1930s

A number of notable events occurred in the mid- to late 1930s. On the legislative front in the United States, the Fair Labor Standards Act of 1938 was passed. This law set a minimum wage and maximum hour limit for workers engaged in interstate commerce and prohibited children under the age of sixteen from working in mining or manufacturing jobs (Yellowitz, Reference Yellowitz, Foner and Garraty1991). As the 1930s progressed, the worldwide economic depression deepened, and countries moved toward a second worldwide military conflict. In the United States, although test use in industry was down compared to the 1920s, employee selection continued to be the primary activity for industrial psychologists. By the mid-1930s, selection research was declining in Europe, and the United States was becoming the center of that activity (Salgado, Anderson, & Hülsheger, Reference Salgado, Anderson, Hülsheger, Farr and Tippins2010). There were refinements made in the statistics used in testing and selection. In Reference Kuder and Richardson1937, Kuder and Richardson developed an improvement of split-half estimates of test reliability they named KR-20. Taylor and Russell (Reference Taylor and Russell1939) published a series of tables that considered the selection ratio, the base rate, and the validity coefficient when determining the utility of a selection test.Footnote ¹ The US Employment Service published the initial volume of the Dictionary of Occupational Titles (DOT) in 1939, the first comprehensive listing of occupations in the United States. Also in 1939, work began on the Army General Classification Test (AGCT), a replacement for the Army Alpha and Beta tests of World War I vintage (Harrell, Reference Harrell1992). Walter Van Dyke Bingham, who played a pivotal role in the personnel testing program in World War I, was appointed chair of the Committee on Classification of Military Personnel.

By 1936, applied psychology research and practice was prohibited in those countries controlled by the USSR. A particular point of contention was the practice of distinguishing among individuals based on psychological differences, which conflicted with communist ideology. This began a long period of isolation from Western psychologists for Soviet psychologists that lasted until the end of the 1950s (Warr, Reference Warr and Koppes2007). The first industrial psychology text in China was published in 1935. Industrial psychology work there concentrated on working conditions and employee selection and guidance (Wang, Reference Wang, Triandis, Dunnette and Hough1994).

After immigrating to the United States from Germany in 1933, Kurt Lewin (1890–1947) conducted groundbreaking work in social psychology, particularly in the study of group dynamics. While at the University of Iowa, Lewin and his colleagues conducted a classic study contrasting authoritarian, democratic, and laissez-faire leadership styles (Lewin, Lippitt, & White, Reference Lewin, Lippitt and White1939). The same year that study was published, Lewin was invited to the Harwood Manufacturing Company by Alfred J. Marrow, a company vice president who also held a doctorate in psychology (Highhouse, Reference Highhouse and Koppes2007).Footnote ² The Harwood Company was experiencing a turnover problem. Lewin and his students Alex Bavelas and John R. P. French Jr. conducted a series of studies to attempt to alleviate this problem. For a time, what became known as the Harwood Studies rivaled the Hawthorne Studies in notoriety, and similar to those studies, they were subjected to criticism and reinterpretation (Highhouse, Reference Highhouse and Koppes2007; Hilgard, Reference Hilgard1987). Lewin’s work at Harwood was the genesis of his action research model, an approach that proved influential in organizational development (OD) practice (French, Reference French1982).

In 1938, New Jersey Bell executive Chester Barnard published The Functions of the Executive. In this influential book, Barnard described organizations as cooperative systems, with leaders who valued cohesion and power that originated from the bottom instead of the top of the organizational hierarchy. He also discussed upward communication and the formation of natural groups in the organization (Perrow, Reference Perrow and Landy1986). Perrow believed Barnard’s book was the first effective counterargument to the scientific management system.Footnote ³ Lawrence (Reference Lawrence and Lorsch1987) viewed The Functions of the Executive as a plausible starting point for the organizational behavior perspective, along with two other works from the 1930s that described the Hawthorne Studies, Elton Mayo’s Human Problems of Industrial Civilization (Reference Mayo1933) and Fritz Roethlisberger’s and William Dickson’s Management and the Worker (Reference Roethlisberger and Dickson1939). By the early 1930s, the psychology of advertising and selling, one of the seminal applications of psychology to business, had grown so specialized and distinctive that Viteles (Reference Viteles1932) chose to omit that topic from his landmark text Industrial Psychology. In 1938, of the 2,318 members and associates of the APA, 1,923 reported employment in psychological work (the remainder were students or retirees or did not list a position). Of those employed, sixty-one identified as industrial psychologists. This did not include individuals who identified as “guidance and personnel workers” or as consultants (Finch & Odoroff, Reference Finch and Odoroff1939).

1940s

World War II took place from 1939 to 1945. Like World War I, the Second World War gave psychologists an opportunity to apply their expertise to large numbers of individuals. In particular, there was an enormous need for an expansion of the military’s selection and classification systems. To illustrate, the United States had a failure rate for aviators of 40 to 60 percent before the war; however, this was not a significant problem because so few pilots were needed. For example, only twelve pilot applicants were accepted in 1937, but by 1942, that total had increased to more than 293,000 (Napoli, Reference Napoli1981). By 1940, the Army General Classification Test (AGCT) was ready for use (Harrell, Reference Harrell1992). Two years later, the US Navy established its Applied Psychology Panel, which developed more than 250 tests on topics such as reading, arithmetic reasoning, mechanical knowledge, and aptitude. The Navy’s aviation psychology program was directed by John G. Jenkins (Napoli, Reference Napoli1981). In the Army Air Force, John C. Flanagan (Reference Flanagan1948) supervised the construction of the Aviation Cadet Qualifying Exam. The military created research centers that evolved after the war’s end into the Army Research Institute (ARI), the Navy Personnel Research and Development Center (NPRDC), and the Air Force Human Resources Laboratory (AFHRL) (Katzell & Austin, Reference Katzell and Austin1992).

