Projective Methods

views updated

Projective Methods

I. Projective TechniquesGardner Lindzey and Joseph S. Thorpe


II. The Rorschach TestSamuel J. Beck


III. The Thematic Apperception TestWilliam E. Henry



Projective techniques and their application represent a point of significant intersection for psychoanalytic theory, clinical psychology, social psychology, and cultural anthropology. These instruments include a wide array of materials, instructions, and interpretative rules, but in spite of this diversity, there are certain common qualities that distinguish them from other personality measures, such as inventories, ratings, or situational tests.

The most distinctive feature of projective techniques concerns the presentation to the subject of a task that is relatively unstructured. This lack of structure involves ambiguous or vaguely defined stimuli and a relatively unrestricted set of response alternatives by means of which the subject may assign meaning to the stimulus material. It is generally assumed that in the process of choosing between the many alternatives for structuring or interpreting the ambiguous stimulus material, the subject reveals significant and fundamental aspects of himself. Ordinarily such devices are considered to be particularly responsive to latent or unconscious components of the person, and consistently there is a minimum of subject awareness concerning the purpose of the test. The response data are typically profuse and often involve fantasy or imaginary creations. The instruments are highly multidimensional; that is, they are intended to provide information concerning a large number of personality dimensions.

Historical development

The origins and early development of projective tests are closely intertwined with developments within psychoanalysis. Indeed, Freud’s theory of dreams and technique of dream interpretation may be considered the most important single source of the dominant ideas in this area. At about this time Jung’s Studies in Word-association (1904–1909) appeared, which also has been heavily influential in the development of projective testing. The Rorschach test, best known of all projective tests, was developed by a student of Jung’s and was undoubtedly influenced by psychoanalytic theory and practice. The term “projective technique” came to be associated with these instruments primarily as a result of the writings of Henry A. Murray and L. K. Frank during the 1930s; however, the relationship between the concept of projection and these instruments is anything but precise (see Lindzey 1961). Indeed, one can say that “projection,” as the term is conventionally used within psychoanalysis today, is seldom directly involved in projective-technique responses.

In the two decades following World War i there was a very rapid development and application of aptitude and ability tests as exemplified by the Stanford-Binet and various group tests. Following World War ii there were comparable developments in the area of personality testing, with projective techniques receiving the most prominent attention. It was during this period that the Society for Projective Techniques, made up of persons involved in the study and application of these devices, was formed. At the same time the Journal of Projective Techniques, a journal that contains only papers dealing with these instruments, appeared. In addition, an extraordinarily large number of articles and books concerning these techniques have been published. Even for a single instrument, such as the Rorschach test, a total bibliography would now include more than a thousand items. Moreover, within most modern graduate programs in psychology the inclusion of one or more courses dealing with projective techniques has become standard.

The development and application of these instruments have been associated closely with clinical activities. For the most part, individuals drawn to these devices have been working in clinical settings, often with disturbed persons and typically under circumstances where individual subjects could be subjected to a relatively intensive and detailed study. Moreover, the techniques have been considered most attractive and useful by persons who are heavily influenced by psychoanalytic theory or some variety of holistic theory. Seldom has a heavy emphasis upon objectivity, precision, and psychometric elegance been characteristic of work with these devices.

Major techniques and their classification

Given these general remarks concerning the nature and origins of projective techniques, it may be helpful to provide concrete identification of the instruments by means of representative illustrations. There are many different approaches to classifying these tests (Lindzey 1959), but the most useful approach focuses on the nature of the response evoked from the subject. On this basis projective techniques may be grouped broadly into the following categories: (a) association, (b) construction, (c) completion, (d) choice or ordering, and (e) expression.

Association techniques

Included under association techniques are those instruments that tend to minimize ideation and emphasize immediacy of response. The subject is instructed not to reflect but rather to respond to the stimulus with the first word, image, or percept that occurs to him.

Word association. An early and well-known example is the word association technique, which involves the oral presentation of a series of disconnected words, one at a time, to which the subject is instructed to respond with the first word that occurs to him. Typically, both the subject’s association and reaction time to each word are recorded. While many lists of words have been used, the most frequently encountered are the early list by Kent and Rosanoff (1910), which was designed as a psychiatric screening device, and the more recent list by Rapaport, Schafer, and Gill (1945–1946). The latter is similar to Jung’s (e.g., 1904–1909), which contains many words of psychoanalytic significance, particularly with regard to psychosexual conflict. Depending upon the particular purpose of the user, scoring and interpretation may involve an analysis of the content of the associations, in terms of their relevance to particular motives or conflict areas such as dominance or aggression, or if a judgment concerning the degree of general adjustment is to be obtained, the subject’s responses may be compared with the typical performance of psychiatric and normal groups on the same list of words. In addition to their content, associations are often analyzed in terms of proportion of common responses, temporal deviations, and associative disturbances, such as repeating the stimulus word or giving extremely remote associations.

Rorschach and other inkblot techniques. By far the most popular projective technique, in terms of frequency of clinical and research application, is the Rorschach technique (Rorschach 1921), which consists of ten symmetrical inkblots. Typically the cards are presented individually and in a set order, with instructions to the subject that he report what the figures resemble or suggest to him. Responses to each card are recorded verbatim and reaction times are noted, following which an inquiry is conducted wherein the subject identifies the characteristics of the stimuli which determined or contributed to his associations. [SeeRorschach.]

There are a variety of scoring systems in use, the most popular of which are by Beck (Beck et al. 1944) and Klopfer (Klopfer et al. 1954–1956). Those aspects of responses which are usually scored are location, the relative size of the blot area incorporated (including the use of white space); determinants, the use of color and shading in comparison with the form or shape of the blot; and the presence of movement—either human, animal, or inanimate—in the response. Also noted are the accuracy of form perception, the incidence of popular and original responses, and the total number of responses. The content of responses is also classified, at least in terms of human, animal, anatomical, nature, and abstract classes, although some scoring conventions are quite elaborate in this respect. Various scores, representing the number of responses falling into each category, are usually further transformed into ratios of scores which are considered essential for interpretation.

Most systems of interpretation of these quantitative scores and ratios prescribe inferences to be drawn regarding the more general organization of personality, rather than specific motives or conflicts. It is usually emphasized that the interpreter must consider the entire profile of scores rather than isolated scores.

Modifications of this technique have involved developing new stimuli and more objective scoring procedures, as well as alternate forms and group versions of the test. Perhaps the most interesting modification is the Holtzman Inkblot Technique (Holtzman et al. 1961), which introduced desirable psychometric properties to the basic Rorschach approach.

Construction techniques

Construction techniques require the subject to go beyond simple association to a stimulus and to create or construct a more elaborate product, which is typically a complete art form, such as a story or picture.

Thematic Apperception Test. Next to the Rorschach, the most widely used of all projective techniques is the Thematic Apperception Test (Murray 1943). The TAT consists of cards containing black-and-white pictures of vague scenes, and the instructions request the subject to compose a story to fit each picture, describing what the people are thinking and feeling, what led up to the scene depicted, and what the outcome will be. Analysis of TAT stories, which are usually recorded verbatim, is far from standardized, although a number of general scoring systems have been proposed (e.g., Aron 1949; Bellak 1947; Stein 1948; Tomkins 1947). Most of these systems focus on the content of behavior and experiences described in the story, although they differ considerably in the unit of analysis employed and the degree of inference required. As used clinically, interpretation typically has not rested upon a set of objective scores for specific variables. The basic assumption of most schemes is one of a psychological isomorphy between dispositions attributed to the part of the major character in the story and those that exist in the storyteller.

The Blacky Pictures. Closely related to the TAT is the Blacky Pictures test (Blum 1949), which is intended to investigate 11 specific psychoanalytic variables, including oral eroticism, oedipal intensity, and castration anxiety. The test consists of 12 cartoons concerned with experiences in the life of a dog named Blacky, including his relationships with Mama, Papa, and Tippy, a sibling. As its central feature the procedure involves story construction in response to the pictures but adds indications of preference and a series of direct questions for each picture. While administration is highly standardized, scoring and interpretation are not.

Completion techniques

Completion techniques present some type of incomplete product, with the requirement that the subject complete it in any manner he wishes and the restriction that the completions meet certain standards of good form or rationality. Best known of these devices are the sentence-completion tests (e.g., Rotter & Wilierman 1947), which are widely used by both clinicians and personality investigators. Typically such a device consists of thirty to one hundred brief sentence stems which the subject is instructed to complete with the first words that come to mind. In some cases the instructions to the subject emphasize that the completions should reveal his own feelings. The type of scoring employed and the customary inferences derived from this technique suggest its close kinship with the TAT. The technique is considered most efficient in assessing the content of personality (attitudes, motives, and conflicts) at a more conscious or manifest level than such instruments as the Rorschach or TAT. Other examples of completion tests are story-completion and argument-completion tests and the Rosenzweig Picture-Frustration Study (Rosenzweig 1949).

Choice or ordering techniques

Choice or ordering techniques require the respondent to choose from a number of alternatives the item or arrangement that fits some specified criterion, such as meaningfulness, relevance, or attractiveness. In some cases, such as multiple-choice versions of the Rorschach (Harrower & Steiner 1945) and the TAT (Goodstein 1954), the subject is required to select from a number of hypothetical responses the one that seems most appropriate to him.

