TIII: RELATIIIIISHIT BETWEEN MYERS-BRIGGS ,. g ,3; PERSONALITY CHARACTERISTICS AND - » TEACHING PREFERENCESOF _ . , If? ” PROSPECTIVE TEACHERS: A Dissertoflon g}; {or the chno a"? PLUS 7:: :. MICHIGAN STATE UNIVERSITY . ,1 .- "" “2’ ~ - . V I.“ . I ., , “I, M‘ .< ..I ‘ , u v ”an.” ‘ ‘ M ,_ :,.,.1.vr9'.I-m. _‘ .4% W H M“ -. :~-:.~.e-.~:;-¢.S “it This is to certify that the thesis entitled THE RELATIONSHIP BETWEENVMYERS-BRIGGS PERSONALITY CHARACTERISTICS AND TEACHING PREFERENCES OF PROSPECTIVE TEACHERS presented by Marcia Carlyn has been accepted towards fulfillment of the requirements for Ph.D. degree in Secondary Education and Curriculum (Social and Philosophical Foundations) kafl, £¢g4445g§74 Major professor Date May 7, 1976 0-7 639 THE RELATIONSHIP BETWEEN MYERS—BRIGGS PERSONALITY CHARACTERISTICS AND TEACHING PREFERENCES 0F PROSPECTIVE TEACHERS By Marcia Carlyn AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1976 ABSTRACT THE RELATIONSHIP BETWEEN MYERS—BRIGGS PERSONALITY CHARACTERISTICS AND TEACHING PREFERENCES 0F PROSPECTIVE TEACHERS By Marcia Carlyn The major problem of the study was to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. Personality type was measured by the Myers— Briggs Type Indicator and teaching preferences were measured by the Teacher Preference Questionnaire, an instrument constructed for the study. The sample consisted of 200 pre—service teachers, all sen— iors majoring in elementary or secondary education at Michigan State University. The Myers—Briggs Type Indicator (MBTI) is a self—report instru— ment based on Carl Jung's theory of personality. The Indicator con— sists of 166 items designed to measure the following personality pref- erences: (1) Extraversion or Introversion, (2) Sensation or Intuition, (3) Thinking or Feeling, and (4) Judgment or Perception. The instru— ment yields four continuous scores and four dichotomous scores (type categories) for each person. MBTI scores for the present sample appeared to be satisfactorily reliable, with internal—consistency reliability coefficients ranging from .77 to .90 for continuous scores and from .67 to .92 for type categories. Marcia Carlyn The Teacher Preference Questionnaire (TPQ) is a self—report instrument which was constructed for the study. The questionnaire consists of 103 items designed to measure a teacher's attitude toward a variety of teaching situations. Cluster analysis was per— formed on the Likert—scale TPQ items and resulted in the following six clusters: (1) Preference for teaching lower grades, (2) Interest in administrative functions, (3) Need for independence and creativity, (4) Interest in planning school projects, (5) Interest in working with small groups of students, and (6) Commitment to classroom teaching. Each person's questionnaire responses were transformed into six com— posite cluster scores. The TPQ clusters appeared to be satisfactorily reliable, with Coefficient Alphas ranging from .70 to .81. Six hypotheses were formulated in order to test whether signi— ficant relationships exist between the personalities of prospective teachers and their preferences toward certain teaching situations. The hypotheses were tested in two ways, first by treating MBTI scores as continuous data and then by treating MBTI scores as dichotomous type categories. In both cases, scores on each of the four MBTI in— dices were compared with scores on a particular TPQ cluster. When MBTI scores were treated as continuous data, Pearson product—moment correlation coefficients were computed in order to estimate the relationship between personality and teaching preference. When MBTI scores were treated as dichotomous type categories, three statistical procedures were used: (1) Point-biserial correlation coefficients were computed to produce minimum estimates of relation— ship between personality type and teaching preference; (2) Biserial '—————~‘ Marcia Carlyn correlation coefficients were computed to produce maximum estimates of relationship between personality type and teaching preference; and (3) One—way analysis of variance was used to approximate the significance level of the F statistic obtained when MBTI type cate— gory was regarded as the independent variable and TPQ cluster score was regarded as the dependent variable. An alpha level of .05 was set for rejecting the null hypotheses. Alternate hypotheses were accepted when the statistical measures produced significant results in the predicted direction. The study produced a number of significant findings. Results of the statistical tests supported the following general conclusions: 1. Feeling types are more interested in teaching at lower levels than thinking types. 2. Extraverted and thinking types are more interested in administrative functions than introverts and feeling types. 3. Intuitive teachers have a stronger need for inde— pendence and creativity than sensing types. 4. Extraverted and intuitive teachers are more inter— ested in planning school projects than introverts and sensing types. 5. Intuitive teachers are more interested in working with small groups of students than sensing types. The findings indicate that teachers of different personality types prefer different kinds of teaching situations. THE RELATIONSHIP BETWEEN MYERS—BRIGGS PERSONALITY CHARACTERISTICS AND TEACHING PREFERENCES 0F PROSPECTIVE TEACHERS By Marcia Carlyn A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1976 © Copyright by MARC IA CARLYN 1976 m To Marvin S iegel a“. ACKNOWLEDGMENTS I wish to express my sincerest appreciation and gratitude to the following persons: to Dr. Keith Anderson from whom I've learned so very much, including a truer appreciation for the differences in people; to Dr. William Farquhar for his high standards, his careful analysis, and his crazy sense of humor; to Dr. Max Raines for his great enthusiasm; to Dr. Marvin Grandstaff who provided a balance with his natural skepticism; to Dr. Maryellen McSweeney for her un— failing patience with my statistical queries; to Isabel Briggs Myers for her kind nature and her inspiring curiosity about people; and to the many others who helped along the way. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . vii CHAPTER I° INTRODUCTION . . . . . . . . . 1 Importance of the Study . . . . . . Statement of the Problem . . . . . Hypotheses . . . . . . Organization of the Study . . . . . CHAPTER II. JUNG'S TYPOLOGICAL THEORY . . . . . . 7 The Myers—Briggs Type Indicator . . . . 8 Descriptions of the Preferences . . . . 10 Extraversion—Introversion . . . . 1O Sensation—Intuition . . . . . . 11 Thinking—Feeling . . . . . . 13 Judgment—Perception . . . . . . 14 Summary . . . . . . . . . . 15 CHAPTER III. REVIEW OF RELATED LITERATURE . . . . . 17 Personality of In—Service Teachers . . . 18 Personality of Pre—Service Teachers . . . 23 Personality and Grade Level . . . . . 27 Personality and Subject Area . . . . 29 Personality of School Administrators . . 33 Personality and Teaching Style . . . 36 Personality and Satisfaction with Teaching . 39 Personality and Teacher Effectiveness . . 42 Summary . . . . . . . . . . 47 CHAPTER IV. DESIGN OF THE STUDY . . . . . . . . 50 Sample . . . . . . . . . . 50 Hypotheses . . . . . . . 53 Statistical Analysis . . . . . 54 Teacher Preference Questionnaire . . . 54 Pilot Study . . . . . . . . 54 Content Analysis . . . . . . 55 Myers— Briggs Type Indicator . . . . . 60 Summary . . . . . . . . . 60 CHAPTER V. MYERS-BRIGGS TYPE INDICATOR . . . . . . 62 Scoring 1 . . . . . . 62 Type Categories . . . . . . . 63 Continuous Scores . . . . . . 66 Frequency Distributions . . . . . 67 Analyzing MBTI Scores . . . 70 Statistical Procedures for Type Categories 70 Statistical Procedures for Continuous Scores . . . 72 Intercorrelations of MBTI Scores . . . 73 Type Categories . . . . . . . 74 Continuous Scores . 77 Reliability of the Myers—Briggs Type Indicator 81 Measures of Internal Consistency . 81 Type Categories . . . . . . 81 Continuous Scores . . . . . 85 Measures of Stability . . . . . 88 Type Categories . . . . . . 88 Continuous Scores . 90 Validity of the Myers—Briggs :Type Indicator . 92 Content Validity . . . . 92 Predictive Validity . . . . . . 94 Construct Validity . . . . . . 96 E—I Validity . . . . . . 98 S—N Validity . . . . . . 99 T—F Validity . . . . . . 101 J—P Validity . . . . . . 103 Validity of Type Combinations . . 105 Summary . . . . . . . . . . 106 CHAPTER VI. FINDINGS . . . . . . . . . . 110 Presentation of the Data . . . . . 112 Hypothesis 1 . . . . . 112 Hypothesis 2 . . . . . 113 Hypothesis 3 . . . . 115 Hypothesis 4 . . 117 Hypothesis 5 . . . . . 121 Hypothesis 6 . . . 123 Additional Findings . . . . . . . 125 Summary . . . . . . . . . . 129 Page CHAPTER VII. SUMMARY AND CONCLUSIONS . . . . . . . 132 Summary of the Study . . . . . . 132 Discussion of the Findings and Recommendations for Future Research . . 134 Hypothesis 1 . . . . . . . 134 Hypothesis 2 . . . . . . . 136 Hypothesis 3 . . . . . . o 138 Hypothesis 4 . . . . . . . 139 Hypothesis 5 . . . . . . . 141 Hypothesis 6 . . . . . . 142 Additional Recommendations . . . . 143 Conclusions . . . . . . . . . 145 APPENDIX A. PRELIMINARY FORM OF TEACHER PREFERENCE QUESTIONNAIRE . . . . . . . . 1216 .APPENDIX B. TEACHER PREFERENCE QUESTIONNAIRE . . . . 153 APPENDIX C. PROCEDURE FOR GRAPHING THE FREQUENCY DISTRIBUTIONS 0F CONTINUOUS MBTI SCORES . . . 165 APPENDIX D. PROCEDURE FOR ESTIMATING THE INTERNAL—CONSISTENCY RELIABILITY OF MBTI SCORES . . . . . . 167 BIBLIOGRAPHY . . . . . . . . . . . . . 169 V1 LIST OF TABLES Table Page 2.1 Sixteen MBTI Personality Types . . . . . . . 9 2.2 Basic Preferences Measured by the Myers—Briggs Type Indicator . . . . . . . . . 16 3.1 MBTI Type Preferences of In—Service Teachers (From Previous Studies) . . . . . . . . . . 20 3.2 MBTI Type Preferences of Male and Female In—Service Teachers (From Previous Studies) . . . . . . 21 3.3 MBTI Type Preferences of Pre-Service Teachers (From Previous Studies . . . . . . . u . . 24 3.4 MBTI Type Preferences of Male and Female Pre—Service Teachers (From Previous Studies) . . . . . . 25 3.5 MBTI Type Preferences Associated with Different Subject Areas . . . . . . . . . . 31 3.6 MBTI Type Preferences Associated with Different Teaching Majors . . . . . . . . . . . 32 3.7 MBTI Type Preferences of School Administrators (From Previous Studies) . . . . . . . . . . 35 4.1 Distribution of the Sample by Grade Level . . . . 51 4.2 Distribution of the Sample by Subject Area . . . . 52 4.3 Results of Cluster Analysis with the Teacher Preference Questionnaire . . . . . . . . . . . 57 4.4 Means and Standard Deviations of TPQ Cluster Scores . . 59 5.1 MBTI Type Table for the Present Sample . . . . . 64 5.2 MBTI Type Table for Males and Females in the Present Sample . . . . . . . . 65 Table Page 5.3 Frequency Distributions of Continuous MBTI Scores for the Present Sample . . . . . . . . . 68 5.4 Frequency Distributions of Continuous MBTI Scores for Males and Females in the Present Sample . . . . 69 5.5 Intercorrelations of MBTI Type Categories (From Previous Studies) . . . . . . . . . . 75 5.6 Intercorrelations of MBTI Type Categories for the Present Sample . . . . . . . . . . . 76 5.7 Intercorrelations of Continuous MBTI Scores (From Previous Studies) . . . . . . . . . . 79 5.8 Intercorrelations of Continuous MBTI Scores for the Present Sample . . . . . . . . . . . 80 5.9 Estimates of Internal-Consistency Reliability of MBTI Type Categories (From Previous Studies) . . . 83 5.10 Estimates of Internal—Consistency Reliability of MBTI Type Categories for the Present Sample . . . . 84 5.11 Estimates of Internal—Consistency Reliability of Continuous MBTI Scores (From Previous Studies) . . . 86 5.12 Estimates of Internal—Consistency Reliability of Continuous MBTI Scores for the Present Sample . . . 87 5.13 Comparison of Original and Retest Type Category Scores (From Previous Studies) . . . . . 89 5.14 Test—Retest Correlation Coefficients for Continuous MBTI Scores (From Previous Studies). . . . . . 91 6.1 Correlations Between Teacher Personality and Preference for Teaching Lower Grades . . . . . . . . 114 6.2 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 1 . . . . . . . . . . . 114 603 Correlations Between Teacher Personality and Interest in Administrative Functions . . . . . . . . 116 6.4 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 2 . . . . . . . . . . . 116 viii s. . . .. .. . . . I . . O . .. .. I l n .1 u . . .L . . h I a .. u . . u . . . . . o .. . . o . . .. I i n t l Table Page 6.5 Correlations Between Teacher Personality and Need for Independence and Creativity . . . . . . . . 118 6.6 One-Way Analysis of Variance for Mean Differences on TPQ Cluster 3 . . . . . . . . . . . 118 6.7 Correlations Between Teacher Personality and Interest in Planning School Projects . . . . . . . . 120 6.8 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 4 . . . . . . . . . . . 120 6.9 Correlations Between Teacher Personality and Interest in Working with Small Groups . . . . . . . 122 6.10 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 5 . . . . . . . . . . . 122 6.11 Correlations Between Teacher Personality and Commitment to Classroom Teaching . . . . . . . . . 124 6.12 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 6 . . . . . . . . . . . 124 6.13 Comparison of TPQ Cluster Scores of Males and Females . 126 6.14 Correlations Between Continuous MBTI Scores and TPQ Cluster Scores of Males and Females . . . . . . 127 6.15 Summary of the Findings . . . . . . . . . 130 CHAPTER I INTRODUCTION Since the beginning of the century, educational researchers have been attempting to identify the personality correlates of successful teaching behavior. Most researchers and administrators believe that personality factors contribute significantly to success in teaching, but no concensus seems to exist on which specific person— ality factor or factors are characteristic of a successful teacher. In their classic summary of research on teacher personality, Getzels and Jackson concluded: Despite the critical importance of the problem and a half—century of prodigious research effort, very little is known for certain about the nature and meas— urement of teacher personality, or about the relation between teacher personality and teaching effectiveness. Lieberman's conclusion in 1956 appears to be still valid today: "Insofar as our present instruments and evaluation can tell us, many different kinds of persons have been good teachers."2 If such a large quantity of research has yielded such negligible results, perhaps a different approach to the problem is warranted. 1J. W. Getzels and P. W. Jackson, ”The Teacher's Personality and Characteristics," Handbook of Research on Teaching, ed. N. L. Gage (Chicago: Rand McNally & Co., 1963), p. 574. 2Myron Lieberman, Education as a Profession (Englewood Cliffs, N.J.: Prentice—Hall, 1956 , p. 255. f0 Andrews suggests that it may be more useful to investigate the differences between teachers than to attempt generalizations to all teachers.3 And Brim urggs researchers to "move away from the stul— tifying conception of some single optimum type of teacher," and examine instead the possibility that teachers of given character— istics may be more effective in certain situations and less effective in other situations.“ Shifting from the simplistic "ideal—teacher" approach to a situational approach introduces theoretical and methodological com— 5 plexities, but holds promise of yielding more valuable results. It would seem logical that the interaction of teacher personality and teaching situation, largely overlooked in previous studies of teacher effectiveness, can be a significant factor affecting teacher performance. The present study was designed to investigate the proposed interaction of teacher personality and teaching situation. Speci— fically, it is an examination of whether prospective teachers of different personality types indicate preferences for different kinds of teaching situations. The study was based on the assumption that teachers are likely to be more effective when placed in situations which they prefer. 3John H. M. Andrews, ”Administrative Significance of Psychological Differences Between Secondary Teachers of Different Subject Matter Fields" (unpublished doctoral dissertation, University of Chicago, 1957). p- 143- llOrville G. Brim, Jr., Sociology and the Field of Education (New York: Russell Sage Foundation, 1958), p. 31. 5Getzels and Jackson, p. 533. v1 Importance of the Study The Getzels—Guba theory of administration supports the propo— sition that teacher performance is dependent on the personality of the individual teacher as well as the particular teaching situation. According to this model, personality and situational variables are the "primary determinants" of behavior in any organization.6 When these two variables are in conflict, the individual must decide whether to fulfill institutional requirements or individual needs. Under such circumstances dissatisfaction is inevitable, and the person will probably be less effective and less efficient. 0n the other hand, when institutional expectations are compatible with individual personality, morale and productivity are likely to increase.7 If it were established that certain personality types have certain specific preferences as teachers, this information could be used by educators involved in teacher training as well as by admin— istrators in the field. The Getzels—Guba theory implies that such knowledge of individual teaching preferences could be very useful in improving teacher morale and effectiveness. Educators could use the research findings to help pre—service 6J. W. Getzels, James M. Lipham, and Roald F. Campbell, Educational Administration as a Social Process (New York: Harper and Row, 1968), p. 106. 7J. W. Getzels and E. G. Guba, ”Social Behavior and the Admin- istrative Process," The School Review, Vol. 65 (Winter, 1957), p, 435. teachers plan programs that are congruent with their interests and aptitudes. For example, the findings could be used to predict which aspects of teaching would be most rewarding and which would be most frustrating for students of different personality types. Awareness of research findings as well as practical experience in a variety of teaching situations would hopefully help students choose permanent teaching positions compatible with their individual needs. Also, school administrators could use the research findings to provide more satisfying teaching environments for teachers of dif- ferent personality types. Getzels and Guba state that one of the most important tasks of administration is "to integrate the demands of the institution and the demands of the staff members in a way that is at once organizationally productive and individually ful— filling."8 The goal is not always achieved, however, for many teachers have voiced their dissatisfaction with classroom teaching. It is hoped that the present study will provide information which can be used by educators and school administrators to encourage teachers of various personality types to stay in the classroom. Knowledge of individual teaching preferences may help to produce a teaching force both more effective and more contented in their chosen profession. 8Getzels and Guba, p. 430. Statement of the Problem The major problem of the present study was to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. Jung's theory of personality type is the underlying construct upon which the study was based. Jungian theory postulates that different personality types have quite different preferences. In order to investigate the validity and usefulness of this theory with regard to the field of education, six hypotheses were tested, using the Myers—Briggs Type Indicator (MBTI) to measure Jungian personality type. Teaching preferences with regard to various teaching situations were measured by the Teacher Preference Questionnaire (TPQ), a self—report instrument constructed for the study. A sample of 200 prospective teachers completed both instruments, and their scores were analyzed to deter- mine whether certain personality types indicated preferences for certain kinds of teaching situations. Hypotheses Six research hypotheses were tested, each one postulating that significant relationships exist between the personalities of pro- spective teachers and their attitudes toward certain teaching sit- uations. The six hypotheses are listed below: 1. Sensors, feelers, and judging types show stronger preferences for teachipg lower grades than intuitors, thinkers, and perceptive types. 2. Extraverts, thinkers, and judging types show more interest in administrative functions than introverts, feelers, and perceptive types. :Ij .47 1:.- '1'! 3. Intuitors and perceptive types express a stronger need for independence and creativity than sensors and judging types. 4. Extraverts, intuitors, feelers, and judging types show more interest in plannipg school projects than introverts, sensors, thinkers, and perceptive types. 5. Extraverts, intuitors, feelers, and perceptive types show more interest in workipg with small groups of students than introverts, sensors, thinkers, and judging types. 6. Sensors and judging types express more commitment to classroom teachipg than intuitors and perceptive types. Definitions of the various Jungian personality types specified in the hypotheses are included in Chapter II. Organization of the Study The study is composed of seven chapters. Jung's typological theory is summarized in Chapter II. Research studies involving the Myers—Briggs Type Indicator and teacher populations are reviewed in Chapter III. The design of the study, the sample, procedure, hypotheses, statistical analysis, and instrumentation are presented in Chapter IV. The Myers—Briggs Type Indicator is described and analyzed in Chapter V. The results of the study are presented in Chapter VI. And a summary of the study, a discussion of the findings, and recommendations for future research are included in Chapter VII. *- CHAPTER II JUNG'S TYPOLOGICAL THEORY Early in his career as an analytical psychologist, Carl G. Jung began to notice specific patterns of behavior among his patients. After further investigations he published Psychological Types in 1921, in which he identified eight distinct personality types. He theorized that each person has a natural preference for either extraversion or introversion. In addition, he postulated that each person has a natural preference for one of the following four mental functions: sensation, intuition, thinking, or feeling. The various combinations of these two attitudes and four functions account for his eight personality types. Whether or not these peculiarities of personality are present at birth is not known, but Jung was convinced that the differentia— tion of types is noticeable in very early childhood, like a natural right— or left—handedness. In Psychological Types he stated, "In the last analysis, it may well be that physiological causes, inaccessible to our knowledge, play a part in this."1 In any case, the theory assumes that children, unless hindered, rely on their natural preferences whenever they can, developing and 1 Carl G. Jung, Ps cholo ical T es, trans. H. G. Baynes (New York: Harcourt, Brace & Co., 1926i, p. 416. strengthening them through use. With the advantage of constant practice, the preferred processes grow more controlled and more trustworthy. Individuals with different preferences, therefore, develop along divergent lines, each displaying a certain one-sided- ness of personality. Persons are not limited exclusively to their preferred ways of functioning, but they usually do their best work when they rely on their natural preferences. Jung believed that with a few people there follows a more advanced stage of individuation in the second half of life, in which the other functions are consciously developed more extensively in order to compensate for the earlier one—sidedness of personality.2 The Myers—Briggs Type Indicator Twenty years after the publication of Psychological Types, Isabel Briggs Myers and her mother, Katherine C. Briggs, began designing an instrument based on Jungian type theory. The aim of the Myers—Briggs Type Indicator (MBTI) was to determine a person's basic preferences with regard to each of the following dimensions: 1. Extraversion (E) or Introversion (I) 2. Sensation (S) or Intuition (N) 3. Thinking (T) or Feeling (F) 4. Judgment (J) or Perception (P) The E-I, S—N, and T—F dimensions of personality were specifically defined by Jung. The J—P index was added by Myers and Briggs 2Marie—Louise von Franz, "The Inferior Function,” Jppg's Typology (New York: Spring Publications, 1971), p. 16. , 9 to indicate a person's preferred way of dealing with the outer world. The basic assumption underlying the Myers—Briggs Type Indicator was that each person has a natural preference for one or the other pole on each of the four dimensions. Sixteen combinations of preferences are therefore possible, as shown in Table 2.1. After Table 2.1 Sixteen MBTI Personality Types ISTJ ISFJ INPU INTJ ISTP ISFP INFP INTP ESTP ESFP ENFP ENTP ESTJ ESFJ ENFU ENTJ completing the Indicator, each person's responses are analyzed in order to classify the respondent as one of these sixteen possible types. Although type classifications are used, the Indicator was not designed to "put people into boxes," but rather to help indi— viduals recognize their own natural preferences, styles, and capa— bilities.3 Myers has emphasized that "there is no right or wrong 3Mary H. McCaulley et al., "Myers-Briggs Type Indicator Applications" (Gainesville, Fla.: University of Florida Typology Laboratory, 1973), p. 4. J I_" _:_l-'. . .13....“ 10 to these basic preferences,“it each type having its own set of strengths and abilities. In addition to helping individuals recog- nize their own preferences, the Indicator was designed to support Jung's theory that "much apparently random variation in human be- "5 havior is actually quite orderly and consistent. Descriptions of the Preferences Descriptions of the personality preferences are presented below. The typological definitions of the E—I, S—N, and T—F pref— erences are based on Jung's original descriptions of the various personality types6 and van der Hoop's more detailed accounts of each preference.7 The definitions of the J—P preferences are based on Myers' descriptions in the Manual.8 Extraversion—Introversion The E—I index was designed to reflect the person's preferred orientation to life. Jung defined extraversion as that mental state where the hIsabel Briggs Myers, "Introduction to Type" (privately pub— lished, Swarthmore, Penn., 1970), p. 1. 5Isabel Briggs Myers, The Myers—Briggs Type Indicator Manual (Princeton, N.J.: Educational Testing Service, 1962 , p. 1. 6Jung, Psychological EXEGS’ pp. 412—517. 7J. H. van der Hoop, Conscious Orientation, trans. Laura Hutton (London: Kegan Paul, Trench, Trubner & Co., 1939), pp. 3-91; see also J. H. van der Hoop, Character and the Unconscious, trans. E. Trevelyan (London: Routledge & Kegan Paul, 1923), pp, 132—198. 