a decompression chamber to show the effects on the human body of sudden changes in atmospheric pressure and of oxygen lack encountered at high altitudes; and a dark adaptation apparatus (Livingston's rotating hexagon) was used to demonstrate the conditions of night vision.
The Unit consisted of a medical staff and nine trained psychological workers who carried out testing, statistical work, and maintained full records of the subsequent training and disposal of all aircrew who had passed through the Unit.
Immediately on entry candidates were given an aircrew medical examination. Then followed a detailed interview at which biographical information was collected and the man's general background, qualities of temperament and personality structure were the subject of investigation. At the same time the candidate was asked to state the order of his preference for the five aircrew categories: pilot, navigator, air bomber, wireless operator/air gunner, and air gunner.
The next step was the administration of three group pencil-and-paper tests. These were the Classification Test, which assesses general intelligence; the Arithmetical Reasoning Test (elementary mathematics is an important subject in aircrew training); and the Space Perception Test. This last-named test was validated by the British National Institute of Industrial Psychology as a measure of practical and mechanical ability.
Following the group tests came a performance test of sensori-motor coordination. All applicants were given this test, but special attention was paid to the results of those who had given pilot as their first choice of aircrew trade.
The results of all psychological tests were summarized on a rating scale which indicated the chances of the candidate's success in his chosen trade.
The total information on the man thus collected went to the Senior Medical Officer in charge of the Unit, who made a final assessment. In cases where this final rating was below average care was taken to check all relevant data on which it was based. These men would be recalled for a further interview, a re-test, or special supplementary testing, and the original grading confirmed or modified accordingly.
The completed Medical Investigation Unit assessment then went forward to the Aircrew Selection Committee, which interviewed all candidates at the end of the Initial Training Wing course. Here the Unit's findings were used, along with other relevant material supplied by the staff of the Initial Training Wing, in the final selection and placement of candidates.
Over the three-year period in which the Medical Investigation Unit was operating numerous tests were tried and rejected because they failed to measure up to the practical facts of aircrew training. It is intended, therefore, to describe only those which were fully validated. No technique was used in grading assessment until it had been tried out, the results compared with success and failure in training, and the extent to which it was of predictive value carefully analysed. This “follow-up” work was inevitably a most important preoccupation of the Unit, as only in this way could the techniques themselves be tested and checked against the yard-stick of practical fact.
The aims of the Medical Investigation Unit in regard to psychological assessment may be summarized as follows:—
On the basis of the aptitudes and personality pattern of the candidate, to recommend the category of aircrew in which his chances of success were judged to be highest.
On the same basis to assess the candidates so that those who were recommended for training were the ones most likely to succeed.
To discover those men likely to fail in the educational work, or “ground subjects.”
In this paper attention is directed mainly to the assessment of pilot candidates. In doing so I do not want to leave the impression that the other aircrew trades were relatively any less important, but within the confines of a brief exposition it is impossible to deal adequately with the whole field of investigation.
Group Psychological Tests.
The Classification Test.—This test conforms to the generally accepted pattern of intelligence tests. There are eighty questions of graded difficulty of the following types: verbal correlates, vocabulary questions, problems in logical reasoning, number series, mathematical problems and spatial correlates. The mean score on this test with a sampling of 3,222 cases was 56.4 with a standard deviation of 11.6. These figures indicate a satisfactory scatter of scores on which to base test ratings. This test, together with the Arithmetical Reasoning Test, was of greatest use in the assessment of educational success and failure. The following figures indicate the extent of the Classification Test's utility in this direction:—
|Classification Test Score.||Percentage of Candidates scoring more than. 65% in the Pre-entry Examination.|
|53 and above||94.3%|
|Less than 35||16.1% 294 cases|
Again, a survey was made of the disposal of 200 men who scored less than 35 in the Classification Test. The figures show:
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Number trained as pilots||15|
|Number trained as air gunners||27|
|Number trained as air bombers||1|
|Number being held over for a further trial on the Initial Training Wing educational course with disposal still pending when this survey was taken||5|
|Number who did not reach the required educational standard||152|
(It is to be noted that these cases were accumulated over a period of time and are not indicative of the standard of any one Course.)
In this table of low scorers on the Classification Test, of the 15 who went under training as pilots, 5 were subsequently terminated in their training. Thus out of a total of 200 candidates only 10 (5%) were successful as pilots. The comparatively high percentage of air gunners in this group is accounted for by the fact that the training of an air gunner, while just as exacting in its own skills, involves less theoretical educational work than the other aircrew trades.
The correlation between the Classification Test and the Pre-entry Examination results was .71.
