Showing papers in "American Psychologist in 1978"

The self system in reciprocal determinism.

Abstract: Explanations of human behavior have generally favored unidirectional causal models emphasizing either environmental or internal determinants of behavior. In social learning theory, causal processes are conceptualized in terms of reciprocal determinism. Viewed from this perspective, psychological functioning involves a continuous reciprocal interaction between behavioral, cognitive, and environmental influences. The major controversies between unidirectional and reciprocal models of human behavior center on the issue of self influences. A self system within the framework of social learning theory comprises cognitive structures and subjunctions for perceiving, evaluating, and regulating behavior, not a psychic agent that controls action. The influential role of the self system in reciprocal determinism is documented through a reciprocal analysis of self-regulatory processes. Reciprocal determinism is proposed as a basic analytic principle for analyzing psychosocial phenomena at the level of intrapersonal development, interpersonal transactions, and interactive functioning of organizational and social systems. Recent years have witnessed a heightened interest in the basic conceptions of human nature underlying different psychological theories. This interest stems in part from growing recognition of how such conceptions delimit research to selected processes and are in turn shaped by findings of paradigms embodying the particular view. As psychological knowledge is converted to behavioral technologies, the models of human behavior on which research is premised have important social as well as theoretical implications (Bandura, 1974). Explanations of human behavior have generally been couched in terms of a limited set of determinants, usually portrayed as operating in a unidirectional manner. Exponents of environmental determinism study and theorize about how behavior is controlled by situational influences. Those favoring personal determinism seek the causes of human behavior in dispositional sources in the form of instincts, drives, traits, and other motivational forces within the individual. Interactionists attempt to accommodate both situational 344 • APRIL 1978 • AMERICAN PSYCHOLOGIST Copyright 1978 by the American Psychological Association, Inc. 0003-066X/78/3304-0344$00.7S and dispositional factors, but within an essentially unidirectional view of behavioral processes. The present article analyzes the various causal models and the role of self influences in behavior from the perspective of reciprocal determinism. Unidirectional environmental determinism is carried to its extreme in the more radical forms of behaviorism. It is not that the interdependence of personal and environmental influences is never acknowledged by advocates of this point of view. Indeed, Skinner (1971) has often commented on the capacity for countercontrol. However, the notion of countercontrol portrays the environment as the instigating force to which individuals can counteract. As will be shown later, people create and activate environments as well as rebut them. A further conceptual problem is that having been acknowledged, the reality of reciprocal interdependence is negated and the preeminent control of behavior by the environment is repeatedly reasserted (e.g., \"A person does not act upon the world, the world acts upon him,\" Skinner, 1971, p. 211). The environment thus becomes an autonomous force that automatically shapes, orchestrates, and controls behavior. Whatever allusions are made to two-way processes, environmental rule clearly emerges as the reigning metaphor in the operant view of reality. There exists no shortage of advocates of alternative theories emphasizing the personal determination of environments. Humanists and existentialists, who stress the human capacity for conscious judgment and intentional action, contend that individuals determine what they become by their own free choices. Most psychologists find conceptions of human behavior in terms of unidirectional personal determinism as unsatisfying as those espousing unidirectional environmental determinism. Preparation of this article was facilitated by Public Health Research Grant M-S162 from the National Institute of Mental Health and by the James McKeen Cattell Award. Requests for reprints should be sent to Albert Bandura, Department of Psychology, Stanford University, Stanford, California 9430S. To contend that mind creates reality fails to acknowledge that environmental influences partly determine what people attend to, perceive, and think. To contend further that the methods of natural science are incapable of dealing with personal determinants of behavior does not enlist many supporters from the ranks of those who are moved more by empirical evidence than by philosophic discourse. Social learning theory (Bandura, 1974, 1977b) analyzes behavior in terms of reciprocal determinism. The term determinism is used here to signify the production of effects by events, rather than in the doctrinal sense that actions are completely determined by a prior sequence of causes independent of the individual. Because of the complexity of interacting factors, events produce effects probabilistically rather than inevitably. In their transactions with the environment, people are not simply reactors to external stimulation. Most external influences affect behavior through intermediary cognitive processes. Cognitive factors partly determine which external events will be observed, how they will be perceived, whether they have any lasting effects, what valence and efficacy they have, and how the information they convey will be organized for future use. The extraordinary capacity of humans to use symbols enables them to engage in reflective thought, to create, and to plan foresightful courses of action in thought rather than having to perform possible options and suffer the consequences of thoughtless action. By altering their immediate environment, by creating cognitive self-inducements, and by arranging conditional incentives for themselves, people can exercise some influence over their own behavior. An act therefore includes among its determinants self-produced influences. It is true that behavior is influenced by the environment, but the environment is partly of a person's own making. By their actions, people play a role in creating the social milieu and other circumstances that arise in their daily transactions. Thus, from the social learning perspective, psychological functioning involves a continuous reciprocal interaction between behavioral, cognitive, and environmental influences. Reciprocal Determinism and Interactionism Over the years the locus of the causes of behavior has been debated in personality and social psychology in terms of dispositional and situational UNIDIRECTIONAL

A thirty-year follow-up of treatment effects.