By the beginning of World War II, the German military employed approximately 200 psychologists. These psychologists conducted job analyses and developed selection tests for aviators, sound detectors, tank drivers, marksmen, and others. Psychologists were also involved in training, morale-building, and propaganda. One of the main activities of military psychologists was the selection of officer candidates. The candidates went through a two-day selection process that assessed characteristics such as leadership ability, willpower, and practical intelligence. Among the procedures used were interviews, life history items, a leaderless group discussion procedure, handwriting analysis, and analysis of speech and facial expressions. Assessment of the candidate’s entire personality was the objective (Ansbacher, Reference Ansbacher1941). The whole process laid the foundation for what after the war became known as an assessment center. This combination of pencil-and-paper tests and situational tests was pioneered by the German military following World War I (Ansbacher, Reference Ansbacher1951) and by the Hungarian Army in the 1930s (Salgado, Anderson, & Hülsheger, Reference Salgado, Anderson, Hülsheger, Farr and Tippins2010). The assessment center concept made the leap from the German to the British military and then to the American military. Among the Americans involved was Henry Murray of Harvard University, who was instrumental in constructing situational tests for applicants of the Office of Strategic Services, a forerunner of the Central Intelligence Agency (CIA).

From a present-day perspective, the German military system was a mix of potentially valid procedures (e.g., aptitude tests, leaderless group discussion) and ones whose validity was questionable at best (e.g., analysis of handwriting, speech, and facial expressions). Davis (Reference Davis1947) reported that the emphasis was on the clinical assessment of character, not on more psychometrically sound approaches. No serious attempts at conventional validation of the selection procedures were made. There was a great deal of conflict between the psychology program and the German High Command, which decided to terminate the Army and Air Force programs in December 1941, early in the war. The Navy’s psychology program limped along in a much reduced state after that time. About 1943, industrial psychology in Germany ground to a halt (Warr, Reference Warr and Koppes2007). The German ally Japan began using tests for military officers in 1942 (McCollom, Reference McCollom1968).

The largest military selection program in Europe during World War II took place in Great Britain (Salgado, Anderson, & Hülsheger, Reference Salgado, Anderson, Hülsheger, Farr and Tippins2010). Many of the tests used were adapted from those developed in the United States, although British psychologists took a less quantitative approach to scoring than the Americans and preferred interviews and biographical data to standardized tests. A nonverbal cognitive ability test, the Progressive Matrices Test, was administered to approximately 3 million British Army and Royal Navy recruits. At least 2 million women and men took a battery of five or more tests (Vernon, Reference Vernon1947).

In the United States after the war, Hubert E. Brogden (Reference Brogden1946, Reference Brogden1949; Brogden & Taylor, Reference Brogden and Taylor1950) extended previous work on selection utility by demonstrating that the size of a validity coefficient is directly proportional to the percentage of gain one would expect to see if selection was based on the criterion itself; that is, the maximally effective method.Footnote ⁴ Salgado (Reference Salgado2001) noted that all subsequent utility analysis work is based on Brogden’s insights. Wagner (Reference Wagner1949) reviewed the employment interview literature and noted that structured interviews showed promise for improving the traditionally low reliability and validity of unstructured ones. The journal Personnel Psychology began publication in 1947, and toward the end of the decade, Edwin Ghiselli and Clarence Brown (Reference Ghiselli and Brown1948) and Robert L. Thorndike (Reference Thorndike1949) published influential books on employee selection.

In performance appraisal, Wiener’s (Reference Wiener1948) cybernetic theory about how systems adapt as a result of prior performance information likely influenced the use of performance feedback in I-O psychology (Farr & Levy, Reference Farr, Levy and Koppes2007). Robert L. Thorndike’s (Reference Thorndike1949) text included his popular classification of criteria into immediate, intermediate, and ultimate levels, with the ultimate criterion representing the final goal of selection or training (Austin & Villanova, Reference Austin and Villanova1992).

By the 1940s, the shortcomings of the trait approach to leadership were becoming evident. A shift from these trait approaches to a behavioral approached occurred, most notably at Ohio State University and the University of Michigan. The Ohio State Leadership Studies, begun in 1945, were a major program of the Personnel Research Board, an interdisciplinary research group that conducted organizational research for industry, the military, and the government (Meyer, Reference Meyer and Koppes2007).Footnote ⁵ In motivational research, the humanist psychologist Abraham Maslow published his influential needs theory in 1943. Developed in the 1930s and based on Maslow’s observations of individuals struggling in the Great Depression (Latham & Budworth, Reference Latham, Budworth and Koppes2007), Maslow postulated a hierarchy of needs, starting with physiological needs; ranging through security, social, and esteem needs; and cumulating with self-actualization. Once a lower set of needs is more or less satisfied, the set above it motivates the individual. Although not particularly well supported empirically, Maslow’s need theory was popular in organizations.

In 1944, Kurt Lewin founded the Research Center for Group Dynamics at the Massachusetts Institute of Technology (MIT) (Salas, Priest, Stagl, Sims, & Burke, Reference Salas, Priest, Stagl, Sims, Burke and Koppes2007). In 1946, Lewin, Kenneth Benne, Leland Bradford, and Donald Lippitt attempted to reduce racial tension in New Britain, Connecticut. Their efforts to change attitudes there resulted in the T-group method, also known as laboratory training and sensitivity training (Benne, Reference Benne, Bradford, Gibb and Benne1964; Highhouse, Reference Highhouse and Koppes2007). The following year, Lewin (Reference Lewin1947) published his influential three-step process for changing group standards: unfreezing, moving, and freezing. In 1948, the Center for Group Dynamics joined Rensis Likert’s Michigan Survey Research Center and the Center for Utilization of Scientific Knowledge to form the University of Michigan Institute for Social Research (ISR). Topics such as organizational conflict, power, and decision making were receiving increased attention in the 1940s (Perrow, Reference Perrow and Landy1986). An example was the work on goal conflict by the sociologist Philip Selznick (Reference Selznick1949) in his study of the Tennessee Valley Authority.

