Air University Review, January-February 1967
Colonel James H. Ritter
The concept of personnel management as a more or less exact science, susceptible to a high degree of mathematical validation, is relatively new in American enterprise. As late as World War II the methods devised to select, train, promote, and retain employees, both in industry and in government, were largely intuitive, based on the personal experience and skill of the manager. Psychological testing had been established as a procedural aid, but the tests reflected mainly the insights of the psychologist into human characteristics, plus results obtained previously through trial and error by supervisors who had successfully directed great numbers of people.
The last ten to fifteen years have brought a radical change in managerial techniques. Increasingly precise measurements of aptitude, ability, and motivation have been introduced into the process of judging applicants and employees. These measurements have been developed by scientific research into human behavior. The tools that have made it possible to discover and apply these more rigorous management standards have been provided by advances in the techniques of statistical analysis, as a means of predicting performance, and by the use of computers to extract significant values from exhaustive files of recorded data.
The Air Force has been a leader in evolving and adopting these new methods.
Although the ultimate aim of the Air Force in the management of personnel
resources is to protect the
In the interest of its own effectiveness, within the budgetary limitations
imposed upon it, the Air Force has had to look for ways to improve performance
through selection, classification, and performance evaluation of all its
people. The scientific agency that conducts these studies is the Personnel
Research Laboratory, situated at Lackland Air Force Base,
The Personnel Research Laboratory has a close working relationship with the
Air Training Command (ATC) and its
Indeed, the original group of psychological test experts from which the Personnel Research Laboratory has grown was established by the Air Training Command on 15 November 1941, to develop and administer tests for selection of aircrews. Although having gone through several reorganizations and name changes through the years, the laboratory completed 25 years of service to the Air Force this past November. Its professional staff has been composed of psychologists, educators, sociologists, mathematicians, and specialists in related fields, many of whom had attained or later attained positions of eminence in academic or industrial circles. During the war most of them were commissioned officers, but now about 61 percent of the 219 authorized spaces are civilian, roughly 7 percent officer, and 32 percent airman.
Almost all the laboratory’s personnel concentrate on research and testing in one capacity or another. Of the 69 members of the professional staff, 24 have Ph.D. degrees and 35 more have master’s. Only 11 airmen are assigned in strictly administrative duties. The preponderance of scientific and technical staffing (95 percent) is made possible by unusually streamlined organization. Overall management is by the Commander, his Executive Officer, and the Technical Director, a civilian scientist. Both technical and administrative services are combined in the Operations Office, which handles comprehensive research planning, publications, and preservation of records, in addition to routine administrative functions. The five specific areas into which the laboratory’s work naturally falls are distributed among five research divisions:
Selection and Classification Division. In this the oldest area, going back to 1941, the laboratory develops and evaluates tests for the selection, classification, assignment, and performance of officers and airmen. On the basis of individual differences, the tests are designed to fit the subject into the position where he will be most effective.
Occupational Structures Research Division. Here the emphasis is on the job to be performed. The laboratory describes and evaluates Air Force occupations; structures them into career fields, specialties, and positions; analyzes the work accomplished in terms of the grade and pay which it merits; and from these factors determines the qualifications required to perform it properly.
Specialty Knowledge Test Division. This is the only division of the laboratory that administers an operational testing program for the Air Force. It develops and revises specialty knowledge tests to determine the fitness of airmen for advancement in grade; distributes the tests to supervisors in the field; and scores the results in accordance with grade quotas established from time to time by Headquarters USAF.
Adaptability and Quality Evaluation Division. The interest here is in the broad area of personal adjustment to Air Force life and responsibilities. The division identifies, measures and evaluates such individual factors as the subject’ s background, motivation, morale, aptitude, and response to specific situations, in a search for patterns that will predict the character of his performance and the likelihood of his promotion and retention. This research also supports the Human Reliability Program of the Air Force.
Statistical Methodology and Analysis Division. With a staff of 92 people, this is the largest division of the laboratory. Its studies are concerned with advanced mathematical and statistical methods in personnel management, with particular emphasis on computer techniques. These analytical procedures are applied extensively also in support of research carried on by the other divisions.
