Spotlight on History
Profiles in military psychology: Raymond E. Christal
By Patrick C. Kyllonen
Editor's note: This is the first in what I hope will be a long series of profiles in military psychology that will appear in this column. Many thanks to Pat Kyllonen for being the first to do a profile and for doing such a great job of writing this important inaugural.
Raymond E. Christal was a giant in modern military psychology. He made important and lasting contributions to occupational measurement, personality testing (“the true father of the Big Five”), abilities testing, policy capturing, and judgment analysis, as well as cross-national collaborations in testing and measurement. His occupational measurement research was the basis for the Comprehensive Occupational Data Analysis (CODAP) system, which is still used by the U.S. Air Force, other U.S. and international militaries, as well as the civilian sector.
He was the father of the now widely used “Big Five” personality model, based on a study he conducted with Ernest Tupes in the late 1950s on Air Force officer peer reports. He wrote an influential and controversial paper on the use of policy capturing methodology long before multiple regression modeling was the prediction staple it is today. He developed a number of cognitive abilities testing programs in the U.S. Air Force, still in operation today, and he initiated a major research program, the Learning Abilities Measurement Program (LAMP). Novel at the time, LAMP used an information processing approach to cognitive abilities testing. Research from this program contributed to similar programs throughout the U.S. Department of Defense (e.g., the Army's Project A and the Navy's Enhanced Computer Administered Testing [ECAT]), and its influence is still felt in today's Armed Services Vocational Aptitude Battery (ASVAB) and its research agenda.
He was the U.S. Air Force representative to the Human Resources and Performance group of The Technology Cooperation Program (TTCP). This work led to collaborations with other militaries, was honored with a recognition for his Trait-Self-Description (TSD) battery, was used by several military services (including Canada and Australia), and influenced the development of the U.K.'s British Army Recruit Battery, still in operational use today (Irvine, 2013).
Born in 1924, Christal served 4 years in the U.S. Navy's Pacific Fleet during World War II. He began his career as a military research psychologist in the San Antonio office of the Human Resources Research Organization in 1948 after receiving a bachelor's degree in history and mathematics and a master's degree in psychology from North Texas State University. He then became a civilian psychologist in government service, and over the next 30 years, he rose through the ranks at a series of predecessors to the Air Force Research Lab—Air Force Personnel and Training Research Center, Personnel Research Laboratory, the Human Resources Research Laboratory, and the Armstrong Laboratory.
Christal completed his PhD degree in Educational Psychology in 1958 at the University of Texas, Austin. For his dissertation, he conducted a comprehensive study on the factorial structure of human visual and spatial memory, which was published as a Psychological Monograph in 1958. After a productive career spent mostly in the area of abilities testing and occupational analysis, he retired as a GS-15, Step 10, in 1980 for less than a month. There had not even been time for his office to be reassigned when he returned as a reemployed annuitant to pursue his passion for human abilities measurement. Colleagues at the time, such as Bill Phalen, joked that they wanted a refund on the retirement gift they had given him just a few weeks earlier. Nevertheless, he was welcomed back, and over the next 5 years, he established the LAMP research program (which is described in more detail below). In 1985, he left government service altogether. However, again he returned to work on his second passion, personality measurement. This time he worked for several contracting firms, including Universal Energy Systems and Metrica for the next 8 years, until his untimely death in 1995.
During his career, Christal was honored with numerous awards, including the Exceptional Civilian Service Award (Secretary of the Air Force), Meritorious Civilian Service Award (Department of the Air Force), Certificate of Commendation (Commandant of the U.S. Marine Corps), Donald B. Haines Award, Harry H. Greer Award, and Federal Employee of the Year Award (Society for Personnel Administration). He served as the Air Force Representative to the Military Human Resources Technical Panel of TTCP for 5 years, and he was a fellow of three divisions of the American Psychological Association, including Division 19. He was a member of the American Educational Research Association and was a charter member of the American Psychological Society (now, the Association for Psychological Science). He was the first American to be awarded an Honorary Doctorate of Science Degree from Plymouth University, United Kingdom, in 1994, where he had consulted for the university's Human Assessment Laboratory since 1985.