In 1946, the Tavistock Institute of Human Relations was established in London through a grant from the Rockefeller Foundation. At the Institute, psychologists, psychiatrists, and anthropologists worked together using the socio-technical systems model to conduct research on topics such as group processes, worker health and well-being, and conflict (Trist, Emory, Murray, & Trist, Reference Trist, Emory, Murray and Trist1997). One of the better-known studies conducted was the Tavistock coal-mining study (Trist & Bamforth, Reference Trist and Bamforth1951). Management changed the mining process from the “short-wall” method, where miners worked in small, autonomous groups, to the highly specialized “long-wall” method. While the latter was more efficient from an engineering perspective, use of the long-wall method disbanded the autonomous work teams that provided social support to the miners. The long-wall method was not initially a success, emphasizing the importance of taking social factors into account when making technological changes; that is, adopting a socio-technical approach. Erich Trist, a leading spokesperson for the Institute, was influenced by psychoanalysis, Kurt Lewin, open systems theory, and the humanist approach to research (John B. Miner, Reference Miner2002). Because traditional academic publications could be unsympathetic to psychiatrically based social science, the Tavistock Institute began publishing its own journal Human Relations in 1947 to provide an outlet for this type of research (Warr, Reference Warr and Koppes2007).

In 1947, a year prior to the founding of the state of Israel, Louis Guttmann established a behavioral unit in the Israeli military. This unit later became the Israel Institute for Applied Social Research. Also in 1947, an English translation of Max Weber’s The Theory of Social and Economic Organization was published, initiating interest in bureaucratic organizational structure. Germany divided into East and West Germany in 1949. From that point, psychology in East Germany came under the influence of the USSR, while the United States was the main influence on psychology in West Germany (Warr, Reference Warr and Koppes2007).

1950s

In contrast to the early years of industrial psychology, in the 1950s, selection research was concentrated in the United States, and there was very little collaborative activity between American academics and those in other countries. This situation would last until the mid-1970s (Salgado, Anderson, & Hülsheger, Reference Salgado, Anderson, Hülsheger, Farr and Tippins2010). The assessment center technique made the transition from the military to private industry, most prominently in the mid-1950s at AT&T. Management studies on assessment centers at AT&T that demonstrated the usefulness of the technique were conducted by Douglas Bray, Richard Campbell, and Donald Grant (Bray & Campbell, Reference Bray and Campbell1968; Bray & Grant, Reference Bray and Grant1966; Bray, Campbell, & Grant, Reference Bray, Campbell and Grant1974; see Howard, Reference Howard, Farr and Tippins2010, for an historical summary). Bernard Bass conducted notable work on a technique often used in assessment centers, the leaderless group discussion (Bellows, Reference Bellows, Stone and Taylor1951; Heron, Reference Heron, Stone and McNemar1954). Another procedure used primarily for selecting managers, individual assessment, saw an increase in research and practice during the 1950s (Highhouse, Reference Highhouse2002; Prien, Schippmann, & Prien, Reference Prien, Schippmann and Prien2003).Footnote ⁶

Advances in measurement, test validation, and test construction relevant for I-O psychology included the introduction of Item Response Theory (IRT) in 1952 by Frederic Lord (Austin, Scherbaum, & Mahlman, Reference Austin, Scherbaum, Mahlman and Rogelberg2002). IRT was an attempt to bring ratio-level measurement to psychological testing. Other noteworthy developments included the following: Lee Cronbach (Reference Cronbach1951) developed a reliability coefficient with more general application than KR-20, his coefficient alpha. Cronbach, along with Paul Meehl, published an influential article on the concept of construct validity (Cronbach & Meehl, Reference Cronbach and Meehl1955). A year earlier, Meehl (Reference Meehl1954) published a controversial book that demonstrated the clear superiority of statistical prediction versus clinical prediction. The statistical approach to selection continued in the 1950s to be the most popular approach in the profession (Katzell, Reference Katzell, Farnsworth and McNemar1957). Cronbach and Gleser (Reference Cronbach and Gleser1957) extended Brogden’s 1940s work on utility analysis. The critical incident technique, a procedure that proved popular in both job analysis and performance appraisal, was introduced by John C. Flanagan in Reference Flanagan1954. The importance of appropriate criteria in selection received attention, although as Wallace and Weitz (Reference Wallace, Weitz, Stone and McNemar1955) noted, actual criterion research about concerns such as criterion relevance, deficiency, and contamination was lacking. Robert Wherry’s work on the effect of psychological, situational, and procedural variables on the accuracy of ratings was an exception (Farr & Levy, Reference Farr, Levy and Koppes2007).Footnote ⁷ Important selection-related publications included the collaborative Technical Recommendations for Psychological and Diagnostic Techniques (APA, AERA, & NCMUE, 1954) in the United States and, on the international level, the first edition of the International Classification of Occupations for Migration and Placement in 1952 and the first edition of the International Standard Classification of Occupations in 1958 (both cited in Salgado, Reference Salgado2001).

Despite criticism from outside the profession (e.g., Whyte, Reference Whyte and LaFarge1954), research on and the use of personality tests increased during the 1950s (Brown & Ghiselli, Reference Brown, Ghiselli, Stone and Taylor1952). A review of studies conducted on biographical data inventories found considerable predictive validity for them (Taylor & Nevis, Reference Taylor, Nevis, Farnsworth, McNemar and McNemar1961). Long criticized for its questionable reliability and validity, the employment interview continued to be the most popular selection technique used by employers (Kendall, Reference Kendall, Farnsworth and McNemar1956). Because employers were resistant to giving up the interview, Heron (Reference Heron, Stone and McNemar1954) advised psychologists to stop simply documenting its deficiencies and work on improving its reliability and validity.