The Szondi Test. Perhaps best known of these techniques is the Szondi Test. This test was devised in the late 1930s by Szondi, a Hungarian psychiatrist (1947), and made popular in the United States through the efforts of Susan Deri (1949). The test materials consist of 48 photographs of individuals drawn from eight psychiatric diagnostic categories. A rather complex and lengthy administration procedure elicits from the subject expressions of preference for the different photographs. The test is based on the assumption that a person can be described in terms of eight “needsystems” (presumably reflected in the photographs), with the subject’s selections and rejections indicating the degree of strength or tension in each.

Although the test enjoyed considerable popularity in clinical settings during the 1950s, more recently there has been less evidence of application. Despite some positive features, such as the recommendation of repeated administrations and a highly objective scoring system, the theoretical rationale of the test has appeared to be anything but compelling to most American psychologists. Attempts at empirical verification of the test seem to be overwhelmingly negative (Borstelmann & Klopfer 1953). A recent volume by Szondi, Moser, and Webb (1959) provides a strongly partisan description of the technique and its rationale.

Picture Arrangement Test. A more recent example of an ordering technique is the TomkinsHorn Picture Arrangement Test. This device consists of 25 plates, each of which contains three line drawings that depict the same figure involved in different but related activities. The subject is asked to indicate the order in which these activities took place and to provide a sentence indicating what is going on in the picture. The efforts of the authors (Tomkins & Miner 1957) have been devoted primarily to developing objective methods for dealing with the arrangements selected by the subject, although they provide suggestions for the use of the descriptive statements as well.

This test is unique in its consideration for interpretation of only those responses that are relatively rare in occurrence. Rare response patterns from a subject provide a basis for estimating certain tendencies such as hypochondriasis or avoidance of people and for distinguishing normal from abnormal subjects. The existence of a highly objective scoring procedure, coupled with the demonstration of certain distinct differences in the performances of normal and disturbed subjects, identifies this instrument as possessing objective and normative properties that are not encountered anywhere else in the domain of projective techniques. On the other hand, the data collected are sharply delimited, and the process of demonstrating that the test is sensitive or useful has only begun.

Expressive techniques

Expressive methods differ from construction techniques in that they place as much emphasis upon the manner and style in which the product is created as upon the product itself. They are often considered to be therapeutic as well as diagnostic devices, since the subject is presumed to relieve his difficulties in the process of revealing them.

Play. Most popular among the expressive instruments are play techniques. Originating in play therapy, these methods have been adapted for the diagnosis and measurement of personality, primarily in children. All the approaches present the subject with an array of toys which he is encouraged to use in some manner. Among the objects frequently selected are dolls representing adults and children of both sexes and various age levels. Typically the examiner is responsible for recording as much of the subject’s behavior as possible, including his choice and arrangement of toys, accompanying comments, and expressive behavior. Except for the work carried out by Sears and his associates, which has focused on measures of aggression and dependency (Sears et al. 1953), little has been done with this technique in the way of standardization and specification of just how the process of interpretation shall take place. In most instances it is assumed that the examiner will recognize important motives and conflicts when he sees them.

Drawing and painting. There is also a variety of drawing and painting techniques that have been used in personality assessment. Attention has centered chiefly on procedures using drawings of the human figure, the most publicized of which have been those of Buck (1948) and Machover (1949). When the test is administered individually, the examiner usually notes the subject’s comments, the sequence of parts drawn, and other procedural details. Scoring of human-figure-drawing tests is essentially qualitative, being concerned with such stylistic features as the figure’s stance, size, and position on the page, disproportions, shading, and erasures. Interpretation of both drawing and painting techniques is far from precise, in some cases depending largely upon general clinical wisdom, coupled with a knowledge of some specific rules or generalizations relating certain features of the drawing or painting to personality characteristics.

Evaluation and interpretation

The preceding survey of the more important examples of different types of projective techniques illustrates the extreme diversity of these instruments. Discussions of projective techniques in a critical context will be found in volumes by Anastasi (1961), Cronbach (1949), and Lindzey (1961). General sources of more detailed information are books by Abt and Bellak (1950), Anderson and Anderson (1951), and Zubin, Eron, and Schumer (1965).

The interpretation of projective-technique responses is carried out under conditions that vary from the highly molecular and objective to the most generalized and intuitive. Virtually all the individual tests have interpretative manuals to assist the user in extracting useful psychological information from them, and in the case of the better-known instruments, such as the Tat and Rorschach, there are dozens of books and monographs and hundreds of articles dealing with general and specific problems of interpretation. In spite of the bulk and diversity of this interpretative array, there are certain generalizations that can be made concerning customary interpretative practice.

The interpretation of projective techniques in individual assessment is more often qualitative and subjective than it is quantitative and objective. Even in those instances where the responses may be subjected to objective scoring that leads to numerical values, the translation of these scores into psychological statements usually involves complex and unspecified processes (Levy 1963). In research use, on the other hand, it is much more common to find the response data subjected to objective scoring, which in turn leads to quantitative scores on specified personality dimensions.

If we make the reasonable assumption that the chief interest of the person using projective techniques is in enduring dispositions or in personality characteristics, we find that the customary task of the interpreter is enormously complicated by the sensitivity of these instruments to a wide variety of determinants, only a small proportion of which are relevant to enduring motivational or personality dimensions. Thus, it is well established (Gleser 1963; Lindzey 1961; Masling 1960) that projectivetechnique responses are influenced by temporary or fleeting psychological states, by the nature of the stimuli that are presented, by a number of response sets, by individual differences unrelated to personality factors (for example, differences in intelligence), and by a wide variety of situational factors (for example, examiner-subject interaction, past testing experiences, reason for taking the test). Given these observations, it is clear that the interpreter must make every effort to demonstrate that those aspects of the response data being used in interpretation are not products of variation in nonpersonality factors. In order to be reasonably certain of this, it is almost always necessary that the examiner know a great deal about the circumstances under which the test was administered, as well as a good deal about the subject in addition to his test responses. For this reason, most projective techniques administered in clinical settings form only part of a larger battery of intelligence and personality tests, with interpretation leaning heavily on the consistency of performance among the several tests. Despite this practice, one of the major shortcomings in the customary application of these techniques is the tendency on the part of many users to assume that all test responses are determined solely by personality variables; consequently, they exhibit a readiness to overinterpret. This caution is somewhat less important in research settings, where it is possible to control or randomize some of the factors that might otherwise distort interpretation.

As we have indicated, the most extensive use of projective techniques has been in clinical situations. In the hands of clinical psychologists these instruments have played a key role in psychodiagnosis and personality description in a wide variety of psychiatric and psychological settings. These techniques have also been used extensively in connection with motivation research, or the study of consumer personality or motivation as a means of predicting buying behavior. Not surprisingly, little of the work in this area has achieved very high standards of rigor and control.

Of more scientific interest has been the use of these instruments in personality research. A sizable number of the controlled studies relevant to psychoanalysis that have been conducted have involved the use of projective techniques (Hilgard 1952; Rickers-Ovsiankina 1960). The extensive use of projective measures of achievement and affiliation motivation in a wide variety of social psychological and personality research represents another major area of application (Atkinson 1958; McClelland et al. 1953). Experimental studies of psychopathology have often made use of projective measures. The use by psychologists and cultural anthropologists of projective techniques in crosscultural studies is another area of extensive application (Kaplan 1961; Lindzey 1961; Zubin et al. 1965). Indeed, a large proportion of the studies of modal personality or national character, psychological correlates of acculturation, and personality development in nonliterate settings has involved more or less extensive dependence upon projective techniques.

In view of this widespread application and the frequently partisan enthusiasm that has been associated with these techniques, it is not surprising that they have attracted a good deal of critical attention. Criticisms have generally focused upon the psychometric flaws of the instruments, including dubious reliability and validity and the frequent lack of objectivity or specificity in the administration and interpretation of the tests. There is no doubt that much of this criticism is reasonable and based upon relatively convincing data. On the other hand, certain projective instruments under certain conditions have been demonstrated to attain reasonable standards of objectivity and standardization (e.g., Atkinson 1958; Murstein 1963).

The matter of validity is always complicated and particularly so when we are dealing with complex motivational concepts, especially when these are presumed to relate to covert or latent aspects of personality. While there are numerous studies (Ainsworth 1954; Lindzey 1959; Murstein 1963) that demonstrate limited predictive and concurrent validity—ability to predict future behavior or performance and to relate to concurrent criterion measures—most findings have not held up well upon cross-validation. Further, such studies do not ordinarily deal adequately with the covert personality characteristics for which these tests are supposed to be singularly appropriate. At risk of oversimplification, we may state that when we are concerned with the validity of projective techniques, their presumed association with covert variables, complex motivational processes, and psychoanalytic concepts forces the investigator and theorist into the tangled jungle of construct validity, and this guarantees the absence of easy answers. The relation between projective techniques and psychological theory is of key importance in any consideration of the construct validity of these devices. It seems clear on the basis of a number of past examinations (e.g., Holt 1954; Conference on Contemporary Issues . . . 1961; Lindzey 1961) that, while there are some important links between these tests and several varieties of psychological theory, particularly psychoanalytic theory, there is little in the way of firm and formal association between theory and method. Relatively detailed discussion of the validity of these instruments may be found in Ainsworth (1954), Cronbach (1949), Harris (1960), and Lindzey (1961). At the present time, no simple statements can be made concerning the over-all validity of even a single projective test, let alone the entire class of tests. Indeed the question of general validity is meaningless; one must always ask, Valid for what purpose? and, perhaps, Valid when used by whom and under what conditions?