8Myers, MBTI Manual, p. 58. * individual's conscious libido flows toward the object and away from the subject. Extraverted types are therefore oriented primarily to the outer world of objects, people, and action, and have a tendency to get caught up with whatever is happening around them. They are strongly influenced by facts and circumstances, the opinions of others, and ideas in current use. When in difficulty, they seek support first of all from the external world, for they feel more secure in contact with others than alone. Extraverts are usually open and friendly, able to fit into existing conditions with relative ease. They do their best work when they are actively applying them— selves to actual objective situations. Jung defined introversion as that mental state where there is an attempt to withdraw conscious libido from the object and concen— trate it on the subject. Introverted types therefore have an inward orientation, and tend to detach themselves from the world around them. They are strongly influenced by their own subjective reactions to external events. When in difficulty, they prefer to be alone with themselves, where they feel more secure and can consult the demands of their own being in order to restore internal harmony. They do their best work inside their heads, in activities requiring reflec— tion and concentration. Sensation-Intuition The S-N index was designed to reflect the person's preferred way of perceiving. Jung defined sensation as a non-rational function which includes 12 all the perceptions received through the sense—organs. It is thus pre—occupied with the concrete details rather than the theoretical possibilities of any situation. Sensing types have a here—and—now orientation toward life and react strongly to simple sensory exper- ience. They tend to be observant and practical, able to absorb an immense number of simple facts. They rely on common sense, making sound estimations with regard to the usefulness of things in their environment. Sensing types do not care for complexity or abstrac- tions, for they like to see all around a thing. They enjoy tradi— tion and routine, and tend to have fixed habits. They are generally wary of the new, and work best in familiar environments. Jung defined intuition as a non—rational function which looks at things rather vaguely, so as to get a certain hunch from the un— conscious. It gives a capacity for suddenly seeing the connections and hidden possibilities in a situation, suddenly understanding the meaning of the whole. Intuitive people are guided by inspiration. They are imaginative and spontaneous, preferring to make as few preparations as possible. They have a facility for dealing with abstractions, symbols, and inferred meanings, and are often bored with concrete reality and everyday routine. Intuitive types like to see every situation as a fresh problem, and they have an intensely personal way of attacking problems, relying on a flash of insight. They are strongly convinced of the validity of their insight at that moment, but they rarely feel bound to it at a later time. Intuition demands freedom to see things at any moment in a new connection. Thinkipg—Feelipg The T—F index was designed to reflect the person's preferred way of judging. Jung defined thinking as a rational function which encompasses the cognitive realm of human experience. It involves reducing objective experience to its essential elements and restructuring these elements into a logical and consistent framework which is assumed to have universal validity. Thinking establishes order and judges whether something is true or false, based on an impersonal system of thought—forms about the relationships and facts of reality. Most of these ideas are taught to the individual in ready—made forms. Thinking types use these fixed mental structures to observe and analyze the world. They are inclined to see everything that happens as a complex web of systems and sub—systems, interacting in accordance with definite rules and laws. They rely on these logical structures to put clarifying order into a particular situation, and are skilled at objectively organizing material, weighing the facts, predicting consequences, and making plans. Thinking types like to state things clearly and can nearly always justify the reason for their behavior. Emotion and impulse are apt to be denied, unless they can see a way of considering them systematically. Jung defined feeling as a rational function which encompasses the affective realm of human experience. It is a process of appre- ciation which involves collecting all the various emotional tendencies and impulses centering around some object or situation, and molding 14 them into a certain feeling—attitude which is in harmony with the person's existing value system. Feeling establishes order and judges whether something is good or bad, agreeable or disagreeable, usually relying on attitudes taught to the individual by society. Feeling types use these various forms of expression to maintain contact and achieve a good feeling rapport with their environment. They usually dislike conflict, preferring to balance extremes by smoothing over the differences. They are skilled at analyzing sub— jective impressions, weighing the importance of various alternatives for themselves and others, and expressing their final judgment as a highly differentiated feeling—attitude. They operate best in the realm of human relationships, ethics, and aesthetic taste, always seeking modes of expression which are in harmony with their own inner ideals. Their tendency is to accept logical thought only so far as it conforms with feeling. Judgment—Perception The J—P index was added by Myers and Briggs to reflect the person's preferred way of dealing with the outer world. This dimension of personality was implied in Jung's theory, but judgment and perception were never explicitly defined by Jung as independent functions. Judgment aims to regulate life and control it. It involves shutting off perception in order to come to a conclusion. Judging types are organized and systematic. They like to plan their lives, deciding what needs to be done and then doing it. ”—_———_——- 15 Perception aims to understand life and adapt to it. It involves delaying judgment in order to allow more evidence to come in. Per— ceptive types are curious, open—minded, and flexible. They go through life in a spontaneous way, awaiting events and adapting to them. Summary The general purpose of the Myers-Briggs Type Indicator was to implement Jung's theory of personality type, a theory constructed "to give some order to the apparently limitless variations of human ”9 individuality. More specifically, the Indicator was designed to determine a person's basic preferences with regard to four dimen- sions of personality. The four pairs of preferences are summarized in Table 2.2. The Myers-Briggs Type Indicator has been employed to examine the personalities of various populations, including many samples of pre—service and in—service teachers. A review of research studies involving teachers is presented in the following chapter. 9Carl G. Jung (ed.), Man and his Syr_nbols (New York: Dell Pub— lishing Co., 1968), p. 118. Table 2.2 Basic Preferences Measured by the Myers—Briggs Type Indicator Preferred Orientation to Life Extraverted types are oriented Introverted types have an inward toward the outer world of objects, orientation, and tend to detach people, and action. themselves from the outer world. Preferred Way of Perceiving Sensipg types rely on perceptions Intuitive types look at things received through the sense-organs, more vaguely, so as to get a noticing the concrete details and certain spontaneous hunch from practical aspects of a situation. the unconscious. Preferred Way of Making Decisions Thinkipg types use logic to put Feeling types are skilled at clarifying order into a particular understanding others' feelings, situation, impersonally judging if and base their judgments on something is true or false. personal values. Preferred Way of Dealing with the Outer World Juggipg types live in a planned, Perceptive types live in a flex— orderly way, aiming to regulate ible, spontaneous way, aiming to life and control it. understand life and adapt to it. * CHAPTER III REVIEW OF RELATED LITERATURE Research with the Myers—Briggs Type Indicator has only recently begun to gain recognition in the literature on teacher personality. One reason may be that the Indicator is a relatively new instrument. Another reason may be that many of the MBTI studies with teacher populations are relatively inaccessible, being reported in disser— tations and not in professional journals. A third reason may be that the "type" approach to personality was violently attacked and generally abandoned in the literature for many years. It has taken an impressive number of MBTI studies to convince many researchers of the potential usefulness of Jungian type theory. In the majority of studies involving the Myers—Briggs Type Indicator and teacher populations, one or more of the following ques— tions have been examined: 1. D0 certain personality types predominate in teacher populations? 2. Do pre—service teachers differ in personality from in—service teachers? 3. Do certain personality types prefer teaching lower levels and other types prefer upper levels? 4. Do certain types prefer teaching specific subjects? 5. Do teachers differ in personality from school administrators? ___—_——‘ 18 Most of these studies have been descriptive in nature, noting the type patterns found with particular samples of teachers. In addition to the descriptive studies, a number of quasi—experimental studies have been conducted with the Indicator in order to examine the relationship between teacher personality and teaching style, teacher satisfaction, and teacher effectiveness. The following review of the literature includes both descriptive and quasi—exper— imental studies involving the Myers—Briggs Type Indicator and teacher populations. Personality of In—Service Teachers The question of whether certain personality types predominate in teacher populations has been investigated in numerous studies, but most researchers have not sampled teachers in the field. Instead, sample populations have been composed of teachers enrolled in graduate courses or teachers who have volunteered to take part in research experiments, thus biassing the results. The findings summarized in the present section have been limited to field studies, and it is assumed that these teacher populations are fairly representative of public school teachers in general. The following researchers have used the Myers—Briggs Type Indica— tor to assess the personalities of teachers in the field: von Fange,1 1Erich A. von Fange, "Implications for School Administration of the Personality Structure of Educational Personnel” (unpublished doctoral dissertation, University of Alberta, 1961), pp. 106—110, — 19 Stricker and Ross,2 Wright,3 Moralesf1 Hoffman,5 and Cage.6 Results of these investigations proved to be quite consistent and are shown in Table 3.1. The majority of teachers in each sample were extra- verted (51% to 62%, depending on the sample), sensing (53% to 74%), feeling (55% to 66%), and/or judging types (65% to 82%). When male teachers were compared with female teachers, as was done in the studies of von Fange,7 Stricker and Ross,8 Wright,9 and Hoffman,10 an interesting finding emerged: Male teachers were pre— dominantly thinking types and female teachers were predominantly feeling types at both elementary and secondary levels. The actual 2Lawrence J. Stricker and John Ross, "A Description and Evalua— tion of the Myers—Briggs Type Indicator,” Research Bulletin 62—6 (Princeton, N.J.: Educational Testing Service, 1962), pp. 164-169. 3Judith Anne Wright, "The Relationship of Rated Administrator and Teacher Effectiveness to Personality as Measured by the Myers— Briggs Type Indicator" (unpublished doctoral dissertation, Claremont Graduate School, 1966), pp. 126—142. itCarmen A. Morales, "Teacher Personality Types and Teacher Morale in Selected Open Space and Self—Contained Elementary Schools" (unpublished doctoral dissertation, University of Florida, 1975). 5Jeffrey L. Hoffman, "Personality Relationships Between Super— vising Teachers and Student Teachers as Determined by the Myers—Briggs Type Indicator" (unpublished doctoral dissertation, University of Florida, 1974), pp. 59-64- 6Based on personal correspondence between Dr. B. N. Cage, Bureau of Educational Research, University of Mississippi, and the writer, December 9, 1975. 7Von Fange, p. 93. 8Stricker and Ross, "A Description and Evaluation of the MBTI,” pp . 164—169 . 9Wright, p. 129. 10Hoffman, p. 62. 20 Table 3.1 MBTI Type Preferences of In—Service Teachers From Previous Studies Sample MBTI Type Categories 169 Elementary Teachers in Alberta, Canada (Von Fange, 1961) 334 Elementary Teachers in Covina, Cal. Form E) (Stricker & Ross, 1962) 257 Elementary Teachers in Covina, Calif. (wright, 1964) 150 Elementary Teachers in Tallahassee, Fla. (Morales, 1975) 113 Elem. & Sec. Teachers in Northern Florida (Hoffman, 1974) 39 Elem. & Sec. Teachers in Northern Mississippi (Cage. 1975) 54 62 56 51 57 57 46 38 44 49 43 43 69 55 53 61 67 74 31 45 47 39 33 26 42 40 35 34 45 43 58 60 65 66 55 57 82 68 65 69 80 82 18 32 35 31 20 18 aAll figures are rounded percentages. 21 Table 3.2 MBTI Type Preferences of Male and Female In—Service Teachers From Previous Studies) Sample MBTI Type Categories 123 Female Elem. Teachers a in Alberta, Canada 53 47 67 33 36 64 82 18 (Von Fange, 1961) 248 Female Elem. Teachers in Covina, Cal. (Form E) 61 39 56 44 33 67 65 35 (Stricker & Ross, 1962) 225 Female Elem. Teachers in Covina, Calif. 56 44 53 47 31 69 65 35 (Wright, 1964) 81 Female Elem. & Sec. Teachers in N0. Fla. 62 38 69 31 41 59 79 21 (Hoffman, 1974) 46 Male Elem. Teachers in Alberta, Canada 52 48 71 29 69 31 69 31 (Von Fange, 1961) 86 Male Elem. Teachers in Covina, Cal. (Form E) 64 36 53 47 59 41 79 21 (Stricker & Ross, 1962) 32 Male Elem. Teachers in Covina, Calif. 56 44 53 47 60 4O 66 34 (Wright, 1964) 32 Male Sec. Teachers in Northern Florida 44 56 63 37 56 44 85 15 (Hoffman, 1974) aAll figures are rounded percentages. — 22 type distributions for these four samples of male and female in- service teachers are shown in Table 3.2. As might be expected, the modal type (the predominant type) in samples of male teachers was usually ESTJ, and the modal type in samples of female teachers was usually‘ESFJ. To determine if the distribution of personality types in teacher samples is typical of the general population, it is necessary to have MBTI data on the general population, with the criterion of random selection having been met. At the present time, no such comprehensive studies of the general population have been reported. Nevertheless, Myers11 has used existing MBTI data compiled from various populations to estimate type distributions in the United States, and she has con— cluded that there are more extraverts than introverts, more sensors than intuitors, more judging types than perceptive types, and the majority of females are feeling types and the majority of males are thinking types. The ESFJ type appears to be the most common type for women in America, and the ESTJ type appears to be the most common type for men. Comparing general population estimates with type distributions found in teacher samples, it would appear that teachers as a group are quite similar in personality to the majority of Americans. There may be a somewhat higher proportion of introverts and judging types among teachers than the general public, but the question has never been directly examined. 11Based on personal correspondence between Isabel Briggs Myers and the writer, December 1, 1975; see also Isabel Briggs Myers, ”Type as the Index to Personality" (privately published, Swarthmore, Penn., 1945); see also Myers, MBTI Manual, pp. 12—19. _71j 23 Personality of Pre—Service Teachers When pre—service teacher samples have been compared with in— service teacher samples, consistent differences between the two groups have emerged. The specific type distributions of various groups of pre—service teachers are shown in Table 3.3. Von Fange12 found sig— nificantly more intuitive, feeling, and perceptive types among female students enrolled in educational degree programs than among teachers in the field; the modal type for female pre—service teachers was ENFP, a type which was held by only 1% of the in—scrvice teachers. Hoffman13 found similar differences when he compared student teachers with their supervising teachers. There were significantly higher proportions of extraverted, intuitive, feeling, and perceptive types among the group of male and female prospective teachers, with ENFP being the modal type for pre—service teachers. In another study, Cagelh also found more extraverted, intuitive, feeling, and per— ceptive types among student teachers than among their supervising 15 teachers. Studies of education majors conducted by McCaulley, 12Von Fange, pp. 89-116. 13Hoffman, pp. 64-77. 14 Cage, personal correspondence. 15Mary H. McCaulley, "Presenting MBTI Information on Indivi— duals and Groups" (paper read at the First National Conference on the Uses of the Myers—Briggs Type Indicator, October, 1975, Gainesville, Fla.), p. 7- 24 Table 3.3 MBTI Type Preferences of Pre—Service Teachers From Previous Studies Sample MBTI Type Categories 570 Education Majors at Univ. of Alberta 57 (Von Fange, 1961) 8, 43 62 38 43 57 66 34 117 Education Majors at Florida State U. 67 33 41 59 29 71 57 43 (Hoffman, 1974) 48 Education Majors at Univ. of Mississippi 65 35 58 42 25 75 61 39 (Case. 1975) 246 Education Majors at Univ. of Florida 58 42 55 45 19 81 56 44 (McCaulley, 1975) QAll figures are rounded percentages. Table 3.4 MBTI Type Preferences of Male and Female Pre—Service Teachers From Previous Studies) Sample MBTI Type Categories 342 Female Educ. Majors at Univ. of Alberta 58 (Von Fange, 1961) a 42 54 46 34 66 63 37 78 Female Educ. Majors at Florida State U. 62 38 35 65 26 74 55 45 (Hoffman, 1974) 228 Male Educ. Majors at Univ. of Alberta 56 44 73 27 58 42 68 32 (Von Fange, 1961) 39 Male Educ. Majors at Florida State U. 76 24 54 46 36 64 62 38 (Hoffman, 1964) éAll figures are rounded percentages. '5""""""""""""""""""""""""""""""""'""""""""‘IIII|II 26 Taylor,16 and Richek17 produced similar results. And in a study comparing pre—service and in—service mathematics teachers, Story18 found that the pre—service teachers had stronger preferences for intuition and feeling than did the career math teachers in the field. When male education majors were compared with female education 19 and Hoffman,20 the majors, as was done in the studies of von Fange females emerged as consistently more intuitive, feeling, and percep- tive in personality type than their male counterparts. Education majors of both sexes, however, were different in personality from in-service teachers. Type distributions for these two samples of male and female pre—service teachers are shown in Table 3.4. Taken together, the studies indicate that there are signifi- cantly more intuitive, feeling, and perceptive types being trained as teachers than are found in the schools, particularly among females. There is also some evidence that pre-service teachers are more extra— verted than in—service teachers. The types who are most likely to 16Robert E. Taylor, "An Investigation of the Relationship Between Psychological Types in the College Classroom and the Student Percep— tion of the Teacher and Preferred Teaching Practices" (unpublished doctoral dissertation, University of Maryland, 1968), p. 72. 17Herbert G. Richek, "Personality Characteristics of Male and Female Prospective Teachers: A Multivariate Analysis," Alberta Journal of Educational Research, Vol. 19 (September, 1973), p. 211. 18Garth E. Story, ”An Analysis of the Relationships Between Personality Types of Mathematics Teachers (7-12), as Measured by the Myers—Briggs Type Indicator, and Selected Factors Related to Teaching" (unpublished doctoral dissertation, University of Florida, 1972), pp. 46—47. 19Von Fange, p. 95- 20Hoffman, p. 63. ’3". r3 - if? l'.1--"'"| . . .i, -. .5: .".' "3' 5.13101 27 stay in teaching appear to be the sensing and judging types, the most fact—oriented, scheduled, organized types. It is possible, as Hoffman mentions, that student teachers may experience a change in personality as they are forced to cope with the practical aspects of full—time classroom teaching.21 This explanation contradicts Jungian theory, but it should nevertheless be tested experimentally. Longitudinal studies should be conducted to see if and why certain types of pre—service teachers later leave classroom teaching. If it is confirmed that the most enthusiastic and innovative types are leaving the classroom, the implications appear serious in a period when rapid educational innovation is required.22 Personality and Grade Level The question of whether certain personality types prefer teach— ing lower levels and other types prefer upper levels has been examined by comparing samples of elementary school teachers with samples of high school teachers. In two different field studies, Hoffman23 and Cage24 found a higher proportion of intuitive, thinking, and percep— tive types at the secondary level than at the elementary level. Hoffman's samples were notably larger than those of Cage, and re— vealed a striking difference with regard to the T—F dimension: 21Hoffman, p. 69. 22Mary H. McCaulley, "Type and Education" (Gainesville, Fla.: University of Florida Typology Laboratory, 1971), p. 5. 23Hoffman, p. 76. 2 4 Cage, personal correspondence. 28 76% of the elementary teachers but only 48% of the secondary teachers were feeling types. This difference may be due to the fact that all of the elementary teachers in Hoffman's sample were women and one— third of the secondary teachers were men. In a field study involving lower and upper elementary school teachers, Wright25 eliminated sex as a possible confounding variable by comparing male and female teachers separately. Her results were consistent with the studies of Hoffman and Cage with regard to the T—F dimension: There were significantly more feeling types among the lower elementary teachers than among the upper elementary teachers, for males as well as females. In a field study at the secondary level, Story26 asked 241 junior and senior high mathematics teachers to indicate the grade level they would most prefer teaching. Results showed that a significantly higher proportion of intuitive math teachers preferred to teach grades 9 to 14 than grades 5 to 8. Research with pre—service teachers has produced a variety of findings with regard to grade level preference. Bown, Fuller, and Richek27 in 1966 and Richek28 in 1973 compared the continuous MBTI scores of prospective elementary teachers with the scores of pro— spective secondary teachers. The one finding which remained consistent 25Wright, pp. 128—129. 26Story, pp. 54~55- 27O. H. Bown, F. F. Fuller, and H. G. Richek, ”Differentiating Prospective Elementary and Secondary School Teachers," Alberta Journal of Educational Research, Vol. 12 (June, 1966), pp. 127-130. 28Richek, "Personality Characteristics,” pp. 208—215. 29 in both studies was that pre—service elementary teachers were signifi— cantly more extraverted than pre—service secondary teachers. This 29 finding was replicated by Cage, who found in addition significantly more sensing types preparing to teach lower levels. In a Canadian 30 study, however, von Fange found that introverted female students were more likely to prefer teaching at the elementary level and extraverted females were more likely to prefer teaching high school. He also found a very high proportion of SF types among females preparing to teach at the elementary level. Among male students, von Fange found that males who wanted to teach at the high school level (approximately 75%) were quite similar in personality to males who wanted to teach lower levels. 31 In another study of pre-service teachers, Hoffman found the same proportion of extraverts and sensors at both levels, but a higher pro— portion of feeling and judging types among those preparing to teach elementary school. The question of whether certain personality types are found more frequently among teachers of certain grades has not been settled, although there is some evidence that sensors, feelers, and judging types are found in greater proportion at lower grade levels. Personality and Subject Area The question of whether certain personality types prefer teach— ing certain subjects has been examined directly by two researchers. 29 Cage, personal correspondence. 30, Von Fange, pp. 130—137. 31Hoffman, p. 76. 30 Hoffman32 tested 84 secondary supervising teachers and Story33 tested 241 mathematics teachers, each sampling teachers in the field. Both studies found math teachers to be predominantly sensing, thinking, and judging types. In Hoffman's sample, all of the 18 math teachers were judging types, and all but one were sensing types. The teachers of other subjects were not as homogeneous in personality as the math teachers, but certain personality types were found to be teaching certain subjects in higher proportion than would be expected by chance. Table 3.5 summarizes Hoffman's findings with his sample of 84 secondary school teachers, and shows which personality types were associated most strongly with each subject area. In studies involving non-teacher populations, results have generally been compatible with the subject area preferences illus— trated in Table 3.5. Using the Strong Vocational Interest Blank with a sample of 727 male freshmen, Stricker and Ross34 found interest in being a social science teacher to be correlated with extraversion and feeling. In two studies involving prospective teachers, McCaulley35 found that students preparing to teach physical education tended to be sensing types, and students preparing to teach art were more introverted, intuitive, and perceptive than other education majors. 32Hoffman, p. 76. 33Story, p. 42. 34Lawrence J. Stricker and John Ross, "Some Correlates of a Jungian Personality Inventory,” Psychological Reports, Vol. 14 (April, 1964), p. 631. 35M°Caulley et al., "MBTI Applications." pp. 42—45; see also McCaulley, "Presenting MBTI Information,” p. 7, . . . _‘ _‘_,._ "an.“ 3. , 'uta ' 1". ' -o'--‘. Table 3.5 MBTI Type Preferences Associated with Different Subject Areas Subject Area Personality Type Math / Science —STJa English / Foreign Language —NFP Social Studies E——— Fine Arts —N—P Applied Arts ——T— Physical Education _s__ Note. Table constructed from the data of Jeffrey L. Hoffman, ”Personality Relationships Between Supervising Teachers and Student Teachers as Determined by the Myers—Briggs Type Indicator" (unpublished doctoral dissertation, University of Florida, 1974), pp. 72-74. aDashes indicate no strong preferences; e.g., math/science teachers appear to be similar to other secondary teachers with regard to the E—I dimension. 32 Table 3.6 MBTI Type Preferences Associated with Different Teaching Majors N = 88 Teaching Major Personality Type Math / Science —S—Ja English / Foreign Language —NFP Social Studies ENF— Fine Arts —NFP Applied Arts -qu Physical Education _s__ Note. Table constructed from the data of Jeffrey L. Hoffman, "Personality Relationships Between Supervising Teachers and Student Teachers as Determined by the Myers—Briggs Type Indicator" (unpublished doctoral dissertation, University of Florida, 1974), pp. 72—74. aDashes indicate no strong preferences; e.g., math/science majors appear to be similar to other pre—service teachers with regard to the E—I and T—F dimensions. 