George Homans’s (Reference Homans1950) The Human Group analyzed groups in terms of systems and established the paradigm for analyzing group behavior in organizations (Lawrence, Reference Lawrence and Lorsch1987). In other human relations–based theory and research, Douglas McGregor (Reference McGregor1957, Reference McGregor1960) discussed managerial beliefs about subordinates. Two examples, Theory X managers (who believe that because of the aversive nature of work employees must be tightly controlled) and Theory Y managers (who believe that if employees find the work intrinsically motivating, they will be self-motivated and can be trusted) have shown considerable staying power in organizational theory. In a similar human relations vein, Chris Argyris (Reference Argyris1957) introduced his developmental model of organizational behavior, demonstrating conflict between the policies of modern organizations and the personalities of mature adults. By the middle of the decade, the human relations or organizational behavior (OB) perspective focused on the fit between the individual employee and the organization, with the assumption that a good fit benefits both (Shafritz & Ott, Reference Shafritz and Ott1996).

In Great Britain, Joan Woodward (Reference Woodward1958) discovered that the type of structure most effective for an organization depended on the type of technology the organization employed. Large-batch or mass production organizations were most successful when they were bureaucratic in structure, whereas small-batch and continuous process organizations tended to have more humanistic or organic structures.

Progress was made in understanding power and decision making in organizations. Dorwin Cartwright, in a 1953 address, argued that organizational variables such as leadership and attitude change could only be understood if viewed through the prism of power (cited in Ott, Reference Ott1989). At the end of the decade, Richard Cyert and James March discussed how power affects organizational goals (Shafritz & Ott, Reference Shafritz and Ott1996), and John R. P. French, Jr. and Bertram Raven (Reference French, Raven and Cartwright1959) introduced their five bases of power: reward, coercive, legitimate, expert, and referent. In decision theory, Cyert, March, and Herbert Simon discussed human cognitive limits when making decisions. They noted that rather than always working toward the optimal outcome, people will employ a “satisficing” strategy to obtain a suboptimal but good enough solution (March & Simon, Reference March and Simon1958; Simon, Reference Simon1947).

By the mid-1950s, motivation had become a central issue in industrial psychology (Ryan & Smith, Reference Ryan and Smith1954). Motivation was explicitly discussed in the industrial psychology literature, and theories of motivation were being developed in what had previously been a largely atheoretical area (Latham & Budworth, Reference Latham, Budworth and Koppes2007). Many of these theories combined motivation and job satisfaction. Some, influenced by Freud’s conception of the unconscious and by Henry Murray’s (Reference Murray1938) work on personality, focused on employee needs. Maslow’s need theory, described in the previous section, is the most notable example. Frederick Herzberg and his colleagues developed a need theory specifically for work situations. In their two-factor theory, the two sets of needs are hygiene factors, such as pay, and motivator factors that have to do with the intrinsic nature of the work itself. Meeting only hygiene needs will result in a worker who is in a neutral state: neither satisfied nor dissatisfied, neither motivated nor unmotivated. Only meeting the intrinsic motivator needs through job enrichment results in a satisfied, motivated employee (Herzberg, Mausner, & Snyderman, Reference Herzberg, Mausner and Snyderman1959). Attuned to the increased emphasis on motivation in industrial psychology, Morris Viteles’s attempted revision of his classic 1932 text resulted in a book focused on motivation and attitudes: Motivation and Morale in Industry in Reference Viteles1953 (Viteles, Reference Viteles, Boring and Lindzey1967). Peter Drucker (Reference Drucker1954) introduced management by objectives, a performance appraisal technique based on the motivational concept of goal setting. The procedure involved the manager and employee setting mutually agreed-upon, explicit goals and receiving performance feedback regarding progress toward those goals.

The 1950s saw the following additional developments. In training, Donald Kirkpatrick (Reference Kirkpatrick1959) introduced four levels of training program evaluation criteria: reactions, learning, behavior change, and performance change. A review by Brayfield and Crockett (Reference Brayfield and Crockett1955) found little evidence for a significant relationship between employee attitudes and job performance, casting doubt on the “satisfied worker is a productive worker” axiom. In Great Britain, the professional category of Psychologist was established in the Civil Service in 1950 (Warr, Reference Warr and Koppes2007). Among notable publications from this decade were the Handbook of Applied Psychology by Fryer and Henry, which appeared in Reference Fryer and Henry1950, and the journal Administrative Science Quarterly, which was first published in 1955. Due in part to explosive growth in military applications, human factors or engineering psychology was now distinct from industrial psychology; its own APA division was created in 1956. In 1958, the journal Human Factors was first published, and the Human Factors Society was formed. Vocational counseling was now considered part of counseling psychology instead of industrial psychology, continuing a trend that began in the 1930s (Savickas & Baker, Reference Savickas, Baker, Walsh and Savickas2005). In the United States, at least 1,000 psychologists were employed full time in industry by the end of the decade (McCollom, Reference McCollom1959).

1960s

In the United States, the decade opened with a lament from Taylor and Nevis (Reference Taylor, Nevis, Farnsworth, McNemar and McNemar1961) that with the expansion of industrial psychology topics, the venerable practice of personnel selection was losing it cachet. It would in fact be a decade of challenges for psychologists involved in selection research and practice. The US Congress passed legislation outlawing discrimination against employees. This required industrial psychologists to substantially reevaluate their modus operandi. The Equal Pay Act of 1963 banned compensation differences based solely on employee sex, the Age Discrimination in Employment Act of 1967 prohibited age discrimination, and Title VII of the Civil Rights Act of 1964 banned discrimination in employment decisions based on race, sex, religion, color, or national origin. Industrial psychologists increasingly had to consider not only professional standards and employer needs but also the legal ramifications of their practice.