All things considered, it is fair to say that these techniques have not lived up to early expectations in regard to their stable sensitivity to covert motivational processes. On the other hand there is little doubt that in the hands of the sophisticated and energetic investigator they have fulfilled satisfactorily a number of important functions. As the previous discussion has indicated, there are some important areas of personality research that have been heavily dependent upon these devices, and there is agreement that some of these areas have involved significant research contributions. It is more difficult to assess the use of the instruments as they have been employed typically in clinical settings, since their function there is highly variable, depending upon the particular setting and the individual clinician. For example, it is difficult to examine Schafer’s discussion (1954) of Rorschach interpretation without being convinced that the test can serve important assessment and diagnostic functions when employed in the manner he recommends and by a person of his ability. At the same time, one must admit that few clinical applications are likely to attain these high standards.

One may conclude that if projective techniques represent the “royal road to the unconscious,” the necessary road maps have yet to become common knowledge. Nevertheless, the instruments seem to have found a useful place within the psychologist’s array of assessment tools.

Gardner Lindzey and Joseph S. Thorpe

[For discussion of other techniques of personality assessment, seeFactor analysis, article on psychological applications; Interviewing, article Onpersonality appraisal; Personality measurement; Sociometry; Traits. Additional relevant material may be found inAchievement Motivation; Fantasy; Personality: contemporary viewpoints, article oncomponents of an Evolving personological system; Psychometrics; and in the biography ofRorschach.]


Abt, Lawrence E.; and Bellak, Leopold (editors) 1950 Projective Psychology: Clinical Approaches to the Total Personality. New York: Knopf.

Ainsworth, Mary D. 1954 Problems of Validation. Volume 1, pages 405–500 in Bruno Klopfer et al., Developments in the Rorschach Technique. New York: World.

Anastasi, Anne 1961 Psychological Testing. 2d ed. New York: Macmillan. → The first edition was published in 1954.

Anderson, Harold H.; and Anderson, Gladys L. (editors) (1951) 1956 An Introduction to Projective Techniques & Other Devices for Understanding the Dynamics of Human Behavior. Englewood Cliffs, N.J.: Prentice-Hall.

Aron, Betty R. 1949 A Manual for Analysis of the Thematic Apperception Test: A Method and Technique for Personality Research. Berkeley: Berg.

Atkinson, John W. (editor) 1958 Motives in Fantasy, Action, and Society: A Method of Assessment and Study. Princeton, N.J.: Van Nostrand.

Beck, Samuel J. et al. (1944) 1961 Rorschach’s Test. Volume 1: Basic Processes. 3d ed. New York: Grune.

Bellak, Leopold 1947 Bellak Tat Blank. New York: Psychological Corporation. → A guide to the interpretation of the Thematic Apperception Test.

Blum, Gerald S. 1949 A Study of the Psychoanalytic Theory of Psychosexual Development. Genetic Psychology Monographs 39:3–99.

Borstelmann, L. J.; and Klopfer, W. G. 1953 The Szondi Test: A Review and Critical Evaluation. Psychological Bulletin 50:112–132.

Buck, John N. 1948 The H-T-P Test. Journal of Clinical Psychology 4:151–159.

Conference On Contemporary Issues in Thematic Aperceptive Methods, Fels Research Institute, 1959 1961 Contemporary Issues in Thematic Apperceptive Methods. Edited by Jerome Kagan and Gerald Lesser. Springfield, 111.: Thomas.

Cronbach, Lee J. (1949) 1960 Essentials of Psychological Testing. 2d ed. New York: Harper.

Del Torto, John; and Corneytz, Paul 1944 Psychodrama as Expressive and Projective Technique. Sociometry 7:356–375.

Deri, Susan 1949 Introduction to the Szondi Test: Theory and Practice. New York: Grune.

Gleser, Goldine C. 1963 Projective Methodologies. Annual Review of Psychology 14:391–422.

Goodstein, Leonard D. 1954 Interrelationships Among Several Measures of Anxiety and Hostility. Journal of Consulting Psychology 18:35–39.

Harris, Jesse G. Jr. 1960 Validity: The Search for a Constant in a Universe of Variables. Pages 380–439 in Maria A. Rickers-Ovsiankina (editor), Rorschach Psychology. New York: Wiley.

Harrower, Mary R.; and Steiner, Matilda E. (1945) 1951 Large-scale Rorschach Techniques. 2d ed. Springfield, 111.: Thomas.

Hilgard, Ernest R. (1952) 1956 Experimental Approaches to Psychoanalysis. Pages 3–45 in Eugene Pumpian-Mindlin (editor), Psychoanalysis as Science. New York: Basic Books.

Holt, Robert R. 1954 Implications of Some Contemporary Personality Theories for Rorschach Rationale. Volume 1, pages 501–560 in Bruno Klopfer et al., Developments in the Rorschach Technique. New York: World.

Holtzman, Wayne H. et al. 1961 Inkblot Perception and Personality. Austin: Univ. of Texas Press. Journal of Projective Techniques. > Published since 1936.

Jung, Carl Gustav (1904–1909) 1918 Studies in Wordassociation: Experiments in the Diagnosis of Psychopathological Conditions Carried Out at the Psychiatric Clinic of the University of Zurich, Under the Direction of C. G. Jung. London: Heinemann. → First published in German.

Kaplan, Bert (editor) 1961 Studying Personality Crossculturally. Evanston, 111.: Row, Peterson.

Kent, Grace H.; and Rosanoff, A. J. 1910 A Study of Association in Insanity. American Journal of Insanity 67:37–96,317–390.

Klopfer, Bruno et al. 1954–1956 Developments in the Rorschach Technique. 2 vols. New York: World. → Volume 1: Technique and Theory. Volume 2: Fields of Application. See especially Volume 1.

Levy, Leon 1963 Psychological Interpretation. New York: Holt.

Lindzey, Gardner 1959 On the Classification of Projective Techniques. Psychological Bulletin 56:158–168.

Lindzey, Gardner 1961 Projective Techniques and Cross-cultural Research. New York: Appleton.

Lindzey, Gardner et al. 1959 Thematic Apperception Test: An Interpretive Lexicon for Clinician and Investigator. Journal of Clinical Psychology Monograph Supplement, No. 12.

Mcclelland, David C. et al. 1953 The Achievement Motive. New York: Appleton.

Machover, Karen (1949) 1961 Personality Projection in the Drawing of the Human Figure. Springfield, 111.: Thomas.

Masling, Joseph 1960 The Influence of Situational and Interpersonal Variables in Projective Testing. Psychological Bulletin 57:65–85.

Moreno, Jacob L. 1946 Psychodrama. Volume 1. New York: Beacon.

Murray, Henry A. 1943 Thematic Apperception Test Manual. Cambridge, Mass.: Harvard Univ. Press.

Murstein, Bernard I. 1963 Theory and Research in Projective Techniques, Emphasizing the Tat. New York: Wiley.

Rapaport, David; Schafer, R.; and Gill, M. M. 1945–1946 Diagnostic Psychological Testing: The Theory,Statistical Evaluation, and Diagnostic Application of a Battery of Tests. 2 vols. Chicago: Year Book Publishers.

Rickers-Ovsiankina, Maria A. (editor) 1960 Rorschach Psychology. New York: Wiley.

Rorschach, Hermann (1921) 1942 Psychodiagnostics: A Diagnostic Test Based on Perception. 3d ed. Bern: Huber; New York: Grune. → First published in German.

Rosenzweig, Saul 1949 Psychodiagnosis. New York: Grune Rotter, Julian B.; and Willerman, Benjamin 1947 The Incomplete Sentences Test as a Method of Studying Personality. Journal of Consulting Psychology 11:43–48.

Schafer, Roy 1954 Psychoanalytic Interpretation in Rorschach Testing: Theory and Application. Austin Riggs Foundation, Monograph Series, No. 3. New York: Grune.

Sears, R. R. et al. 1953 Some Child Rearing Antecedents of Aggression and Dependency in Young Children. Genetic Psychology Monographs 47:135–236.

Stein, Morris I. (1948) 1955 The Thematic Apperception Test: An Introductory Manual for Its Clinical Use With Adults. 2d ed. Reading, Mass.: AddisonWesley.

Swenson, Clifford H. Jr. 1957 Empirical Evaluations of Human Figure Drawings. Psychological Bulletin 54:431–466.

Szondi, Lipot (1947) 1952 Experimental Diagnostics of Drives. New York: Grune. → First published as Experimentelle Triebdiagnostik.

Szondi, Lipot; Moser, Ulrich; and Webb, Marvin W. 1959 The Szondi Test: In Diagnosis, Prognosis, and Treatment. Philadelphia: Lippincott.

Tomkins, Silvan S. 1947 The Thematic Apperception Test: The Theory and Technique of Interpretation. New York: Grune.

Tomkins, Silvan S.; and Miner, John B. (1957) The Tomkins-Horn Picture Arrangement Test. New York: Springer.

Zubin, Joseph; Eron, Leonard D.; and Schumer, Florence 1965 An Experimental Approach to Projective Techniques. New York: Wiley.


The Rorschach test is one attempt at concurrently examining an individual’s functioning intelligence, his emotions, and his unconscious. Aside from theoretical considerations, this has very important practical meaning for at least three disciplines—clinical psychology, psychiatry, and social work. By the time Hermann Rorschach first published his test, in 1921, all three had been forced to grope for multidimensional concepts of personality. While they were ready for objective tests that were sophisticated enough to penetrate to the deeper recesses of human behavior, they were not, to be sure, prepared to have this accomplished by ten inkblots.