33 Hulbert36 found that intuitive, feeling, and perceptive types pre- dominated in various art classes at a community college. -And in-a 37 study of prospective secondary teachers, Hoffman found certain personality types to be associated with choice of teaching major. The results with Hoffman's sample of 88 pre—service teachers are summarized in Table 3.6 and are consistent with the other studies cited. From the studies cited above, it may be concluded that certain personality types tend to prefer teaching certain subjects. Intui— tive and feeling types seem to be attracted to literature, social studies, and fine arts. Sensing types, on the other hand, seem to be attracted to applied subjects, such as mathematics, physical education, home economics, and industrial arts. Personality of School Administrators The question of whether certain personality types are attracted to school administration has been examined by two researchers. The most comprehensive MBTI study involving school administrators was conducted by von Fange38 in 1961. Examining a large sample of Canadian teachers and administrators, he found that school principals as a group were quite homogeneous in personality, with fewer types 36James C. Hulbert, "A Brief Study Showing the Frequency of Types in Art and Creativity Classes" (paper read at the First National Conference on the Uses of the Myers~Briggs Type Indicator, October, 1975, Gainesville, Fla.), pp. 1—2. 37Hoffman, pp. 72—74. 38Von Fange, pp. 110—119. JL — 34 found at this level than at any other level in the educational hierarchy. Of the 63 principals sampled, 92% were judging types, with ESTJ clearly being the most common type pattern. Von Fange found similar results with a sample of 66 superintendents, although there appeared to be a somewhat broader representation of types among superintendents than among principals. A second study of school administrators was conducted by Wright39 in 1964. Her study involved two samples of elementary school principals from a school district in California. Results showed that two—thirds of the principals were thinking types, two—thirds were judging types, slightly over half were extraverts, and there was an even split on the S—N dimension. Both researchers examined populations which were predominantly male, and the results of their studies are shown in Table 3.7. Taken together, the studies indicate that school administration attracts thinking judging types, persons who value impersonal truth, thought—out plans, and orderly efficiency. There is also some evidence that administrators tend to be more extraverted than intro— verted. Intuitive types, the types who are most interested in future possibilities, were found with more frequency among administrators than among teachers in the field, but they did not outnumber sensing— type administrators. 39Wright. pp. 133—135. or: iula'll3.l'_-.'IHL-‘ HIM” 35 Table 3.7 MBTI Type Preferences of School Administrators From Previous Studies) Sample MBTI Type Categories 66 Superintendents in a 7 Canadian Provinces 59 41 61 39 52 48 81 19 (Von Fange, 1961) 63 School Principals in Alberta, Canada 59 41 54 46 59 41 92 8 (Von Fange, 1961) 36 Elem School Principals in Covina, Calif. 53 47 50 50 67 33 61 39 (Wright, 1964) aAll figures are rounded percentages. A 36 . Personality and Teaching Style The majority of research involving the Myers-Briggs Type Indicator and teacher populations has been descriptive in nature, noting the type patterns found with particular samples of teachers. Quasi—experimental studies have been more difficult to conduct, but they have produced significant findings concerning the role of teacher personality in the educational process. Perhaps the most rigorous MBTI study involving teachers was conducted by Lawrence and DeNovellis40 in 1974 in an effort to find possible relationships between teacher personality and overt behavior patterns in the classroom. The researchers administered the Indicator to 74 middle school teachers after six separate observations had been made of their classrooms over a two—day period. Correlation and regression analysis revealed that certain styles of teaching were preferred by teachers of certain personality types. Lawrence and DeNovellis found that the classrooms of introverted teachers were significantly more teacher—centered than were the classrooms of extraverts. In addition, introverts used more struc— tured learning material and had a significantly higher incidence of organized activities in which pupils had no choice. The classrooms of sensing teachers were also teacher—centered and characterized by pupil no—choice activities. In contrast, quordon D. Lawrence and Richard DeNovellis, "Personality Variables and the Middle School Teaching Learning Process" (paper read at American Educational Research Association convention, April, 1974, Chicago, Illinois). 37 intuitive teachers were more pupil-centered and tended to move freely around the room, attending to pupils briefly. Intuitors were much more likely than sensors to encourage pupil choice and self-expression, but they also showed significantly more non—verbal negative affect than did sensing teachers. One consequence of these two different teaching styles was that students exhibited more hostile—aggressive behavior in intuitive teachers' classrooms and students showed more passive-withdrawn behavior in sensing teachers' classrooms. Feeling—type teachers differed from thinking-type teachers in two respects: Feeling teachers exhibited significantly more affect (both positive and negative) and they also employed more indivi— dualized instruction than thinking teachers. The classrooms of feeling teachers were more likely to be student—oriented than subject—oriented, with pupils working independently at various activities and the teacher attending individual pupils closely. Correlations on the judgment—perception dimension were very similar to those on the sensation-intuition dimension, but with fewer correlations reaching the .05 level of significance. In summary, the study by Lawrence and DeNovellis revealed that introverted, sensing, thinking, and/or judging—type teachers were likely to use a traditional approach, with the teacher structuring the material and playing a central role in its presentation. Extraverted, intuitive, feeling, and/or perceptive-type teachers were likely to use informal individualized instruction, with students making more choices and playing a central role in the educational process. The study indicates that specific teaching behavior may be 38 strongly related to teacher personality. Two other studies have been reported in which classroom obser— vations were used to compare teaching style with teacher personality type. Williamsl‘1 administered the Indicator to 58 elementary teachers after observations had been made of their classroom behavior during a 30—minute visit. He found that extraverts, intuitors, feelers, and perceptive types appeared to be the most experimental in their ap- proaches, but results were not significant at the .05 level. In another investigation of teacher classroom behavior, Castellili2 found that intuitive types showed significantly more supportive behavior, but also expressed their negative feelings more overtly than sensing teachers. Rudisill43 examined mathematics teachers' preferences for various teaching strategies by analyzing their responses to a ques— tionnaire. He found that extraverted teachers were more likely than introverts to give high ratings to ”self—paced instruction," extra— verted sensing teachers were more likely than introverted intuitors thavid L. Williams, "An Analysis of the Interrelationships Among Elementary School Teachers’ Personality Types, Beliefs, Observed Classroom Practices, and Reports of How Broadcast Instructional Tele— vision Should be Used" (unpublished doctoral dissertation, University of Florida, 1972). 42Charles D. Castelli, "An Exploration of the Relationship Between Teacher Creative Ability and Teacher-Pupil Classroom Behavior" (unpublished doctoral dissertation, State University of New York at Buffalo, 1964). 43Edwin M. Rudisill, "An Investigation of the Relationship Between Mathematics Teachers' Personality Characteristics, as Measured by the Myers—Briggs Type Indicator, and Their Preferences for Certain Teaching Strategies" (unpublished doctoral dissertation, University of Florida, 1972). 39 to give high ratings to the "laboratory approach," and perceptive—type teachers were more likely than judging types to give high ratings to "questioning techniques." Rudisill concluded that the personality types of the teachers in his sample were significantly related to their preferences for particular teaching strategies. In another study involving mathematics teachers, Storyké asked 241 teachers to evaluate their experiences with the use of manipula— tive materials in developing mathematiCS concepts. He found that extraverted intuitive types gave the highest ratings to manipulative materials and introverted intuitive thinking types had used them the least. Story also asked the math teachers to choose from a list of characteristics the three qualities that they considered most desir— able in a teacher. Results showed that the sensing teachers preferred sensing characteristics (such as "working steadily”), and the intui~ tive teachers preferred intuitive characteristics (such as "using insight to form conclusions”) with higher than expected frequencies. From the studies cited above, it may be concluded that certain personality types tend to prefer certain teaching styles. Personality and Satisfaction with Teaching Teacher satisfaction has been measured in a variety of ways. 45 Using a questionnaire, von Fange asked 122 Canadian teachers to report how satisfied they were with their current teaching situations. 44Story, pp. 58—61. 45Von Fange, pp. 137—142.. 40 He found that extraverts expressed the most all-around satisfaction, particularly ESFU and ESTJ types. In a study of mathematics teachers, Storyli6 asked 241 teachers if they would recommend teaching to a friend. Results showed that nearly all of the extraverts responded affirmatively, and introverted thinking types appeared to be the least likely to recommend teaching to a friend. Collinsli7 conducted a study to see whether teachers with dif— ferent personality types preferred different school climates. Two hundred elementary teachers in twelve school districts volunteered to complete the Myers—Briggs Type Indicator and two questionnaires which were designed to measure teacher satisfaction and teacher per— ception of the organizational climate of their schools. Schools with many rules and regulations were characterized as having "closed" cli— mates, and schools with fewer rules and a principal who was actively involved in the educational process were characterized as having "open" climates. Collins' results showed that intuitors, especially intro— verted intuitorS, favored "open" climates and had a lower satisfaction with ”closed” climates. Sensing feeling types appeared to be quite satisfied in either "open" or ”closed" climates, and sensing thinking types generally preferred "closed" climates. Individual personality was thus shown to be significantly related to preferences for specific school climates and job satisfaction. 46Story, pp. 54—56. 47James A. Collins, ”Individual Personality and Organizational Climate" (un ublished doctoral dissertation, Claremont Graduate School, 1965 - 41 Using a different approach, Stricker and Ross48 correlated the ages of male and female elementary teachers with their personality preferences, and found that female teachers who were sensing types tended to be significantly older than female teachers who were intui— tive types. This finding may indicate that sensing—type women re— main in elementary teaching longer than intuitive—type women, and are more satisfied with teaching as a career. The results of Stricker and Ross's investigation are compatible with results of studies com- paring pre—service teachers with in—service teachers. As mentioned previously, the types that tend to stay in teaching appear to be the sensing and judging types. Taken together, these studies indicate that certain personality types may be more satisfied with classroom teaching than other types. Extraverts, sensors, and judging types usually express more satis— faction with teaching than introverts, intuitors, and perceptive types. In addition, there is some evidence that teacher satisfaction is related to the interaction of teacher personality and school cli— mate. Future researchers should attempt to measure more explicitly the relationship between teacher satisfaction, teacher personality, and school climate. Whether or not school climate is a contributing factor, existing research indicates that intuitors and perceptive types are more likely to drop out of teaching than other types. The types who are most likely to stay in teaching appear to be the sensing and judging types. 48 p. 145. Stricker and Ross, "A Description and Evaluation of the MBTI," 42 Personality and Teacher Effectiveness A number of studies have investigated the question of whether student satisfaction and student achievement are related to the similarity between the personality of the student and the personality of the teacher. MBTI findings have been rather surprising but should be interpreted with caution, for most of the studies have been con— ducted at the college level and results may not be valid for lower levels. Jungian theory holds that persons of similar type generally have similar interests, values, and ways of learning}:9 It might therefore be expected that students would learn more from teachers with personalities similar to their own, and would give higher ratings to teachers with personalities similar to their own. However, Hoffman50 asked 106 education students to rate their relationship with their supervising teacher and found that the ratings did not appear to be related to whether a student and teacher were similar or dissimilar in personality. An experiment conducted by Cage, Blackburn, and Peaster51 with a smaller sample of 29 education stu- dents and their supervising teachers revealed similar results: Student ratings of the qualifications and competencies of their 49Myers, MBTI Manual, p. 75. 5OHOffman, ppo 86—89. 51B. N. Cage, M. Blackburn, and M. Peaster, "The Effect of Personality on Student/Cooperating Teacher Relationships" (paper read at the First National Conference on the Uses of the Myers—Briggs Type Indicator, October, 1975, Gainesville, F1a.). .':': ".‘ 'u'fl'l' “ fl 43 supervising teachers were not higher when students and teachers had similar personalities. In fact, the researchers found an inverse relationship with regard to the thinking-feeling dimension. Higher ratings were given to supervising teachers who had an opposite T—F preference to that of the student teacher. 52 In another study, Taylor asked 156 education majors to rate their instructors in a human development class and found no corre— lation between a student's evaluation of the teacher and the differ— ence between the student's and teacher's psychological type. Blank53 asked 311 undergraduate students to rate the teaching ability of their instructors and also found no correlation between ratings and teacher—student similarity. Perhaps the relationship between student attitude and teacher— student similarity is not linear. In a study involving counselors and their clients, Mendelsohn and Geller54 found a curvilinear re— lationship between client attitude and counselor—client similarity. Clients who were moderately similar to their counselors in personality type evaluated their counseling experience more positively than clients who were very similar or very dissimilar to their counselors. 52Taylor, pp. 1—157. 53Logan F. Blank, "The Relationship Between Student Instruc— tional Ratings and Student—Faculty Psychological Types” (unpublished doctoral dissertation, Michigan State University, 1969). 52kG. A. Mendelsohn and M. H. Geller, "Similarity, Missed Sessions, and Earl Termination,” Journal of Counseling Psychology, Vol. 12. (May, 1967 , pp. 210—215. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII______________”’i:::]|lllllll 44 Schmidt and Fretz55 tested the effect that similarity of teacher—student personality had on student achievement as well as attitude. The sample consisted of over 400 college students en— rolled in an educational skills course, and results showed that student performance and satisfaction did not appear to be improved by teacher—student similarity. In fact, students least like their teachers in extraversion—introversion showed significantly more improvement in reading skills. Also, students indicated higher satisfaction when their teachers were least like themselves on the J—P dimension. Certain teacher characteristics, however, appeared to affect student performance and attitude. Students showed the most achievement and satisfaction when their teachers were highly intuitive and feeling oriented. Levell56 found similar results with a group of 117 school coun— selors. Those who were rated as most effective by their staff super— visors scored Significantly higher in intuition and feeling than did 57 those rated least effective. And in another study, McNamara found that counselor trainees who were intuitive and feeling types generally received higher rankings on measures of counseling performance. 55Lyle D. Schmidt and Bruce R. Fretz, ”The Effects of Teacher— Student Similarity in an Educational Skills Course," Cooperative Research Project No. s—217 (Washington, D.C.: U. 8. Office of Education, 1965). 56James P. Levell, "Secondary School Counselors: A Study of Differentiating Characteristics" (unpublished doctoral dissertation, University of Oregon, 1965). 57Thomas C. McNamara, ”A Study of Philosophical Identities in a Counseling Practicum" (unpublished doctoral dissertation, University of Illinois, 1967). '. f .Ti ".-_ 45 Von Fange's study58 also produced similar results. Female student teachers who were rated as most effective by their super— vising teachers tended to be extraverted, intuitive, feeling, and perceptive types. Male student teachers rated most effective tended to be introverted, intuitive, feeling, and judging types. In both cases, intuition and feeling were correlated with teacher effec— tiveness. Ratings often depend on who is evaluating the teacher and on what criteria are being used. In a study by Wright,59 18 elementary principals were asked to choose the three most effective and the three least effective teachers in their respective schools. Results showed that ENFU types received the highest ratings, and Wright con— cluded that intuition appeared to be the key personality factor related to effectiveness in teaching.60 There was a tendency in this study, however, for principals to rate teachers with personality types similar to their own as more effective than teachers of opp05ite typeso In interpreting such findings, it is important to recognize that subjective ratings of teachers are frequently highly unreliable, and teachers, students, and supervisors tend to differ on what con— stitutes teacher effectiveness.61 Perhaps the most effective teacher personality type in any given 58Von Fange, pp. 144—148. 59Wright, pp. 152—156. 60Wright, p. 186. 61Getzels and Jackson, p. 575. 46 case depends on the grade level and the subject area being taught. Blank62 found that engineering students at the college level consis— tently gave the best ratings to the INTJ instructor type. In this case, thinking types were rated higher than feeling types. Little research has been done at lower levels to see whether teacher personality type is related to student attitude and achieve- ment. In fact, only one MBTI study has been reported in which elementary and secondary students were asked to rate their teachers. Paraskevopoulos63 asked the students of 102 teachers to evaluate the classroom behavior of their teachers, and he concluded that the relationship between teachers‘ personality preferences and the stUr dents' evaluations was not impressive. He did find, however, that sensing male teachers were rated as friendlier and happier than intuitive male teachers and that intuitive males were rated as more flexible, open—minded, and understanding than sensing male teachers. No attempt was made in this study to evaluate student satisfaction or student achievement. The research cited above indicates that different personality types have distinctly different teaching styles and may be more effective in teaching certain subjects or certain types of students than others. More studies are needed at the elementary and secon— dary level, however, to determine whether teacher effectiveness is related to teacher personality type. 62Blank, pp. 80—81. 63Ioannis Paraskevopoulos, "How Students Rate Their Teachers," Journal of Educational Research, Vol. 62 (September, 1968), pp. 25—29. '.---_-:...'. z-ni'w Is”- 47 Summary An impressive amount of research has been conducted with the Myers—Briggs Type Indicator. MBTI studies involving various teacher populations were reviewed in the present chapter. Although none of these studies was designed to test specific predictions about teachers, a number of consistent findings were reported and are summarized below. 1. The majority of public school teachers are sensing and judging types, the most fact—oriented, systematic, and organized of the types. 2. Female teachers are predominantly feeling types and male teachers are predominantly thinking types. 3. Teachers as a group appear to be quite similar in person— ality to the general population. 4. There are significantly higher proportions of intuitive, feeling, and perceptive types among groups of male and female pre— service teachers than among teachers in the field, which may indicate that the most enthusiastic and innovative types are dropping out of teaching. The types who tend to stay in teaching appear to be the sensing and judging types. 5. There is some evidence that sensors, feelers, and judging types are found in greater proportion at lower grade levels, and that intuitors, thinkers, and perceptive types are more likely to prefer teaching at the secondary level. .-"l IIZ::_______________________________________________——————_____—————————————————vvi:::::qllllll 48 6. Certain personality types appear to be attracted to certain subject areas. Compared to teachers with opposite preferences, a high— er proportion of (a) extraverts teach social studies; (b) sensors teach math, applied arts, and physical education; (c) intuitors, feelers, and perceptive types teach language and fine arts; (d) thinkers teach math and science; and (e) judging types teach math and applied arts. 7. Certain personality types appear to be attracted to school administration, particularly extraverts, thinkers, and judging types. 8. Teachers of different personality types tend to prefer different teaching styles. Introverted, sensing, thinking, and judging teachers are more likely to use a traditional classroom approach, with the teacher structuring the material and playing a central role in its presentation. In contrast, extraverted, intui— tive, feeling, and perceptive teachers are more likely to prefer in— formal individualized instruction, with students making more choices, and are also more likely to express their own positive and negative feelings overtly. 9. Intuitive teachers tend to prefer schools with ”open" cli— mates, sensing thinking types tend to prefer schools with ”closed” climates, and sensing feeling types appear to be quite satisfied in either type of school. 10. The question of whether teacher effectiveness is related to teacher personality has not been settled. There is some evidence that certain personality types may be more effective in teaching certain subjects and certain types of students, but more research is needed at both elementary and secondary levels. 49 The findings summarized above are significant, for the bulk of research on teacher personality has yielded negligible results. The studies reported in this chapter suggest that the Myers—Briggs Type Indicator measures important dimensions of personality and is a re- markably useful instrument for studying the role of teacher person— ality in the educational process. These findings were used in con— junction with Jungian theory to formulate the hypotheses of the present study. CHAPTER IV DESIGN OF THE STUDY The study was designed to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. The sample, procedure, hypotheses, statistical analysis, and instrumentation used in the study are described in the present chapter. The Myers—Briggs Type Indicator is further analyzed in Chapter Vo Sample The sample consisted of 200 pre—service teachers, all seniors majoring in elementary or secondary education at Michigan State Universityo 0f the entire sample, 68% were women and 32% were men, 40% were majoring in elementary education and 60% were majoring in secondary education. Their ages ranged from 20 to 44, with 94% in their early 20's. All of the students had completed at least one term of student teaching. Tables 4.1 and 4.2 illustrate the com- position of the sample according to grade level and subject area specialty. Two self-report instruments were used: the Myers—Briggs Type Indicator (MBTI) and the Teacher Preference Questionnaire (TPQ). The pre—service teachers completed the instruments during their last 51 Table 4.1 Distribution of the Sample by Grade Level Grade Level Males Females Total Pre—School to K 1 20 21 Grades 1 to 3 5 27 32 Grades 4 to 6 5 18 23 Grades 7 to 8 10 16 26 Grades 9 to 12 43 55 98 Total: 64 136 200 Table 4.2 Distribution of the Sample by Subject Area Subject Area Males Females Total Math / Science 16 16 32 Reading / English 3 9 12 Foreign Language 2 7 9 Social Studies 13 9 22 Fine Arts (Art, Music, Drama) 4 15 19 Applied Arts (Home Ec., Indus. Arts) 9 8 17 Physical Education 8 10 18 General Elementary 9 46 55 Special Education 0 16 16 Total: 64 136 200 53 required education course. Of those students enrolled in this course during the 1975 winter term, 88% completed both instruments. The sample is assumed to be representative of a typical graduating class of Michigan State University's College of Education. Hypotheses Six research hypotheses were tested, each one postulating that significant relationships exist between the personalities of pro— spective teachers and their attitudes toward certain teaching situa— tions. The six hypotheses are listed below: 1. Sensors, feelers, and judging types show stronger preferences for teachipg lower grades than intuitors, thinkers, and perceptive types. 2. Extraverts, thinkers, and judging types show more interest in administrative functions than introverts, feelers, and perceptive types. 3. Intuitors and perceptive types express a stronger need for independence and creativity than sensors and judging types . 4. Extraverts, intuitors, feelers, and judging types show more interest in plannipg school projects than introverts, sensors, thinkers, and perceptive types. 5. Extraverts, intuitors, feelers, and perceptive types show more interest in workigg with small groups of students than introverts, sensors, thinkers, and judging types. 6. Sensors and judging types express more commitment to classroom teachipg than intuitors and perceptive types. The hypotheses were based on previous findings with the Indicator as well as Jungian theory. 