The following notable activity occurred in employee selection research and practice. The problem of unqualified testers administering nonvalid tests, present since the beginning of industrial psychology, received renewed attention (Dunnette, Reference Dunnette, Farnsworth, McNemar and McNemar1962; Taylor & Nevis, Reference Taylor, Nevis, Farnsworth, McNemar and McNemar1961). The basic selection validity model that had been relatively unchanged for close to fifty years was now criticized for being overly mechanistic and simplistic (Dunnette, Reference Dunnette, Farnsworth, McNemar and McNemar1962; Guion, Reference Guion, Farnsworth, McNemar and McNemar1967; Porter, Reference Porter, Farnsworth, McNemar and McNemar1966). Efforts to expand those models included Edwin Ghiselli’s (Reference Ghiselli1963) use of moderator variables to improve selection for homogenous subgroups and Marvin Dunnette’s (Reference Dunnette1963) examination of variables such as job behaviors, job situations, and subgroups of applicants that intervene between the predictor and the criterion. Wernimont and Campbell (Reference Wernimont and Campbell1968) proposed an alternative to the traditional validity model. Their behavioral consistency model preferred measures of behavior, or “samples,” over “signs,” such as tests, as predictors. In the 1960s, there was an increased interest in a broader view of the selection process, of integrating selection into the larger personnel system (Sells, Reference Sells, Farnsworth, McNemar and McNemar1964), and of taking a systems approach to selection itself (Dudek, Reference Dudek, Farnsworth, McNemar and McNemar1963). Predictors of job performance continued for the most part to be cognitive ability tests, biographical data inventories, interest inventories, and personality tests, although the latter continued to be criticized for lack of validity by industrial psychologists (Guion & Gottier, Reference Guion and Gottier1965) and for violation of privacy by individuals outside psychology (e.g., Gross, Reference Gross1962). Guion and Gottier’s (Reference Guion and Gottier1965) critique led to the perception that personality test selection research was not a fruitful activity, although that was not their intent (Guion, Reference Guion, Farnsworth, McNemar and McNemar1967, Reference Guion, Dunnette and Hough1991). In the 1960s, Edwin Fleishman began his longtime research program on a taxonomy of human motor performance (Fleishman, Reference Fleishman1988).

Notable publications this decade include Robert Guion’s (Reference Guion1965) text on personnel testing, the first edition of Standards for Educational and Psychological Tests and Manuals (APA, 1966), the second edition of International Standard Classification of Occupations (1968, cited in Salgado, Reference Salgado2001), and the third edition of the DOT (1968). The historian Loren Baritz wrote The Servants of Power (Reference Baritz1960), a book-length critique of how social scientists were co-opted by management. Also in 1960, applied psychologists in the United States interested in advertising and other consumer behavior established their own APA division. Over the objections of the industrial psychology division, the APA Council approved the new Division of Consumer Psychology, establishing for consumer psychologists an identity separate from industrial psychology (Schumann & Davidson, Reference Schumann, Davidson and Koppes2007).

Performance appraisal research emphasized behavioral measurement in the 1960s. There was a move away from trait ratings to behavioral ratings and a focus on measurement issues, accuracy of ratings, and rating format development (Farr & Levy, Reference Farr, Levy and Koppes2007). The trend toward behavior ratings was exemplified by the Behaviorally Anchored Rating Scale (BARS) introduced by Smith and Kendall in 1963. Also evident by the early 1960s was an increased interest on performance appraisal feedback and on the effect these evaluations have on employees’ careers and lives (Farr & Levy, Reference Farr, Levy and Koppes2007). Herbert Meyer’s work at General Electric on performance appraisal feedback suggested that two separate sessions should be held with employees. One should focus on development, the other on administrative concerns such as salary and promotion (Meyer, Kay, & French, Reference Meyer, Kay and French1965). Despite cognitive approaches surfacing in other areas of industrial psychology, psychologists interested in training still favored a behavioral approach (Kraiger & Ford, Reference Kraiger, Ford and Koppes2007). McGehee and Thayer (Reference McGehee and Thayer1961) published their text Training in Business and Industry, which included a description of training needs analysis.

In the 1960s, there was the beginning of a shift from need-based theories to more cognitive approaches to motivation. Victor Vroom’s (Reference Vroom1964) expectancy theory of motivation had its roots in the work of Kurt Lewin and the purposive behaviorism of Edward Tolman along with earlier work in organizations in the 1950s by Basil Georgopoulos, Gerald Mahoney, and Nyle Jones (John B. Miner, Reference Miner2002). Expectancy theory posits that effort is a function of an individual’s expectations that her or his effort will lead to successful performance, the expectation that performance will lead to a successful outcome, and the value or valence of that outcome for the individual. Another cognitive theory to emerge in this decade was equity theory, which views both motivation and satisfaction as the result of a comparison that each individual makes between two ratios: his or her ratio of perceived outcomes over inputs compared to a referent or relevant other’s ratio. If, for example, an employee believes that she or he is putting in twice the amount of effort as a coworker yet only receiving half the reward, that employee will be dissatisfied. The employee will be motivated to either increase the amount of reward received or reduce his or her amount of effort in an attempt to bring the ratios of the employee and coworker closer to equity. J. Stacey Adams (Reference Adams and Berkowitz1965), a developer of this approach, was influenced both social justice theories and Leon Festinger’s cognitive dissonance theory (John B. Miner, Reference Miner2002). A third cognitive motivational theory that received a great deal of attention was goal-setting theory, developed into a modern theory of motivation by Edwin Locke (Reference Locke1968), who later collaborated with Gary Latham.

Leadership research made the transition from behavioral approaches to situational or contingency approaches in the 1960s (Day & Zaccaro, Reference Day, Zaccaro and Koppes2007). Rather than search for behaviors that characterize effective leaders across situations, contingency approaches recognize that leader behavior is moderated by the situation. What may be effective leader behavior in one situation may be ineffective in another. Fred Fiedler’s (Reference Fiedler1967) contingency theory is a classic early example. Fiedler’s theory had two components: leadership style, which can be relationship- or task-oriented; and situation favorableness, based on how strong leader–member relations are, how structured the task is, and how much position power or formal authority the leader holds. Task-oriented leaders are most effective when situation favorableness is either very high or very low; with moderate situation favorableness, relationship-oriented leaders do best. Paul Hersey and Kenneth Blanchard (Reference Hersey and Blanchard1969) proposed their life cycle situational theory of leadership, positing that to be an effective leader, one should match leadership style to the maturity level of the subordinates, where maturity is defined as the ability to perform the job unaided.