Occupational Measurement and CODAP
Christal spent the greater part of his career working in the area of occupational measurement developing a system called the CODAP system. Christal's work established the Air Force's Occupational Research division and then the Occupational Measurement Squadron at Randolph Air Force Base. The system was also used in the other U.S. military services and government agencies; the Armed Forces in Canada, Australia, and the United Kingdom; and in the civilian sector. Its main uses included establishing aptitude requirements, preparing a training curriculum, and designing jobs.
The particular form of job analysis Christal developed focused on job attributes rather than on individual attributes and has come to be known as the job task inventory approach. Over his career, Christal developed a system of computer programs for analyzing, organizing, and reporting this information to management and decision makers.
His job analysis data collection system also included an extensive background questionnaire, which addressed such topics as demographic information, job description and location, prior experience, certifications, career goals, and other personal attributes. Hundreds of studies were conducted over a 20+ year period beginning in the late 1950s on how best to collect and report out information. It is beyond the scope of this article to summarize that research (see Christal & Weissmuller, 1988, for a review). However, a couple of findings have general applicability beyond even job analysis. One is the importance for quality data collection of having a multiple-choice as opposed to constructed-response job-task inventory. Constructed response (free text) is used for the initial preparation of the task form, but a multiple-choice version is used for actual data collection. Multiple-choice eliminates the requirement for otherwise having to determine equivalences between different characterizations and spellings of job tasks. The second finding was that the nine-category time spent scale (from “very small” to “very large” amount) predicted actual time spent better than did direct judgments of time spent.
Personality Testing and a “True Father” of the Big Five
In the late 1950s and early 1960s, Christal collaborated with Ernest Tupes on a research program for measuring the abilities, including personality, of Air Force pilots for possible use in a pilot selection and classification system. To assess personality, they used a peer-rating approach, being concerned that self-ratings would be too easily and too often faked. The work never resulted in an operational selection battery because there was fear that the validity of peer ratings would be downward biased because some examinees would likely down-rate peers in order to boost their own relative standing. In Christal's words, “some applicants would stab their mother to be a pilot.”
Nevertheless this work is now universally acknowledged to have established the Big Five factors of personality, and Tupes and Christal have been declared “the true fathers” of the Big Five theory of personality by one of its main proponents, popularizers, and researchers, Lew Goldberg (1993). Tupes and Christal (1961/1992) reported that they could not replicate Cattell's 16-factor personality model (the famous 16PF) with an Air Force sample. Instead they identified five personality dimensions, which they labeled as follows: surgency (i.e., extroversion), agreeableness, dependability (i.e., conscientiousness), emotional stability, and culture (i.e., openness/intellect). Warren Norman, a visiting University professor at the Air Force lab, replicated their findings with a university sample and published the results (Norman, 1963). Until recently, Norman received most of the credit (over 1,500 citations) for being the father of the Big Five.
Soon after Tupes and Christal's seminal finding, dimensional analysis of personality became unfashionable due to the influence of Walter Mischel, who emphasized the role of situational over personal factors in governing behavior. The dimensional analysis of personality went into a dormant period until the late 1980s, when the five-factor structure of personality was rediscovered through the efforts of Digman (1990), Goldberg (1993), McCrae and John (1992), and others. Then in 1992, in a special issue of the Journal of Personality devoted to papers on the five-factor model of personality, Christal was invited to have the original 1961 technical report republished in the journal and was delighted to have the work so recognized.
After Christal left government service in 1990 and came back as a contractor, he devoted his efforts to further explorations of personality, examining the interrelationships between life experiences, activity preferences, personality traits, self-image, and measured abilities. This work resulted in the TSD inventory, which was used by the British, Canadian, and Australian Forces through the TTCP (e.g., see Collis & Barucky, 1999).
As was the case with the original Tupes and Christal report, and more broadly with much of Christal's work, it was never published in the mainstream scientific literature. Christal's priority was applications. Although he understood the importance of documenting interesting and important findings, it was not one of his personal priorities. As a case in point, much of what Christal found on personality from his work in the 1990s is documented in the humbly titled (and difficult to find) R&D Summary Report (Christal, 1993). Findings include (a) best items—both trait terms and statement items—for measuring the Big Five; (b) the cross-validation method of splitting items into two halves and comparing factor scores as a way of validating the necessity and sufficiency of five dimensions; (c) the finding that negative agreeableness items measure neuroticism better (and therefore defining neuroticism as a negative-pole factor per se); (d) the importance of ability for understanding item content and therefore in the factor structure obtained; (e) the importance of the rating scale per se (Christal experimented with all formats of computer administered rating scales, including slider bars), particularly in governing the frequency of neutral (0 point) and extreme responses; (f) that peer- and self-ratings provide independent information, and that peer ratings change with the length and quality of observation time; and (g) single words are scarce for measuring certain behavioral facets such as culture, scientific curiosity, and helpfulness, which require longer statements for their measurement.