The Rorschach test’s basic data consist of a person’s free associations to ten inkblot figures. These associations are the behaviors from which the examiner can conclude that the person is a hysteric, an obsessive neurotic, a schizophrenic, and so on, or is in sound mental health. These associations are behaviors just as are any other visible activities of an individual by which we judge his character, such as combing his hair or wearing certain kinds of clothes.

The term “projective” was originally attached to this kind of test by Frank (1939). It can be justified etymologically but not clinically. The person does put forth, or project, elements of his own personality onto the stimulus. However, the word “project” had already acquired its meaning, specifically with reference to paranoid behavior, in clinical language. So “projection” now has two meanings in the languages of two closely related disciplines. Frank’s designation has become affixed to these tests and appears likely to remain so. In order to attain an understanding of the behaviors which constitute the Rorschach test data, the investigator needs to know the grammar and the syntax of the test’s language. The present article undertakes to sketch them.

Psychological processes measured

Rorschach found the inkblot data informative regarding three major psychological processes: intellectual control, emotional life, and fantasy life, and his own thinking pivoted around these three areas (1921). Rorschach’s initial interpretation of responses to inkblots explicitly follows psychoanalytic theory, and the subsequent evolution of the test as a clinical instrument can readily be traced within the framework of that theory. We are enabled by means of the Rorschach test to describe a person in terms of his ego functioning, his emotional state, and his unconscious ideation, that is, the core areas of Freudian psychology. We are also able to identify that other set of dynamic traits, the defenses.

Intellectual controls and good form

Concerning the intellectual controls, the subtitle of Rorschach’s monograph (1921) indicates that diagnosis is based on perception. The person looks at inkblot forms and tells what he “sees.” These perceptions of forms are the basic behavioral data derived from the test. If now we administer the test to, say, one hundred presumably normal individuals, we infer from the distribution of their responses what a normal population “sees” in the inkblots, that is, what their perceptual behaviors are like in reaction to the test. Administering it to a person about whom we know nothing, we compare his responses with those of the normal population. To the extent that this person responds as they do, one can infer that he sees as they do and that he is to that degree normal.

This is the foundation on which the test rests. What I have described in the preceding paragraph is Rorschach’s concept of good form (F+). The patient’s ability to respond with such “good forms” within a certain empirically determined range is an index of his ability to see accurately; the criteria for accurate perceptions are assumed to be set by one’s peers.

What the test does then is to set up ten situations which establish their own frames of reference for what is accurate and inaccurate, good form or bad, reality or deviation from it. The inkblots are a cosmos in themselves. Not having any conventional meanings or reality, they can start de novo to trace out their own realities. This is the great advantage of the test’s using “unstructured” stimuli.

With this advantage, there also exists a corresponding responsibility for the dependability of the frame of reference, that is, the normative foundation.

“Good forms” (or F+) are the nucleus for ego evaluation. The logic for this rests on the fact that the persons by whom Rorschach established his good-form criterion were assumed to be “normals,” who take care of their affairs in life with the usual respect for law, custom, and morals as well as for the social graces. Such living is assumed to be motivated by the wish for the respect of one’s fellow citizens. The respect that we want and that we know we derive from others is also considered to be a measure of our self-respect, of ego. Accordingly, an F+ score can be considered as a measure of ego.

This deductive reasoning has been borne out inductively by clinical experience. Among persons who are very disturbed neurotically, many are also very erratic in their form perceptions. Neurotics are assumed to be persons in conflict. They have emotional urges which are inconsistent with their standards of good and bad. That is, they are motivated by urges that are ego-alien. In most cases of neurosis, the conflictual stress drains the person of his critical control power, and so his perceptual accuracy breaks down. The further observation is that the milder (clinically) the neurosis, the higher the F+ percentage; the reverse is also true. This relationship is also obtained in the study of development; the lower the age range, the lower the percentage of accurate form perception. Accuracy of form perception progresses with maturation, which is to say, with the ego’s growth. A plateau is reached at the early adolescent years, and the percentage remains stable thereafter, assuming the relative stability of the personality. [SeeNeurosis; Perception, article onperceptual development; Sensory and motor development.]

When an individual’s response deviates considerably from expectancy, we look for the explanation in psychological factors, usually emotional, that minimize or enhance the accuracy of perceptual responses. Excitability or agitation may break down perception. Obsessive anxiety may heighten it. Or the person’s maturation may have been arrested, perhaps as a result of inadequate or damaged brain tissue, mental deficiency, or a pathological environmental climate—a family setting that has contributed toward halting the personality’s growth within a childish or even infantile phase, which often occurs in schizophrenia. In these patients the Rorschach perceptual range is usually below the critical minimum both for the normal and the neurotic ranges.

Such are the actual empirical data in clinical groups. To restate: the percentage of good-form perception varies predictably according to what is known clinically of the ego’s growth level and of its stamina in the clinical condition. Findings of heightened perceptual accuracy agree with this logic. The responses of depressives and some very severely disturbed neurotics are very close to, and sometimes at, one hundred per cent accuracy. This is due to their crushing anxiety. It makes them incapable of seeing anything except in the most rigidly correct fashion. They overreact to their tensions and pain. They develop a character armor that severely disables them for effective adjustment to life. These conflicts are just as stressful as those of the patients whose conflicts lead to distorted perception. Such is the clinical logic. The empirical fact is that healthy individuals, and this includes those of highest intelligence, never rate at one hundred per cent F+, or form accuracy. They are liberated emotionally, and they do make mistakes.

Perceptual accuracy is not the only index of the individual’s ego functioning. Ego functioning includes other intellectual attributes. Principally these are attentiveness, the ability to synthesize one’s percepts, the intact or pathological quality of the thought processes, and thematic content in thinking as a clue to the personal values which the patient holds. These variables and their numerous nuances are also assessed by the Rorschach test.

Emotional life and color responses

Emotional life is tapped by the Rorschach test principally through the person’s reactions to the color and to the shading tones of the blot figures. Rorschach himself described the variety of the color reactions and their psychological significances. Both his technical observations and his interpretative principles on this point have remained essentially unchanged. He had only begun to glimpse the meanings of the reactions to the variations in gray, as a posthumous paper shows (see Rorschach & Oberholzer 1923). Binder (1932) first described these in detail; Oberholzer (1944) has elucidated an especially important response to the grays, that is, the vista percept; and Klopfer (1942) was the first to describe another nuance, the texture response, that is, the person responds as though he feels the stimulus. What any of these variables tell us psychologically would take us far beyond the scope of the present article.

Fantasy and the movement response

The most original of Rorschach’s discoveries is the “movement” (Bewegung) response. As he described it, both in his monograph (1921) and in a posthumous paper (see Rorschach & Oberholzer 1923), it is a perception of a human form as if in movement. The patient may see this in the whole blot figure or in any portion of it. While the content in the perceived form may be that of some animal, the movement must nonetheless be one which is anatomically feasible for a human being (e.g., “two bears standing on their hind feet, as if boxing,” or even “two birds talking together”).

Rorschach was both explicit and definitive in excluding from his fantasy concept any animal movement that is a specific characteristic activity of that animal (e.g., “a dog snapping at a butterfly”). His caution about identifying the movement determinant accurately was dictated by the critical psychological significance of these associations. They are the language by which the test reports fantasy activity in which the person is engaging, fantasy which is a wish or a fear. Fantasy is considered to be introverted emotion, that is, a repressed fear or wish. Nearly all Rorschach fantasy associations also have a significance for the patient’s unconscious that is not evident in the overt associational content. The association, like the dream, expresses an idea which the patient cannot entertain in his conscious awareness. It is thus the lead to the unconscious. [SeeFantasy.]

The importance of such an invention in the field of personality study can hardly be overestimated. The test becomes a depth instrument. Like the dream, it communicates in an idiom seemingly innocent because it is only responses to inkblots. The test thus enables the material to bypass censorship of the ego, enabling the examiner to discern conflictual needs in the patient and to relate them to the entire personality structure—schizophrenic, neurotic, or healthy. It can be seen too why Rorschach was so circumspect in defining his fantasy response. The fantasy concept has since his time been extended, principally under Klopfer’s influence (Klopfer 1942; Klopfer et al. 1954–1956) to include apparent movement associations that do not conform to Rorschach’s definition. They have been identified by the symbol Fm, that is, “form tending to movement.” Examples of Fm provided by one of Klopfer’s co-workers (Klopfer 1942, chapter 12) are “spiders crawling up on a leaf”; “a bat turned upside down, a bat flying with his wings spread out”; “a squirrel on a stub, or a nest; just the back, with the tail sticking up.” The present writer has not been convinced that these Fm responses have the psychological significance of the Rorschach Bewegung, and he follows Rorschach’s original technique.

Application and evaluation

The test currently is being used in almost all parts of the world. Aside from psychology proper, the disciplines that are much interested in it are social work, psychiatry, and anthropology. Social work agencies having recourse to it as an exploratory instrument find it an aid in assessing the severity of a client’s disorder and thus are able to arrive more quickly at a treatment plan. In psychiatry the test provides important leads and thus an economical approach to differential diagnosis. To the degree of its validity the Rorschach test does, in comparatively short time, what might otherwise require many exploratory clinical sessions.