Abstract: VioLit summary: OBJECTIVE: This study by McCord was designed to follow up the subjects who had participated in the Cambridge-Somerville Youth study after 1942. Criminal behavior, health, lifestyle, and beliefs/attitudes were primary subjects of inquiry. METHODOLOGY: The original study was a quasi-experimental design which followed 253 problem boys and 253 matched controls who were selected by various public sources and identified as "difficult" or "average." Boys were paired by age, delinquency-prone histories, family background, and home environments and then randomly assigned to the experimental and the control group. Treatment began in 1939 when the subjects were 5-13 (median, 10.5) and continued for an average of 5 years (except those dropped from the program during 1941 because of counselor shortage). Counselors visited the home an average of twice a month and encouraged families to call on the program for assistance. This assistance included family counseling, tutoring, medical and/or psychiatric attention, summer camps, and connection with Boy Scouts, YMCA, or other community programs. The control group only participated insofar as they gave information about themselves. Each group included boys from "difficult" and "average" groups. The study this author conducted was a follow-up, thirty years later, on the men who had participated in the original study. Official records (court records, mental hospital records, statistics from alcohol treatment centers, and vital statistics) and personal contacts were used to obtain information about the long-term effects of the Cambridge-Somerville Youth Study. The follow-up took place in 1975 and 1976. The men were located using telephone calls, city directories, motor-vehicle registration, marriage and death records, and researcher intuition. 480 (95%) of the men were located; 48 (9%) of these had died and 340 (79%) were living in Massachusetts. Questionnaires were mailed to 208 men from the treatment group and 202 men from the control group. The items in the questionnaire covered the topics of family, occupation, drinking, health, and attitudes. Former members of the treatment group were asked how (if at all) the treatment program had been useful to them. Responses to the questionnaire were received from 113 men in the treatment group (54%) and 122 men in the control group (60%). Men who lived outside of Massachusetts were significantly (p= FINDINGS/DISCUSSION: The treatment and control groups were compared on a variety of measures for criminal behavior. With the exception of Crime Prevention Bureau records for unofficial crimes committed by juveniles, court convictions served as the standard by which criminal behavior was assessed. Almost equal numbers of each groups had committed crimes as juveniles on both official and unofficial measures. On dimensions of "difficult" and "average" distinctions, there were also no differences. As adults, equal numbers (168) had been convicted for some crime. From the treatment group, 119 committed minor crimes, and 49 had committed serious crimes against property or person as adults. For the control group, 126 had committed minor crimes, and 42 had committed serious crimes. 29 of the treatment group and 25 from the control group had committed serious crimes after age 25. When comparing adult criminal records controlling for juvenile records, there was no evidence that the treatment program had any effect. There were no differences in number of serious crimes committed, age at which the first crime was committed, age at which first serious crime was committed, or age after which no serious crime was committed. Chi-square revealed that a significantly higher proportion of criminals from the treatment group committed more than one crime. 78% of the treatment group committed at least two crimes versus 67% of the control group who had. Dimensions of health that were evaluated were alcoholism, stress-related diseases, and early death. There were no differences in mental hospitalization/alcohol treatment between the groups. A significantly higher proportion of the treatment group mentioned that they were alcoholic or were judged by the CAGE test be alcoholic when compared to the control group (17% versus 7%). Of those men who had received treatment in mental hospitals for non-alcoholic problems, a significant majority of the men who had been in the treatment group were diagnosed with more serious diagnoses (71%) whereas the control group was diagnosed with less serious diagnoses (67%). A significantly higher proportion of the men from the treatment group reported stress-related diseases than the control group, particularly for heart trouble. In comparisons of family and work, the only significant difference was in the number who were white collar professionals (43% for the control group, 29% for the treatment group) and, in general, the control group had a significantly higher level of prestige in their jobs. Among blue-collar workers, a significantly greater number of the treatment group expressed dissatisfaction (p=.02). There were no significant differences in authoritarianism, political orientation, or identification of the best periods of their lives. The majority of the men (2/3) who had been in the treatment program reported that it had been helpful to them, and some responded that the program had helped keep them out of trouble and made them more law-abiding. In general, the author concluded that the program did not, according to objective measures, improve the lives of the treatment group. AUTHOR'S