Scientists and technicians in Statistical Methodology are mainly responsible for operation of the largest and most valuable piece of equipment maintained by the Personnel Research Laboratory—the IBM 7040 electronic computer, with its files of research data now stored on some 5400 reels of magnetic tape. This computer already is outmoded for the work of the laboratory, so rapid is the increase in computer technology. Within the next year or so it will need to be replaced by a larger system with quadruple its capacity. In addition to the coded files now in use, approximately 15,000 square feet of records wait to be reduced to tape, at a cost of about $150,000. The 7040 computer, valuable though it is as a means of obtaining significant results from large collections of complex data, is limited to sequential operations; it can be programmed for only one research project at a time. When the laboratory wishes to obtain answers for one of a variety of continuing programs, drawing on a body of relevant data that have been previously assembled and stored, it turns to another, more sophisticated computer, located 1220 air miles away.
The massive Q-32 computer, owned by the Air Force and operated under
contract by the Systems Development Corporation in
Scientists at the laboratory in
In most computer operations the prime factor affecting the rapidity with which
answers are received is the efficiency of input and readout devices. Computer
people define this problem by saying that the electronic brain is “tape-bound.”
The 7040 computer at Lackland is less tape-bound than the line to
The laboratory’s unique file of officer and airman records, reaching back in many instances to World War II, constitutes its main source of raw material. They include, for example, test-score information on all airmen entering the Air Force since 1947; airman reenlistment and separation actions for the past decade; annual officer effectiveness reports, beginning in 1954; and similar follow-up material on airmen for the last several years.
If the laboratory wants comparative information on the educational levels of officers or airmen, the areas of the nation in which they grew up, their social or economic status, their skills, aptitudes, achievements in civilian life, or attitudes toward their work, it has only to consult these records, to choose a representative sample showing the characteristics whose effects are to be compared, and then to process the resulting mass of data through the computer, reducing it to statistical tables. The inferences drawn from these tables will help to predict the behavior of future officers or airmen having the same characteristics.
This presentation is of course a gross oversimplification of the studies compiled by the laboratory. It neglects, for instance, the rigorous and often abstruse mathematical procedures that must be deduced and then followed in assigning the proper statistical weights to the varied influences affecting a typical subject’s performance. Nevertheless, this is in general the kind of determination which the scientists at the laboratory are making.
The aim of Air Force personnel management is, of course, to obtain the most capable people it can find, to train them in the jobs for which they are best fitted, and then to reward them adequately so that they will remain in the service to the end of their careers. This effort is complicated by many individual and social responses, not the least of which is the public attitude toward the military environment.
Throughout most of human history, from ancient times until perhaps two centuries ago, the military profession was a choice occupation for ambitious people. Military men usually were the heads of state and the leaders of society. They relied upon learned advisers, but policy decisions and actions were normally the responsibility of military persons trained to exercise command. There was then no problem in finding competent military personnel.
Since the industrial revolution, however, the occupations that enjoy the greatest prestige have come to be those in the upper levels of business management and in certain related professions, notably the law. Political leaders are drawn most often from these groups, while military service is relegated more or less to the status of a technical specialty that becomes briefly esteemed in periods of emergency.
As the armed forces have been made responsible for the development and use of vast scientific, technological, human, and material resources, their need for skilled managerial and operational talent has become increasingly acute. The personnel problem of the Air Force, then, is to attract and hold people with these qualifications. Upon the Personnel Research Laboratory devolves the task of discovering how this feat is to be accomplished. It is a basic objective in all the laboratory’s programs, most directly in selection and classification, which evaluate the incoming officer or airman and, later on, his fitness for promotion.
A recent study in this area aimed at the development of a management control system for Air Force Reserve Officer Training Corps detachments at civilian colleges and universities. The Department of Defense was concerned about the comparative costs of training AFROTC graduates at 188 institutions around the nation in relationship to their effectiveness and length of service. The cost varied from as little as $15,000 for a career-officer graduate of a small regional college to as much as $153,000 for a career-officer graduate of a large university of great national repute. The training cost by itself was not a reliable predictor of the long-term value of the graduate to the armed services.
The Personnel Research Laboratory turned to its data files as a means of evaluating the products of each of these institutions. Together with the academic standing of the college or university and the intellectual attainments of the graduates, the study considered such factors as the geographic, cultural, and economic backgrounds of the students. Along the way some interesting facts were turned up. One was that the retainability of the graduate depended in no small degree upon his social and economic origins. If he came from a less prominent or affluent background and attended a regional college—perhaps on a scholarship—the outlook for his retention was more likely to be favorable. The reason inferred from this finding was that the Air Force then represented an opportunity for advancement, and the student might be expected to work harder in order to achieve it. Graduates of combined military and civilian schools also have been found to have a high retention rate. In such an institution, where I military commander and instructors usually exemplify the ideals of soldierly prestige and competence, the student gains a clearer view of the potential rewards from a military career than in a standard AFROTC unit at a predominantly civilian university.