As a research tool in psychotherapy the test can be administered at various stages in the course of an illness. The progress of the patient (and so the effectiveness of the therapy) can be gauged from changes in the test findings. Developmental growth can be similarly traced by applying the test to children at various ages, something that Ames, Métraux, and Walker have been doing (1959); a similar study was reported by Thetford, Molish, and Beck (1951).

Anthropology is still another field in which the test has been useful in research (see Hallowell 1956). It yields information about how psychological structures vary with cultural differences.

Research problems stemming from the Rorschach test are grouped broadly into three areas. One is general and experimental psychology, dealing with questions of the relations of color and shading to the affective states of the person, the rationale for employing the human movement response as an index of unconscious ideation, and the way in which the intellectual variables of the test provide information about the ego’s functioning.

In the second area are certain theoretical problems that are specifically derived from Rorschach test findings and principles. First among these is Rorschach’s Erlebnistypus, or “experience balance,” the ratio between the number of movement responses and the number of color responses (1921; Rorschach & Oberholzer 1923), and with it its derivative, the “experience actual” (Beck 1960). These concepts bear on the innermost emotional and individualized needs of each person. Yet these are made accessible through the objective stimuli —the Rorschach inkblot figures. The experience balance and the experience actual are known from the relative amount of externalizing or internalizing of the feelings. This recalls of course Jung’s introversion–extraversion hypothesis. But the experience balance data have significance apart from in troversion–extr aversion. Related, too, are concepts of the psychologic structure of personality as a balance of forces—affective and intellectual, id and ego.

Third, in the practical clinical field the research needs are for more and more controlled observations. Here the task is threefold: (1) Statistical data are to be gathered for various diagnostic groups to determine the means and measures of dispersion for each of the Rorschach test variables, for each clinical category. (2) To determine what the entire structure is like in the typical individual of each group, that is, how much F+ per cent, human movement, and color and other test responses occur in the typical individual. Then, assuming the Rorschach test principles are valid psychologically, how does the cofunctioning of the data in any one individual make clinical sense? (3) What are the finer nuances of responses that differentiate individuals in one clinical group from others? The test is sensitive to whether thinking is integrative or pathologic, to whether fantasy is regressive, autistic, or promising of constructive imagination, and to whether emotional life is uninhibited, a painful withdrawing, or controlled release. The goal is to constantly reinforce the test’s major foundations and at the same time to discern the ever finer and finer differential diagnostic features.

Validity and research value

The big question with regard to the test relates naturally to its validity. The essence of the problem is: What can be the criterion for validity of a psychological tool that undertakes to explore the person in four dimensions—the intellectual, the emotional, the unconscious, and the unique personal dynamics (idea content in the test associations)? A single descriptive score that can take account of the myriad features in human behavior has not yet been devised.

Statistics are of course essential in this as in any psychological test. Rorschach noted his dependence on them (1921). The present writer and his colleagues have established the statistical bases in adults (Beck et al. 1950) and in children (Thetford et al. 1951). These are means and measures of dispersion for each Rorschach variable as found in the population samples. These statistics enable us to say whether the person under scrutiny deviates, and how far, from the norm. What he is like in his total adjustment is then inferred from the configuration set up by the several variables and the comparison of the configuration with those established for the several clinical groups.

The most sophisticated view that can for the present be taken is that which has been summed up for the clinical field generally by the American Psychological Association’s Committee on Clinical Tests. According to Cronbach and Meehl, “it is not a question of finding an imperfect criterion but of finding any criterion at all. The psychologist . . . cannot expect to find a clear unitary behavior criterion” ([1955] 1956, p. 180). A correspondence between a response to a Rorschach test picture and some outside, objective measure is not now within this writer’s aspirations, or within those of anyone dealing realistically with the test. The clinical information about each patient, and predictability of test results throughout the course of the patient’s treatment, must for the present be our validating source.

A psychological tool with the objective and promise of the Rorschach test was bound to stimulate new research from a variety of viewpoints. Reports of investigations with the test have increased with the years. The 1965 Mental Measurements Yearbook contains 733 Rorschach references, bringing the total up to 3,030 in the six volumes of this series published to date. As was to be expected, validating studies have been anything but one-sidedly favorable. They range from sharp refutations and rejections to unequivocal acceptance and enthusiasm. A serious difficulty in evaluating the reported publications is that very few replicate others in method, whether of administration or processing results. Almost none publish samples of the raw data, that is, the patients’ associations. Variations in basic procedures and in interpretive principles are in fact very wide. In part these are consequences of the new perspectives that have brought about alterations both in technique and in interpretation. The most radical innovations have been Klopfer’s (1942; Klopfer et al. 1954–1956). He follows the phenomenological orientation in evaluating the test’s data.

The objective of the phenomenologist—one may say his aspiration—is to know and describe the mental process of the other. This assumes the ability to enter the mind of the other. While the behaviorist sees this as a consummation devoutly to be wished, his orientation is that such entry can never be made; the objective is illusory; all that we can know of another, including his mental processes, we know from his behavior. The behaviorist can respect the phenomenologist. The school has its very able exponents. But as Mandler and Kessen put it, phenomenology is “unashamedly subjective” (1959). The behaviorist’s temperament limits him to objective data.

Other investigators who have made important contributions are, in the United States: Piotrowski, on brain pathology (1937); Molish (1958), on Navy, especially submarine, personnel, and also on a South Polar expedition; Schafer, with reference to psychoanalytic concepts (1954); Schachtel (1941; 1943), on theoretical questions; and Ames, Metraux, and Walter (1959), Hertz (1960), and Rabin and Haworth (1960), on childhood and adolescence. The principal American general texts are Klopfer (Klopfer et al. 1954–1956); Piotrowski (1957); Anderson and Anderson (1951); Rickers Ovsiankina (1960); Beck (1960); and Beck, with others (1944). In Europe, the principal names are those of Loosli-Usteri, whose work deals with disturbed children (1938); Oberholzer, on various clinical or theoretical topics (1931; 1944); and Zulliger, who published (posthumously for Behn-Eschenburg) an alternate series of test cards (see Zulliger 1941). An important Danish text is that of Bohm (1951).

Samuel J. Beck

[Directly related are the entriesPersonality measurement; Psychometrics; and the biography ofRorschach.]


Ames, Louise B.; MÉtraux, Ruth W.; and Walker, Richard N. 1959 Adolescent Rorschach Responses: Developmental Trends From Ten to Sixteen Years. New York: Hoeber.

Anderson, Harold H.; and Anderson, Gladys L. (editors) (1951) 1956 An Introduction to Projective Techniques & Other Devices for Understanding the Dynamics of Human Behavior. Englewood Cliffs, N.J.: Prentice-Hall.

Beck, Samuel J. 1960 The Rorschach Experiment: Ventures in Blind Diagnosis. New York: Grune.

Beck, Samuel J. 1966 Emotions and Understanding. International Psychiatry Clinics 3, no. 1.

Beck, Samuel J. et al. (1944) 1966 Rorschach’s Test. Volume I: Basic Processes. 3d ed. New York: Grune.

Beck, Samuel J. et al. 1950 The Normal Personality as Projected in the Rorschach Test. Journal of Psychology 30:241–298.

Binder, Hans 1932 Die Helldunkeldeutungen im psychodiagnostischen Experiment von Rorschach. Zurich: Füssli.

Bohm, Ewald B. (1951) 1958 A Textbook in Rorschach Test Diagnosis for Psychologists, Physicians and Teachers. Translated by Anne G. Beck and Samuel J. Beck. New York: Grune. → First published as Lehrbuch der Rorschach-Psychodiagnostik, für Psychologen, Ärzte, und Pädagogen.

Cronbach, Lee J. 1949 Statistical Methods Applied to Rorschach Scores: A Review. Psychological Bulletin 46:393–429.

Cronbach, Lee J.; and Meehl, P. E. (1955) 1956 Construct Validity in Psychological Tests. Pages 174–204 in Herbert Feigl and Michael Scriven (editors), The Foundations of Science and the Concepts of Psychology and Psychoanalysis. Minneapolis: Univ. of Minnesota Press. → First published in Volume 52 of the Psychological Bulletin.

Frank, Lawrence K. 1939 Projective Methods for the Study of Personality. Journal of Psychology 8:389–413.

Hallowell, A. Irving 1956 The Rorschach Technique in Personality and Culture Studies. Volume 2, pages 458–544 in Bruno Klopfer et al., Developments in the Rorschach Technique. New York: World.

Hertz, Marguerite R. 1960 The Rorschach in Adolescence. Pages 29–60 in Albert I. Rabin and Mary R. Haworth (editors), Projective Techniques With Children. New York: Grune.

Klopfer, Bruno (1942) 1946 The Rorschach Technique: A Manual for a Projective Method of Personality Diagnosis.… New York: World. → A supplement of 44 pages is added in the 1946 edition.

Klopfer, Bruno et al. 1954–1956 Developments in the Rorschach Technique. 2 vols. New York: World. → Volume 1: Technique and Theory. Volume 2: Fields of Application.

Loosli-Usteri, Marguerite (1938) 1958 Manuel pratique du test de Rorschach. 3d ed. Paris: Hermann. → First published as Le diagnostic individuel chez Venfant au moyen du test de Rorschach.

Mandler, George; and Kessen, William 1959 The Language of Psychology. New York: Wiley. Mental Measurements Yearbook. → See especially Volume 6, published in 1965.

Molish, Herman B. 1958 Schizophrenic Reactive Types in a Naval Hospital Population as Evaluated by Rorschach Test. Unpublished manuscript.