Policy Capturing, Judgment Analysis, and Clustering Technology
Although it would be inappropriate to characterize Christal as a pure methodologist per se, he contributed to the development of several statistical methods focusing on specific applications, and he is considered the originator (with Bob Bottenberg and Joe Ward) of the methods of policy capturing and judgment analysis (e.g., see Cooksey's, 2011, note on that history.) Policy capturing is the use of multiple regression, primarily, to apprehend the policy of a decision maker. The decision maker could be selecting pilots, for example, and the predictor variables could be whatever pilot attributes (e.g., test scores, eye color, visual acuity, height, athleticism, physical attractiveness) the decision maker would have available to make the decision. Policy capturing analysis finds the variable weights that the decision maker implicitly uses in making a go–no-go decision, for example. Christal (1968) provided a description of the concept and a tutorial of the method in his article, “Selecting a Harem—And Other Applications of the Policy-Capturing Model.” According to a personal conversation I had with Christal in about 1985, this article was actually considered for one of the late Senator William Proxmire's “Golden Fleece awards”—an award intended to highlight instances of wasteful government spending. However, a phone call with one of Proxmire's staffers was sufficient to convince the Senator that this was in fact a serious scientific activity.
Judgment analysis—or JAN as Christal, Bottenberg, and Ward originally called it—referred to a technique for clustering judges based on their policies. To go back to the pilot example, if there was a board making the decision, then each board member's policy could be captured (using policy capturing), and then judges with similar policies could be clustered together using JAN techniques. The essential case for this method was presented in Bottenberg and Christal's (1961) report, in which they discussed the necessity of clustering equations for prediction. The idea is that it is not feasible to have separate equations developed to predict each individual's success in every context (interestingly, Schmidt's validity generalization later addressed the same issue, but from the standpoint of lack of evidence that separate equations are necessary). Clustering equations (objectively identifying similar situations) provides a more feasible approach, and the value of the degree of clustering (from many small clusters to a few large ones) can be determined in a cost-benefit analysis. Ward (1963) later generalized this idea in his highly cited (over 7,000 citations) Journal of the American Statistical Association article entitled “Hierarchical Grouping to Optimize an Objective Function.”
Over the years, these methods were applied to person clustering (a forerunner to today's latent class analyses), jobs clustering to minimize retraining time for transfers, job descriptions to describe a wide variety of jobs with a smaller number of descriptions, and regression equation clustering to identify job clusters. Another application for the methodology and the technology was to establish the present Air Force “Weighted Airman Enlisted Promotion System,” which has been operational for several years.
Aptitude Measurement and LAMP
I met Christal when I joined the Air Force Human Resources Laboratory in 1982, shortly after he established LAMP at the Lackland Air Force Base. I worked with him for the next 14 years on ability and personality testing. However, prior to that time and throughout his career, he had been developing abilities testing programs for the Air Force. He was a key contributor to the development of the original Airmen Classification Battery, the first enlisted differential aptitude battery used by the Air Force. It provided the framework for all subsequent enlisted selection and classification test batteries, including today's ASVAB. He developed the first Air Force Officer Qualifying Test. This test and its subsequent forms have been used continuously for selection and classification of all Air Force officer personnel going back to 1957. Christal also developed the Air Force Pilot Instructor Selection Battery, which was used for selecting pilot instructors during the Korean War buildup.
In 1979, Christal established LAMP, a basic research program funded by the Air Force Office of Scientific Research and by the (then) Air Force Human Resources Laboratory. The program was designed to capitalize on research that had been conducted by the Office of Naval Research on an information processing approach to abilities measurement (led by Marshall J. Farr)—with prominent researchers such as Earl (Buz) Hunt, Robert Sternberg, and Richard Snow—on determining how individuals learn, remember, and process information when acquiring new knowledge and skills. At the time, there were several related programs in the U.S. military, including the Army's Project A and the Navy's ECAT program. Unlike these other programs, LAMP was specifically focused on a comprehensive information-processing-based overhaul of conventional abilities testing. Although LAMP was never implemented, there were many findings and ideas that emerged from the research, such as automatic item generation, that are still being evaluated in current research programs and that have been recommended as research priorities for future research on the ASVAB (Drasgow, Embretson, Kyllonen, & Schmitt, 2006).