54 Statistical Analysis The hypotheses were tested in two ways, first by treating MBTI scores as continuous data and then by treating MBTI scores as dicho— tomous type categories. In both cases, MBTI scores were compared with TPQ scores. A variety of statistical techniques were employed, including Pearson product—moment correlations, point—biserial correlations, biserial correlations, and one—way analysis of variance. The rationale for using these specific statistical measures with the Myers-Briggs Type Indicator is presented in Chapter V, following a detailed description of MBTI scoring procedures. An alpha level of .05 was set for rejecting the null hypotheses. Alternate hypotheses were accepted when the statistical measures produced significant results in the predicted direction. Teacher Preference Questionnaire The Teacher Preference Questionnaire (TPQ) is a self~report instrument which was constructed especially for the present study. The questionnaire consists of 103 items designed to measure a teacher's attitude toward a large number of teaching situations. Pilot Study A preliminary form of the Teacher Preference Questionnaire was given to a group of 24 pre—service teachers enrolled in their last required education course during the 1974 fall term. The preliminary version of the questionnaire is presented in Appendix A. at 55 The pilot study revealed that students were able to complete the Myers—Briggs Type Indicator and the Teacher Preference Ques— tionnaire within one hour and fifteen minutes. In a followeup discussion, many of the students said that they enjoyed completing the questionnaire because it encouraged them to re—examine their student—teaching experiences with respect to their own individual preferences. A few questions were reported as ambiguous, particu— larly questions 2, 3, 7, 46 to 60, and 67 to 71. In addition, some of the students reported difficulty translating their various prefer— ences into numbers on a six—point scale. It was decided to follow Likert's procedure,1 replacing the six—point scales with attitude scales containing five—point statements. The final version of the Teacher Preference Questionnaire con— sisted of 103 items, 96 of which were Likert—scale items, and it incorporated the revisions suggested by the pilot study. The revised questionnaire is presented in Appendix B and is assumed to have reasonable content validity. Content Analysis After the data were collected, a first approximation to defining the cluster structure of the variables was obtained using principal components factor analysis followed by varimax rotations. The multiple—group method was then employed to evaluate the resulting 1Rensis Likert, "A Technique for the Measurement of Attitudes,” Archives of Psychology, Vol. 22 (June, 1932), p. 26. 56 clusters of variables. The end result was a set of six clusters, each one dealing with a different aspect of teaching. The six clusters are listed below: 1. Preference for teaching lower grades 2. Interest in administrative functions 5. Need for independence and creativity 4. Interest in planning school projects 5. Interest in working with small groups of students 6. Commitment to classroom teaching Descriptions of the items comprising each of the six clusters are listed in Table 4.3. The questionnaire items which did not appear to form clusters were disregarded. The TPQ clusters appear to be satisfactorily reliable, with Coefficient Alphas ranging from .70 to .81, as shown in Table 4.3. After the clusters were determined, each person's scores on the individual items comprising each cluster were totalled. Ques— tionnaire answers were thus transformed into six composite cluster scores. The six scores varied in range, depending on the number of items comprising each cluster, and provided data for testing the six hypotheses. Mean scores and standard deviations for each TPQ cluster are shown in Table 4.4. Item analysis of the questionnaire revealed that standard deviations for individual items were rather low, due to the tendency of the respondents to answer most questions positively. It is sug— gested that many of the statements be rephrased in negative form if further research is conducted with the questionnaire, since research rill. I'll. . III: 57 Table 4.3 Results of Cluster Analysis with the Teacher Preference Questionnaire N = 200 Coeff. Cluster Item Questionnaire Alpha Loading Number Items .77 Cluster 1: Preference for Teachipg Lower Grades .82 96. I've had the most teacher training at lower levels .73 *38. I would not like teaching grades 9 to 12 .60 36. I'd like teaching grades 1 to 5 .59 35. I'd like teaching pre—school or kindergarten .46 *39. I would not like teaching college .43 *21. The opportunity to teach a certain subject is not important to me .70 Cluster 2: Interest in Administrative Functions .60 80. I'd like to serve on faculty committees .56 83. I'd like to speak to community groups .54 72. I'd like to supervise students during lunch time .53 77. I'd like to participate in PTA activities .50 88. I'd like to serve as a department chairman .48 87. I'd like to observe and evaluate other teachers .47 73. I'd like to supervise students during study hall .46 8. I'm likely to be a school administrator 10 years from now .75 Cluster 3: Need for Independence and Creativity .64 32. Few opportunities to be creative might make me leave teaching .63 26. Lack of independence and freedom might make me leave teaching .59 28. Lack of intellectual stimulation might make me leave teaching .53 27. Having to work with adults who do not share my values might make me leave teaching .52 34. Too much paperwork and clerical duties might make me leave teaching .52 30. Having to spend all class time on subject matter might make me leave teaching .43 33. Too many ineffective and lazy teachers might make me leave teaching .37 *50. I would not like teaching in school D, which is traditional, stresses 3 R's, has good reputation *Item was weighted negatively for cluster analysis. 58 Table 4.3 (cont'd.) Coeff. Cluster Item Questionnaire Alpha Loading Number Items .81 Cluster 4: Interest in Planning School Projects .71 85. I'd like to work with students on community action projects .61 84. I'd like to organize class plays, talent shows, art shows, etc. .60 74. I'd like to organize a multi—media center .60 89. I'd like to plan teacher workshops .56 81. I'd like to attend professional conferences .54 670 I'd like to work with project committees .54 70. I'd like to serve as advisor to a student club .49 820 I'd like to arrange bulletin boards and displays .45 86. I'd like to set up a special math lab, reading lab .32 9. I'm likely to be a subject area specialist 10 years from now .78 Cluster 5: Interest in Working with Small Groups .92 61. I'd like working with 2 to 5 students regularly .77 62. I'd like working with 1 student regularly .69 60. I'd like working with 5 to 10 students regularly .78 Cluster 6: Commitment to Classroom Teachipg .76 1. I'd go into teaching again .72 99. I'm very satisfied right now with teaching as a career .57 *24. A heavy workload would not make me leave teaching .52 *25. Little opportunity to get away from children would not make me leave teaching .50 *12. I'm not likely to hold a job in another field 10 years from now .49 7. I'm likely to be a classroom teacher 10 years from now .46 14. The opportunity to spend most of the day with children is important to me .39 *31. Unruly or bored students would not make me leave teaching *Item was weighted negatively for cluster analysis. 59 Table 4.4 Means and Standard Deviations of TPQ Cluster Scores N=2OO TPQ Cluster Mean Std. Dev. Cluster 1: Preference for Teaching 14.360 5.251 Lower Grades Cluster 2: Interest in Administrative 14.910 4.735 Functions Cluster 3: Need for Independence and 15.100 5.151 Creativity Cluster 4: Interest in Planning School 12.625 5.469 Projects Cluster 5: Interest in Working with 2.600 2.925 Small Groups of Students Cluster 6: Commitment to Classroom 10.075 5.270 Teaching Earlrr- - 60 suggests that negative items tend to be more differentiating and have greater validity than positive items.2 Nevertheless, in its present form the Teacher Preference Questionnaire appears to be an adequate instrument for measuring a teacher's attitude toward a variety of teaching situations. Myers-Briggs Type Indicator The Myers—Briggs Type Indicator (MBTI) is a self—report instru— ment based on Carl Jung's theory of personality. The Indicator con— sists of 166 items designed to measure the following preferences: (1) Extraversion or Introversion, (2) Sensation or Intuition, (3) Thinking or Feeling, and (4) Judgment or Perception. A detailed description of the Myers-Briggs Type Indicator is presented in Chapter V, including an analysis of scoring techniques, statistical procedures, reliability data, and validity studies conducted with the Indicator. Summary The major problem of the present study was to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. The sample, procedure, hypo— theses, and statistical analysis used in the study were described 2N. L. Gage, G. S. Leavitt, and G. C. Stone, "The Psychological Meaning of Acquiescence Set for Authoritarianism," Journal of Abnormal Social Psychology, Vol. 55 (July, 1957), p. 101. 61 in the present chapter. A detailed description of the Teacher Preference Questionnaire was also presented, including results from the pilot study and final content analysis. An analysis of the Myers—Briggs Type Indicator is presented in the following chapter. CHAPTER V MYERS—BRIGGS TYPE INDICATOR The Myers-Briggs Type Indicator is a self—report instrument developed in the 1950's to measure personality type. It is based on Carl Jung's theory that much apparently random variation in human behavior is actually quite orderly and consistent, due to certain basic differences in the way people approach life. Form F of the Indicator was used in the present study. It consists of 166 items designed to measure the following preferences: 1. Extraversion (E) or lntroversion (I) 2. Sensation (S) or Intuition (N) 3. Thinking (T) or Feeling (F) 4. Judgment (J) or Perception (P) Both the full names and the one—letter abbreviations of the prefer— ences are used in the present study. Detailed descriptions of the preferences may be found in Chapter II. Scoring Because the aim of the Myers—Briggs Type Indicator was to determine habitual choices between opposites, the questions were set up in forced—choice form. Each scored item has one answer weighted in favor of one of the eight preferences and the other answer weighted 62 _ 63 in favor of the opposing preference. Different weights have been assigned to certain answers in an attempt to offset social desira— bility bias.1 The Indicator yields two types of scores for each person. It classifies respondents on four dichotomous type categories, and it also produces eight numerical scores which can be transformed into four continuous scores. MBTI scores may therefore be regarded as either dichotomous or continuous data. Type Categories To determine the person's type, the points for each preference are totalled, yielding eight numerical scores. These eight scores are interpreted as four pairs of scores, with the larger of each pair indicating the preferred pole. When the two poles appear to be equally preferred, males are arbitrarily assigned to the I, N, T, or P classification and females are arbitrarily assigned to the I, N, F, or P classification. For example, a male or female with an E score of 12 and an I score of 12 is typed as an introvert. The final result is that a person is classified as one of 16 possible types: ISTJ, ISFJ, INFJ, INTJ, ISTP, ISFP, INFP, INTP, ESTP, ESFP, ENFP, ENTP, ESTJ, ESFJ, ENFJ, or ENTJ. The distribution of types for the present sample is shown in Table 5.1. It is quite similar to distributions obtained in other studies of pre-service teachers, with ENFP being the modal type (the predominant type). As expected, there were significantly more buyers, MBTI Manual, p. 86. |. .! i 64 Table 5.1 MBTI Type Table for the Present Sample N = 200 ISTJ ISFJ INFU INTJ N = 14 N = 18 N = 9 N : 2 75 = 7 75 = 9 75 = 4 % = 1 UUUUUUU UUUUUUUUU UUUU [I ISTP ISFP INFP INTP N = 3 N = 19 N = 14 N : 8 70 : 1 % : 10 % : 7 71,; 2 it [I UUUUUUUUUU UUUUUUU UUUI] ESTP ESFP ENFP ENTP N = 7 N = 14 N : 28 N = 11 % = 3 % = 7 % = 11. % = 6 [IUD UUUUUUU DEDUUUUUUU UUUUUU UUDU ESTJ ESFU ENFJ ENTJ N = 10 N = 12 N = 19 N = 12 % = 5 % = 6 % = 10 % = 6 UUUUU UUUUUU UDUUUUUUUU HUUUUU Note. [] represents 1% of sample. a a z m h . .- . a u , A. . . ‘ - - ..- G n .. . . I ’_ . . . . 1 . . . \. ’ 7 I 4| 70 Analyzipg MBTI Scores A significant problem in dealing with the Myers—Briggs Type Indicator is whether to restrict analyses to dichotomous data based on type classifications, or whether to regard scores as continuous quantitative data. The Manual presents the Indicator primarily as an instrument to determine type, but the instrument also measures the reported strength of each type preference. Whether it is preferable in a particular study to treat scores as dichotomous or continuous data will ultimately depend on the nature of the study. In the pres— ent study a theoretical case could be made for both approaches, so both methods were employed, permitting a comparison of the relative utility of the two approaches. Statistical procedures for type categories. Throughout the study, two statistical procedures were used with type categories: phi coefficients and tetrachoric coefficients. The phi correlation coefficient, ¢ , is most appropriate for dichotomous variables where there is no distribution underlying the dichotomies, such as the categories "male" and "female.” Estimates of relationship derived with phi coefficients are nearly always lower than estimates derived with Pearson product—moment coefficients, and difficulties arise when these two measures of relationship are com— pared with each other.5 When used with MBTI type categories, the 5Gene V. Glass and Julian C. Stanley, Statistical Methods in Education and Ps cholo (Englewood Cliffs, N.J.: Prentice—Hall, 1970 . pp. 1161—162. ——_ ::lllll'! 71 phi correlation is likely to underestimate relationships because there is always E222 kind of frequency distribution underlying the dichotomous type categories. As Myers states in the Manual, phi coefficients are low (conservative) estimates of true values because most of the score information has been sacrificed to the coarsest possible grouping.6 The tetrachoric correlation coefficient, r is used with tet’ dichotomous variables when there is a normal distribution underlying the dichotomies, such as "children shorter than 4 feet and "children taller than 4 feet." When used with MBTI type categories, the tetrachoric correlation is likely to overestimate relationships because the underlying frequency distributions are generally not normal. True correlations involving type categories cannot be directly estimated with existing statistical procedures, but can be expected to fall somewhere between estimates derived with phi coefficients and estimates derived with tetrachoric coefficients. In the present study, both statistical procedures were used whenever type categories were involved, yielding a low and a high estimate of correlation. When type categories were compared to continuous quantitative data, variations of these two procedures were employed: The point—biserial correlation coefficient was used to produce a low estimate and the biserial correlation coefficient was used to produce a high estimate. It is logical to assume that the true correlation falls somewhere 6Myers, MBTI Manual, po 20a. 72 between these two estimates. Such an approach is recommended for future research with the Indicator. Analysis of variance is another possible technique when MBTI scores are treated as dichotomous data, and it has been used by a number of researchers with the Indicator. A one—way ANOVA design might include the E—I type category as the independent variable; a two—way ANOVA design might include the E—I and T—F type categories as independent variables. Analysis of variance is not an ideal method for comparing personality types, however, because subjects are not randomly assigned to treatment groups (personality types) and significant differences between groups may be due to confounding variables. Nevertheless, analysis of variance can be a useful took for testing for differences between personality types if causal inferences are not drawn from the results. In the present study, both correlational statistics and analysis of variance were used to test the hypotheses. Statistical procedures for continuous scores. Most researchers have used the Pearson product—moment correlation coefficient in their investigations involving continuous MBTI scores. The Pearson coef— ficient would seem to be an appropriate statistic because it does not require the distributions to be normal. As Carroll states: "No assumptions are necessary for the computation of a Pearson coef— ficient, but the interpretation of its meaning certainly depends upon the extent to which the data conform to an appropriate statistical m-I ': -. 73 ,7 model for making this interpretation.‘ In other words, normality is an assumption not so much for using the Pearson coefficient, but for interpreting the results. Research has shown that if the two distributions being compared have similar shapes, departures from normality have little effect on the correlation, but when the two distributions have quite different shapes, the size of the correla- tion is restricted, particularly with very high correlations. With moderate correlations, however, differences in the shapes of the distributions tend to have little effect.8 From these findings, Nunnally has concluded that "the results of correlating two con— tinuous variables in most studies in psychology would be about the same whether the distributions were shaped the same or somewhat dif— ferently.9 The Pearson correlation coefficient would thus appear to be an appropriate statistical procedure for continuous MBTI scores, and it was used throughout the present study whenever MBTI scores were treated as continuous data. Intercorrelations of MBTI Scores Many researchers have examined the intercorrelations of the four indices, some treating MBTI scores as dichotomous type cate— gories and some treating scores as continuous data. A summary of 7John B. Carroll, "The Nature of the Data, or How to Choose a Correlation Coefficient," Psychometrika, Vol. 26 (December, 1961), p. 349. 8Jum C. Nunnally, Ps chometric Theor (New York: McGraw—Hill Book Co., 1967) pp. 128—129 "‘“Z 9 Nunnally, p. 129. 7 4 intercorrelation studies is presented below, including data obtained with the sample of 200 prospective teachers. The following summary is limited to studies using Form F of the Indicator with samples of college students and in—service teachers. Type Categories The relative independence of the dichotomous MBTI type cate— gories was examined in studies by Webb10 and by Stricker and Ross.11 In both of these experiments, phi coefficients were used to estimate the intercorrelation among type categories, a statistical procedure which is likely to underestimate the actual correlation. The results of these studies are shown in Table 5.5 and reveal that the E—I, S—N, and T—F categories appear to be fairly independent of each other. The J—P category, however, was found to correlate rather consistently with the S—N category, indicating that sensors tend to be judging types and intuitors tend to be perceptive types. The apparent inde— pendence of the first three scales lends support to Jung's theory that there are actually only three typological dimensions: E—I, S-N, and T—F. In the present study, phi coefficients and tetrachoric coeffi— cients were computed to determine low and high estimates of inter— correlation between type categories. Results with the sample of 10Sam C. Webb, “An Analysis of the Scoring System of the Myers— Briggs Type Indicator,” Educational and Ps cholo ical Measurement, V01. 24 (Winter, 1964), pp. 769—770, 11Lawrence J. Stricker and John Ross, “Intercorrelations and Reliability of the Myers—Briggs Type Indicator Scales," Psychological Reports, Vol. 12 (February, 1963), pp. 288—289. 75 Table 5.5 Intercorrelations of MBTI Type Categories (From Previous Studies) Sample MBTI Type Categories a EI/SN EI/TF EI/JP SN/TF SN/JP TF/JP Phi Coefficients 300 Male Freshmen at L.I.U. .08 —.04 .14* .07 .23** .09 (Stricker & Ross, 1963) 184 Female Freshmen at L.I.U. —.03 .05 .13 —.02 .31** —.05 (Stricker & Ross, 1963) 422 Freshmen at Emory —.04 —.03 —.01 —.05 .19** .08 (Webb, 1964) aPhi coefficients tend to yield low estimates of true values. *Significant at the .05 level. **Significant at the .01 level. 76 Table 5.6 Intercorrelations of MBTI Type Categories for the Present Sample Sample MBTI Type Categories a EI/SN Phi Coefficients 136 Female Pre—Service —.21* Teachers 64 Male Pre—Service —.28* Teachers Tetrachoric Coefficientsb 136 Female Pre—Service —.33** Teachers 64 Male Pre—Service —.43** Teachers Total Group: 200 Pre—Service Teachers —.24** (Phi Coefficients) 200 Pre—Service Teachers —.37** (Tetrachoric Coefficients) EI/TF —.05 .28* .10 . 4396* .05 ELAJP -.01 —.09 -.01 —.14 —.02 —.04 SN/TF SN/JP TE/IP .00 .69** —.02 o 3396* .49** .13 .20 .21* .32** 015* .23** .07 .32** .12 . 4996* .12 . 2096* aPhi coefficients tend to yield 10w estimates of true values. Tetrachoric coefficients tend to yield high estimates of true values. *Significant at the .05 level. **Significant at the .01 level. 77 200 prospective teachers are shown in Table 5.6, and reveal signi— ficant correlations between many of the type categories: The S—N, T—F, and J—P categories were found to be positively correlated with each other, and E—I was found to be negatively correlated with S—N. For this sample of prospective teachers, the E, N, F, and P classi— fications appear to be related and the I, S, T, and J classifications appear to be related. Continuous Scores A considerable number of researchers have investigated the relay tive independence of continuous MBTI scores, all using Pearson 2 product—moment correlations: Richek,1 Schmidt and Fretz,13 Stricker, 15 Schiffman, and Ross,14 Webb, Stricker and Ross,16 Madison, Wilder, 17 18 and Suddiford, and Myers. The results of these studies are shown 12Herbert G. Richek, "Note on Intercorrelations of Scales of the Myers—Briggs Type Indicator,” Psychological Reports, Vol. 25 (August, 1969) , pp. 28—30. 13Schmidt and Fretz, p. 27. 1lltLawrence J. Stricker, Harold Schiffman, and John Ross, ”Predic— tion of College Performance with the Myers—Briggs Type Indicator,” Educational and Psychological Measurement, V01. 25 (Winter, 1965), pp. 1087—1089. 15webb, pp. 769—770. 16Stricker and Ross, "Intercorrelations and Reliability of the MBTI Scales,” p. 289. 17Peter Madison, David H. Wilder, and Walter B. Suddiford, "The Myers—Briggs Type Indicator in Academic Counseling," Statistical Research Report No. 13 (Princeton, N.J.: Princeton University Counseling Service, 1963), pp. 5—6. I 18Myers, MBTI Manual, p. 11. _———_———-_ ': Hun": 9'13 -. ' .- _..1'n:_ Ir_-I_;-..- — 78 in Table 5.7 and are similar to the previously mentioned type— category results obtained by Webb and by Stricker and Ross: The E-I, S—N, and T—F scales appear to be relatively independent of each other, and the J—P scale appears to correlate consistently with the S—N scale and less consistently with the T—F scale. Similar results were obtained by both Saunders19 and Ross20 using factor analysis. In the present study, Pearson product—moment correlations were computed to estimate the interdependence of the four continuous scales. Results with the sample of 200 prospective teachers are shown in Table 5.8, and reveal a significant correlation between the S—N scale and the J—P scale, especially for females. Other significant correlations were also found, generally consistent with results obtained with the same sample when scores were treated as dichotomous type categories. The EI/SN correlation is most surpris- ing, for it has rarely occurred with other populations, and was not found by Richek21 in his study of 435 prospective teachers (see Table 5.7). The numerous intercorrelations found with the present sample indicate that additional research is needed to investigate the interrelationships of MBTI scores with samples of teachers. 19D. R. Saunders, "Evidence Bearing on the Existence of a Ra— tional Correspondence Between the Personality Typologies of Spranger and Jung," Research Bulletin 60—6 (Princeton, N.J.: Educational Testing Service, 1960), pp. 11—15. 20John Ross, "The Relationship Between a Jungian Personality Inventory and Tests of Ability, Personality, and Interest," Australian Journal of Psychology, Vol. 18 (April, 1966), pp. 9—10, 21Richek, ”Note on Intercorrelations,” p. 28. 79 Table 5.7 Intercorrelations of Continuous MBTI Scores (From Previous Studies) ‘ Sample MBTI Scales EI/SN EI/TF EI/JP SN/TF SN/JP TF/UP 300 Female Pre-Service Secon— .03 .06 .03 .08 .28** .14* dary Teachers (Richek, 1969) 65 Female Pre—Service Elemen— .13 —.23 .10 —.08 .29** .29** tary Teachers (Richek, 1969) 70 Male Pre—Service Teachers —.06 —.26* .07 .02 .09 .20 (Richek, 1969) 591 Male Freshmen at Brown —.03 —.03 .05 .02 .30** .10* (Myers, 1962) 300 Male Freshmen at L.I.U. .06 .03 .08 .02 .33** .18** (Stricker & Ross, 1963) 225 Male Freshmen at Wesleyan .00 —.08 .02 .14* .34** .24** (Stricker et al., 1965) 201 Male Freshmen at Caltech —.13* —.15* —.12 .00 .24** .23** (Stricker et al., 1965) 102 Male Freshmen at Princeton .17 —.09 .08 _,20* .48** _‘06 (Madison et al., 1963) 240 Female Freshmen at Pem— —.09 —.19** —-05 .05 ,39** .12 broke (Myers, 1962) 184 Female Freshmen at L.I.U. —.14 .04 .01 .06 .47** —.02 (Stricker & Ross, 1963) 422 Freshmen at Emory .00 —.04 .01 —.13** .25** —.06 (Webb, 1964) 403 Students at Ohio State -.07 .07 .02 .08 .39** .17** (Schmidt & Fretz, 1965) *Significant at the .05 level. **Significant at the .01 level. 1- 80 Table 5.8 Intercorrelations of Continuous MBTI Scores for the Present Sample Sample MBTI Scales EI/SN EI/TF EI/JP SN/TF 136 Female Pre—Service —.23** —.06 —.11 .16 Teachers 64 Male Pre—Service —.26* .20 .18 .04 Teachers Total Group: 200 Pre—Service Teachers -.25** .01 —.02 .13 SN/JP TF/JP .46** .17* .27* .33** .39** .21** *Significant at the .05 level. **Significant at the .01 level. :i- r...» -. weir: 81 Reliability of the Myers—Briggs Type Indicator A number of researchers have investigated the reliability of the Myers—Briggs Type Indicator, some treating MBTI scores as dicho— tomous type categories and some treating scores as continuous data. A variety of statistical procedures have been employed, making it rather difficult to compare findings. A summary of studies involving the reliability of the Indicator is presented below. Reliability data obtained with the present sample of 200 prospective teachers is included. Two aspects of reliability are examined: measures of in- ternal consistency and measures of stability. The following summary is limited to studies using Form F of the Indicator with samples of college students and in—service teachers. Measures of Internal Consistency Internal—consistency reliability is based on the average corre— lation among items within a test. [ype categories. A variety of statistical procedures have been used to measure the internal consistency of the dichotomous type cate- gories of the Myers—Briggs Type Indicator. One approach, reported by Webb22 and by Stricker and Ross,23 has been to compute lower—bound estimates of reliability with Guttman's two procedures, but such estimates are not very informative without corresponding upper-bound 22Webb, pp. 773—779. 23Stricker and Ross, "Intercorrelations and Reliability of the MBTI Scales," pp. 289—290. 