Interest in job satisfaction continued, although little evidence was found for a strong relationship between satisfaction and performance. A review by Vroom (Reference Vroom1964) found an average correlation of only 0.14 between those two variables. The Job Descriptive Index (JDI) for measuring job satisfaction was introduced by Smith, Kendall, and Hulin in Reference Hersey and Blanchard1969. In time, the JDI became the “gold standard” for measuring job satisfaction (Balzer, Locke, & Zedeck, Reference Balzer, Locke and Zedeck2008). Lofquist and Dawis (Reference Lofquist and Dawis1969) introduced their theory of worker adjustment, which resulted in the Minnesota Satisfaction Questionnaire.

Other notable 1960s publications on organizational psychology topics include Rensis Likert’s (Reference Likert1961) New Patterns of Management that described his “linking pin” model for small group integration and his four systems of management ranging from the exploitive-authoritarian System 1 to the participative System 4 (Hilgard, Reference Hilgard1987). That same year, Burns and Stalker (Reference Burns and Stalker1961) contrasted mechanistic and organic systems of management; the following year formal and informal components of organizations were described by Blau and Scott (Reference Blau and Scott1962). In 1964, Robert Kahn and his colleagues discussed the effect of role conflict and role ambiguity on individuals in organizations (Kahn, Wolfe, Quinn, Snoek, & Rosenthal, Reference Kahn, Wolfe, Quinn, Snoek and Rosenthal1964), and in 1966, Kahn and Daniel Katz introduced open-systems theory to organizations in The Social Psychology of Organizations (Katz & Kahn, Reference Katz and Kahn1966). One final development worth mentioning from the 1960s was the importation of quality circles (QC) from Japan to other countries. These small volunteer groups that analyze and solve organizational problems were thought to have originated in the United States in the 1950s; however, the modern form is most associated with Japanese professor Kaoru Ishikawa (Salas, Priest, Stagl, Sims, & Burke, Reference Salas, Priest, Stagl, Sims, Burke and Koppes2007).

1970s

Employee discrimination lower court cases triggered by the equal employment opportunity legislation of the 1960s eventually made their way to the US Supreme Court in the 1970s, resulting in landmark decisions that had an impact on personnel selection practices. In Griggs v. Duke Power (1971), for example, the justices ruled that when a selection procedure has differential impact on a protected minority group, that procedure must be job-related; that is, the employer must demonstrate that the procedure is valid. In line with the employer guidelines distributed by the Equal Employment Opportunity Commission (EEOC), the government agency charged with enforcing Title VII of the 1964 Civil Rights Act, job-relatedness came to imply that the employer needed a professional validation study. In Griggs and other Supreme Court decisions, in the basic Title VII paradigm, the initial burden of proof was on the plaintiff, the person or group suing for discrimination, who had to demonstrate that an employer procedure had a differential impact on a protected group (Arvey & Faley, Reference Arvey and Faley1988).Footnote ⁸ The burden of proof then shifted to the defendant, usually the employer, who had to show a sound business reason for using the procedure; that is, that the procedure is valid. In 1975, the APA Division for Industrial-Organizational Psychology issued guidelines for test validation in employee selection: Principles for the Validation and Use of Personnel Selection Procedures. In 1978, four agencies of the federal government – the EEOC, the Civil Service Commission, the Department of Justice, and the Department of Labor – issued their own guidelines – the Uniform Guidelines on Employee Selection Procedures – where selection is viewed broadly enough to include virtually any employer decision (e.g., you can be selected for a job, a raise, or a promotion or for termination).

Federal legislation and subsequent court decisions drove a fair amount of selection research in the 1970s on statistical models of test fairness and on topics such as differential validity, where tests demonstrate different validity coefficients for different groups (Bray & Moses, Reference Bray, Moses, Mussen and Rosenzweig1972). One related line of research had far-reaching ramifications for selection research. Validity coefficients were long held to be mostly situation specific; that is, specific to the time, the organization, and/or the group used when they were initially determined. Predictors therefore had to be validated for each new situation. Contrary to that belief, the validity generalization (VG) research of Frank Schmidt and John Hunter (Reference Schmidt and Hunter1977) demonstrated that, at least for cognitive ability tests, much or in some cases virtually all of the variability in validity coefficients across groups and situations was not due to substantive factors but rather was due to statistical artifacts such as sampling error, unreliability of measures, and restriction of range.

Other notable developments in test validation and employee selection included the following. There was increased interest in job analysis (Ash & Kroeker, Reference Ash, Kroeker, Rosenzweig and Porter1975), beginning with Prien and Ronan’s (Reference Prien and Ronan1971) review article and highlighted by the first comprehensive book on job analysis by McCormick (Reference McCormick1979). Functional Job Analysis was developed (Fine & Wiley, Reference Fine and Wiley1971), as was the Position Analysis Questionnaire (PAQ) (McCormick, Jeanneret, & Mecham, Reference McCormick, Jeanneret and Mecham1972), which was destined to be one of the most researched job analysis inventories (Wilson, Reference Wilson and Koppes2007). Notable work on utility analysis included Schmidt, Hunter, McKenzie, and Muldrow’s (Reference Schmidt, Hunter, McKenzie and Muldrow1979) method for estimating the standard deviation of job performance in a dollar metric, which simplified the use of what was now known as the Brogden-Cronbach-Gleser utility models. The use of assessment centers increased dramatically during the 1970s (Ash & Kroeker, Reference Ash, Kroeker, Rosenzweig and Porter1975). A theoretical foundation for biographical data was proposed by Owens (Reference Owens and Dunnette1976). And the traditional validity model continued to undergo modification, which was viewed by Tenopyr and Oeltjen (Reference Tenopyr, Oeltjen, Rosenzweig and Porter1982) as one of the most important developments in selection research in recent years. Robert Guion (Reference Guion and Dunnette1976) emphasized that validity is a unitary concept. The traditional trio of criterion-related, content, and construct validation represent three different strategies for determining the meaningfulness, usefulness, and appropriateness of test score inferences, not three different types of validity. While much of the work in selection over the previous decades had been centered in the United States, beginning in the mid-1970s, Europe underwent a resurgence in employee selection research and practice (Salgado, Anderson, & Hülsheger, Reference Salgado, Anderson, Hülsheger, Farr and Tippins2010).