TTCP and Cross-National Work
During a 5-year period in which Christal was the Air Force's representative to TTCP, he forged many connections with services in Australia, the United Kingdom, Canada, and New Zealand, as well as the other U.S. services. It was partly through this work, but also other cross-national work he was involved in, that led to the use of occupational analysis methods, including the CODAP system, in those other military services. He also formed a strong collaborative bond with Sidney Irvine and members of his team at Plymouth University, who had been working with the British Ministry of Defence to create a new information-processing abilities battery for U.K. enlistees, the British Army Recruit Battery, currently in operational use. Irvine (2013) has devoted whole chapters in a recent book describing the development of that system to his collaboration with Christal.
Christal made tremendous contributions to military psychology through his career at Brooks and Lackland Air Force Bases. His contributions were wide ranging, from methodological to applied, and included occupational analysis, aptitude testing for enlisted personnel and officers, and personality as well as ability assessment. He had a relentless energy and an infectious enthusiasm, and many who spent time with him remarked on how he seemed never to run out of ideas, studies he wished to try out, variables he wanted to investigate, computer programs and statistical analyses he wanted to learn, and his new ways of looking at old problems. It is not surprising then to learn that when Christal passed away in 1995, he was at work in his beloved Air Force lab. He made his mark on military psychology not only through his research but also especially through the operational programs that emerged from it. But, perhaps most importantly, he made a lasting impact through the people he met and influenced, and the people who remember him even today.
The author thanks Sidney Irvine, who provided bibliographic material and the photograph taken when Christal received the honorary doctorate of science degree from Plymouth University.Comments and questions may be directly addressed to the author .
Bottenberg, R. A., & Christal, R. E. (1961). An iterative technique for clustering criteria which retains optimum predictive efficiency (WADD-TN-61-30, AD-261 615). Lackland Air Force Base, TX: Personnel Research Laboratory, Wright Air Development Center.
Christal, R. E. (1968). Selecting a harem—And other applications of the policy-capturing model. The Journal of Experimental Education, 36, 35–41.
Christal, R.E. (1993). R&D summary report (F33615-91-D-0010). Brooks Air Force Base, TX: Armstrong Laboratories.
Christal, R. E., & Weissmuller, J. (1988). Job-task inventory analysis. In G. Gael (Ed.), The job analysis handbook for business, industry, and government (Vol. II, pp. 1036–1050). New York, NY: Wiley.
Collis, J. M., & Barucky, J. M. (1999). “Big Five” personality factors research using Christal's Self Description Inventory (Metrica, Inc., Technical Report). Retrieved from DTIC Online website: http://oai.dtic.mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADA364039
Cooksey, R. W. (2011). Comments on the distinction between “policy capturing” and “judgment analysis.” Retrieved from http://www.brunswik.org/notes/japc.html
Drasgow, F., Embretson, S. E., Kyllonen, P. C., & Schmitt, N. (2006). Technical review of the Armed Services Vocational Aptitude Battery (ASVAB) (FR-06-25). Alexandria, VA: Human Resources Research Organization.
Goldberg, L. R. (1993). The structure of phenotypic personality traits. American Psychologist, 48, 26–34.
Irvine, S. H. (2013). Tests for recruitment across cultures: A tactical psychometric handbook . Amsterdam, the Netherlands: IOS Press.
McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its applications. Journal of Personality, 60 , 175–215. doi:10.1111/j.1467-6494.1992.tb00970.x
Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer nomination personality ratings . Journal of Abnormal and Social Psychology, 66, 574–583.
Tupes, E. C, & Christal, R. E. (1992). Recurrent personality factors based on trait ratings. Journal of Personality, 60, 225–251. (Original work published 1961)
Ward, J. H., Jr. (1963). Hierarchical grouping to optimize an objective function. Journal of the American Statistical Association , 58, 236–244.