82 estimates. Another approach, reported by Webb24 and by Myers,25 has been to estimate split—half reliabilities by calculating phi correlation coefficients and applying the Spearman—Brown prophecy formula. The phi coefficient, however, is likely to underestimate the reliability of a particular type category. Myers' recommendation is to estimate split—half reliabilities by calculating tetrachoric correlation coefficients and applying the Spearman—Brown prophecy formula.26 This approach yields higher reliabilities, for it assumes additional information: that the underlying scores are normally distributed across each index. Re— liability estimates based on tetrachoric r's will probably be too high, however, because the tetrachoric correlation depends on a strict assumption of normality,27 and frequency distributions of Indicator scores are generally not normal. For any particular sample, actual type—category reliabilities can be expected to fall somewhere between estimates derived with phi coefficients and estimates derived with tetrachoric coefficients. Results of previous studies estimating the reliability of the type categories with phi and/or tetrachoric coefficients are shown in Table 5.9. Reliabilities appear to be satisfactory, although there is a rather wide range between conservative and liberal estimates 2tween, pp. 773-779. 0 I5Myers, MBTI Manual, pp. 20a—20b. 26Myers, MBTI Manual, pp. 20a—20b. f) ‘7Nunnally, p. 123. 83 Table 5.9 Estimates of Internal—Consistency Reliability of MBTI Type Categories (From Previous Studies) Sample MBTI Type Categories Phi Coefficientsa 100 Male Freshmen at Brown .55 .73 .75 .58 (Myers, 1962) 100 Female Freshmen at Pembroke .65 .64 .67 .68 (Myers, 1962) 99 Male Freshmen at Emory .58 .64 .68 .84 (Webb, 1964) 100 Female Freshmen at Emory .58 .64 .43 .84 (Webb, 1964) . ... . b Tetrachoric Coeffic1ents 100 Male Freshmen at Brown .74 .88 .90 .76 (Myers, 1962) 100 Female Freshmen at Pembroke .81 .83 .84 .84 (Myers, 1962) 117 Pre-Service Teachers .75 .92 .81 .82 (Hoffman, 1974) 113 In—Service Teachers .70 .82 .66 .80 (Hoffman, 1974) aPhi coefficients tend to yield low estimates of true values. Tetrachoric coefficients tend to yield high estimates of true values. Table 5.10 Estimates of Internal—Consistency Reliability of MBTI Type Categories for the Present Sample Sample MBTI Type Categories Phi Coefficientsa 136 Female Pre—Service Teachers 64 Male Pre-Service Teachers Tetrachoric Coefficientsb 136 Female Pre-Service Teachers 64 Male Pre—Service Teachers Total Group: 200 Pre—Service Teachers (Phi Coefficients) 200 Pre-Service Teachers (Tetrachoric Coefficients) E—I .68 .86 .92 .76 .88 .89 .87 .62 .67 .83 .67 .78 .86 .73 .88 aPhi coefficients tend to yield low estimates of true values. Tetrachoric coefficients tend to yield high estimates of true values. 85 of internal consistency. In the present study, a method similar to Myers' split—half procedure was used to estimate the reliability of each type category. The specific procedure is described in Appendix D. Low and high estimates of reliability are shown in Table 5.10, and reveal reason— ably high reliability coefficients, averaging in the high seventies. Type—category scores appear to be adequately reliable for the present sample of prospective teachers. Continuous scores. Researchers have used various methods to estimate the reliability of continuous scores. Myers28 developed a split—half procedure involving Pearson product—moment correlations, 29 Webb used a split—half procedure similar to Myers' method, and 30 Stricker and Ross used Cronbach's Coefficient Alpha. The three procedures have produced similar results, which are presented in Table 5.11. Most of the estimated reliabilities of continuous scores fall within a narrow range, in the high seventies and low eighties. They are somewhat higher than estimates of type—category reliability because information is lost in changing from continuous scores to dichotomous categories. In the present study, estimates of the reliability of continuous MBTI scores were computed using Myers' split—half procedure, as 28Myers, MBTI Manual, p. 20. 29Webb, pp. 767—768. 30Stricker and Ross, "Intercorrelations and Reliability of the MBTI Scales,” pp. 289—290. Table 5.11 Estimates of Internal—Consistency Reliability of Continuous MBTI Scores (From Previous Studies) Sample MBTI Scales 300 Male Freshmen at L.I.U. (Stricker & Ross, 1963) 100 Male Freshmen at Brown (Myers, 1962) 99 Male Freshmen at Emory (Webb, 1964) 184 Female Freshmen at L.I.U. (Stricker & Ross, 1963) 100 Female Freshmen at Pembroke (Myers, 1962) 100 Female Freshmen at Emory (Webb, 1964) 117 Pre—Service Teachers (Hoffman, 1974 113 In—Service Teachers (Hoffman, 1974) .71 .63 .85 .80 .87 .85 .81 .82 .78 .71 083 .69 .68 .67 .80 .82 .81 .84 .82 .80 87 Table 5.12 Estimates of Internal—Consistency Reliability of Continuous MBTI Scores for the Present Sample Sample MBTI Scales 136 Female Pre—Service Teachers 64 Male Pre—Service Teachers .87 Total Group: 200 Pre—Service Teachers 88 described in Appendix D. Results with the sample of 200 prospective teachers are shown in Table 5.12, and reveal relatively high relia— bility coefficients, mostly in the eighties. Measures of Stability Test-retest experiments involving the Myers—Briggs Type Indi— cator have provided information regarding the relative stability of the type preferences over a period of time. Type categories. Test—retest data for MBTI type categories have been reported for four samples: Studies by Levy, Murphy, and Carlson,31 Stalcup,32 and Stricker and Ross33 involved college stu— dents, and a study by Wright34 involved elementary school teachers. The proportion of agreement between the original and the retest type classifications for each of these experiments is shown in Table 5.13. In every case, the proportion of agreement was significantly higher than would be expected by chance. The college populations appear to have maintained reasonably 31Nissim Levy, Glennie Murphy, Jr., and Rae Carlson, ”Personality Types Among Negro College Students," Educational and Psychological Measurement, Vol. 32 (Autumn, 1972), pp. 644—645. 32Donna L. K. Stalcup, "An Investigation of Personality Charac— teristics of College Students Who Do Participate and Those Who Do Not Participate in Campus Activities" (unpublished doctoral dissertation, Auburn University, 1967), pp. 45—52. 33Lawrence J. Stricker and John Ross, "An Assessment of Some Structural Properties of the Jungian Personality Typology," Journal of Abnormal and Social Psychology, Vol. 68 (January, 1964), p. 64. 3[*Wright, pp. 81—83. _. _.——_—a- l ill-“'15- ..l' IrJul‘u C ' ' :H'T'l-MI'I' 89 so so meos< a an owseno an an meon< n sh ewqeno sum sea mewn¢ N an owsdno sac son Ram Ram awn awn seam. owns H an and an ammono owdcao 02 sea sea see see see sac enemas sesame mlb an mtg :H ewddno owaeno 02 oz saw smw sum assess 21m :H ewfleno oz Aeemfi .ssmanzv amw macaw.o popes wepmopom whenceee khepqefiofim so Aroma .nnom s bosoneemv awe masses sH seems sonnoeom ensoasa es necessem as Anwmfl .mdoflepwv §Nw masow m seams woemopem ghdps< we mesowdpm awn “mama ..Ho so aeonv mnemoz N Hopes wewmoeom unseen so ossossnm mms wsflmha Him an owqafio oz seseoosmsnneno ease Ham: as newness Amoflwdpm mSOHbohm Eonmv momoom hhowoeuo cth poopem use fidqflwflso we nomflnemaoo mH.m canoe camsem IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII! 90 stable scores over a period of time, with a clear majority of the subjects displaying either complete stability or a shift in only one of the four basic variables. Wright's study of 94 elementary teachers is even more impressive. When the teachers were retested six years after they had originally taken the Indicator, 61% of them remained in the same category on all four scales. Also, each of the four scales appeared highly stable when considered separately. Wright's findings indicate that the test—retest correlations for type categories may be exceptionally high for teachers who remain in the same school district for at least six years. Perhaps stability of scores is a function of occupation or age. Further research is needed to examine test-retest data with different occupational groups and different age groups. Continuous scores. In two of the four retest studies, test— retest correlation coefficients were computed for continuous MBTI scores: the study by Levy, Murphy, and Carlson35 36 and the study by Stricker and Ross. The Pearson product—moment correlations were all significant at the .01 level and appear reasonable for an instru— ment of this kind (see Table 5.14). One interesting finding is that the female students scored more consistently over a period of time than their male counterparts. Future researchers should examine test— retest correlations with larger populations of males and females. 35 36Stricker and Ross, ”An Assessment of Some Structural Proper— ties," p. 64. Levy, Murphy, and Carlson, pp. 644—645. 91 Table 5.14 Test—Retest Correlation Coefficients for Continuous MBTI Scores (From Previous Studies) Sample MBTI Scales E—I S-N T—F J—P 287 Female Students at Howard Retested after 2 Months .83 .78 .82 .82 (Levy et al., 1972) 146 Male Students at Howard Retested after 2 Months .80 .69 .73 .80 (Levy et al., 1972) 41 Male Students at Amherst Retested after 14 Months .73 .69 .48 .69 (Stricker & Ross, 1964) 92 Validity of the Myers—Briggs Type Indicator The validity of the Myers—Briggs Type Indicator is dependent on how well it measures what it was intended to measure: the theoretical constructs of Jung's typology. Selecting criteria by which to quan— titatively measure Jung's concepts is not an easy task. Nevertheless, an impressive number of studies have compared Indicator scores with criteria which is assumed to reflect Jung's constructs, and the findings have shown that different personality types do tend to be— have in consistently different ways. A summary of research on the validity of the Indicator is presented below. Using Nunnally's definitions,37 three types of validity are examined: content validity, predictive validity, and construct validity. Content Validity The content validity of an instrument rests mainly upon an appeal to reason regarding the propriety of the content and the way that it is presented in the form of test items. Myers' extensive account38 of the construction of the Indicator includes the criteria used for choosing and scoring items, and provides considerable evi— 39 dence for the instrument's content validity. Stricker and Ross have also examined item content. They concluded in 1962 that the 37Nunnally, pp. 75-99- 38Myers, MBTI Manual, pp. 83—87. 39Stricker and Ross, “A Description and Evaluation of the MBTI," pp. 59—65. 93 S—N and T—F scales seem largely consistent with their corresponding conceptual definitions, but the E—I and J-P scales may measure some- thing quite different from the dimensions postulated by Myers. The question has not been settled, but it would appear from an inspection of the scored items that the E—I, S-N, and T—F scales are generally consistent with the content of Jung's typological theory. Another type of evidence for content validity was obtained by Bradwayli0 in a study involving 28 Jungian analysts. The analysts were asked to classify themselves according to the E—I, S—N, and T—F type categories, and comparisons were then made between self-typing and MBTI typing. There was 100% agreement on E—I classification, 68% agreement on S—N classification, 61% agreement on T—F classifi— cation, and 43% agreement on all three dimensions. The E—I index thus proved to be remarkably valid for this sample, and the T—F index was shown to be the least valid. Additional evidence for content validity has been obtained by correlating subjectS' MBTI scores with their scores on the Gray— Wheelwright Questionnaire, another instrument designed to identify Jungian types. The Gray—Wheelwright is similar to the Indicator in that it uses continuous scores to assign subjects to type cate— gories, but it has no J-P scale. Bradway's studyli1 of 28 Jungian analysts compared their scores 40Katherine Bradway, ”Jung's Psychological Types: Classification by Test Versus Classification by Self,“ Journal of Analytical Psychology, V01. 9 (July, 1962.), pp. 129-135. 41Bradway, p. 132. 94 on the two instruments, and found that 96% of the analysts received the same E—I classification with both tests, 75% received the same S—N classification, 72% received the same T—F classification, and 54% received identical classifications on all three dimensions with both tests. The proportion of agreement between the tests was significantly higher than would be expected by chance. Grant!12 administered the two tests to college freshmen, however, and obtained a much lower propor- tion of agreement. 43 Stricker and Ross compared continuous scores received on the Myers—Briggs Type Indicator and the Gray—Wheelwright using a sample of 47 male college students. The two E—I scales exhibited a .79 correlation, the S—N scales showed a .58 correlation, and the T—F scales showed a .60 correlation. All three correlations were signi— ficant at the .01 level, and lend support to Myers' contention that "both tests are reflecting the same basic realities, that is, the Jungian opposites which both were designed to reflect.”l£lk Predictive Validity Predictive validity is at issue when an instrument is used to estimate some form of behavior, the latter being referred to as the ”criterion." Predictive validity is especially important when the 42W. Harold Grant, ”Comparability of the Gray—Wheelwright Psychological Type Indicator and the Myers—Briggs Type Indicator" (Auburn, Ala.: Auburn University Student Counseling Service, 1965). 43Stricker and Ross, "Some Correlates of a Jungian Personality Inventory,” p. 628. 44Myers, MBTI Manual, p. 22. 95 instrument is used to help make decisions about people. Four studies have examined the Indicator‘s ability to predict choice of major and success in college. Goldschmidli5 derived regres— sion equations to forecast college major along a science continuum and along a humanities continuum with a small sample of undergrad— uates, and found that the Indicator's scales had moderate predictive validity. In another study, Conary46 predicted that certain specific personality types in his sample of 1709 entering freshmen would be more likely than other types to receive good grades and to make specific curricula choices during their freshman year, and the pre— dictions were substantiated. Stricker, Schiffman, and Ross47 con— ducted a similar study with three samples of entering freshmen and concluded that the Indicator's scales had some ability to predict GPA and dropout, but this ability varied considerably with the nature of the sample. They found that a contingency measure combining all four type categories generally had greater predictive validity than did the individual scales. Using a very small sample, Saunderslk8 predicted the relative 45Marcel L. Goldschmid, "Prediction of College Majors by Person— ality Type," Journal of Counseling Psychology, Vol. 14 (July, 1967), pp. 302—308. 46Franklin M. Conary, "An Investigation of the Variability of Behavioral Response of Jungian Psychological Types to Select Educa— tional Variables" (unpublished doctoral dissertation, Auburn Univer— sity, 1965). £17Stricker, Schiffman, and Ross, pp. 1081—1095. 48D. R. Saunders, "Evidence Bearing on Use of the Myers—Briggs Type Indicator to Select Persons for Advanced Religious Training,” Research Bulletin 57—8 (Princeton, N.J.: Educational Testing Service, 1957), pp. 11—12. 96 likelihood that 13 Rockefeller Theological Fellows would adjust to their role as divinity students. The predictions, based on MBTI scores and a measure of "strain," were recorded in a letter in the form of a rank—ordering of types. Results showed that five of the first six students in rank order decided to continue their theo— logical education the following year, whereas six of the last seven drapped out, a finding which is significant beyond the .01 level. However, Saunders‘ study is based on a small sample, and contaminating variables may have influenced the results. The studies cited above suggest that the Indicator has moderate predictive validity in certain areas. Additional studies are needed, however, if the instrument is to be used to help make decisions about people. Construct Validity Construct validity is at issue when an instrument purports to measure abstract variables, referred to as "constructs." In order to evaluate the construct validity of an instrument, observable behaviors which are related to the construct must be specified. Studies may then be conducted to determine how well the test corre— lates with the related behaviors. Numerous studies have been conducted with the Myers—Briggs Type Indicator comparing MBTI scores with scores on other instruments, but there have been few attempts to specify beforehand the behaviors associated with each personality type. Nevertheless, a wealth of circumstantial evidence has been gathered, and results appear to be quite consistent with Jungian theory. 97 Several researchers have used factor analysis to investigate the relationship between the constructs measured by the Indicator, and constructs measured by other tests. Saunderslk9 compared the continuous MBTI scores of 1132 subjects with their scores on the Allport—Vernon—Lindzey Study of Values, an instrument based on Spranger's theory of types. Factor analysis revealed that the four Jungian type dimensions formed a good simple structure and both instruments appeared to be measuring related constructs. In studies 50 and by Madison, Wilder, and Suddiford,51 factor analysis by Ross was used to relate a variety of tests to the Myers—Briggs Type Indi— cator. In all of the studies, the four MBTI scales tended to have substantial loadings on different factors, lending support to Myers and Briggs' premise of a four—dimensional interlocking structure of personality. The construct validity of the Indicator has also been assessed by correlating continuous MBTI scores with scores on other person— ality tests, such as the Edwards Personal Preference Schedule (EPPS) and the Allport—Vernon—Lindzey Study of Values (AVL). In addition, MBTI scores have been correlated with such variables as choice of occupation, job turnover rate, faculty ratings, learning preferences, and academic performance. A summary of research findings is presented kgsaunders, "Personality Typologies of Spranger and Jung," Pp. 11—15. 50Ross, "Jungian Personality Inventory," pp. 9—10; see also Stricker and Ross, "A Description and Evaluation of the MBTI," pp. 30, 123. 51Madison, Wilder, and Suddiford, pp. 3-9. 98 below. In order to compare results, each index is considered separately and only those correlations significant at the .01 level are reported. E—I validity. Extraverted types appear to like action and getting involved in new situations. Their scores on a number of tests indicate that they tend to be talkative, gregarious, and im— pulsive, with underlying needs for dominance, exhibition, and affil— iation.52 They are generally more competitive than introverted types, and are much more likely to be rated as ”active" by faculty members.53 In high school, extraverts report that they like to work on group projects, make reports to the class, and engage in team 54 competitions. Vocationally, they prefer active jobs in which they can interact with others, such as salesman, personnel director, social worker, public administrator, or athletic director.55 Studies indicate that introverted types would rather reflect before acting and they enjoy working alone. They score high on 52Stricker and Ross, ”A Description and Evaluation of the MBTI," pp. 105, 120; see also Myers, MBTI Manual, pp. B4—B6; see also Ross, ”Jungian Personality Inventory," p. 13; see also Webb, pp. 771,775. 53John Ross, "Progress Report on the College Student Character— istics Study,” Research Memorandum 61-11 (Princeton, N.J.: Educational Testing Service, 1961), p. 13. 54Mary H. McCaulley and Frank L. Natter, "Psychological Type Differences in Education," The Governor's Task Force on Disruptive Youth, ed. F. L. Natter and S. A. Rollin (Tallahassee, Fla.: Office of the Governor, 1974), p. 151. 55Stricker and R0ss, ”Some Correlates of a Jungian Personality Inventory," pp. 630—631. 99 scales of self—sufficiency and are rated by faculty as more solitary 56 and less carefree than extraverts. 0n tests of aptitude, abstract reasoning, reading ability, and aesthetic values, they usually score significantly higher than average.57 Introverts as a group are at— tracted to technical—scientific professions and other vocations which allow them to work alone, such as mathematician, engineer, dentist, artist, writer, printer, farmer, or carpenter.58 S—N validity. Research indicates that a preference for sensing 59 leads to an interest in what is solid and real. Sensors tend to be attracted to practical vocations, such as business administration, sales, office management, banking, veterinary medicine, farming, and police work.60 One study revealed that sensing types were much more likely to stay employed with a utility company than were intuitive 61 types, particularly in jobs requiring mechanical ability. As a 56Stricker and Ross, "A Description and Evaluation of the MBTI," p. 120; see also Myers, MBTI Manual, pp. B5—B6; see also Ross, "College Student Characteristics Study," p. 48. 57Myers, MBTI Manual, pp. B3, B8—B9; see also Stricker and Ross, "Some Correlates of a Jungian Personality Inventory,” p. 627; see also Madison, Wilder, and Suddiford, p. 8; see also Webb, pp. 771, 775; see also Saunders, "Personality Typologies of Spranger and Jung," p. 13; see also Stricker, Schiffman, and Ross, p. 1088. 58Stricker and Ross, “Some Correlates of a Jungian Personality Inventory,” pp. 630—631. 59McCaulley and Natter, pp. 102, 154. 6OStricker and Ross, "Some Correlates of a Jungian Personality Inventory,” pp. 630—631. 61Arthur R. Laney, ”Occupational Implications of the Jungian Personality Function—Types as Identified by the Myers—Briggs Type Indicator" (unpublished master's thesis, George Washington University, 1949), pp, 24—26. 4U 100 group, sensors score extremely high on the Economic scale of the AVL.62 On other personality inventories they consistently place a high value on authority and work, and are likely to be rated by faculty as cooperative, pragmatic, and willing to take direction.63 In school, sensing types report that they like to proceed orderly toward well—defined goals and they learn especially well with audio— visual aids, such as films and television.64 These findings support the theory that sensors have a factual orientation and a strong need for order. Research indicates that intuitors like to use their minds and have considerable tolerance for complexity.65 As a group, they score high on the Aesthetic and Theoretical scales of the AVL, and excel on tests of reading ability, concept mastery, arithmetic reasoning, and general knowledge.66 Compared with sensing types, they spend more time each week on non—required reading, and they prefer teachers who give open—ended instructions and allow them to 62Myers, MBTI Manual, p. B3; see also Saunders, "Personality Typologies of Spranger and Jung," p. 13. 63Ross, "College Student Characteristics Study,” p. 48. 6liMcCaulley and Natter, p. 155. 65Myers, MBTI Manual, pp. B5-B6; see also Stricker and Ross, "A Description and Evaluation of the MBTI," p. 120; see also Madison, Wilder, and Suddiford, p. 6. 66Myers, MBTI Manual, pp. B3, B8—B9; see also Saunders, "Person— ality Typologies of Spranger and Jung,” p. 13; see also Stricker and Ross, “Some Correlates of a Jungian Personality Inventory," p. 627; see also Madison, Wilder, and Suddiford, p. 8; see also Webb, pp. 771, 775; see also Stricker, Schiffman, and Ross, p. 1088; see also McCaulley and Natter, p. 139. 101 use their own initiative.67 Intuitors generally express a strong need for autonomy and have a positive attitude toward change.68 They are more likely than sensing types to be rated by faculty as imaginative,69 which is consistent with MacKinnon's findings that over 90% of the creative writers, architects, research scientists, and mathematicians studied at the Institute of Personality Assessment and Research were intuitive types.70 Intuition appears to be stimu— lated by novelty and change, and research indicates that one conse- quence may be a high job turnover rate when intuitors are placed in mechanical and clerical jobs.71 They usually prefer professional vocations which allow them more autonomy, such as psychologist, architect, artist, musician, minister, writer, chemist, or physi- cist.72 These findings support the theory that intuitive types would rather consider the possibilities of a situation than the practical facts. T—F validity. Thinking types tend to be objective, analytical, and logical in making decisions. EPPS scores indicate that thinkers 67McCaulley and Natter, pp. 156—157. 68Myers, MBTI Manual, p. B4; see also Madison, Wilder, and Suddiford, p. 6. 69Ross, "College Student Characteristics Study," p. 48. 70Donald'W. MacKinnon, "The Nature and Nurture of Creative Talent," Personality Research, ed. D. Byrne and M. L. Hamilton (Englewood Cliffs, N.J.: Prentice—Hall, 1966), p. 160. 71Myers, MBTI Manual, p. 16. 72Stricker and Ross, "Some Correlates of a Jungian Personality Inventory,” pp. 630—631. 102 have strong needs for order, autonomy, dominance, achievement, and endurance.73 They are more likely than feeling types to have a posi- 74 tive attitude toward work and good study habits. In high school, thinking types report that they learn best from the lecture approach and do well on exams.75 As a group, they score very high on the Theoretical scale of the AVL, and excel on tests of mechanical apti— 76 tude. They are generally attracted to vocations which require logical thinking, such as scientific, technical, and business pro— 77 fessions. Feeling types appear to be extremely interested in human values and interpersonal relationships. As a group, they score high on the Social and Religious scales of the AVL, and their EPPS scores indicate strong needs for nurturance and affiliation.78 They are generally rated by faculty as ”pleasant," yet they report having 73Myers, MBTI Manual, p. B4. 711Myers, MBTI Manual, p. B4; see also Stricker and Ross, "A Description and Evaluation of the MBTI," p. 120; see also Ross, "Jungian Personality Inventory," p. 14; see also Stricker and Ross, ” Some Correlates of a Jungian Personality Inventory," p. 627; see also McCaulley and Natter, p. 159, 75McCaulley and Natter, pp. 159, 173. 76Myers, MBTI Manual, p. B3; see also Saunders, "Personality Typologies of Spranger and Jung," p. 13; see also McCaulley and Natter, p. 140. 77Stricker and Ross, ”Some Correlates of a Jungian Personality Inventory," pp. 630—631. 78Myers, MBTI Manual, pp. B3—B4; see also Saunders, ”Personality Typologies of Spranger and Jung," p. 13. 103 more free—floating anxiety than thinking types.79 In school, feeling types enjoy working on group projects and learn best in a setting that is harmonious.80 They tend to be attracted to the helping professions, such as teaching, preaching, counseling, nursing, . . 81 s001al work, and customer relations. J—P validity. Judging types are generally rated as being responsible, industrious, and steady workers.82 They tend to have much better study habits than perceptive types, perhaps because of their positive attitude toward work.83 In high school, judging types report that they usually follow a study schedule, apportion their time each day, and are seldom late in completing assignments.84 As might be expected, they generally get higher grades than perceptive 85 types, and tend to be "over—achievers." Judging types appear to 79Ross, "College Student Characteristics Study," p. 48; see also Myers, MBTI Manual, pp. B5—B6; see also Stricker and Ross, "A Descrip— tion and Evaluation of the MBTI,” p. 120. 