The Arithmetical Reasoning Test. This test is made up of 30 questions to be done in 35 minutes. For each question five answers are suggested, and the candidate is required to mark the correct one. The solutions require no special knowledge of formulae or methods. As one might expect, the intercorrelation between this test and the Classification Test was high (.734).
In considering the combined results of these two tests the following figures are significant. Of a total of 110 cases with below average scores on both the Classification and the Arithmetical Reasoning Tests:
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Number trained as pilots||7|
|Number trained as air gunners||13|
|Number held over for further educa-||1|
|Number who did not reach the required educational standard||89|
Within this low-scoring group the test results of the 7 who proceeded with pilot training were all within the higher limits of the group.
The Space Perception Test. This pencil-and-paper test involves rapid and accurate observation of the differences between closely resembling designs and geometrical patterns. In contrast with the Classification Test this test is nonverbal and does not involve abstract reasoning. A high score on the Space Perception Test is indicative of good practical and mechanical ability.
A comparison made between Space Perception Test scores and success in pilot training yielded the following results:
|Space Perception Test Rating.||Percentage successful in pilot training.|
These ratings were taken over 1,383 cases.
Sensorimotor Tests of Pilot Aptitude.
I shall next consider the practical test of sensorimotor co-ordination, the S.M.A.3. (This short reference is, by the way, an abbreviation for Sensorimotor Apparatus, Mark III.) It is designed as a test of pilot aptitude, calling for fine and co-ordinated movements of the arms and legs.
The following comments will, I hope, give you some idea of the apparatus. On a screen facing the candidate a small light moves in an irregular course. By means of hand and feet controls somewhat similar to those used in an aircraft the candidate tries to counteract the movement of the light and keep it within the bounds of a small square marked on the centre of the screen. At the same time two distraction lights below the screen flash intermittently and must be turned off immediately by the appropriate backward or forward movement of a small lever. A single test “run” on this apparatus takes 90 seconds.
Every action on the controls is registered on a series of dials. Separate dials record the time the light is outside the central square, and the reaction time for switching off the distraction lights.
The candidate was given three tests. Between the second and third test the tester discussed the candidate's performance with him and gave advice on how to correct any difficulties he had encountered.
The following data were used in arriving at a performance assessment:
The score registered on the dials of the machine for the three tests.
The reaction time for the distraction lights.
A co-ordination rating; this was classified on a five-point scale on the machine score recorded in the second test.
Emotional attitude of the candidate to the test.
Method of approach and handling of the controls.
Distribution of attention.
Control of the distraction lights. (4–7 inclusive were also assessed on a five-point scale.)
The Centration Rating; this was classified on a seven-point scale and depended on the score, in the second test, on the dial which recorded the time the light was left outside the central square.
The total assessment rating arrived at by combining the separate ratings already enumerated.
Any relevant general remarks on the performance of the test.
Where the tester contributed his own ratings and opinions, great care was taken to define precisely the nature of the phenomena that he was to observe and comment on. The terms used in his report were standardized with equal care.
The following figures indicate the relationship between the S.M.A.3 assessment and the percentage failure in flying training. Of those assessed as Excellent only 12% failed; Above Average, 18%; Average, 28%; Below Average, 38%; and Poor, 47%. The total on which these figures were taken was 1,013.
In addition, the S.M.A.3 result was indicative of the standard of flying ability attained by successful trainees at the end of the Elementary Flying Training School course. The relevant correlation here was .32 (tetrachoric correlation).
This was carried out by specially selected Medical Officers attached to the Unit, and was designed to assess the personality pattern and the temperamental suitability of the candidate for aircrew duties. Inquiry was also made into the motivation of the candidate in volunteering for aircrew and the reason for preferring the particular aircrew trade which he had chosen. At the same time a fairly comprehensive history of the man was obtained: previous occupations, educational history, general background, hobby interests, etc. In this way a picture was built up of the man against the background of his own environment.
A full report was written of this interview and, after stating the facts upon which he based his opinion, the Medical Officer concluded with an assessment of the candidate's suitability for the trade he preferred and also, possibly, an assessment for other trades in aircrew for which he appeared a likely candidate.
The interview, too, was subjected to analysis against the results in flying training. Over a group of 1,546 cases which was analysed in sections according to the courses to which the men belonged, positive correlations were obtained which varied between .2 and .434.
There are two points to bear in mind in considering these correlations. The data do not include any eliminations subsequent to the period of early flying training and it is then, in operational training, etc., that the demands on personality adjustment are the more exacting. Secondly, it is relevant to add that although a few of the more patently unsuitable were eliminated by the Aircrew Selection Board, these cases have not been included in this survey.