Molish, Herman B. 1959 Contributions of Projective Tests to Psychological Diagnoses in Organic Brain Damage. Pages 190–230 in Samuel J. Beck and Herman B. Molish (editors), Reflexes to Intelligence: A Reader in Clinical Psychology. Glencoe, 111.: Free Press.

Oberholzer, Emil 1931 Zur Differentialdiagnose psychischer Folgezustände nach Schädelträumen mittels des Rorschachschen Formdeuteversuchs. Zentralblattfür die gesamte Neurologie und Psychiatrie 136:596–629.

Oberholzer, Emil (1944) 1960 Rorschach’s Experiment and the Alorese. Volume 2, pages 588–640 in Cora DuBois, The People of Alor: A Social-psychological Study of an East Indian Island. Cambridge, Mass.: Harvard Univ. Press. → A paperback edition was published in 1961 by Harper.

Piotrowski, Zygmunt A. L. 1937 The Rorschach Ink Blot Method in Organic Disturbances of the Central Nervous System. Journal of Nervous and Mental Disease 86:525–537.

Piotrowski, Zygmunt A. L. 1957 Perceptanalysis: A Fundamentally Reworked, Expanded, and Systematized Rorschach Method. New York: Macmillan.

Rabin, Albert I.; and Ha Worth, Mary R. (editors) 1960 Projective Techniques With Children. New York: Grune.

Rickers-Ovsiankina, Maria A. (editor) 1960 Rorschach Psychology. New York: Wiley.

Rorschach, Hermann (1921) 1942 Psychodiagnostics: A Diagnostic Test Based on Perception. 3d ed. Bern: Huber; New York: Grune. → First published in German.

Rorschach, Hermann: and Oberholzer, E. (1923) 1924 The Application of the Interpretation of Form to Psychoanalysis. Journal of Nervous and Mental Disease 60:225–248, 359–379. → Published posthumously. First published as “Zur Auswertung des Formdeutversuches für die Psychoanalyse.” Reprinted in 1942 in Hermann Rorschach’s Psychodiagnostics.

Schachtel, Ernest 1941 The Dynamic Perception and the Symbolism of Form. Psychiatry 4:79–96.

Schachtel, Ernest 1943 On Color and Affect: Contributions to the Understanding of Rorschach’s Test. Psychiatry 6:393–409.

Schafer, Roy 1954 Psychoanalytic Interpretation in Rorschach Testing: Theory and Application. Austin Riggs Foundation, Monograph Series, No. 3. New York: Grune.

Stark, Stanley 1966a Role-taking, Empathic Imagination and Rorschach Human Movement Responses: A Review of Two Literatures. Perceptual and Motor Skills 23:243–256.

Stark, Stanley 1966b Toward the Psychology of Knowledge Ii: Two Kinds of Foresight and Foresight Theorizing. Perceptual and Motor Skills 23:547–574.

Thetford, William N.; Molish, Herman B.; and Beck, Samuel J. 1951 Developmental Aspects of Personality Structure in Normal Children. Journal of Projective Techniques 15:58–78.

Zulliger, Hans (1941) 1952 Einführung in den Behn-Rorschach Test. 3d ed. Bern: Huber.


The Thematic Apperception Test (TAT) is a projective method for the investigation of personality that consists of a series of 20 pictures about which individuals are instructed to tell stories (Murray 1943). While instructions vary somewhat among users of the test, they all request that the person being tested relate a plot with a beginning, a middle, and an end. They ask the individual to use his imagination freely in order to tell what the people are thinking and feeling. Individuals tell a story for each picture presented to them, with minimal intervention of the tester except mild encouragement and occasional instruction to be sure to include any crucial part (the beginning, the intervening middle events, or the outcome), should it be omitted.

The pictures were chosen by Henry A. Murray, Christiana D. Morgan (Morgan & Murray 1935), and the staff of the Harvard Psychological Clinic involved in the research work resulting in the volume Explorations in Personality (1938). These pictures were derived from a wide range of sources according to the judgment of that staff as to which were most consistently revealing of personality trends in subjects in clinical study. As originally presented the pictures consist of a series of 10 pictures used for all subjects and an additional group of 20 that are coded for age (prepubescent and postpubescent) and sex.

The examiner draws from these 20 an additional 10 appropriate to the age and sex of the subject being tested. Pictures are numbered and are given in a standard order reflected by this numbering.

Underlying assumptions

The underlying logic of the full test, and of the particular pictures which compose the test, rests upon the assumption that when persons are asked to assign meaning to a stimulus which is essentially ambiguous, they will tend to reveal their own basic assumptions about people and human interaction; from these statements, underlying personality trends may be deduced. This assumption leads to the selection of pictures covering a fairly wide range of usual, and some highly unusual, kinds of personal or interpersonal situations. Some pictures are considered more structured—less ambiguous in their apparent meaning—than others; that is, the situations as described are quite clear-cut, and there is little question about the identification of the scenes. The specific actions in which the persons are engaged, however, what brought them about, how they will come out, what the persons are thinking and feeling, do remain ambiguous. The subject is asked to imagine these elements in his story, and his decisions are seen as reflecting his underlying personality traits.

Elsewhere in the series are pictures that are more ambiguous in their basic identification. These less-structured pictures are normally seen as calling for greater imagination on the part of the subject. While they provide information on the response of the subject to less definite stimuli, they are not necessarily more revealing of the subject’s personality. A totally blank white card is also included.


Variations upon administration methods revolve around the questions of the number of cards to be used, whether they are given in one session or two, and the use of various laborsaving procedures. Murray originally proposed that the set be divided into two sets of 10, the second being administered one day after the first and with increased emphasis in the instructions upon full use of the imagination. Some investigators prefer this method (Stein 1948), while others (Henry 1956) utilize a single session and reduce to about 15 the total number of cards administered. Administration in groups, with projected slides, is generally not preferred but has been shown to be a useful method (Murstein 1963, pp. 45–48). Self-administered forms have been devised and found useful, at least for adult subjects of average to above-average education (Henry & Gardner 1949). Other labor-saving devices have also been used successfully during the preferred individual administration, though in each instance effort is made to maintain good rapport between subject and examiner and to reduce any possible distractive effect of such devices as shorthand or recording equipment.


Any of a variety of psychodynamic personality theories are appropriate to the analysis of the TAT for its original purpose—the study of unconscious and underlying aspects of personality. Murray, in his manual for testers (1943), suggested a framework based upon his conceptual analysis of personality in need-press terms. This analysis provides a series of categories of motivational states. Needs refer to inner states that can possibly be expressed, for example, N achievement (need for achievement), N affiliation, N sex, N order. Press refers to inner states, but to those involving a perception of some outer event or force acting upon the individual. The analysis of TAT stories in this system involves the translation, using a set of definitions, of story content into the appropriate need or press and the interconnecting of these categories into a thema descriptive of the entire story. The need-press or the combination of them into a thema is thought to represent motivational trends in the subject. To derive a personality description or a clinical diagnosis from such statements of need, press, and thema still requires that the examiner interpret the material along the lines of some more comprehensive personality theory. This is commonly done in some form of psychoanalytically based theory. Some workers with the test (e.g., Bellak 1954) see their analyses as dealing with unconscious elements of personality or follow a strongly psychoanalytic form of analysis. Some, still within this essentially clinical framework of individual case analysis, see their analyses as, in addition, reflecting upon preconscious, ego-related aspects of behavior. This latter focus of analysis would include more attention upon the cognitive aspects of personality and social role preferences and would lead more directly to examining underlying motivations in relation to behavioral events.

Procedures in analysis and theoretical concepts used depend somewhat upon the conception of the examiner as to the basic nature of TAT data. Most interpreters who have published statements of their methods assume that the TAT data represent interactions between deeper personality trends and current images of interpersonal situations. While the TAT data are clearly prompted in part by underlying dynamic forces, it is apparent that the stories as told are also influenced by various other factors. These include some cognitive abilities of the subject, his understanding of the formal instructions, his assumptions about the nature of a story, the unique properties of the pictures themselves, and the fact that the test-taking act and circumstances are themselves social interactions. Holt (1961) provides an important discussion of this topic. For documentation of specific methods and their logic and conceptual background, see Bellak (1954), Henry (1956), Tomkins (1947), Stein (1948), Holt (1958), Schafer (1958), and Wyatt (1947). An interpretive lexicon is provided by Lindzey and his associates (1959). This lexicon is not a system of analysis but rather a summary of findings from a variety of empirical studies or of differentiating properties of TAT stories proposed by various interpreters. A summary and analysis of common assumptions regarding interpretive logics is to be found in Lindzey (1952).