80McCaulley and Natter, pp. 160—161. 81Stricker and Ross, "Some Correlates of a Jungian Personality Inventory,” pp. 630—631. 82Ross, ”College Student Characteristics Study," p. 48. 83Herbert G. Richek and Oliver H. Bown, ”Phenomenological Correlates of Jung's Typology," Journal of Analytical Psychology, Vol. 13 (January, 1968), p. 61; _see also Stricker and Ross, "Some Correlates of a Jungian Personality Inventory," p. 627; see also Myers, MBTI Manual, pp. B5, B8. 84McCaulley and Natter, pp. 162—163. 85 Myers, MBTI Manual, pp. 39—43; see also McCaulley and Natter, pp. 141, 162—163; see also Webb, p. 771. hi. ' 1".” .-'n-: 5.2. l _ .- --_._‘ -I'- is" "'--'5' ::..:0H‘ ‘12 E : .' .-__ ..._ - ,- .. _-' ,.- :-.... r”. '3": ‘ . . . . . . r". . - . ‘ ' ' . '. ' ;' '. "En.-Jfll|3‘-’.‘ih :- . 51".: _ . . .- too: 104 have a strong need for order, liking to have things decided and settled.86 Research indicates that they have a remarkable capa— city for endurance and would be more likely to stick with a job than change jobs.87 They are generally attracted to vocations requiring administrative skills, particularly business—oriented . 88 profes51ons. Research supports the theory that perceptive types are spon— taneous, flexible, and open—minded. They usually score high on 89 measures of impulsiveness and express a strong need for autonomy. 0n tests of abstract reasoning ability and scholastic aptitude, perceptive types generally do better than judging types, but they 90 tend to get lower grades in school. Performing below capacity may be due to their tendencies to procrastinate, to place a higher value on play than work, and to be less competitive than judging types.91 Research indicates that perceptive types enjoy change, have a high 86Ross, "Jungian Personality Inventory," p. 15; see also McCaulley and Natter, p. 106. 87Laney, pp- 32—33. 88Stricker and Ross, "Some Correlates of a Jungian Personality Inventory," pp. 630—631. 89MyerS, MBTI Manual, pp. B4—B6; see also Madison, Wilder, and Suddiford, pp. 6-7; see also Webb, pp. 771—775; see also Stricker and Ross, ”A Description and Evaluation of the MBTI," p. 120. 90Myers, MBTI Manual, pp. 39—43, B9; see also McCaulley and Natter, p. 141; see also Webb, pp. 771—775; see also Stricker, Schiffman, and Ross, p. 1088. 91McCaulley and Natter, pp. 164—165; see also Webb, pp. 771, 775; see also Ross, "College Student Characteristics Study," p. 48; see also Myers, MBTI Manual, pp. B5—B6. 105 tolerance for complexity, and tend to be attracted to such pro- fessions as writing, art, music, psychology, architecture, and 92 advertising. Validity of type combinations. Myers' research93 with dif— ferent populations has revealed that certain type combinations are attracted to certain fields. For example, ST types predominate in business and administration, SF types are attracted to sales and service professions, NF types generally outnumber other types in fields involving counseling or writing, and NT types tend to go into 94 science and research. In addition, Miller's study 95 Myers and Davis's study of law students, 96 of medical students, and Saunders' studies of theology students have revealed that certain types tend to drop out of these professions. As might be expected, the types that drop out are usually the types that are least represented in the parti- cular field. Whether or not the individual indices of the Indicator, when 92Myers, MBTI Manual, p. B4; see also Webb, pp. 771, 775; see also Madison, Wilder, and Suddiford, p. 6; see also Stricker and Ross, "Some Correlates of a Jungian Personality Inventory," pp. 630—631. 93Myers, ”Introduction to Type," p. 2. 94Paul Van Reed Miller, Jr., ”The Contribution of Non—Cognitive Variables to the Prediction of Student Performance in Law School" (unpublished doctoral dissertation, University of Pennsylvania, 1965). 95Isabel Briggs Myers and Junius A. Davis, ”Relation of Medical Students' Psychological Type to Their Specialties Twelve Years Later," Research Memorandum 64—15 (Princeton, N.J.: Educational Testing Service, 1964). 96 Saunders, "Advanced Religious Training." 106 taken in combination, tend to modify each other and produce unique effects is a question which has not been settled. Studies97 which have examined interaction effects between the scales have shown inconsistent results. The numerous studies of construct validity summarized above suggest that the individual scales of the Myers-Briggs Type Indicator measure important dimensions of personality which are quite similar to those postulated by Jung. Findings indicate that MBTI scores relate meaningfully to a large number of variables including person— ality, ability, value, aptitude and performance measures, academic choice, and behavior ratings.98 The Indicator appears to be a reasonably valid instrument, which is potentially useful for a variety of purposes. Summary A detailed description of the Myers—Briggs Type Indicator was presented in this chapter, including an analysis of scoring tech— niques and appropriate statistical procedures. The chapter also included a summary of intercorrelation studies, reliability studies, and validity studies conducted with the Indicator. 97Stricker and Ross, "An Assessment of Some Structural Prop— erties," p. 67; see also Lawrence and DeNovellis, pp. 6—7; see also Saunders, "Personality Typologies of Spranger and Jung;” see also Ross, "Jungian Personality Inventory,” p. 15; see also Stricker, Schiffman, and Ross, p. 1093. 98Gerald A. Mendelsohn, "Review of the Myers—Briggs Type Indicator," Sixth Mental Measurements Yearbook, ed. 0. K. Buros (3d ed.; Highland Park, N.J.: Gryphon Press, 1965), p. 322. '- i " Jun 01' 107 The Myers—Briggs Type Indicator is a rather unique instrument, yielding four continuous scores which may or may not be regarded as four dichotomous type—category scores. Frequency distributions of continuous scores are generally not normal distributions, which presents a problem for researchers regarding appropriate statistical procedures. It was suggested in the present chapter that Pearson product—moment correlation coefficients be used with continuous scores since the Pearson coefficient is fairly robust with respect to de— partures from normality. Correlations involving type categories cannot be directly estimated with existing statistical techniques, but can be expected to fall somewhere between estimates derived with phi coefficients and estimates derived with tetrachoric coefficients. It was therefore suggested that both phi and tetrachoric coefficients be used whenever type categories are compared with dichotomous data, and that point—biserial and biserial coefficients (variations of phi and tetrachoric) be used whenever type categories are compared with continuous data. Numerous researchers have examined the relative independence of the four MBTI indices, and these intercorrelation studies have produced rather consistent results. The E—I, S—N, and T—F indices appear to be relatively independent of each other, and the J—P index appears to correlate consistently with the S—N index and less con— sistently with the T—F index. For the present sample, however, sig— nificant correlations were found between other indices as well. 108 Further research is needed to investigate the intercorrelations of MBTI scores with larger samples of teachers. Considerable evidence was presented regarding the reliability of the Myers—Briggs Type Indicator. Estimates of the internal consistency of continuous MBTI scores have been similar to relia~ bilities generally obtained from self—report instruments, ranging with few exceptions between .70 and .90. Estimating the internal consistency of the type categories has been more difficult because existing statistical procedures can provide only low and high estimates. Nevertheless, the type categories appear to be adequately reliable, particularly for the present sample. In all of the reported test—retest experiments, the proportion of agreement between the original and retest type classifications has been significantly higher than would be expected by chance. One study reported extremely stable scores for a sample of elementary teachers, lending substantial quantitative support to Jung's theory that psychological type is relatively stable over a period of time. Test-retest data is scanty, however, and there is an urgent need for additional long-range studies with the Indicator. Researchers have investigated the content validity, predictive validity, and construct validity of the Myers—Briggs Type Indicator with a variety of populations. One obvious advantage of the Indicator is that it was derived from a well—known conceptual formulation (Jung's typological theory) to which empirical findings may be related. 109 The findings have been quite consistent with Jungian constructs, and taken as a whole, they give strong support to the validity of the Indicator. There is a need, however, for additional experi— ments designed to test more rigorously the instrument's ability to predict specific behavior. The present study is one such attempt to further examine the predictive validity and the construct validity of the Myers—Briggs Type Indicator. CHAPTER VI FINDINGS The present study was designed to examine the relationship be- tween the personality characteristics and teaching preferences of prospective teachers. The sample was composed of 200 pre—service teachers, all seniors majoring in elementary or secondary education at Michigan State University. Personality preferences were measured by the Myers—Briggs Type Indicator (MBTI), a self—report instrument which yields four continuous scores and which also classifies respondents on four dichotomous type categories. The following personality preferences were measured: (1) Extraversion or Introversion, (2) Sensation or Intuition, (3) Thinking or Feeling, and (4) Judgment or Perception. Teaching preferences were measured by the Teacher Preference Questionnaire (TPQ), a self—report instrument which was designed especially for the study. The questionnaire yields six composite cluster scores, each cluster representing a different teaching preference. The following teaching preferences were measured: (1) Preference for teaching lower grades, (2) Interest in admin— istrative functions, (3) Need for independence and creativity, (4) Interest in planning school projects, (5) Interest in working with small groups of students, and (6) Commitment to classroom teaching. 110 Six hypotheses were formulated in order to test whether signi— ficant relationships exist between the personalities of prospective teachers and their preferences toward certain teaching situations. The hypotheses were tested in two ways, first by treating MBTI scores as continuous data and then by treating MBTI scores as dichotomous type categories. In both cases, scores on each of the four MBTI indices were compared with scores on a particular TPQ cluster. When MBTI scores were treated as continuous data, Pearson product—moment correlation coefficients were computed in order to estimate the relationship between personality and teaching pref— erence. When MBTI scores were treated as dichotomous type cate— gories, three statistical procedures were used: (1) Point—biserial correlation coefficients were computed to produce minimum estimates of relationship between personality type and teaching preference; (2) Biserial correlation coefficients were computed to produce maximum estimates of relationship between personality type and teach- ing preference; and (3) One—way analysis of variance was used to approximate the significance level of the F statistic obtained when MBTI type category was regarded as the independent variable and TPQ cluster score was regarded as the dependent variable. An alpha level of .05 was set for rejecting the null hypotheses. Alternate hypotheses were accepted when the statistical measures produced significant results in the predicted direction. 112 Presentation of the Data The study produced a number of significant findings. The null hypotheses, the directional alternate hypotheses, and the results of the hypothesis tests are presented below. Hypothesis 1 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and the grade level they most prefer to teach. Alternate hypothesis: Sensors, feelers, and judging types show stronger preferences for teaching lower grades than intuitors, thinkers, and perceptive types. The alternate hypothesis was based on findings presented in Chapter III. Most researchers of teachers in the field have found a significantly higher proportion of intuitors, thinkers, and per- ceptive types teaching at upper levels than at lower levels. Studies of pre—service teachers have been less consistent, but researchers have generally found a higher proportion of sensors, feelers, and judging types preparing to teach lower levels than upper levels. It was therefore predicted that sensors, feelers, and judging types would show the strongest preferences for teaching lower grades. Hypothesis 1 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 1. In order to establish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and a preference for teaching lower grades was regarded as the negative pole of Cluster 1. A low cluster score indicated a strong preference for teaching lower grades. -- oath;- -.~.;r nae 113 Statistical results are shown in Tables 6.1 and 6.2, and indi— cate that the T—F dimension was the only index which correlated significantly with scores on Cluster 1. In the sample of 200 prospective teachers, feeling types showed more interest in teaching at lower levels than thinking types. The null hypothesis was there— fore rejected. Because there were no significant correlations be— tween scores on Cluster 1 and scores on the S—N or J—P scales, the alternate hypothesis was supported for only one index: the E—I index. Hypothesis 2 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and their interest in administrative functions. Alternate hypothesis: Extraverts, thinkers, and judging types show more interest in administrative functions than introverts, feelers, and perceptive types. The alternate hypothesis was based primarily on findings pre- sented in Chapter IIIo In each of two reported studies1 of princi— pals and superintendents, extraverts outnumbered introverts, thinkers outnumbered feelers, and judging types outnumbered perceptive types. These findings are consistent with results from a vocational interest study2 summarized in Chapter V, which found that extraverts, sensors, thinkers, and judging types were attracted to administrative work. From the vocational study, one would expect school administrators to be sensors as well as extraverts, thinkers, and judging types, but 1Von Fange, pp. 110—119; see also Wright, pp. 133—135. 2Stricker and Ross, ”Some Correlates of a Jungian Personality Inventory," pp. 630-631. 114 Table 6.1 Correlations Between Teacher Personality and Preference for Teaching Lower Grades N = 200 Statistical Procedure MBTI Scores E—I S—N T—F J—P Continuous MBTI Scores Correlated with Scores on TPQ Cluster 1 —.03 —.01 —.30** -.08 (Pearson Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 1 —.02 .02 —.26** —.06 (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 1 —.03 .03 —.34** —.08 (Tetrachoric Coefficients) *Significant at the .05 level. **Significant at the .01 level. Table 6.2 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 1 MBTI Type Category N Mean F Probability Extraversion 112 14.455 .084 .773 Introversion 88 14.239 Sensation 97 14.247 .086 .769 Intuition 103 14.466 Thinking 68 16.250 14.238 <.0005** Feeling 132 13.386 Judgment 96 14.688 .717 .398 Perception 104 14.058 *Significant at the .05 level. **Significant at the .01 level. 115 researchers have found a higher proportion of intuitors among administrators than among teachers in the field. Both sensors and intuitors appear to be attracted to educational administration. It was therefore predicted that extraverts, thinkers, and judging types would show the most interest in administrative functions. Hypothesis 2 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 2. In order to establish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and an interest in administrative functions was regarded as the negative pole of Cluster 2. A low cluster score indicated a strong interest in administrative functions. Statistical results are shown in Tables 6.3 and 6.4, and indi— cate that the E—I, T—F, and J—P indices all correlated significantly with scores on Cluster 2. Extraverts, thinkers, and judging types showed considerably more interest in administrative functions than introverts, feelers, and perceptive types. The null hypothesis was therefore rejected and the alternate hypothesis was fully supported. Hypothesis 3 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and their need for independence and creativity. Alternate hypothesis: Intuitors and perceptive types express a stronger need for independence and creativity than sensors and judging types. No previous researchers have specifically examined the relation— ship between the personality preferences of teachers and their need for independence and/or creativity. The alternate hypothesis was therefore based on MBTI studies involving non—teacher populations, 116 Table 6.3 Correlations Between Teacher Personality and Interest in Administrative Functions N = 200 Statistical Procedure MBTI Scores E—I S—N' T—F J—P Continuous MBTI Scores Correlated with Scores on TPQ Cluster 2 .42** —.03 .23** .16* (Pearson Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 2 .36** —.04 .27** .15* (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 2 .45** —.05 .35** .18* (Tetrachoric Coefficients) *Significant at the .05 level. **Significant at the .01 level. Table 6.4 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 2 MBTI Type Category N Mean F Probability Extraversion 112 13.411 29.122 <.0005** Introversion 88 16.818 Sensation 97 15.103 .312 .577 Intuition 103 14.728 Thinking 68 13.132 15.583 <.OOO5** Feeling 132 15.826 Judgment 96 14.188 4.370 .038* Perception 104 15.577 *Significant at the .05 level. **Significant at the .01 level. 117 which have been summarized in Chapter V. Intuitors and perceptive types were found to be the most creative and the most independent types. Introverts and thinkers were found to be more independent than extraverts and feelers, but not necessarily more creative. It was therefore predicted that intuitors and perceptive types would ex- press the strongest need for independence and creativity in teaching. Hypothesis 3 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 3. In order to establish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and a need for independence and creativity was regarded as the negative pole of Cluster 3. A low cluster score indicated a strong need for independence and creativity. Statistical results are shown in Tables 6.5 and 6.6, and indi— cate that the S—N and J—P indices both correlated significantly with scores on Cluster 3. Intuitors and perceptive types expressed a con— serably stronger need for independence and creativity than sensors and judging types. The null hypothesis was therefore rejected and the alternate hypothesis was fully supported. Hypothesis 4 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and their interest in planning school projects. Alternate hypothesis: Extraverts, intuitors, feelers, and judging types show more interest in planning school projects than introverts, sensors, thinkers, and perceptive types. No previous researchers have directly examined the relationship between teacher personality and interest in planning school projects. 118 Table 6.5 Correlations Between Teacher Personality and Need for Independence and Creativity N = 200 Statistical Procedure ' MBTI Scores E-I S—N T-F J—P Continuous MBTI Scores Correlated with Scores on TPQ Cluster 3 .00 —.32** .03 —.25** (Pearson Coefficients) ' ' MBTI Type Categories Correlated with Scores on TPQ Cluster 3 —.02 —.24** .02 —.17* (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 3 —.03 —.30** .03 —.22** (Tetrachoric Coefficients *Significant at the .05 level. **Significant at the .01 level. Table 6.6 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 3 MBTI Type Category N Mean F Probability Extraversion 112 15.214 .125 .724 Introversion 88 14.955 Sensation 97 16.381 12.320 .001** Intuition 103 13.893 Thinking 68 14.926 .1716 .733 Peeling 132 15.189 Judgment 96 16.031 6.192 .014* Perception 104 14.240 *Significant at the .05 level. **Significant at the .01 level. 119 The alternate hypothesis was therefore based on MBTI studies in- volving non—teacher populations, which have been summarized in Chapter V. These studies indicate that extraverts and feeling types enjoy working on group projects, intuitors like to solve new prob- lems, and judging types enjoy organizing activities and carrying out plans. It was therefore predicted that extraverts, intuitors, feelers, and judging types would show the strongest interest in planning school projects. Hypothesis 4 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 4. In order to establish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and an interest in planning school projects was regarded as the negative pole of Cluster 4. A low cluster score indicated a strong interest in planning school projects. Statistical results are shown in Tables 6.7 and 6.8, and indi— cate that the E—I and S-N indices correlated significantly with scores on Cluster 4. Extraverts and intuitive types showed con— siderably more interest in planning school projects than introverts and sensing types. The null hypothesis was therefore rejected. Because there were no significant correlations between scores on Cluster 4 and scores on the T-F and J—P scales, the alternate hypothesis was supported for only two indices: the E—I and S—N indices. Table 6.7 120 Correlations Between Teacher Personality and Interest in Planning School Projects N = 200 Statistical Procedure MBTI Scores E—I S—N T—F J—P Continuous MBTI Scores Correlated with Scores on TPQ Cluster 4 .39** -.34** —.09 .06 (Pearson Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 4 .31** —.29** .01 .07 (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 4 .39** -.36** .01 .09 (Tetrachoric Coefficients) *Significant at the .05 level. Table 6.8 **Significant at the .01 level. One—Way Analysis of Variance for Mean Differences on TPQ Cluster 4 MBTI Type Category N Mean F Probability Extraversion 112 11.116 21.356 (.0005** Introversion 88 14.545 Sensation 97 14.237 17.740 <.0005** Intuition 103 11.107 Thinking 68 12.559 .015 .903 Feeling 132 12.659 Judgment 96 12.229 .967 .327 Perception 104 12.990 *Significant at the .05 level. **Significant at the .01 level. Hypothesis 5 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and their interest in working with small groups of students. Alternate hypothesis: Extraverts, intuitors, feelers, and perceptive types show more interest in working with small groups of students than introverts, sensors, thinkers, and judging types. The alternate hypothesis was based on findings presented in Chapter III regarding teacher personality and teaching style. Teachers who were extraverts, intuitors, feelers, and/or perceptive types were found to use informal individualized instruction more than introverts, sensors, thinkers, and judging types. It was therefore predicted that extraverts, intuitors, feelers, and per— ceptive types would show the strongest interest in working with small groups of students on a regular basis. Hypothesis 5 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 5. In order to establish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and an interest in working with small groups of students was regarded as the negative pole of Cluster 5. A low cluster score indicated a strong interest in working with small groups. Statistical results are shown in Tables 6.9 and 6.10, and indicate that the S—N index correlated significantly with scores on Cluster 5. Intuitive types showed more interest in working with small groups of students than sensing types. The null hypothesis 122 Table 6.9 Correlations Between Teacher Personality and Interest in Working with Small Groups N = 200 Statistical Procedure MBTI Scores E—I S-N T—F J—P Continuous MBTI Scores Correlated with Scores on TPQ Cluster 5 .03 —.26** -.10 —.03 (Pearson Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 5 .07 —.21** —.11 —.01 (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 5 .08 —.27** —.15* -.01 (Tetrachoric Coefficients) *Significant at the .05 level. **Significant at the .01 level. Table 6.10 0ne~Way Analysis of Variance for Mean Differences on TPQ Cluster 5 MBTI Type Category N Mean F Probability Extraversion 112 2.429 .874 .351 Introversion 88 2.818 Sensation 97 3.237 9-312 .003** Intuition 103 2.000 Thinking 68 3.059 2.556 .111 Feeling 132 2.364 Judgment 96 2.635 .027 .870 Perception 104 2.567 *Significant at the .05 level. **Significant at the .01 level. 123 was therefore rejected. Because there were no significant corre— lations between scores on Cluster 5 and scores on the E—I, T—F, and J—P scales, the alternative hypothesis was supported for only one index: the S—N index. Hypothesis 6 Null hypothesis: There is no relationship between the personality preferences of prospective teachers and their commitment to classroom teaching. Alternate hypothesis: Sensors and judging types express more commitment to classroom teaching than intuitors and perceptive types. Studies comparing pre—service and in—service teachers have indicated that sensors and judging types are more likely to remain in teaching than intuitors and perceptive types. It was therefore predicted that sensors and judging types would express the most commitment to classroom teaching. Hypothesis 6 was tested by comparing scores on each of the four MBTI indices with scores on TPQ Cluster 6. In order to estab— lish directionality, the E, S, T, and J preferences were regarded as the negative poles of the MBTI indices, and a commitment to class— room teaching was regarded as the negative pole of Cluster 6. A low cluster score indicated a strong commitment to classroom teaching. Statistical results are shown in Tables 6.11 and 6.12, and indicate that the E—I index correlated significantly with scores on Cluster 6. Extraverts expressed considerably more commitment to classroom teaching than introverts. The null hypothesis was there— fore rejected. Because there were no consistent significant 124 Table 6.11 Correlations Between Teacher Personality and Commitment to Classroom Teaching N z 200 Statistical Procedure MBTI Scores Continuous MBTI Scores Correlated with Scores on TPQ Cluster 6 .25** .02 —.21** .13* (Pearson Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 6 .23** .00 -.12 .09 (Phi Coefficients) MBTI Type Categories Correlated with Scores on TPQ Cluster 6 .29** —.01 —.15* .11 (Tetrachoric Coefficients) *Significant at the .05 level. **Significant at the .01 level. Table 6.12 One—Way Analysis of Variance for Mean Differences on TPQ Cluster 6 MBTI Type Category N Mean F Probability Extraversion 112 9.018 10.746 .001** Introversion 88 11.420 Sensation 97 10.103 .005 .942 Intuition 103 10.049 Thinking 68 10.926 2.713 .101 Feeling 132 9.636 Judgment 96 9.594 1.544 .215 Perception 104 10.519 *Significant at the .05 level. **Significant at the .01 level. 