RECOMMENDATIONS: The author suggested a cautious stance to intervention programs, though it is also suggested that new programs should be developed with attention to the problems of potential damage through the use of pilot projects with mandatory evaluations. EVALUATION: This study serves as an interesting follow-up to one of the best known experiments in delinquency prevention. It is surprising to see the negative effects the author found in the treatment group thirty years after being in the program. This study, however, cannot make a strong empirical claim that the program did not work. First, the initial selection and identification of youth casts doubts on the original methodology. Additionally, the present study relies upon mailed questionnaires. It has been shown that there is a likely self-selection bias which makes findings difficult to generalize to the population which confounds the usual difficulties with mailed questionnaires. Face-to-face interviews or even telephone interviews would add some credibility. Thirdly, the data collection could have been tightened up (for example, using FBI rap sheets as an additional measure of criminal activity). Finally, the data analysis techniques do not provide for an adequate level of control to establish the causal linkages which one would want to see. Additional analyses using regression techniques would have been helpful. This study was published in 1978, and methodology has improved since then. It would be insightful to do another follow-up or a reanalysis of the original data to see if the lack of effect or opposite effects of treatment hold. (CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado) KW - Massachusetts KW - Geographic Location KW - Long-Term KW - Criminal Behavior KW - Delinquency KW - Diagnosis KW - Follow-Up Studies KW - Problem Behavior KW - Life Course KW - Prevention KW - Nutrition-Health KW - Ideology KW - Perceptions KW - Lifestyle KW - Juvenile Offender KW - Male Offender KW - Adult Offender

IQ, Social Competence, and Evaluation of Early Childhood Intervention Programs.

Abstract: The IQ score has been without question the most often utilized outcome measure in evaluations of early childhood intervention programs. Reasons for the popularity of the IQ as an assessment tool are discussed, and problems raised by employing the IQ in this manner are noted. The importance of accurate outcome evaluation of programs with clearly defined goals is related to both the social science and policy- making arenas. The authors argue that social com- petence, rather than IQ, should be the primary mea- sure of the success of intervention efforts. Difficulties in defining and assessing social competence are dis- cussed. An index of social competence is suggested that includes measures of physical health, IQ, school achievement, certain motivational and emotional vari- ables, and such molar social expectancy variables as school attendance and incidence of juvenile delin- quency.

Productivity and scholarly impact (citations) of British, Canadian, and U.S. departments of psychology (1975).

Abstract: The psychology departments at 180 universities in Canada, the United Kingdom, and the United States were evaluated in terms of their productivity and the impact of their scholarly research. The 1975 Social Science Citation Index (SSCI) was used as the basis for counting both citations of, and publications by, each of the 5,597 faculty members. Psychologists at the 180 evaluated departments received a total of 76,189 citations in 1975 (M —13.6) and produced a total of 4,977 publications (M = .89). Data are presented for the top 100 departments. These include the total, mean, and median number of citations of each department as well as the total and mean number of publications. The citation measures correlated more highly with reputational ranks taken on 76 American schools in an earlier study than did the publication measures. The departments of psychology at four universities were consistently in the top eight on the total, mean, and median citation measures of scholarly impact: Stanford University, Harvard University, Yale University, and the University of Pennsylvania. Eighteen of the top 100 departments in terms of total citations were from either Canada or the United Kingdom. Methodological difficulties in using the SSCI and possible uses and limitations of citation counts are discussed. Psychology departments can be evaluated on a number of dimensions: scientific contributions (research productivity and impact), teaching excellence, applied contributions, and contributions to the community. In the present article we are concerned with providing objective measures of the relative scientific contributions of 180 graduate psychology departments in Canada, the United Kingdom, and the United States. Several attempts have been made in the past to compare psychology departments in the United States using ratings. In one study (Keniston, 1959), chairpersons in 25 leading universities rank 1064 • DECEMBER 1978 • AMERICAN PSYCHOLOGIST Copyright 1978 by the American Psychological Association, Inc. 0003-066X/78/3312-1064$00.75 ordered the 15 \"strongest\" departments in their fields. For psychology, the first few, in order, were Harvard University, the University of Michigan, Yale University, the University of California, Berkeley, Stanford University, the University of Minnesota, and the University of Illinois. In 1966, the American Council on Education conducted an extensive survey, gathering data from 4,000 faculty members in 30 disciplines at 106 major