In general these systems of interpretation rest upon the earlier stated assumptions that the story as told is, in large part, the invention of the storyteller and that the properties characteristic of that invention stem from tendencies characteristic of the subject. Interpreters vary in the emphasis which they give to the determining influence of the stimuli pictures themselves, though in principle all acknowledge that influence. The analysis of the properties of the stories thought relevant to the personality analysis and the particular relevance proposed may best be seen in the references given in the preceding paragraph. For illustrative purposes, however, the properties of stories suggested by Henry (1956) will suffice. In broad outline, it is suggested that stories may be analyzed in two basic categories with various subdivisions. These are content and form. Content calls attention to the fact that stories differ in topic, in plot, in characteristics attributed to persons, in actions and emotions reported, in the interrelations of persons, and in the outcome and fate of all of these. Form refers to the manner in which any particular story is told. It would call attention to the length and linguistic facility, to the sophistication or crudeness of language, to the order and organization of plots, and to the utilization of large inclusive concepts or observations of small details. The nature of the sequence of time (references to past, present, future events), the clarity of portrayal of the actual recognizable elements of the stimulus, and the degree of conformity to the instruction as given would also be considered. Having observed such properties as these for all stories told, the interpreter must then attribute meaning to them in personality terms and order those terms into the conceptual framework he finds appropriate. It is at this point that the most skill is called for and from this process that the widest variation in resulting analyses may arise. It is perhaps appropriate here to suggest that the TAT is more a method of investigating personality than a test of it. In this sense, and in this individual personality use, it is a clinical investigatory aid, not an instrument with known absolute properties reflective of definite and circumscribed personality attributes. For this clinical use the TAT provides an order of data on inner processes and on their relation to outer events, which yields its fullest meaning only upon study and analysis by the interpreter already skilled in personality theory and experienced with wide varieties of social and personality groupings. In this sense the TAT response constitutes a social and a personal document as complex and variegated as the subject providing the data, and as tortuous and subtle as most efforts to condense and systematize the life career of any individual.

Reliability and validity

Complexity by no means suggests that issues of reliability and validity are to be resolved by estimates of the professional competence of the interpreter. The real issues of reliability or validity for the TAT are complex ones; they involve far more than the usual definition of reliability as reproducibility either of specific elements of stories on the part of a subject or of analytic statements from such data on the part of interpreters—or the definition of validity as a correlation between analytic statements and an external criterion of the same attribute. The problems are methodological ones of a nature as yet unresolved. They involve both procedural and statistical matters, as well as decisions regarding the specificity or abstractness of statements from both TAT and criterion data. In general these problems must be resolved for projective data as a type of data before any specific projective instrument can soundly be assayed. To date, the evidence for the reliability and validity of the TAT is extremely ambiguous. Jensen (1959) and Eron (1959), in reviewing the test for the fifth Mental Measurements Yearbook, report many negative or inconclusive findings. Studies reporting validity have varied greatly in method, in kinds of criteria used, and in the kinds of variables from the TAT used. The best summary of the state of affairs regarding these topics is provided by reviewing the general analyses and proposals regarding the problem in the works by Cronbach (1948; 1956) and Murstein (1963). A recent study by Shipman (1964) is of particular interest in that it deals with several kinds of units of analysis from the TAT and provides evidence of differential degrees of congruence with an external criterion. These differences depend in some instances upon the specific versus the global nature of the units dealt with in the TAT and in part upon the personality constructs available from the criterion instrument. The study which most directly parallels the normal clinical use of the TAT and which attempts to reflect directly upon varying types of validity is that of Henry and Farley (1959). This study is based upon the design proposed by Cronbach (1948). It relates blind analysis of TAT tests of 36 adolescents from a homogeneous social environment to a variety of criterion instruments—life history from interviews, the Rorschach test, objective tests, intelligence tests, sociograms—representing data from overt levels of behavior, subjective levels, and projective levels. The authors believe that a crucial element in the substantially significant results obtained is the conscious and exhaustive attention given to the understanding and definition of the system of personality constructs in terms of which all instruments were analyzed and interpreted. [SeePsychometrics.]

Modifications and applications

The Murray TAT was developed for use with subjects from the mainstream of American social life. While the pictures used were intended to refer to personality issues that are not culture-bound, the fact that these pictures contain specific objects, clothes, and physical settings has tended to lend them a cultural specificity which restricts their use. In some instances, therefore, the TAT principles have been maintained but new pictures designed. There have been instances where it was felt that there was some inappropriateness in the pictures or where a shorter and more focalized set was desired. The Children’s Apperception Test (Bellak 1954) and the Michigan Picture Test (Andrew et al. 1951) are examples designed for younger age groups on the grounds that the general tone of the Murray is too adult. In the Michigan Picture Test, in addition, the focus was upon elementary school adjustment, and thus the pictures were developed to reflect many school settings. One special set has been developed by Thompson (1949) for use with American Negroes. It duplicates the regular Murray cards, but all figures are Negro. Special sets have also been composed for the study of elderly persons (Lieberman & Larkin 1963) and for the study of personality in work adjustment of business executives (Henry & Moore 1950). The Blacky pictures (Blum 1949) consist of animals in humanlike scenes and are designed for the study of children. Lennep (1948; 1951) has prepared a 4-picture set used in educational and industrial settings. A special set by Phillipson (1955) consists of particularly vague pictures aimed at the study of object relations in the psychoanalytic sense.

Cross-cultural, role, and group studies

The principal use of the TAT for cross-cultural purposes has focused upon the study of nonliterate societies. Here it is normally assumed that the modern, middle-class tone of most Murray TAT pictures is inappropriate. For these reasons special sets have been developed for several nonliterate societies. These tend to follow the principle of design as illustrated by the Murray but are modified for special problems of the society and for general cultural appropriateness (Henry 1956, pp. 47–53; Henry 1951). The experience with such sets has varied. Generally they have been felt to make a useful and positive contribution to cross-cultural study, though the formal evidence of their validity is mixed. The best presentation and analysis of such use may be found in Lindzey’s Projective Techniques and Cross-cultural Research (1961).

These special sets are commonly used in connection with problems having anthropological or sociological emphasis, as well as personality concerns. In some instances, particularly the special sets for nonliterate societies, the intention has been to examine personality factors as they relate to cultural differences or as they form the psychic correlates for intracultural variations. Increasingly, the TAT is also being used in contexts where social roles are at issue, especially where the investigator sees the role behavior as having personality components. This was the focus of Henry’s study of business executives, where the aim was to clarify certain personality components of the behavior of executives in large organizations rather than to analyze the personality of the individual adult working in that setting. In some instances, other social role concerns have directed the selection of pictures to be used and indicated the mode of analysis. In a study of adults in Kansas City (Neugarten et al. 1964), four of the regular Murray cards were given, but the fifth was a specially designed card. It contained four persons in a nondescript living-room setting. One was clearly an older male, one an older female, one a younger male, and one a younger female. They could readily be seen as the older parents of a younger couple. In designing this picture and in planning its analysis, Neugarten and Guttman (1958) were interested in age-graded perceptions of male and female figures. Their interest was in the differential perception of these age roles and sex roles by subjects of differing ages. With such pictures the instructions normally remain the same—tell a story—though these may be supplemented by specific directive questions. Similarly modified sets might be used to study supervisor-worker situations in industry, doctor-nurse interactions, teams inspecting new automobiles, interactive small groups, racial conflict situations—in fact, almost any social interaction that lends itself to portrayal in this ambiguous pictorial form.

The use of the TAT in the study of groups has had several focuses. One revolves around the study of interactive processes in groups themselves, usually groups of modest size. Another revolves around interest in properties common to various natural social groups, sometimes in cross-cultural contexts and sometimes in cross-class or cross-age contexts within a single society. The Group Projection Sketches (Henry & Guetzkow 1951) is a special set of five cards designed to reflect various concepts for the analyses of small-group interaction. The analysis of TAT data in various racial or social groups has a somewhat different focus. It is best seen as overlapping with cross-cultural studies in which properties of stories common to particular cultures is the issue (see Lindzey 1961). A closely related use involves the social-psychological study of subgroups of our own society. The monograph of Warner and Henry (1948) is one illustration. In the effort to provide data on the functional significance of the daytime radio serial for the normal listeners of those programs, the authors attempt to identify common psychological concerns and modes of adjustment in their principal group of subjects, middle-class American housewives. A more extensive study utilizing this and other methods is that of Rainwater, Coleman, and Handel (1959) on lower-class American housewives. A special system of analysis has been developed by Schaw and Henry (1956); they illustrated its use by comparing three groups of business executives differing in age.

Personality variables

A common use of the TAT has been the study of specific personality variables under a wide range of circumstances. Here the interest is on the TAT as a measuring device of personality variables thought to be of unconscious or “beyond awareness” nature—hostility, anxiety, achievement drives, and other motives such as affiliation and power. The most extensive work has been that of McClelland and his associates (1953) on need-achievement as defined in the Murray need-press system of motivational categories. The general logic of this approach is that if achievement imagery is present in stories elicited by selected TAT cards, then the subject may be said to be motivated to achieve. Efforts to demonstrate that logic have varied considerably: some show that persons with high achievement imagery do indeed attempt to improve their performance or that they are in real life “high achievers,” while others clearly disprove these same points. There has also been substantial work on the variable of hostility. Complex relationships tend to be found, which suggests that there is a positive but by no means clear-cut relation between such variables and any particular form of overt behavior. The most exhaustive treatment of studies in this area may be found in Murstein (1963, pp. 84–106, 300–322). [SeeAchievement motivation; Field theory; Role.]

Special scoring systems which deal with somewhat more complex units have been devised for the study of particular variables. Arnold (1962) has been interested in achievement and has developed a system of analysis focused upon story themes and a scoring method dealing with motivation to achieve. This method shows some interesting differences between high-achievement and lowachievement students in school, among schoolteachers and business executives, and among welladjusted and poorly adjusted men in the Navy. Dana (1959) has devised a system based upon degrees of approximation to various expected story form and content properties and uses it to distinguish normals, neurotics, and psychotics.