125 correlations between scores on Cluster 6 and scores on the S—N and J—P scales, the alternate hypothesis was not supported. Additional Findings In addition to testing the six hypotheses, exploratory analyses were conducted with the data to examine whether any of the findings were sex—related. The teaching preferences of the 136 females in the sample were compared with the teaching preferences of the 64 males in the sample, and no predictions were made regarding the out— come of these exploratory investigations. First, the mean scores on each TPQ cluster were computed for the group of females and for the group of males, and a two—tailed Z—test was used to determine if there were significant differences between the two groups. Results are shown in Table 6.13, and indicate that females showed significantly more interest in teaching at lower levels and in planning school projects than males. Males showed signifi— cantly more interest in administrative functions than females. To further investigate the relationship between personality and teaching preferences, the hypotheses were re—examined after grouping the pre—service teachers by sex. Pearson correlations were used to compare continuous MBTI scores with TPQ cluster scores. The results are shown in Table 6.14 and are summarized in the following para— graphs. Regarding Cluster 1, feeling types of both sexes showed stronger preferences for teaching lower grades than did thinking types. In fact, this correlation was stronger among the males than among the 126 Table 6.13 Comparison of TPQ Cluster Scores of Males_and Females TPQ Cluster N Mean Z Probability Cluster : Preference for Teaching Lower Grades Females 136 13.140 5.673 <.001** Males 64 16.953 Cluster : Interest in Administrative Functions Females 136 15.441 2.087 .045* Males 64 13.781 Cluster : Need for Independence and Creativity Females 136 15.390 1.217 .190 Males 64 14.484 Cluster : Interest in Planning School Projects Females 136 11.809 3.115 .003** Males 64 14.359 Cluster Interest in Working with Small Groups of Students Females 136 2.596 .032 .399 Males 64 2.609 Cluster Commitment to Classroom Teaching Females 136 9.757 1.060 .228 Males 64 10.750 *Significant at the .05 level. **Significant at the .01 level. 127 Table 6.14 Correlations Between Continuous MBTI Scores and TPQ Cluster Scores of Males and Females TPQ Cluster MBTI Scales E—I S-N T—F J—P Cluster 1: Preference for Teaching Lower Grades Females .03 —.02 —.21** —.10 Males —.33** .13 —.35** —.10 Cluster 2: Interest in Administrative thctions Females .47** —.07 .15* .10 Males .38** .01 .33** .33** Cluster 3: Need for Independence and Creativity Females .06 -.39** .03 —.30** Males -.10 —.21* —.04 —.14 Cluster 4: Interest in Planning School Projects Females .41** —.39** -.13 —.02 Males .34** —.24* .12 .21* Cluster 5: Interest in Working with Small Groups of Students Females ~.O6 —.27** —.08 .05 Males .22* -.22* —.15 —.21* Cluster 6: Commitment to Classroom Teaching Females .26** .03 —.23** .12 Males .21* .02 -.11 .16 Note. The sample consisted of 136 females and 64 males. *Significant at the .05 level. **Significant at the .01 level. 128 females. An additional finding was that male introverts tended to prefer teaching at lower levels, and male extraverts tended to prefer teaching at upper levels. No similar E—I correlation was found among the females. Regarding Cluster 2, extraverts and thinkers of both sexes showed more interest in administrative functions than introverts and feelers. Also, males who were judging types showed much more interest in administrative functions than males who were percep— tive types, but no similar J—P correlation was found among the females. Regarding Cluster 3, intuitors of both sexes expressed a stronger need for independence and creativity than sensors. Also, females who were perceptive types expressed a much stronger need for independence and creativity than females who were judging types, but no similar J—P correlation was found among the males. Regarding Cluster 4, extraverts and intuitors of both sexes showed more interest in planning school projects than introverts and sensors. Also, males who were judging types showed more inter— est in planning school projects than males who were sensing types. No similar J—P correlation was found among the females. Regarding Cluster 5, intuitors of both sexes showed more interest in working with small groups of students than sensors. Also, males who were extraverts and perceptive types showed more interest in working with small groups of students than males who were introverts and judging types. No similar E—I or J—P correla— tion was found among the females. i 129 Regarding Cluster 6, extraverts of both sexes expressed more commitment to classroom teaching than introverts. Also, females who were feeling types expressed more commitment to classroom teaching than females who were thinking types. No similar T—F correlation was found among the males. The main purpose for conducting exploratory investigations with the data was to examine whether any of the findings were sex- related. Results indicated that most of the original findings were valid for both males and females in the sample, although a number of other relationships between personality and teaching preferences appeared to be sex—related. Results of these exploratory investi— gations are promising, but they should not be generalized to other populations until cross—validation studies produce compatible results. Summary In order to examine the relationship between personality char— acteristiCS and teaching preferences of prospective teachers, the MBTI scores of 200 pre—service teachers were compared with six cluster scores derived from responses to the Teacher Preference Questionnaire. Each TPQ cluster score represented a different teaching preference and was used to test a different hypothesis. Six hypotheses were tested and the results are summarized in Table 6.15. Each of the null hypotheses was rejected because signi- ficant correlations were found between MBTI scores and TPQ cluster Table 6.15 Summary of the Findings Hypothesis Results ° S, F, and J types show stronger preferences for teaching lower grades than N, T, and P types. , : E, T, and J types show more interest in administrative functions than I, F, and P types. : N and P types express a stronger need for indepen— dence and creativity than S and J types. : E, N, F, and J types show more interest in planning school projects than I, S, T, and P types. : E, N, F, and P types show more interest in working with small groups of students than I, S, T, and J types. : S and J types express more commitment to Classroom teaching than N and P types. H was partly supported: F types showed stronger preferences for teaching lower grades than T types. H2 was fully supported for the sample as a whole, but further analyses revealed that the J—P correlation was not significant for the females. H was fully supported for the sémple as a whole, but further analyses revealed that the J—P correlation was not significant for the males. H4 was partly supported: E and N types showed more interest in planning school projects than I and S types. H was partly supported: N types sgowed more interest in working with small groups of students than S types. H was not supported. It was found that E types expressed more commitment to classroom teaching than I types. ) 131 scores in every case. Half of the predictions stated in the alter— nate hypotheses were substantiated, as indicated in Table 6.15. None of the results appear to contradict previous findings with the Indicator, lending support to the argument that teachers with different personality types prefer different teaching situations. A discussion of the findings is included in the following chapter. CHAPTER VII SUMMARY AND CONCLUSIONS The major problem of the present study was to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. Summary of the Study Personality type was measured by the Myers-Briggs Type Indicator and teaching preferences were measured by the Teacher Preference Questionnaire, an instrument constructed for the study. The sample consisted of 200 pre—service teachers, all seniors majoring in elementary or secondary education at Michigan State University. The Myers—Briggs Type Indicator (MBTI) is a self-report in- strument based on Carl Jung’s theory of personality. The Indicator consists of 166 items designed to measure the following personality preferences: (1) Extraversion or Introversion, (2) Sensation or Intuition, (3) Thinking or Feeling, and (4) Judgment or Perception. The instrument yields four continuous scores and four dichotomous scores (type categories) for each person. MBTI scores for the pres— ent sample appeared to be satisfactorily reliable, with internal— consistency reliability coefficients ranging from .77 to .90 for 132 133 continuous scores and from .67 to .92 for type categories. The Teacher Preference Questionnaire (TPQ) is a self—report instrument which was constructed especially for the study. The questionnaire consists of 103 items designed to measure a teacher's attitude toward a variety of teaching situations. Cluster analysis was performed on the Likert—scale TPQ items and resulted in the following six clusters: (1) Preference for teaching lower grades, (2) Interest in administrative functions, (3) Need for independence and creativity, (4) Interest in planning school projects, (5) Inter- est in working with small groups of students, and (6) Commitment to classroom teaching. Each person's questionnaire responses were transformed into six composite cluster scores. The TPQ clusters appeared to be satisfactorily reliable, with Coefficient Alphas ranging from .70 to .81. Six hypotheses were formulated in order to test whether signi— ficant relationships exist between the personalities of prospective teachers and their preferences toward certain teaching situations. The hypotheses were tested in two ways, first by treating MBTI scores as continuous data and then by treating MBTI scores as dichotomous type categories. In both cases, scores on each of the four MBTI indices were compared with scores on a particular TPQ cluster. A variety of statistical techniques were employed, and an alpha level of .05 was set for rejecting the null hypotheses. Alternate hypotheses were accepted when the statistical measures produced -4n1- 134 significant results in the predicted direction. The following results were obtained: 1. Feeling types show more interest in teaching at lower levels than thinking types. 2. Extraverts, thinkers, and judging types show more interest in administrative functions than intro— verts, feelers, and perceptive types. 3. Intuitors and perceptive types express a stronger need for independence and creativity than sensors and judging types. 4. Extraverts and intuitive types show more interest in planning school projects than introverts and sensing types. 5. Intuitive types show more interest in working with small groups of students than sensing types. 6. Extraverted types express more commitment to class- room teaching than introverted types. In addition to testing the six hypotheses, exploratory analyses were conducted to examine whether any of the findings were sex—related. Results indicated that most of the original findings were valid for both males and females in the sample. Discussion of the Findings and Recommendations for Future Research The study produced a number of significant results. A dis— cussion of the findings and recommendations for future research are presented below. Hypothesis 1 The first hypothesis postulated that sensors, feelers, and judging types show stronger preferences for teaching lower grades than intuitors, thinkers, and perceptive types. The statistical 135 tests of this study revealed that feeling types showed more interest in teaching at lower levels than thinking types, lending partial support to the original hypothesis. When separate analyses were conducted with the females and the males, it was found that females who were feeling types and males who were introverts and feeling types showed the strongest preferences for teaching lower grades. The correlation between the T—F dimension and preferred grade level is consistent with Jungian theory. According to Jung, feeling encompasses the affective realm of human experience, and feeling types are skilled at maintaining contact with others and under— standing their feelings. Thinking, on the other hand, encompasses the cognitive realm of human experience, and thinking types are skilled at using logic to deal with the world. It is not surprising that feeling types are attracted to lower grade levels where an affective approach to teaching has traditionally been emphasized, and thinking types are attracted to upper grade levels where a cog— nitive approach has traditionally been emphasized. Perhaps one reason that there are relatively few male teachers at lower levels is that the majority of males in the general popu— lation are thinking types. The males most interested in teaching lower levels were found to be introverted and feeling types, with personality preferences not commonly found among males in our society. Studies comparing male teachers who are introverted feeling types and male teachers who are extraverted thinking types are needed in order to clarify the issues with regard to this apparent difference in interest. 136 The failure to find significant correlations between the S—N and J—P dimensions and preference for teaching lower grades is some— what surprising because field studies have indicated that a larger proportion of sensors and judging types teach at lower levels than at upper levels. Perhaps many intuitive and perceptive—type pre— service teachers who plan to teach lower levels later transfer to upper levels or drop out of teaching altogether. Longitudinal studies are needed to examine the relationship between teacher per— sonality and career patterns. Hypothesis 2 The second hypothesis postulated that extraverts, thinkers, and judging types show more interest in administrative functions than introverts, feelers, and perceptive types. Hypothesis 2 was fully supported for the sample as a whole, but further analyses revealed that the J—P correlation was not significant for the females. Re— sults indicated that females who were extraverts and thinking types and males who were extraverts, thinkers, and judging types showed the most interest in administrative functions. The correlation between the E—I dimension and interest in admin— istrative functions is consistent with Jungian theory. Extraverts, according to Jung, are oriented primarily to the outer world of ob- jects, people, and action. It is not surprising that extraverts are attracted to school administration, which involves many extra— verted activities such as working with faculty committees, speaking to community groups, and supervising teachers. Future comparisons 137 between introverted and extraverted administrators may reveal that the introverts need more uninterrupted time each day for reflection and concentration. The correlation between the T-F dimension and interest in ad— ministrative functions is also consistent with Jungian theory. Jung believed that mature thinking types are skilled at objectively organ— izing material, weighing the facts, predicting consequences, and making plans. Administration generally requires these skills and would therefore attract thinking types. Thinking-type administrators may have difficulty communicating with teachers, however, since the majority of teachers are feeling types, particularly at lower levels. The administrators may think that many of the teachers are emotional and unreasonable, and the teachers may feel that the administrators are too businesslike and impersonal. Recognition of basic differences may help to improve communication. The correlation between the J—P dimension and interest in admin— istrative functions was found to be significant for males but not for females. This finding is not surprising, however, because the research hypothesis was based on two studies1 of administrators, most of whom were male, and on a vocational interest study2 conducted with a sample of male freshmen. 1Von Fange, pp. 110—119; see also Wright, pp. 133—135. 2Stricker and Ross, "Some Correlates of a Jungian Personality Inventory,” pp. 630-631. 138 Hypothesis 3 The third hypothesis postulated that prospective teachers who are intuitors and perceptive types express a stronger need for inde— pendence and creativity than sensors and judging types. Hypothesis 3 was fully supported for the sample as a whole, but further analyses revealed that the J—P correlation was not significant for the males. Results indicated that females who were intuitive and perceptive types and males who were intuitive types expressed the strongest need for independence and creativity. The correlation between the S—N dimension and need for indepen— dence and creativity is consistent with Jungian theory. According to Jung, intuitive types are imaginative and spontaneous, guided by inspiration. They like to see every new situation as a fresh prob— lem, and they have an intensely personal way of attacking problems, relying on a flash of insight. It is not surprising that intuitive types express strong needs for independence and creativity as teachers. This finding is further supported by classroom observations of in— service teachers. Results show that intuitive teachers tend to be pupil—centered and move freely around the room, encouraging pupil choice and self—expression. They also tend to be more experimental in their approaches than sensing teachers, preferring schools with "open" climates. From an inspection of the items comprising TPQ Cluster 3, it would appear that external restrictions are particularly displeasing to intuitive teachers, and a number of conditions might force them to 139 leave teaching: few opportunities to be creative, lack of indepen— dence and freedom, lack of intellectual stimulation, having to work with adults who do not share their values, too much paperwork and clerical duties, having to spend all of their class time on subject matter. There is in fact substantial evidence that a high proportion of intuitive types are dropping out of classroom teaching, perhaps because their needs for independence and creativity are not being met. Longitudinal studies should be conducted to see if and why in— tuitive teachers are leaving the classroom. The correlation between the J—P dimension and need for indepen— dence and creativity was found to be significant for females but not for males. This finding is somewhat surprising because MBTI validity studies indicate that perceptive types of both sexes tend to be more independent and creative than judging types. Future research should examine the possibility that there are significant differences between male and female perceptive types. Hypothesis 4 The fourth hypothesis postulated that extraverts, intuitors, feelers, and judging types show more interest in planning school pro— jects than introverts, sensors, thinkers, and perceptive types. The statistical tests of this study revealed that extraverts and intuitors showed more interest in planning projects than introverts and sensors, lending partial support to the original hypothesis. When separate analyses were conducted with the females and the males, it was found that females who were extraverts and intuitive types and males who 140 were extraverts, intuitors, and judging types showed the most in— terest in planning school projects. The correlation between the E—I dimension and interest in plan— ning school projects is consistent with Jungian theory. Extraverts, according to Jung, are oriented primarily to the outer world of ob- jects, people, and action. It is therefore not surprising that extraverts expressed an interest in such activities as working with students on community action projects, organizing class plays, organ- izing a multi—media center, and planning teacher workshops. The correlation between the S—N dimension and interest in plan- ning school projects is also consistent with Jungian theory. Jung believed that intuition gives a capacity for suddenly seeing the connections and hidden possibilities in a situation. It is therefore not surprising that intuitors expressed an interest in innovative programs. If it is true, as research suggests, that intuitive types are dropping out of classroom teaching, one reason may be that they are not given opportunities to plan new projects, particularly if they are young teachers. The correlation between the J—P dimension and interest in plan— ning school projects was found to be significant for males but not for females. This finding is surprising because MBTI validity studies indicate that judging types of both sexes tend to enjoy organizing activities and carrying out plans. As was previously sug— gested, future research should examine the possibility that there are significant differences between male and female judging types. 141 Hypothesis 5 The fifth hypothesis postulated that extraverts, intuitors, feelers, and perceptive types show more interest in working with small groups of students than introverts, sensors, thinkers, and judging types. The statistical tests of this study revealed that intuitive types showed more interest in working with small groups of students than sensing types, lending partial support to the original hypothesis. When separate analyses were conducted with the females and the males, it was found that females who were intuitive types and males who were extraverts, intuitors, and perceptive types showed the most interest in working with small groups of students on a regular basis. The correlation between the S-N dimension and interest in work— ing with small groups is consistent with Jungian theory. Intuitors, according to Jung, are imaginative and spontaneous, preferring to make as few preparations as possible. Sensors, on the other hand, enjoy tradition and routine, and they tend to be more observant and practical than intuitive types. It is not surprising that intuitors expressed an interest in working with 2 to 5 students on a regular basis, which would allow them more freedom. And it is not surprising that sensors preferred working with larger groups, which is a more traditional approach to teaching. Correlations between the E—I and J—P dimensions and interest in teaching small groups of students were found to be significant for males but not for females. This finding is surprising because 142 classroom observations of teachers indicate that extraverts and perceptive types of both sexes tend to use informal individualized instruction more than introverts and judging types. There is a need for additional studies involving classroom observations of teachers in order to verify the existing evidence that certain personality types tend to prefer certain teaching styles. Possible interactions between the personality and the sex of the teacher should also be examined. Hypothesis 6 The sixth hypothesis postulated that sensors and judging types express more commitment to classroom teaching than intuitors and perceptive types. Hypothesis 6 was not supported. When separate analyses were conducted with the females and the males, however, it was found that females who were extraverts and feeling types and males who were extraverts expressed the strongest commitment. The failure to find significant correlations between the S—N and J—P dimensions and commitment to teaching is surprising, for field studies indicate that sensors and judging types remain in teaching longer than intuitors and perceptive types. Perhaps the items in TPQ Cluster 6 measure something other than commitment to teaching as a lifetime career; perhaps they measure current satis— faction and enthusiasm toward teaching. If so, it is not surprising that extraverts score high, because researchers have found that extraverted teachers usually express more satisfaction with their current teaching situations than introverts. 143 In any case, the results indicate that intuitive and perceptive pre—service teachers express as much commitment, satisfaction, and/or enthusiasm toward teaching as sensors and judging types. Therefore, if it is true that intuitors and perceptive types later leave class— room teaching, it may be because they are frequently placed in teaching situations which are incompatible with their personalities and not because they were originally uncommitted to teaching as a career. As was previously suggested, there is a need for longitu— dinal studies designed to examine the relationship between teacher personality and career patterns. Additional Recommendations In addition to the recommendations mentioned above, a number of suggestions are offered to future researchers using the Myers—Briggs Type Indicator. Continuous scores. If continuous MBTI scores are used, the following suggestions are offered: 1. Frequency distributions of continuous scores should be illustrated. The method which was used in this study for graphing frequency distributions is described in Appendix C. 2. Intercorrelations of the four continuous indices should be calculated using Pearson product—moment correlations. 3. Internal—consistency reliabilities of MBTI scores should be estimated for the sample being investigated. The split—half procedure which was used in this study for calculating reliability estimates of continuous MBTI scores is included in Appendix Do 144 4. Regarding statistical procedures, the Pearson correlation coefficient appears to be the most appropriate statistic for comparing continuous MBTI scores with other continuous data. Type categories. If dichotomous type categories are used, the following suggestions are offered: 1. Type distributions should be illustrated in a standardized type table, such as Table 5.1 (page 64). 2. Intercorrelations among the type categories should be calcu— lated. Minimum intercorrelation estimates can be calculated using phi coefficients and maximum estimates can be calculated using tetrachoric coefficients, as described in Chapter V. 3. Internal—consistency reliabilities of MBTI scores should be estimated for the sample being investigated. The split—half procedure which was used in this study for calculating reliability estimates of MBTI type categories is included in Appendix D. 4. Regarding statistical procedures, phi and tetrachoric corre- lation coefficients appear to be the most appropriate statistics for comparing dichotomous type categories with other dichotomous data. Point—biserial and biserial correlation coefficients appear to be the most appropriate statistics for comparing dichotomous type categories with continuous data. The rationale for choosing these specific statistical measures for the present study may be found in Chapter V. The majority of studies involving the Myers—Briggs Type Indicator and teacher pOpulations have been descriptive accounts of teacher per— sonality. Considerable data have been accumulated, but there is a need for experiments designed to test specific predictions of teacher behav— ior. 145 Conclusions The present study was designed to examine the relationship between the personality characteristics and teaching preferences of prospective teachers. Six research hypotheses were formulated in order to test specific predictions of teacher behavior. Results of the statistical tests support the following general conclusions: 1. 2. 