Similar questions arise in research amid at increasing the retainability of officers in professional specialties such as medicine, law, science, and engineering. The retention rate for medical officers is particularly low. Most doctors remain in the armed forces only for the years to which they are committed and then go into civilian practice. Even though medical officers receive extra compensation in rank and pay (they enter the Air Force as captains, and flight surgeons draw flying pay as well), the military obviously cannot offer them either the financial return or the professional esteem which they can earn for themselves in private practice. Somewhat the same situation applies to many legal officers, scientists, and engineers.
Offsetting these disadvantages in some degree are the facilities provided by the military services for research and practice in exotic areas of medicine and science—for example, aerospace flight. Again, the armed forces afford opportunities for travel and experience in out of-the-way places that few civilian specialists are likely to visit, such as the Asian countries. These are possibilities that appeal to certain temperaments, though by no means to all.
The Personnel Research Laboratory has been developing criteria by which to isolate and measure these factors, so that the Air Force may be able to attract and retain more professional people. So far, these studies suggest that the military may give too much thought to retention policies, as such, and not enough to finding persons with backgrounds and tastes that fit them for life in the armed forces and thus make them more readily retainable.
A continuing program of the laboratory in this area is known as “Project M.”
Based upon data showing the officer’s aptitude, his education, the source of
his commission, and his effectiveness reports for the past decade, Project M is
a historical data bank identifying the many variable factors that determine his
performance and the probability of his retention. Used for current evaluation
as well as to predict success in the future, the file is revised periodically
to bring the assessment up to date. Among other uses, Project M lends itself to
cost-effectiveness analyses of officer procurement programs such as the AFROTC
detachments and the
An example of this kind is the study of officers entering the Air Force with college majors in science and engineering. From the Project M data file the laboratory can determine how many of these officers actually are employed by the Air Force in science and engineering assignments and how those in this group who are separated from the service after fulfilling their commitment differ from those who remain on active duty.
Closely related to the task of weighing personnel performance is the process of appraising the jobs that the people perform. In the Air Force, as in other organizations, the measure of the man is his ability to handle the work assigned to him. It follows that the work must be assayed with equal care.
Defining the nature of the 600,000 different jobs in the Air Force raises a number of peculiar problems. For one, the jobs are continually changing, sometimes in response to the special needs of a commander, sometimes as a result of technological innovations. A man trained for a specific task may be diverted to another one in a different career field. Or he may find, upon reporting to his unit, that the work for which he was trained is now handled by a computer.
Traditionally, a job was described by having the person who performed the work write down what he did. Close analysis often showed, however, that jobs with different titles were essentially the same. Conversely, jobs with the same specialty code number might vary widely from one command or agency to another. When a man was transferred to the same position in a different organization, he might discover that the duties were outside the scope of his experience.
The Personnel Research Laboratory attacks these problems by developing an inventory of tasks performed in each officer utilization field and on each airman career ladder. Test control officers call in jobholders from the field and ask them to enumerate the amount of time they spend on tasks which they consider normal in their work and also to describe significant tasks which they perform but which are not listed in the description. This information is recorded on magnetic tape for computer processing. The computer produces a consolidated job description of work done by people in particular commands, or at specific locations, or on given levels of experience or skill. Also the computer identifies persons who are performing essentially the same tasks and publishes their composite job descriptions.
Eventually the laboratory hopes to collect information showing the ability of each person to perform—or learn to perform—other tasks to which he might be assigned. Along with the task requirements for each job, this information can be used to establish an automated personnel reassignment system for the Air Force.
Upon request from the Director of Manpower and Organization, Headquarters USAF, in 1964, the laboratory undertook a large-scale effort to formulate a scientific basis for the assignment of grades to officer positions. Job descriptions were obtained for some 80,000 positions in all the grades from lieutenant through colonel. From this file was drawn a criterion sample of 3575 descriptions.
A special Hq USAF policy board, composed of 22 colonels with a wide range of experience, was appointed to determine the appropriate grade for each job in the sample. The rating policy of the board as a whole was translated by the computer into a mathematical equation, giving numerical weights to all the factors that were found to be relevant in determining the grades of the jobs. The equation was then used to assign suitable grades to 10,000 more officer jobs. These results were projected to show the distribution of grades among different officer specialties and specialty groups. The information assembled by the laboratory in this study was a contributing source of data to support the 1966 legislative program delineating officer grade requirements for the Air Force.