institutions (Cartter, 1966). The top departments of psychology from this study were Stanford University and Harvard University (tied for first place); Yale University, the University of Michigan, and the University of California, Berkeley (tied for third place); then the University of Illinois, the University of Wisconsin, and the University of Minnesota. Overall, the results were similar to those of the previous study. In 1970, the American Council on Education updated their 1966 study, rank ordering the top 32 institutions for psychology and providing data on another 44 (Roose & Andersen, 1970). The A preliminary version of this article was presented at the 85th Annual Meeting of the American Psychological Association, San Francisco, August 30, 1977. The research was supported by minor research grants to the three authors from their respective universities and by Social Science and Humanities Research Council of Canada Grant No. 410-78-0108 to J. Philippe Rushton. We wish to thank Brenda Hart, Doris Henschel, Karen Lungmus, Sari Meltzer, Marilyn Okada, and Susan Wareing for their invaluable help with the collection of the data and Jean Edwards for both her help with the data and her comments about the manuscript. We would also like to thank Brenda Hart and Lynne Mitchell for their patient forbearance in typing the manuscript. The third author was on leave at the University of Toronto when this research was conducted. Requests for reprints should be sent to Norman S. Endler, Department of Psychology, York University, Toronto, Ontario, Canada M3J 1P3. top few in this listing were, in order, Stanford University, the University of Michigan, the University of California, Berkeley, Harvard University, the University of Illinois, the University of Pennsylvania, the University of Minnesota, the University of Wisconsin, and Yale University. Once again, we note that the same names keep cropping up, at least at the top of the list. The use of ratings in evaluations does have some value but also may present certain problems. One drawback to these analyses is that they simply survey subjective opinions. Raters may be more familiar with some programs than others, and this may result in one source of bias. Another may arise from halo effects, where the evaluation of a particular department may be influenced by the prestige of the school. For example, we discovered in the course of the present study that the psychology department at one school (California Institute of Technology) that was ranked in the top 48 by Roose and Andersen (1970) actually employed only one full-time psychologist in 1977 (Breger, Note 1). Several other psychologists are employed in other departments but apparently are not even cross-appointed with the psychology department. Presumably, the psychology department there received its high rating because the school is excellent in other areas and possibly because the raters assumed that wellknown psychologists located there were associated with the psychology department. Insufficient knowledge of numerous programs and various other sources of bias may limit the usefulness of ratings as 'a reflection of scholarly impact. One recent study (Cox & Catt, 1977) attempted to avoid the problem of rating procedures by using the more objective method of assessing the productivity of United States psychology departments. This was accomplished by counting the number of journal articles appearing in 13 journals of the American Psychological Association that could be attributed to psychology departments. The top five schools based on total publications over a 6-year period (1970-1975) were the University of Wisconsin, the University of Illinois, Yale University, the University of Michigan, and Ohio State University. Cox and Catt (1977) found that their productivity index based on total publications correlated only .35 (n = 85, p < .001) with the Roose and Andersen (1970) rank orderings based on ratings. In addition, Cox and Catt's total-publications measure correlated .38 with another rank ordering they produced taking faculty size into account. This mean-publicationper-faculty measure correlated, in turn, .21 (p < .005) with the Roose and Andersen (1970) listing. Thus, while these three rank orderings had some degree of commonality, they were to a rather great extent independent. This lack of commonality may in part be due to the several limitations of the Cox and Catt (1977) study. A first problem deals with the representativeness of the journals that were sampled. As Cox and Catt point out, there are many other important psychological journals than those published by the APA. More importantly, simply measuring the sheer number of publications attributable to particular departments says nothing about their quality or impact. Some publications obviously have much more impact than others, and many books and book chapters that have great impact were obviously not included. In addition, Cox and Catt (1977) made no attempt to rank psychology departments outside of those in the United States. The present study was primarily devoted to measuring the impact of psychology departments based on the number of citations accruing to the individuals in those departments in the 1975 Social Science Citation Index (SSCI). The SSCI was also used to gain an estimate of the productivity of psychology departments. Previous studies have measured the impact and productivity of Canadian psychology departments (Buss, 1976; Endler, 1977) and those of the United Kingdom (Rushton & Endler, 1977). In the present study we evaluated a large