Age. The effect of age may be deduced from the following figures: In an 18-year-old group 227 candidates were successful and only 31 unsuccessful; in the 25th year the proportion rose to 87 successful and 42 unsuccessful. This trend is emphasised in the oldest group, aged 32, where 8 were successful and 16 unsuccessful in flying training.
There are two facts which stand out in the relationship between age and the other data investigated. First, the older candidates achieved equal success on the S.M.A.3 and group tests. This is important, since as age and S.M.A.3, for example, are unrelated, they form two distinct factors which taken together give a more useful criterion of assessment. (A multiple correlation combining both gave .41). Secondly, with older men civilian occupation is of much higher significance in pilot training than is the case with the younger age groups. This, one would judge, is due to a more highly developed occupational set and a consequent loss of adaptability in acquiring new and complex skills. I shall say a little more on this topic in dealing with the influence of civilian occupation on pilot training success to which I shall now pass.
Occupation. In considering the occupational background of pilot trainees it is necessary first to outline the classification of occupations which was used. There were four main groupings with sub-divisions as follows:
Group I—Professional, Clerical, and General Administrative.
Category 1. Professional and administrative posts, judged on the possession of relevant qualifications, degrees, diplomas, etc.
Category 2. Clerical workers, where the duties are, in the main, of repetitive nature.
Category 3. Salesmen, shop assistants, etc.
Group II—Skilled, Semi-skilled and Unskilled Trades.
Category 1. Skilled tradesmen.
Category 2. Semi-skilled.
Category 3. Unskilled.
Category 1. Farm managers, owners, etc.
Category 2. Farm cadets.
Category 3. Farm labourers.
Included in this group were youths straight from school and recent school-leavers who had been employed on a casual basis while awaiting entry into the Services.
A set of figures showing the relative success of these occupational groups in pilot training can be briefly summarized. The highest percentage of success fell to the school-leavers, closely followed by the farm cadets: these two groups show 79 successful trainees out of 86 candidates, or 91.8%. The skilled trades followed with 86.5%. Next came the semi-skilled and unskilled with 74.2%. Least success was had by the salesmen, shop assistants, etc., with 64.1%. This survey was made on the somewhat small sampling of 601 cases.
It is significant, and confirms the remarks made previously in discussing age, that those occupational groups noted here as having the greatest chance of success (the school-leavers and the farm cadets) fall within the lower age limits. Thus occupation was relatively less important with the younger and became more important with the older age groups. In those occupations where there was a wide scatter of ages the skilled tradesmen were more successful than the semi-or unskilled tradesmen and the professional, clerical, and salesman groups.
Education. The educational background of successful pilots shows an interesting trend. The following gradings were used: primary schooling only; 1 year, 2 years, and 3 years secondary schooling; more than 3 years without Matriculation; Matriculation; some University work; University graduate.
Greatest success was attained by the Matriculation and the “more than 3 years without Matriculation” groups. Expressed as a curve, there was a gradual rise to this peak and a falling off with the University students. This survey was made on 1,827 cases. The average schooling was 3.2 years post-primary education.
This brief sketch has attempted to describe some of the facts found to be of importance in the assessment of aircrew at the training stage. It is obvious that no one criterion could be used in selection to the exclusion of all other considerations. Each rating and report must be carefully integrated into the total pattern. The question then presents itself, “What are the ‘typical’ or ‘normal’ patterns for the aircrew trades based on the tests and indicators here described?”
In conclusion, I shall try, therefore, to outline to you the general pattern into which these indicators might be expected to fall for the various categories of aircrew.
On the Classification Test and Arithmetical Reasoning Test navigators would be above average, pilots average, and the other trades average to average minus.
With the Space Perception Test air bombers, wireless operator/air gunners, and air gunners would be above average, and pilots and navigators average.
The S.M.A.3 Test. Pilots above average; air bombers, wireless operator/air gunners, and air gunners average, as there is much fine manual adjustment and skill called for in these trades; for navigators the test is not particularly-relevant.
With age groupings the typical pilot would be in the lower groups, navigators in the higher groups, and the other trades would show a wide scatter with a tendency in all to favour the lower groups.
In connection with previous civilian occupation, the skilled trades show up-well in all categories of aircrew. The success of the professional and clerical group appears to be greatest in the navigator category.
The highest educational standard is to be noted in the navigators; pilots require slightly lower. With the other trades it is less essential, though it must be at a reasonable minimum standard.
In the interview assessment, keenness to undertake the very arduous training which is involved in all aircrew trades is quite essential and an adequately balanced and somewhat extroverted personality is desirable.
Finally, there is no need for me to stress that successful aircrew will not be found to adhere rigidly to these patterns. One factor will obviously modify and compensate for another. Thus it is only by considering them as manifestations of human beings against the background of the total personality that they can be most adequately and justly assessed.