Administering the specialty knowledge tests used as a basis for the promotion of airmen is an end product of research conducted by the Personnel Laboratory since World War II, rather than a current experimental program. To the extent that it reflects continual refinements in testing techniques, the experimental work is done by research scientists in other areas, usually selection and classification, and then incorporated into the operational tests as they are periodically revised. The airman promotion system covers 691 specialties in 46 career fields. Some are too small to require written tests. Nevertheless, the laboratory maintains an inventory of about 450 separate tests for successive levels of skill from apprentice up through the supervisory grades to chief master sergeant. Revised everyone to two years, they serve as the basis on which advancement is programmed for approximately 85 percent of the airman population.
To construct a series of revised tests for a specific career ladder, the laboratory calls in a team of three to fifteen senior noncommissioned officers from the field, on temporary duty for six weeks while working with the psychologist in charge of the project. After the tests have been reviewed and approved, they are published by the Government Printing Office and distributed to test control officers throughout the Air Force. The laboratory also scores the tests, averaging roughly 200,000 a year, most of them by machine scoring. Passing rates are established by Hq USAF. Promotion quotas for airmen are limited, as they are for officers, by the number of vacancies in higher grades of specialties and by the effort to maintain an orderly progression through the ranks. However, the scoring system is designed to see that the best-qualified candidates for advancement will be available to meet these quotas.
Retention of highly skilled airmen is, of course, a problem comparable to the one of holding officers with exceptional talents or administrative ability. In some respects the airman problem is more complex because of the larger numbers involved, the greater diversity of jobs, and the fact that the personal and vocational characteristics of recruits often are not so clearly indicated by their backgrounds or education.
Quality controls over the airman population are the particular concern of research in the area of adaptability and quality evaluation. The present program began in 1958, when the Personnel Research Laboratory was asked to make a study of the airmen discharged during that year as chronic offenders, or because their progress or behavior was unsatisfactory for other reasons, and to find out whether the personal factors responsible for their failure could be isolated for screening purposes. At that time the attrition rate for unsuitability was about 18,000 out of 100,000 recruits entering the Air Force each year. The accepted method of dealing with the problem was by counseling, punishment, reassignment to less challenging duties, and, as a last resort, discharge. The individual problem was rarely recognized until after the airman had received his basic training. So the cost to the service in dealing with these nonassimilable people was obviously high.
The laboratory turned to its files of airmen entering the service since 1956 and also to its data on re-enlistments and losses for the same years. Analysis showed that the dropout rate was highest among airmen who had not completed high school, especially in the 17-year age group with relatively low aptitude scores.
One result of this study was that the Air Force Recruiting Service changed its entry standards to give bonus points for recruitment of high-school graduates. From about 55 percent in the late 1950’s, the proportion of high school graduates among Air Force recruits has risen to more than 90 percent. Over the same decade the number of bad-conduct and unsuitability discharges has been reduced by about 5000 per year.
A corollary of this study was a project started in 1959 to find reliable methods of screening recruits for high risk assignments in nuclear weaponry, security services, and intelligence. Some 10,000 new airmen every year were being routed into these sensitive positions, where the safety of the nation could depend upon their responsibility and discretion. The information obtained on their previous emotional or disciplinary problems was often sparse.
Again the laboratory recommended that none but high-school graduates be considered for these positions, not only because they were more likely to be successful but also because more information about them was usually available. An improved screening process was suggested: that if any psychiatric or behavior problems turned up in the airman’s record, either before or after his entry into the Air Force, he would be assigned automatically to other duties of a nonsensitive kind. The alternative jobs were not necessarily less attractive to the airman; often he would find them more agreeable, if less exacting, than the high-risk positions.
These recommendations were adopted. But the report of the laboratory went beyond the obvious improvements in screening. It called for a background investigation of the airman’s disciplinary and adjustment problems before he had entered the Air Force, starting with an airmail letter of inquiry to his high-school principal. (It was found, incidentally, that such letters almost invariably were answered fully, frankly, and promptly.) The mere fact that the investigation was begun would often lead the airman to reveal incidents in his background which he had not mentioned in the routine screening.