number of U.S. departments, and we present data on the top 100 departments in Canada, the United Kingdom, and the United States in terms of both their impact and productivity as derived from the 1975 SSCI. The Science Citation Index (SCI) and the more recently established Social Science Citation Index (SSCI) serve as a valuable data base for the relatively objective assessment of the productivity and scholarly impact of psychology departments and psychologists. The SCI indexes articles from \"hard science\" journals and selectively indexes articles from social science journals. At least 90 psychology journals are fully indexed (see Rushton & Roediger, 1978, for a partial listing). The SSCI completely indexes articles from over 1,400 journals representing virtually every discipline in the social sciences and selectively covers another 1,200 journals representing the natural and physical sciences. At least 180 psychology journals are indexed. Therefore, the SSCI is more compreAMERICAN PSYCHOLOGIST • DECEMBER 1978 • 1065 hensive than the SCI for psychology. Both citation indexes are organized such that one can look up a particular person for a particular year and count the number of times that the person was cited in that year in articles in the journals covered by the index. Thus, citations that appear in books, book chapters, etc., are not counted. However, citations in journal articles of books, chapters, and convention papers do appear in the citation indexes. Thus it is possible to measure the impact of a person's (or department's) scholarly activity by counting the number of times that the person's work is cited. Data on the reliability and validity of citation counts as a measure of impact for psychology have been provided by Myers (1970). In regard to validity, for example, the total number of citations an individual earns has been found to predict such different

Parental loss and genius.

Abstract: A scientific theory is proposed to account for the historically eminent individual or genius by relating his or her development to loss of parents. A parental loss profile is described that rigorously defines orphanhood, and a study of 699 eminent persons that makes use of this profile is reported. Early orphanhood was found to be characteristic of this eminent group. Comparisons with previous work were attempted despite obvious methodological problems. Theoretical considerations indicating the effects of bereavement and orphanhood are offered to explain the relationship between achievement and parental loss as well as that between the genius and the disturbed psychotic. Renewed interest has recently been shown in the study of genius (Albert, 1969, 1971, 1975; Besdine, 1968a, 1968b; Sorell, 1970). Many have tried to explain the development of those who mold civilization. There are leads and there are worthy thoughts on the subject, but few actual facts. Genius was described initially as an act of creativity on the part of the Supreme Creator and until very recently in history was the subject of religious speculation. Beginning in the 1870s, however, scientists attempted to analyze the operational components of genius. Galton (1869/1962) believed that the faculty of genius was transmitted through hereditary principles. Lombroso (1891) believed in a theory of genius that he based on his work as a psychiatrist. He had observed at close range the many forms of mental deterioration, extreme behavioral manifestations, and emotional disturbances of patients in large institutions for the mentally ill. He believed all forms of genius were the result of psychoses and moral degeneracy, and he offered great numbers of cases to prove his point. There have been many examples of actual insanity among the famous, yet Ellis (1904/1926) reported that mental illness is not found among the famous in anywhere near the proportions which Lombroso stated it would be. A great step forward in the study of the genesis of genius was made by Wilhelm Lange-Eichbaum (1928/1956,1932). He explained that psychosis does occur in the lives of many geniuses and that even when psychosis is not found, markedly psychopathic traits can be found in a great majority of the eminent. Aside from (a) Galton and his theory of heredity, (b) Lombroso's degenerative-psychosis hypothesis with its modification by Lange-Eichbaum, and (c) the sociological school that cataloged the characteristics of genius (e.g., Bowerman, 1947; Cattell, 1903; Cox, 1926/ 1959; Ellis, 1904/1926; Goertzel, 1962; Illingworth, 1966; Kenmare, 1960), all three of which are acknowledged to be grossly inadequate theories, there is no theoretical position that can explain the phenomenon of eminence or creative genius, and there are no facts to support any generalized theory. In other words, there is as yet no scientific theory to account for the development of a historically eminent individual. The present study attempts a new viewpoint in discussing genius and its origins by relating creative thinking, historical eminence, administrative prowess, and scientific acumen to the variable of loss of the parents by death. The study was an outgrowth of previous work in the area of creativity (Eisenstadt, 1966). Genius is defined here as the development of an individual to a high degree of competency and superiority in an pccupational field. This is postulated to be due to several factors, including (a) a certain degree of innate, biologically determined characteristics— principally intelligence, physical abilities, and the like; (b) individual development of those capacities by a unique