Campbell (Reference Campbell1971) noted that theory and research in training was substandard at the beginning of the decade. By the mid-1970s, things had improved, as illustrated by Goldstein’s (Reference Goldstein1974) instruction systems design (ISD) model, a systems approach to the training process. Performance appraisal formats continued to build on the behavioral rating approach of Smith and Kendall (Reference Smith and Kendall1963) with variants such as the Mixed Standard Scale (Blanz & Ghiselli, Reference Blanz and Ghiselli1972) which included a check for logical errors in rater judgment and the Behavior Observation Scale (Latham & Wexley, Reference Latham and Wexley1977), in which the rater assessed actual occurrences of behavior rather than comparing the ratee against a hypothetical anchor.

On the organizational side, additional contingency theories of leadership were proposed. Examples include Path-Goal Theory, in which the effectiveness of four leadership styles was viewed as dependent both on the situation and on subordinate maturity levels (House, Reference House1971); and vertical dyad linkage theory, which postulated that leaders respond differently to subordinates depending on whether those employees are perceived as in-group or out-group members (Graen, Reference Graen and Dunnette1976). Toward the end of the decade, transformational leaders who inspire their followers were contrasted with transactional leaders whose leadership is based on rules and social contracts (Burns, Reference Burns1978). Hackman and Oldman (Reference Hackman and Oldman1975, Reference Hackman and Oldman1976, Reference Hackman and Oldham1980) introduced their job enrichment theory, postulating that increasing the interest level, autonomy, responsibility level, and overall significance of a job would increase both motivation and job satisfaction. Regarding group processes, Irving Janis (Reference Janis1972) described groupthink, the situation where concurrence-seeking becomes the dominant mode of thinking in the group and conformity becomes the norm, resulting in poor-quality group decisions. In motivation theory, Albert Bandura (Reference Bandura1977) introduced social learning theory, later renamed social cognitive theory. Bandura viewed behavior as a continuous reciprocal interaction among behavioral, cognitive, and environmental variables (Latham & Budworth, Reference Latham, Budworth and Koppes2007).

In other developments, the first edition of the Handbook of Industrial and Organizational Psychology (Reference Dunnette1976), edited by Marvin Dunnette, was published. By the 1970s, in the United States, the venerable term industrial psychology became industrial-organizational psychology (I/O or I-O psychology). In Great Britain, the National Institute of Industrial Psychology (NIIP), an important center for industrial psychology since its founding in 1921, closed due to financial difficulties in 1973. The NIIP’s journal Occupational Psychology was taken over by the British Psychological Society, was renamed the Journal of Occupational Psychology in 1975, and became international in scope. In China, psychology endured years of neglect during Mao’s Cultural Revolution from 1966 to 1976. Following Mao Zedong’s death in 1976, China became more open to outside ideas, and in 1978, the Chinese Psychological Association created a national committee of industrial psychology consisting of engineering psychology and organizational psychology branches (Warr, Reference Warr and Koppes2007).

1980s

In the United States, John Hunter and Frank Schmidt extended their validity generalization work into an innovative and influential meta-analysis procedure (Hunter, Schmidt, & Jackson, Reference Hunter, Schmidt and Jackson1982).Footnote ⁹ By eliminating the statistical “noise” across research studies, researchers could make sense of often contradictory research results among the studies. They could then estimate population effect sizes and draw strong conclusions if variability across study results could be substantially accounted for by statistical artifacts. Meta-analysis, particularly the Hunter-Schmidt version, rapidly became the standard procedure in I-O psychology for summarizing results across individual studies; that is, for conducting a quantitative literature review. For example, meta-analyses conducted in the 1980s found cognitive ability tests have substantial validity across occupations (e.g., Hunter & Hunter, Reference Hunter and Hunter1984).

Biodata inventories, although underused, continued to show substantial validity (Schmidt, Ones, & Hunter, Reference Schmidt, Ones, Hunter, Rosenzweig and Porter1992). The use of personality tests, long maligned for their lack of validity, underwent a resurgence due to the increased acceptance of the Five-Factor Model of personality (Digman, Reference Digman, Rosenzweig and Porter1990). It appeared that much of the low validity attributed to personality tests in the past was due to confusion over multiple trait names and inconsistency in defining these traits. There was an upswing of interest in the use of integrity tests for identifying dishonest applicants (Schmidt, Ones, & Hunter, Reference Schmidt, Ones, Hunter, Rosenzweig and Porter1992). On the criterion end of validation, the decades-long trend away from output measures and toward behavioral measures continued (Austin & Villanova, Reference Austin and Villanova1992; see Campbell, Reference Campbell, Dunnette and Hough1990a, for an example). A highlight of the 1980s was the seven-year Project A: Army Selection and Classification Project conducted by the US Army, one of the largest selection studies ever attempted. Project A researchers evaluated multiple predictors and job performance constructs and developed a multidimensional model of job performance (Campbell, Reference Campbell1990b; Campbell & Knapp, Reference Campbell, Knapp, Farr and Tippins2010; Borman, Klimoski, & Ilgen, Reference Borman, Klimoski, Ilgen, Borman, Ilgen and Klimoski2003).

In Europe, despite little or no evidence for its validity, the use of graphology as a selection procedure continued to be popular (Guion & Gibson, Reference Guion, Gibson, Rosenzweig and Porter1988). In Canada, Wiesner and Cronshaw’s (Reference Wiesner and Cronshaw1988) meta-analysis demonstrated the superiority of structured over unstructured interviews. In performance appraisal, Landy and Farr (Reference Landy and Farr1980) called for a moratorium on research that focused on formats and a shift toward understanding the cognitive processes that underlie the appraisal process. In the 1980s, that shift began to occur, along with a recognition of the importance of the social context that surrounds the performance appraisal process. There was also interest in performance ratings from multiple sources, sometimes known as 360-degree feedback (Farr & Levy, Reference Farr, Levy and Koppes2007). In job analysis, Gael (Reference Gael1988) published the first handbook of job analysis (Wilson, Reference Wilson and Koppes2007).