This procedure, known to the laboratory as the Human Reliability Program, also has been adopted by the Air Force. The Air Training Command activated the Assessment Branch of the Personnel Processing Squadron at Lackland in January 1966. In effect a specialized investigative service, it saves the Air Force an enormous amount of time, money, and administrative work by identifying potential security risks before they enter technical training or are assigned to sensitive jobs. The airman’s record is in no way harmed by the investigation. If he is diverted to other duties of a less responsible type, he is saved from the possibility of a later security breach that could lead to disciplinary action or a dishonorable discharge. The investigations also provide the laboratory with data on the background problems of recruits, from which more refined selection procedures are being developed. With these procedures it is possible to predict the influence of the airman’s home environment on his future career in the Air Force. A similar program now is being extended to the other military services.
From all these studies in different areas of personnel selection, adaptation, and performance, the Air Force hopes eventually to obtain a computerized model of the entire personnel system, including all ranks, all jobs, and all identifiable factors that could affect the efficiency of an individual unit or the service as a whole. With such a model it will be possible to foresee not merely the probable result of a specific policy or situation in a single area but also the long-term interaction of different policies or situations in many areas. Future personnel problems then may be identified and solved in advance.
The development of this comprehensive model is the task of specialists in statistical methodology and analysis at the laboratory. Theirs is the responsibility for the advanced mathematical techniques by which nearly all the laboratory’s present-day studies are accomplished. Projects in other areas of personnel management often are given to these scientists, particularly when they involve the manipulation of large quantities of complex data to reach an early decision or policy.
One such project in the last two years was the program known as Top-Flow. The problem was how to provide for a great number of earned promotions of airmen from the lower grades, in the face of an apparent stagnation at the upper levels. A change in current promotion policies would involve a vast amount of administrative effort. The laboratory ran a broad sample of the airman personnel structure through the computer, testing hypothetical effects of different promotion plans. It was found that increasing numbers of master sergeants could be expected to retire within the next several years, relieving the pressure at the top. Thus, no action would be necessary; the problem would resolve itself. That this was the correct answer is shown by the latest quota list for airman promotions. From 136,068 in 1965, the quota had risen to 264,246—an increase of more than 94 percent.
The Air Staff handed the laboratory an even more urgent question for statistical analysis in February 1965, on the eve of the annual budget hearings in Congress. What would be the long-term effects of a general increase in pay on recruiting and retention of officers and airmen? Here the key factors were the quality levels of the jobs to be filled and of the people required to fill them, not only at the present time but in the foreseeable future as well. Other factors were the trends in re-enlistment rates, the value of Air Force training and experience in civilian employment, and the cost of the pay increase compared with the cost of recruiting and training new personnel. In only nine days, by an all-out effort at the laboratory, the Air Staff had its data. The pay increase was recommended to Congress, supported with firm figures projecting its economies into the future. The bill was enacted into law and became effective on 1 September 1965.
It takes only a casual acquaintance with the studies carried on by the Personnel Research Laboratory to answer the criticism most often directed at computer techniques: that they take the human element out of personnel management. If by “human element” is meant the recognition of individual identity, the critic is, of course, right. The computer does not examine the unique personality of a single human being in depth, as a novelist or a biographer does. But in a vast collection of people such as the Air Force or any other large organization—or indeed society as a whole—the personality of a single human being rarely is known fully except to the immediate circle of his associates. To achieve this kind of recognition there is no substitute for direct contact between one person and another.
The peculiar virtue of the computer is that it has the capacity to single out, remember, and compare the individual traits of many different people. Before the computer, when a large organization attempted a statistical study of its employees, the tendency was to establish an ideal against which they were to be measured and then consider only their conformance to or departure from this norm, which was assumed to be the standard measure of success. The bias was in the direction of conformity, toward the submergence of individual character in the type. With the computer, on the contrary, it is possible to establish the existence of many divergent types, possessing more or less of certain traits that combine to fit the people into different functions at varying levels of responsibility within the organization as a whole. The computer is a much more flexible instrument with which to view the heterogeneous nature of society, whose every member is in some degree exceptional.
Considered in this light, the computer is not the enemy but the ally of human individuality. It may find a useful place for a person who once would have been considered an eccentric and looked upon askance in consequence.
As one of the few institutions making large-scale studies in this field, the Personnel Research Laboratory has been a pioneer in the art of evaluating the whole man for his employer—the U.S. Air Force.
Personnel Research Laboratory, AMD, AFSC
Colonel James H. Ritter (Ph.D.,
The conclusions and opinions expressed in this
document are those of the author cultivated in the freedom of expression,
academic environment of
Home Page | Feedback? Email the Editor