Bottenberg, R. A., & Christal, R. E. (1968). An iterative technique for clustering criteria which retains predictive efficiency. The Journal of Experimental Education, 36, 28–34.
Christal, R. E. (1952). A factor analysis of Airmen Classification Battery AC-1A and Radar Operator final school grade (Research Note 52-37). Lackland Air Force Base, TX: Human Resources Research Center.
Christal, R. E. (1955). Development of the pilot instructor selection battery (PRL-TM-55-2). Lackland Air Force Base, TX: Personnel Laboratory, Air Force Personnel and Training Research Center.
Christal, R. E. (1958). Factor analytic study of visual memory. Psychological Monographs: General and Applied, 72, 1–24.
Christal, R. E. (1968). JAN: A technique for analyzing group judgment. The Journal of Experimental Education, 36, 24–27. (Original work published 1963)
Christal, R. E. (1970). Inputs to vocational-technical education from occupational research. The Vocational Guidance Quarterly, 18, 253–257.
Christal, R. E. (1971). Stability of consolidated job descriptions based on task inventory survey information (AFHRL-TR-71-48, AD-734 739). Lackland Air Force Base, TX: Personnel Research Division, Air Force Human Resources Laboratory.
Christal, R. E. (1989). Estimating the contribution of experimental tests to the Armed Services Vocational Aptitude Battery (AFHRL-TP-89-30, AD-A216 122). Brooks Air Force Base, TX: Manpower and Personnel Division, Air Force Human Resources Laboratory.
Christal, R. E. (1991). Comparative validities of ASVAB and LAMP tests for logic gates learning (Technical Report No. AL-TR-1991-31). Brooks Air Force Base, TX: Armstrong Laboratory, Human Resources Directorate.
Christal, R. E. (1992). Author's note on “Recurrent Personality Factors Based on Trait Ratings.” Journal of Personality, 60, 221–224. doi:10.1111/j.1467-6494.1992.tb00972.x
Christal, R. E., & Bottenberg, R. A. (1964). Procedure for keying self-report test items (PRL-TR-64-22, AD-608 066). Lackland Air Force Base, TX: Personnel Research Laboratory, Aerospace Medical Division.
Christal, R. E., Madden, J. M, & Harding, F. D. (1960). Reliability of job evaluation ratings as a function of number of raters and length of job descriptions (WADD-TN-60-27 AD-251- 837) Lackland Air Force Base, TX: Personnel Research Laboratory, Wright Air Development Center.
Gael, S. (1988). The job analysis handbook for business, industry, and government (Vol. II). New York, NY: Wiley.
Kyllonen, P. C., & Christal, R. E. (1986). Modeling learning abilities. In R. E. Dillon (Ed.). Advances in testing and training. Carbondale, IL: Southern Illinois University Press.
Kyllonen, P. C., & Christal, R. E. (1987). Cognitive modeling of learning abilities: A status report of LAMP (Technical Paper No. 87-66). Brooks Air Force Base, TX: Air Force Human Resources Laboratory.
Kyllonen, P. C., & Christal, R. E. (1989). Cognitive modeling of learning abilities: A status report of LAMP. In R. Dillon & J. W. Pellegrino (Eds.), Testing: Theoretical and applied issues (pp. 146–173). New York, NY: Praeger.
Kyllonen, P. C., & Christal, R. E. (1990). Reasoning ability is (little more than) working memory capacity?! Intelligence, 14, 389–433.
Kyllonen, P. C., Tirre, W. C., & Christal, R. E. (1991). Knowledge and processing speed as determinants of associative learning. Journal of Experimental Psychology: General, 120, 89–108.
Payne, D. L., Christal, R. E., & Kyllonen, P. C. (1986) Behavioral differences in learning abilities. In T. Sticht, F. Chang, & S. Wood (Eds.), Advances in reading/language research (Vol. IV). Greenwich, CN: JAI Press.
Tupes, E. C., & Christal, R. E. (1954). The Officer Aptitude Test (AFPTR 423, PRL-TM-54-1). Lackland Air Force Base, TX: Air Force Personnel and Training Research Center.
Tupes, E. C., & Christal, R. E. (1958). Stability of personality trait ratings obtained under diverse conditions (WADC-TN-58-61, AD-151 O41). Lackland Air Force Base, TX: Personnel Research Laboratory, Wright Air Development Center.