The concept of organizational culture became popular. Edgar Schein’s (Reference Schein1985) Organizational Culture and Leadership provided an overview. In work motivation, Greenberg (Reference Greenberg1987) extended equity theory by addressing workplace fairness and trust with his organizational justice theory (Latham & Budworth, Reference Latham, Budworth and Koppes2007). Work on group processes and work teams included examination of how teams form, evolve, and perform; work on the construction of team taxonomies and classifications; and research on the measurement of team performance (Salas, Priest, Stagl, Sims, & Burke, Reference Salas, Priest, Stagl, Sims, Burke and Koppes2007).

1990s and Beyond

As we approach the present day, it becomes difficult to summarize major events in the field in a few paragraphs. The continuing expansion of I-O psychology, the increased specialization by I-O psychologists, and lack of sufficient time to gain historical perspective make trying to determine what contributions will stand the test of time more of an exercise in prediction than history. With that caveat in mind, I will mention a few developments that appear worthy of interest. The US Congress passed the Americans with Disabilities Act of 1990, which codified requirements to provide reasonable accommodations for individuals with physical or mental disabilities, including accommodations for selection of employees and in other job-related areas. Meta-analyses estimated the average size of validity coefficients for various predictors, and the procedure was also offered as a component of an alternative research model that avoids the perceived pitfalls of the traditional null hypothesis testing model (see Schmidt, Reference Schmidt1992). Barrick and Mount’s (Reference Barrick and Mount1991) meta-analytic review of the Big Five personality traits found substantial evidence for the validity of the trait conscientiousness across occupations. In a review of meta-analysis research that spanned eighty-five years, Frank Schmidt and John Hunter (Reference Schmidt and Hunter1998) found general cognitive ability, work samples, peer ratings, the personality trait conscientiousness, integrity tests, and interviews to have substantial validity for selection, with structured interviews outperforming unstructured interviews. Poor predictors included graphology, educational level, interest inventories, and age of the applicant. There was an increase in research that evaluated selection from the perspective of the applicant (Borman, Hanson, & Hedge, Reference Borman, Hanson, Hedge, Spence, Darley and Foss1997). The US Army’s Project A continued to have an influence; for example, Campbell, McCloy, Oppler, and Sager (Reference Campbell, McCloy, Oppler, Sager, Schmitt and Borman1993) developed a construct-based theory of job performance based on that research.

The use of computers and the internet was beginning to affect practice in the 1990s. For example, the paper version of the US Department of Labor’s Dictionary of Occupational Titles (DOT) was revised and expanded into an internet version, the occupational information network (O*Net) (Wilson, Reference Wilson and Koppes2007). In training, technology-delivered instruction (TDI) methods such as web-based instruction and distance learning increased in popularity (Kraiger & Ford, Reference Kraiger, Ford and Koppes2007). The international nature of organizations combined with new technology has generated interest in virtual teams, although evidence for their effectiveness was not yet in evidence (Salas, Priest, Stagl, Sims, & Burke, Reference Salas, Priest, Stagl, Sims, Burke and Koppes2007).

In work motivation, Latham and Budworth (Reference Latham, Budworth and Koppes2007) noted the continued endurance of equity and of goal-setting theories of motivation, with the latter garnering considerable evidence for their validity and usefulness. In the 1990s, several trends in leadership research were noted, including interest in transformational leadership and neo-charismatic leadership. Of note is the Global Leadership and Organizational Behavior Effectiveness (GLOBE) study. This study attempted to identify culturally specific and universal leadership behaviors across sixty countries, determining that there were some attributes endorsed by middle managers across countries (Chhokar, Brodbeck, & House, Reference Chhokar, Brodbeck and House2007; Lowe & Gardner, Reference Lowe and Gardner2000).

I-O psychology continued to grow more international in outlook. Many I-O consulting firms had become multinational, with offices spread throughout the globe. In the 1990s, new journals were established with an explicit international perspective, such as the International Journal of Selection and Assessment, founded in 1993, and the International Journal of Training and Development, founded in 1997 (Warr, Reference Warr and Koppes2007). The number of international members in the Society for Industrial and Organizational Psychology (SIOP) was increasing, and SIOP collaborated with both the International Association of Applied Psychology and the European Association of Work and Organizational Psychology to form the Alliance of Organizational Psychology, whose goals include advancing the science and practice of I-O psychology and improving the quality of working life (Peiró, 2009, cited in Koppes Bryan & Vinchur, Reference Koppes Bryan, Vinchur and Kozlowski2012). When the second edition of the Dunnette’s (Reference Dunnette1976) Handbook of Industrial and Organizational Psychology expanded to four volumes (Dunnette & Hough, Reference Dunnette and Hough1990–92; Triandis, Dunnette, & Hough, Reference Triandis, Dunnette and Hough1994), the subject of the fourth volume was cross-cultural I-O psychology.

I-O psychology has entered both the twenty-first century and its second century of existence. It is currently a vital, expanding field with a future that appears bright. I-O psychologists hold academic positions in psychology departments and business schools, they maintain consulting firms, and they are directly employed in business and government. While it is impossible to predict what the future will bring for I-O psychology, it is a safe bet that the same combination of internal and external forces that shaped the field during the past century will continue to do so. Changes in the social, cultural, and economic environments that I-O psychology is embedded within will have an influence. Technological changes and changes in the nature of work – for example, the increasing sophistication of robotics and artificial intelligence – will force corresponding changes in research and practice. Advances in psychology and neuroscience in understanding behavior and the nervous system should lead to advances in our understanding of work behavior.

There will be no shortages of challenges. How will I-O psychologists meet them? The basic scientist-practitioner model has served the field well over the past century. It should do the same in the future. Reliance on empirical evidence, genuine concern for worker welfare as well as the bottom line, and adherence to high ethical and professional standards will ideally guide future efforts.