An additional type of variable has been used by Rosen and Neugarten (1960), who have utilized the TAT as a source of data on ego energy. In this instance they proposed that ego energy would be characterized by (1) the ability to integrate wide ranges of stimuli, (2) the readiness to perceive or deal with complicated, challenging, or conflictful situations, (3) the tendency to perceive vigorous and assertive activity, and (4) the tendency to perceive feelings and affects as related to life situations. As measures of these attributes, they chose certain elements of TAT stories commonly used in clinical analysis but here treated only as scorable indicators of these specific ego elements. The central point of relevance about this study, illustrative of others of related type, is that the TAT was used as a source of data for certain properties of ego functioning selected as a test of a theory. Specific TAT response categories were selected as scorable measures of these properties and the behavior of these measures predicted and tested in an appropriate design. In this format, reliability is handled by the common interjudge-agreement method. Validity is thus subsumed under issues of the viability of constructs used in the theoretical formulation, and the usefulness of the TAT estimates of those constructs is determined by discovering relationships to other data.


The TAT was developed to aid in the “exploration” of personality (Explorations in Personality 1938). It has quite justified its existence in the sense of the access which it provides to personality variables, especially when these are seen as egorelated—as opposed to deep unconscious fantasies —and as they are seen as reflecting the juncture of psychodynamics and social behavioral events. The TAT is best viewed as a source of data on this order of personality event rather than as a test of specific personality attributes of known character. It can, nevertheless, assume testlike properties under certain conditions. These conditions are, generally speaking, outside the instrument itself and reside in the manner in which it is used, the data to which it is related, and the personality constructs in terms of which the basic data are ordered. Its successes and failures, in the usual validity sense, are explained by analysis of these external events. In the usual clinical use, for personality description, considerable variation in stability appears. Such variations may be seen in terms of the specificity–generality of personality units used and in terms of the order of corresponding external criteria. The clarification of the use and misuse of the instrument will rest upon continued exploration of these factors—the personality constructs utilized, the logic of translating data into these constructs, the manner of relating these constructs to other data about the individual. Among the sensitivities, as opposed to the stabilities of the instrument, are those reflecting change and ambiguity in the individual, factors which confound the effort to distinguish error from genuine difference. The future of the instrument will be decided by the study of other conditions of use, by increased sophistication in the manipulation of the basic data, and by increases, from any source, in the user’s knowledge of psychodynamics and of social systems. A crucial element in the future use of the TAT may be found in the degree of richness and flexibility of the basic concept and in the fact that constructs from several fields of psychological and social science investigation may be the focus of its use, in the original as well in as modified forms.

William E. Henry

[See alsoPersonality measurement. Other relevant material may be found inAchievement motivation; Content analysis; Fantasy; Motivation, article onhuman motivation; Psychoanalysis; Psychometrics.]


Andrew, Gwen et al. 1951 The Michigan Picture Test: The Stimulus Values of the Cards. Journal of Consulting Psychology 15:51–54.

Arnold, Magda B. 1962 Story Sequence Analysis: A New Method of Measuring Motivation and Predicting Achievement. New York: Columbia Univ. Press.

Aron, Betty R. 1949 A Manual for Analysis of the Thematic Apperception Test: A Method and Technique for Personality Research. Berkeley (Calif.): Berg.

Bellak, Leopold 1954 The Thematic Apperception Test and the Children’s Apperception Test in Clinical Use. New York: Grune.

Blum, Gerald S. 1949 A Study of the Psychoanalytic Theory of Psychosexual Development. Genetic Psychology Monographs 39:3–99.

Cronbach, Lee J. 1948 A Validation Design for Qualitative Studies of Personality. Journal of Consulting Psychology 12:365–374.

Cronbach, Lee J. 1956 Assessment of Individual Differences. Annual Review of Psychology 7:173–196.

Dana, Richard H. 1959 Proposal for Objective Scoring of the TAT. Perceptual and Motor Skills 9:27–43.

Eron, Leonard D. 1959 Thematic Apperception Test. Mental Measurements Yearbook 5:306–310.

Explorations in Personality: A Clinical and Experimental Study of Fifty Men of College Age, by Henry A. Murray et al. 1938 New York and London: Oxford Univ. Press.

Henry, William E. (1951) 1956 The Thematic Apperception Technique in the Study of Group and Cultural Problems. Pages 230–278 in Harold H. Anderson and Gladys L. Anderson (editors), An Introduction to Projective Techniques & Other Devices for Understanding the Dynamics of Human Behavior. Englewood Cliffs, N.J.: Prentice-Hall.

Henry, William E. 1956 The Analysis of Fantasy: The Thematic Apperception Technique in the Study of Personality. New York: Wiley.

Henry, William E.; and Farley, Jane 1959 The Validity of the Thematic Apperception Test in the Study of Adolescent Personality. Psychological Monographs 73, no. 17.

Henry, William E.; and Gardner, Burleigh B. 1949 Personality Evaluation in the Selection of Executive Personnel. Public Personnel Review 10:67–71.

Henry, William E.; and Guetzkow, Harold 1951 Group Projection Sketches for the Study of Small Groups. Journal of Social Psychology 33:77–102.

Henry, William E.; and Moore, Harriett B. 1950 The Industrial Use of Personality Tests. Advanced Management 15, no. 11:16–20.

Holt, Robert R. 1958 Formal Aspects of the TAT: A Neglected Resource. Journal of Projective Techniques 22:163–172.

Holt, Robert R. 1961 The Nature of TAT Stories as Cognitive Products: A Psychoanalytic Approach. Pages 3–50 in Conference on Contemporary Issues in Thematic Apperceptive Methods, Fels Research Institute, 1959, Contemporary Issues in Thematic Apperception Methods. Springfield, III.: Thomas. → Contains six pages of discussion.

Jensen, A. E. 1959 Thematic Apperception Test. Mental Measurements Yearbook 5:310–313.

Lennep, David J. Van 1948 Four Picture Tests (1930). The Hague: Nijhoff.

Lennep, David J. Van (1951) 1956 The Four Picture Test Pages 149–180 in Harold H. Anderson and Gladys L. Anderson (editors), An Introduction to Projective Techniques & Other Devices for Understanding the Dynamics of Human Behavior. Englewood Cliffs, N.J.: Prentice-Hall.

Lieberman, Morton A.; and Larkin, Martin 1963 On Becoming an Institutionalized Aged Person. Volume 1, pages 475–503 in Richard H. Williams, C. Tibbitts, and W. Donahue (editors), Processes of Aging: Social and Psychological Perspectives. New York: Atherton.

Lindzey, Gardner 1952 Thematic Apperception Test: Interpretative Assumptions and Related Empirical Evidence. Psychological Bulletin 49:1–25.

Lindzey, Gardner 1961 Projective Techniques and Cross-cultural Research. New York: Appleton.

Lindzey, Gardner et al. 1959 Thematic Apperception Test: An Interpretive Lexicon for Clinician and Investigator. Journal of Clinical Psychology Monograph Supplement no. 12.

Mcclelland, David C. et al. 1953 The Achievement Motive. New York: Appleton.

Meehl, Paul E. 1959 Structured and Projective Tests: Some Common Problems in Validation. Journal of Projective Techniques 23:268–272.

Morgan, Christiana D.; and Murray, Henry A. 1935 A Method for Investigating Fantasies: The Thematic Apperception Test. Archives of Neurology and Psychiatry 34:289–306.

Murray, Henry A. 1943 Thematic Apperception Test Manual. Cambridge, Mass.: Harvard Univ. Press.

Murstein, Bernard I. 1963 Theory and Research in Projective Techniques, Emphasizing the TAT. New York: Wiley.

Neugarten, Bernice L.; and Guttman, David L. 1958 Age–Sex Roles and Personality in Middle Age: A Thematic Apperception Study. Psychological Monographs 72, no. 17.

Neugarten, Bernice L. et al. 1964 Personality in Middle and Late Life. New York: Atherton.

Phillipson, Herbert 1955 The Object Relations Technique. London: Tavistock.

Rainwater, Lee; Coleman, R. P.; and Handel, G. 1959 Workingman’s Wife: Her Personality, World, and Life Style. Dobbs Ferry, N.Y.: Oceana.

Rosen, Jacqueline L.; and Neugarten, Bernice L. 1960 Ego Functions in the Middle and Later Years: A Thematic Apperception Study of Normal Adults. Journal of Gerontology 15:62–67.

Schafer, Roy 1958 How Was This Story Told? Journal of Protective Techniques 22:181–210.

Schaw, Louis C ; and Henry, William E. 1956 A Method for the Comparison of Groups: A Study in Thematic Apperception. Genetic Psychology Monographs 54:207–253.

Shipman, William G. 1964 TAT Validity: Congruence With an Inventory. Journal of Projective Techniques 28:227–232.

Stein, Morris I. (1948) 1955 The Thematic Apperception Test: An Introductory Manual for Its Clinical Use With Adults. 2d ed., rev. Reading, Mass.: Addison-Wesley.

Thompson, Charles E. 1949 The Thompson Modification of the Thematic Apperception Test. Rorschach Research Exchange 13:469–478.

Tomkins, Silvan S. 1947 The Thematic Apperception Test: The Theory and Technique of Interpretation. New York: Grune.

Warner, W. Lloyd; and Henry, William E. 1948 The Radio Day Time Serial: A Symbolic Analysis. Genetic Psychology Monographs 37:3–71.

Wyatt, Frederick 1947 The Scoring and Analysis of the Thematic Apperception Test. Journal of Psychology 24:319–330.