and specific psychological mechanism of interaction within the family unit; and (c) training and educational advancement leading to (d) accomplishment. The unique and specific psychological mechanism focused upon in this study is the bereavement experience and its resolution or, more generally, the problem of orphanhood. Requests for reprints should be sent to J. Marvin Eisenstadt, 352 South Oyster Bay Road, Syosset, New York 11791. AMERICAN PSYCHOLOGIST • MARCH 1978 • 211 Copyright 1978 by the American Psychological Association, Inc. 0003-066X/78/3303-0211$00.75 The essential element in orphanhood that uniquely describes it is that no possibility exists for a return to a former family situation. Once a parent dies, whether father or mother, the family unit is permanently altered. A curious fact of the English language is that the word orphan is an inexact term. According to the dictionary definition, an orphan is someone who has lost either one parent or both parents. In this study, orphanhood is denned in three aspects: paternal orphanhood—the loss by death of the father; maternal orphanhood— the loss by death of the mother; and full, total, or double orphanhood—the loss by death of both mother and father. I developed the concept of the parental-loss profile to rigorously define the orphanhood situation of any individual. Thus, Sigmund Freud's profile reads F40, M74, meaning that Freud was 40 years old when his father died and 74 years old when his mother died. Charlotte Bronte's profile is F after, MS, S38, which states that her father was still alive when she died at age 38 and that her mother died when she was 5 years old. In the present study, parental loss by death was the main consideration. Eliminated for the sake of research strategy were sibling loss, the loss of children and its effects on parents, and other loss events including separations, divorces, hospitalizations, mental illness of parents, etc. It seemed expedient from a research point of view to study the most basic form of parental loss—actual loss by death of the parent, or orphanhood. First, when a parental death is studied, it is easier to determine the actual point in time of the loss. Second, the effects should be more prominent and more easily noticed than those of other forms of loss. Third, the information to be obtained is more readily available. What is the specific relationship between the loss of parents by death and the desire for fame, eminence, and occupational excellence? Certainly one of the important considerations is the nature of the family unit prior to the disruption caused by the death of the parent. The individual whose parents provided defective care and a disturbed family background would be affected quite differently by the death of a parent than the individual with a healthy family background whose parents showed genuine concern. It has already been remarked in the developing parental-loss literature that various facets comprise the crisis of bereavement. Such factors as the age at which the death takes place, the composition of the household at the time of death, the previous psychological and economic relationships that have existed before the loss, and the capacity of the family members to absorb the crisis have been mentioned as contributory factors to the traumatic nature of orphanhood. Thus, parental loss is conceived in two ways: (a) Parental loss by death has a direct result, and depending on the age of the child, this result can be specified, and (b) parental loss by death has an indirect result depending on the family dynamics existing before the death occurred. Researching Genius: The Study Group and Parental-Loss-Profile Results The study of eminence and the criteria used to define the eminent has a well-developed history and can be dated for our purposes as beginning with Sir Francis Galton (1869/1962). The selection of eminent individuals was personally decided upon by him, although he was guided in his choice of judges, statesmen, scientists, poets, and artists by standard reference works available at the time. Galton later selected Fellows of the Royal Society who had won medals for scientific work, had been president of a learned society, had attained membership on the counsel of the society, or were professors at important universities. Havelock Ellis (1904/1926) used the 66 volumes of the Dictionary of National Biography. He selected individuals to whom three or more pages were devoted, but he also included those whom he believed to have shown a high order of intellectual ability despite the fewer than three pages of print. He excluded the notorious and members of the nobility regardless of their eminence. Cattell (1903) selected his group of eminent men from six biographical dictionaries or encyclopedias: two French, one German, and three English, including Lippincott's Biographical Dictionary, the Encyclopaedia Britannica, and Rose's Biographical Dictionary. The chosen group was defined by inclusion in at least three of the sources, with the greatest average space allotted determining the magnitude of eminence. The subjects in the present study were derived from listing all individuals who appeared in the 1963 edition of the Encyclopaedia Britannica with 1 column of space (| page) or more and from listing all individuals who were given 1 column of space (£ page) or more in the 1964 edition of the Encyclopedia Americana. A person with at least 1 column in each encyclopedia was included; this resulted in a group of 699 individuals, 20 women and 212 • MARCH 1978 • AMERICAN PSYCHOLOGIST