5. Feeling types are more interested in teaching at lower levels than thinking types. Extraverted and thinking types are more interested in administrative functions than introverts and feeling types. Intuitive teachers have a stronger need for inde— pendence and creativity than sensing types. Extraverted and intuitive teachers are more inter— ested in planning school projects than introverts and sensing typeso Intuitive teachers are more interested in working with small groups of students than sensing types. The findings indicate that teachers of different personality types prefer different kinds of teaching situations, lending support to Getzels and Guba's theory that teacher morale and teacher effec— tiveness are likely to improve when institutional expectations are compatible with individual personality. APPENDI CES APPENDIX A PRELIMINARY FORM OF TEACHER PREFERENCE QUESTIONNAIRE 146 APPENDIX A PRELIMINARY FORM OF TEACHER PREFERENCE QUESTIONNAIRE Mark how strongly you feel (1 to 6) about each question below, using the IBM answer sheet. Do not write on this questionnaire. Your answers will be strictly confidential, so please be completely honest and candid. 1. If you had to do it over again, would you go into teaching? Definitely yes 1 2 3 4 5 6 Definitely not 2. In your opinion, is teaching really as challenging and important as it's made out to be? Definitely yes 1 2 3 4 5 6 Definitely not I 3. When did you first get the idea in your head that you wanted to be a teacher? Before I was 10 1 2 3 4 5 6 In college ‘ 4. Being completely honest, how satisfied are you right now with teaching as a career? Extremely satisfied 1 2 3 4 5 6 Extremely dissatisfied| 147 What was your student—teaching experience like? 50 I Extremely enjoyable 1 2 3 4 5 6 Very frustrating 6. Was your supervising teacher the ideal one for you? l Definitely yes 1 2 3 4 5 6 Definitely not 7. Was your teaching evaluated fairly? [ Definitely yes 1 2 3 4 5 6 Definitely not For each type of job listed below, indicate the probability of your holding such a job 10 years from now (if you are working): [ Very likely 1 2 3 4 5 6 Very unlikely 8. Classroom teacher 9. School administrator 10. Subject area specialist 11. School counselor 12. A different job in the field of education 13. A job in another field 148 For each item, indicate if this is an important reason for your staying in teaching as a career: 6 This is not a reason This is the main why I'm teaching reason I'm teaching 14. Intellectual stimulation 15. Working with children 16. Better vacations than other jobs 17. Independence and freedom 18. Working with adults who share my values 19. Socializing with other teachers, administrators, parents, etc. 20. Many opportunities to be creative 21. Adequate salary For each item, indicate if this would be a reason for you to consider leaving teaching as a career: This is the main This would not be reason why I might 1 2 3 4 5 6 a reason for me to leave teaching leave teaching 22. Lack of intellectual stimulation 23. Unruly or bored students 24. Heavy workload 25. Lack of independence and freedom 26. Working with adults who do not share my values 27. Having to socialize with teachers, administrators, parents, etc. 28. Few opportunities to be creative 29. Inadequate salary 149 For the remainder of the questions, mark your preference for each situation using this scale: I l 1 _ _ I'd definitely 1 I d really like it 1 2 3 4 5 6 dislike it J How would you feel if you had to teach this grade level on a regular basis: 30. Pre—school to K 31. Grades 1 to 5 32. Grades 6 to 8 33. Grades 9 to 12 34. College How would you feel if you had to teach this subject area on a regular basis: 35. Math / Science 36. Reading / Literature 37. Social studies 38. Fine arts 39. Applied arts 40. Physical education How would you feel if you had this teaching arrangement on a regular basis: 41. Totally self-contained classroom 42. Self-contained classroom using teacher aides each day 43. Self—contained classroom using subject specialists each day 44. Team—teaching with one other teacher 45. Team—teaching with 2 or 3 other teachers 150 l , . . I'd really like it 1 2 3 4 5 6 I d definite1y dislike it How would you feel if you had to teach this type of student on a regular basis: 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. Gifted student Slow learner Physically handicapped student Mentally handicapped student Bookwormish student Erratic and daydreaming student Cold and critical student Conforming and peer—dominated student Shallow and narroweminded student Clinging student Stubborn and defensive student Loud and hyperactive student Aloof and cynical student Hyper—sensitive and emotional student Radical and rebellious student How would you feel if you had to work with this size group on a regular basis: 61. 62. 63. 64. 65. 66. Over 30 students 20 to 30 students 10 to 20 students 5 to 10 students 2 to 5 students 1 student J 151 . . I'd definitely 1 I d really like it 1 2 3 4 5 6 dislike it J How would you feel if you had to teach in this school on a regular basis: 67. The school has an organized and effective curriculum program with standard exams. You are expected to follow it. 68. The school has various new innovative programs which are con- stantly discussed, but rarely evaluated. You are expected to try out many of them. 69. The school is one in which you may do whatever you like. Teachers are rarely observed or given any support by the administration. 70. The school has a reputation for being traditional, stressing the 3 R's. 71. The school has a reputation for being non—traditional, with an emphasis on the child's interests and social—emotional development. 72. The school is located in a small town. 73. The school is located in a large town. 74. The school is located in a suburb. school is located in a safe urban area. 75. Th (D 76. Th (D school is located in a rough inner—city area. How would you feel if you had to use this teaching style on a regular basis: 77. Lecturing 78. Leading class discussions 79. Working with project committees 80. Working with students engaged in independent study 81. Supervising desk work assignments 152 . . I'd definitely ' l I d really like it 1 2 3 4 5 6 dislike it I How would you feel if you had to do this non—classroom activity on a regular basis: 82° Tutor students individually 83. Serve as advisor to a student club 84. Coach athletics after school (girls or boys) 85. Monitor students during lunch time 86. Monitor students during study hall 87. Organize a multi—media center 88. Help students resolve personal concerns 89. Attend student events (dances, games, etc.) 90. Participate in PTA activities 91. Serve on faculty committees 92. Attend professional conferences 93. Teach after—school workshops in the arts 94. Speak to community groups 95. Arrange bulletin boards and display cases 96. Organize class plays, talent shows, art shows, etc. 97. Set up a special lab (math lab, reading lab, science lab, etc.) 98. Observe and evaluate other teachers 99. Serve as a department chairman 100. Arrange and participate in field trips 101. Plan teacher workshops 102. Participate in overnight camping trips with students and teachers APPENDIX B TEACHER PREFERENCE QUESTIONNAIRE 153 APPENDIX B TEACHER PREFERENCE QUESTIONNAIRE First fill in your name and student number on the enclosed IHT answer sheet. Then mark how strongly you feel (1 to 5) about each of the following questions, using the answer sheet. Please do not write on the questionnaire. Here is an example of how you might respond: Do you think you will earn an M.A. degree within the next 10 years? r1 F7 VT 57 2 a a 3 I 2 Definitely Probably I have no Probably Definitely yes opinion not not Your answers will be strictly confidential, so please be completely honest and candid. 154 1. If you had to do it over again, would you go into teaching? "1‘ '2‘ 5 52 55‘ LJ LJ LJ LJ LJ Definitely Probably I have no Probably Definitely yes opinion not not 2. When did you first get the idea that you wanted to be a teacher? T T T T T L! |_J |_1 I_J L_l Before During Freshman Sophomore high school high school in college Junior/Senior in college in college 3. What was your student—teaching experience like? T T T T T L4 LJ LJ L4 L4 Extremely Quite I have no Rather Extremely enjoyable enjoyable opinion frustrating frustrating 4. Would you have preferred a different supervising teacher (if this had been possible)? T T T T T [_l LJ l_J L_l Ll Definitely Probably I have no Probably Definitely yes opinion not not 50 155 What is your sex? “T r2 Ll |_l Male Female 6. How old are you? T T T |_J L_l \_J 19 _ 21 22 _ 23 24 _ 29 5? 30—40 [”3 Over 40 For each type of job listed below, indicate whether you think you may hold such a job 10 years from now (if you are working): 57 E? 5‘ 5? E? |_J |—| |_J l_l L1 Very probably Possibly I have no Probably Definitely yes opinion not not 7. Classroom teacher T [—27‘ '3' '4‘ ’5 L_l [_J L_J Ll 1_| 8. School administrator T :31 [3| ‘4] [5 1.1 |_u 1_| L_J LI 9. Subject area specialist L1] 3' 2' 4' 12 r71 [—1 W 1—1 m 10. School counselor i: 3| ‘3] £1 21 _ , , _ . r—l F‘l m r-u 1—1 11. A different job in the field of education L_1_1 ‘2‘ ‘3' ‘4] |_51 . . g . [—1 [—1 m |_l m 12. A job in another field ‘1' E} Q] L44 3 156 How important are the following to your staying in teaching as a career? fl l—‘I 1—1 r—I l_| L 8 E h 3 Extremely Important I have no Rather Definitely important opinion unimportant not important . . . . W m M 1'1 13. Opportunity for intellectual discuss1ons [1' El 3 4.3 14. Opportunity to spend most of the day T [51 r—1 [4' with children L_l L_l [2| L_l , _ [—1 [—1 1—1 [—1 15. Better vacations than other jobs [1‘ La [3‘ ‘4] . [_l I“! r‘l [—1 16. Better hours than other jobs [1' [2' I; [4‘ l—I r—I r—I r—i 17. Independence and freedom £1 31 L3; 121 18. Opportunity to work with adults who T I? l; I? share your values l_l I_J 1.1 |_1 [—1 m r‘: l_| 19. Many opportunities to be creative LL £1 L3] £' 20. Opportunity to work with children who T '3 |—| '2 want to learn L_J L._I l2! L_J _ _ r—1 {—1 [_I r—I 21. Opportunity to teach a certain subject L1J IE, [3‘ L4] . m [—7 r1 l_‘l 22. A stable job —- no threat of a layoff L1l IE, l_3_‘ i1] l—‘I [—1 l—l [—I 23. Adequate salary |_1J [2‘ L1: L3: '57 [VI] [0'] [0'] [“1 [V1] [U1] [W] [W] [m] [m] Would the following be a 157 as a career? reason for you to consider leaving teaching T '2‘ T T 5‘ Ln Ln Ln Ln Ln Definitely Probably I have no Probably Definitely yes opinion not not 24. Heavy workload FF 5? r31 7? 7? Ln Ln Ln Ln Ln 25. Very little opportunity to get away if Q? r1 5? I? from children during the school day L3 L4 L3 L3 Ln 26. Lack of independence and freedom FF 5? r31 7? GT Ln Ln Ln Ln Ln 27. Having to work with adults who do not T 131 r31 '4‘ F57 share your values Ln Ln Ln Ln Ln 28. Lack of intellectual stimulation T ’2 '5‘ T 5 Ln Ln Ln Ln Ln r1 rfi r1 r1 FE . 2 29 Inadequate salary :5 Ln £2 :5 £2 30. Having to spend nearly all class time r11 5? g? 5: 7; on subject matter and cognitive skills Ln Ln Ln Ln Ln 31. Unruly or bored students r11 53 [31 7: 7; LI |._l L_J Ll l_l 32. Few opportunities to be creative r1] 5? 5? 7: 7; Ln Ln Ln Ln Ln 33. Too many ineffective and lazy teachers, FII q? 31 7: g; who cannot be fired Ln Ln Ln Ln Ln 34. Too much paperwork, clerical duties, etc. E: ES 51 fig E2 158 For the following questions, mark your preference for each teaching situation using this scale: T 5' 7 'Z‘ s Ln Ln Ln Ln Ln I'd definitely I'd probably I have no I'd probably I'd definitely like it like it opinion dislike it dislike it How would you feel if you had to teach the following grade levels on a regular basis? 35. 36° 37. 38. 39- How would you feel if you had to teach the following Pre—school to K Grades 1 to 5 Grades 6 to 8 Grades 9 to 12 College regular basis? 40. 41. 42. 43. 44. 45. 460 Math or Science Reading or English Social studies Fine arts (Music, Art, Drama) Applied arts (Home Ec., Industrial Arts) Physical education Special education [H] [H] [H] [H] [H] [H] [H] [...] [H] [H] [H] [H] [w] [M] [w] [N] [N] [W] [W] [W] [W] [W] [p] [p] [p] [is] [in] subjects on ['0] ['0] ['0] [l0] ['0] [1°] ['0] [u] [w] [w] [W] [W] [W] [M] [a] [P] [rs] [P] [P] [P] [is] [W] [UI] [W] [W] [W] [m] [kn] [U1] [U1] [VI] [o1] En] 159 T a a 'Z‘ @ Ln Ln Ln Ln Ln I'd definitely I'd probably I have no I'd probably I'd definitely like it like it opinion dislike it dislike it How would you feel if you had to teach in the following schools on a regular basis? 47. School A has an organized and effective curriculum program which includes r? E? [31 Q: I? ready—made exams. You are required to Ln Ln Ln Ln Ln use this specific program. 48. The principal of School B believes in trying out every new educational program which hits the marketplace, but few of them are ever evaluated in a systematic 1 E? T? 5: 5; way. The principal decides which of LJ LJ LJ LJ these new programs you are to use in your classroom. 49. School C is one in which you are totally r1 r1 rn rfi on your own. Teachers are rarely given £1 53 L3] :5 £2 any support or help by the administration. 500 School D has an excellent reputation. It r1 F1 r1 71 ris considered traditional, stressing the 1 £3 121 Si Si 3 R's. 51. School E has a reputation for being non— r1 r1 r1 r1 traditional and "free," with little £3 £3 £2 £5 £2 importance placed on the 3 R‘s. 52. School F is located in a very small town. E E E E 3 _ , Pu Fn F1 F1 To 53. School G is located in a large town. £3 £5 £2 £5 £2 . . [—I m [_I [—1 [—1 54. School H 18 located in a suburb. p5 £3 £3 £5 £2 . _ r1 r1 r7 r1 r1 55. School I is located in a safe urban area. Ll] E5 L3] &5 £2 56. School J is located in a rough inner— r? E; [7 [7 r1 n n L3] .5 3 city area. 160 T T T Ln Ln Ln I'd definitely I'd probably I have no like it like it opinion fl &. [_I é] I'd probably I‘d definitely dislike it dislike it How would you feel if you had to work with the following group size on a regular basis? 57. 58. 59. 60. 61. 62. How would you feel if you were required to use Over 30 students 20 to 30 students 10 to 20 students 5 to 10 students 2 to 5 students 1 student styles on a regular basis? 63. 64. 65. 66. 67. 68. Lead class discussions Present new subject matter to the class, perhaps using audio—visual aids Supervise desk work assignments Use games and role—playing techniques to teach the subject matter Work with project committees Work with individual students engaged in independent activities [...] [h] [H] [t] [1-] [I-] [10] [Ix] [Ix] [Ix] [1x] [1x] [U] [U] [U] [U] [U] [U] [T [*‘] [*] [P] [*P] ['3] [U] [U] [U] [U] [U] [U] the following teaching [...] DA] [...] [H] Dd] [H] [10] [w] [10] [10] ['0] [UI] [U1] [U1] [U1] [U4] [Is] [P] [s] [P] [P] [11>] [3] [U1] [U1] [U1] T 2T T T T Ln L_J |_l l_} In] I'd definitely I'd probably I have no I'd probably I'd definitely like it like it opinion dislike it dislike it How would you feel if you were required to do the following on a regular basis? 69. Tutor students individually after school T '5' '3‘ '4] I? |_l Ln I_J |_J Ln 70. Serve as advisor to a student club T I? [3] [4' '5] 1_1 Ln Ln [_1 |_l 71. Coach athletics after school (girls or T I? 151 I; I? boys I_J L_l I_J l_J Ll . . _ W 1—1 l—“l r—I r—I 72. Superv1se students during lunch time i. E] L3J in L5] . . l_’l {—l m m r—I 73. Superv1se students during study hall i1 13: ‘3' lit] 3 74. Organize a multi—media center (with equipment for photography, videotaping, E E 3 E E model—building, etc. f‘l 75. Help students resolve personal concerns E E 3 E L5] r—I 76. Attend student events (dances, games) E E 3 [4] E . . ~ _ . _ l—I r—1 I_| f—| l—l 77. Parthlpate 1n PTA act1v1t1es ‘1‘ 3 I; [4‘ [5, 78. Make arrangements for field trips T 'E' '3‘ Z‘ '5' (but not participate in them) u LA L4 L4 14 79. Participate in field trips (but not T ’3 'g' ’1‘ I; make arrangements for them) ‘—1 ‘—J l—‘ L—1 ‘—' [_| [_‘l rfi m m 80. Serve on faculty committees I_1_J 131 I; [4‘ ,_5_1 m r'1 [_I l_l l_l 81. Attend professional conferences LL ‘3, [3] L4_‘ Q] 82. Arrange bulletin boards and display '_‘ {—1 ‘_' '_' V] IL £1 [1. 14. I21 cases 162 {—1 {—1 F_I [_‘1 {—1 1L, :3 3 E, 3 I‘d definitely I'd probably I have no I'd probably I'd definitely like it like it opinion dislike it dislike it How would you feel if you were required to do the following on a regular basis? . r1 1‘1 [*1 m l_1 83. Speak to community groups IL I_2J 121 [it] ‘2 811:. Organize class plays, talent shows, [*1 r-: r‘! 1 2 3 £1 5 art shows, etc. l_1 Ln Ln l.J |_: 85. Work with students on community action I? 721 rgl I? r? projects L! I_l |_n Ln u 86. Set up a special lab (math lab, reading r—1 r—1 l—I m [_I . 1 2 3 lab, sc1ence lab, etc. LI Ln Ln Ll Ll 87. Observe and evaluate other teachers T [3 f? T I? l_J k_J 1_: Ln Ln 88. Serve as a department chairman T :3 I? T I; Ln Ln vn: Ln L_J r—I [_| m rm 1—1 890 Plan teacher workshops [£1 |_2J ‘2' szn a 90. Participate in overnight camping trips T 2 I—l I? W with students and teachers l_l Ln I21 m1 L5n How would you feel if you had the following teaching arrangements on a regular basis? m [_I l—I r—I r—I 91. Totally self—contained classroom I_1_J En I2, £1 Q] 920 Self—contained classroom using teacher T 'E‘ '_' "7 7" aides '—‘ l—J '21 £1 '2' 93. Self—contained classroom using subject T '3 '_' '1‘ 'g‘ specialists each day '—‘ '—' Q‘ '—' L—' [*7 |_1 fl l—l I—I 94. Team—teaching with one other teacher L1_, L; LE, .2] Q, |_l {—1 I_| {—_I m 95. Team—teaching with 2 or 3 other teachers g, Lg [_3. lg, [51 163 96. For which level have you had the most teacher training? (Mark only one. '3‘ a a 7; ? |_l L_l Li L! Ll Pre—school Grades Grades Grades Grades to K 1 to 3 4 to 6 7 to 8 9 to 12 97. In which one of these subject areas have you had the most teacher training? Mark only onec Leave this question blank if you have not been trained to teach any of them.) T 3‘ 3‘ T ? Ln Ln Ln Ln Ln General Math Foreign Social Home Ec, elementary language studies Indus. Arts 98. In which one of these subject areas have you had the most teacher training? (Mark only one. Leave this question blank if you have not been trained to teach any of them.) r1 r1 rfi r1 r1 Lin 3 3 a 3 Special English Art, Music Science Physical education Drama education 99. As honestly as you can judge, how satisfied are you right now with teaching as a career? |'_1 F_l l—l l—I {—1 .5 £1 Li, A in Extremely Quite I have no Rather Extremely satisfied satisfied opinion dissatisfied dissatisfied 164 Rank order each of the following student descriptions, indicating which type of student you would most prefer to teach on a regular basis. Assign each student type a different number, from 1 to 4. "1' g 31 T l_l [_1 l_] l_J Favorite 2nd favorite 3rd favorite Least favorite type type type type 100. Student type A is practical and matter—of—fact. These students are realists, concerned with the here—and—now. In school they are patient and accurate, with a great capacity for facts and details, but not very imaginative. They are more interested in first—hand experience than in new ideas or books. They make decisions with impersonal analysis, weighing logical outcomes. [H] ['0] [W] 101. Student type B is enthusiastic and insightful. These students are more interested in future possibilities than in facts, and always seem to be involved in a new project. In school they love group discussions (especially brainstorming) if but they are not always successful at putting new L— ideas into practice because they are impatient with routine. When making decisions they rely on their intuition, taking people's feelings into account. DO] [w] 1020 Student type C is sociable and friendly. These students enjoy people and are tactful and sympa— thetic with others' feelings. They have an ability to see the needs of the moment and adapt to them. In school they are well—liked, but weak up on self—discipline. They are more interested in Ln applying skills they have already learned than in learning new skills. They make decisions with personal warmth, taking people's feelings into account. I w] [W] DP] 103. Student type D is logical and ingenious. These students trust their own intuitive insight as to the true relationships and meanings of things, and they follow their inspirations, whether good or bad. In school they attack any new problem F7 with determination and perseverance. Once they an have figured out how to solve the problem, how; ever, they tend to get bored, and leave the de— tails of the solution for someone else. They make decisions with impersonal analysis, relying on their own intuition and logic. ['0] [W] Dr] Er] IF] d -l H'TF. tn}. . ... L“.- APPENDIX C PROCEHJRE FOR GRAPHING THE FREQUENCY DISTRIBUTIONS OF CONTINUOUS MBTI SCORES 165 APPENDIX C PROCEDURE FOR GRAPHING THE EREQUENCY DISTRIBUTIONS OF CONTINUOUS MBTI SCORES The first step in graphing the frequency distributions of con— tinuous MBTI scores was to calculate four continuous scores for each person, using the method described in the Manual.1 This method pro— duces continuous scores with specific characteristics: the scores are all odd numbers, ranging from 33 to 161, with 100 serving as the division—point which separates the two opposing preferences. After continuous MBTI scores were determined for each of the four indices, the following procedure was used to graph the frequency distributions of the scores: 1. All of the continuous scores for a particular index were rank—ordered from 33 to 161, and were separated into the following groups: scores ranging from 33 to 43, from 45 to 51, from 53 to 59, from 61 to 67, from 69 to 75, from 77 to 83, from 85 to 91, from 93 to 99, from 101 to 107, from 109 to 115, from 117 to 123, from 125 to 131, from 133 to 139, from 141 to 147, from 149 to 155, and from 157 to 161. 2. The total number of scores contained in each of these 16 groups was then determined. 1Myers, MBTI Manual, pp. 8—10. 166 3. These totals were plotted as points on a graph, with the y—axis representing the total number of scores in a group and the x—axis representing the 16 different groups of scores. The plotted points were then connected with straight lines. APPENDIX D PROCEDURE FOR ESTIMATING TEE INTERNAL-CONSISTENCY RELIABILITY OF MBTI SCORES 167 APPENDIX D PROCEDURE FOR ESTIMATING THE INTERNAL—CONSISTENCY RELIABILITY OF MBTI-SCORES In order to estimate the internal—consistency reliability of MBTI scores, a split—half procedure was used. Following Myers' suggested item assignment which is presented in the Manual,1 each item was assigned to the X—half or the Y—half of the appropriate index. Eight hand-scoring keys were then constructed for scoring items assigned to the X—half of each index, and eight hand—scoring keys were constructed for scoring items assigned to the Y—half of each index. Different weights were given to certain answers, based on the differential weightings specified in the regular hand—scoring keys. For the Thinking and Feeling dimensions, the male keys were used for both sexes because items were not evenly split between the X— and Y—halves of the female keys. Using the newly constructed keys, MBTI responses were then re—scored, producing eight X—half scores and eight Y-half scores for each person. Type categories. To estimate the reliability of the type cate— gories, the following procedure was used: 1. Each person's X—half scores were examined in order to assign the person to four type categories, and then Y—half scores were exam- ined in order to assign the person again to four type categories. 1Myers, MBTI Manual, p. E1. 168 2. Phi coefficients were calculated for each of the four in- dices, comparing thalf type—category assignments with Y-half type— category assignments. 3. The Spearman—Brown prophecy formula was applied to each of the phi coefficients, which produced low (conservative) estimates of internal reliability. 4. Tetrachoric coefficients were then calculated for each of the four indices, comparing X—half type—category assignments with Y—half type—category assignments. 50 The Spearman—Brown prophecy formula was applied to each of the tetrachoric coefficients, which produced high (liberal) estimates of internal reliability. 6. Actual reliability coefficients for the type categories were assumed to fall somewhere between estimates derived with phi coeffi— cients and estimates derived with tetrachoric coefficients. Continuous scores. To estimate the reliability of continuous scores, the following procedure was used: 1. Each person's X—half scores were converted to four continuous scores and each person's Y—half scores were converted to four contin— uous scores, using the method described in the Manual.2 2. Pearson product—moment correlation coefficients were calculated for each of the four indices, comparing X—half scores with Y—half scores. 3. The Spearman—Brown prophecy formula was applied to each of the Pearson coefficients, which produced estimates of the internal relia— bility of the continuous scores. 2Myers, MBTI Manual, pp. 8—10. BIBLI OGRAPHY BIBLIOGRAPHY Andrews, John H. M. "Administrative Significance of Psychological Differences Between Secondary Teachers of Different Subject Matter Fields." Unpublished doctoral dissertation, University of Chicago, 1957. Blank, Logan F. "The Relationship Between Student Instructional Ratings and Student—Faculty Psychological Types." Unpublished doctoral dissertation, Michigan State University, 1969. Bown, 0. H., F. F. Fuller, and H. G. Richek. "Differentiating Prospective Elementary and Secondary School Teachers," Alberta Journal of Educational Research, Vol. 12 (June, 1966), pp. 127— 130. Bradway, Katherine. ”Jung’s Psychological Types: Classification by Test versus Classification by Self,” Journal of Analytical Psychology, Vol. 9 (July, 1964), pp. 129—135. Brim, Orville G., Jr. Sociology and the Field of Education. New York: Russell Sage Foundation, 1958. Cage, B. N., Bureau of Educational Research, University of Mississippi. Personal correspondence to the writer, December 9, 1975. Cage, B. N., M. Blackburn, and M. Peaster. "The Effect of Personality on Student/Cooperating Teacher Relationships." Paper read at the First National Conference on the Uses of the Myers—Briggs Type Indicator, October, 1975, Gainesville, Florida. Carroll, John B. "The Nature of the Data, or How to Choose a Corre— lation Coefficient,” Psychometrika, Vol. 26 (December, 1961), Castelli, Charles D. "An Exploration of the Relationship Between Teacher Creative Ability and Teacher—Pupil Classroom Behavior.” Unpublished doctoral dissertation, State University of New York at Buffalo, 1964. Collins, James A. "Individual Personality and Organizational Climate." Unpublished doctoral dissertation, Claremont Graduate School, 1965- 169 170 Conary, Franklin M. ”An Investigation of the Variability of Behav— ioral Response of Jungian Psychological Types to Select Educa— tional Variables.” Unpublished doctoral dissertation, Auburn University, 1965. Gage, N. L., G. S. Leavitt, and G. C. Stone. "The Psychological Meaning of Acquiescence Set for Authoritarianism," Journal of Abnormal Social Psychology, Vol. 55 (July, 1957), pp. 98—103. Getzels, J. W., and E. G. Guba. "Social Behavior and the Adminis- trative Process," The School Review, Vol. 65 (Winter, 1957), pp. 423—441. Getzels, J. W., and P. W. Jackson. ”The Teacher's Personality and Characteristics." Handbook of Research on Teachi , ed. N. L. Gage. Chicago: Rand McNally & Co., 1963. Getzels, J. W., James M. Lipham, and Roald F. Campbell. Educational Administration as a Social Process. New York: Harper and Row, 1968. Glass, Gene V., and Julian C. Stanley. Statistical Methods in Education and Psychology. Englewood Cliffs, N.J.: Prentice— Hall, 1970. Goldschmid, Marcel L. ”Prediction of College Majors by Personality Type," Journal of Counseling Psychology, Vol. 14 (July, 1967), pp. 302-308. Grant, W. Harold. ”Comparability of the Gray—Wheelwright Psycho— logical Type Indicator and the Myers—Briggs Type Indicator." Auburn, Alabama: Auburn University Student Counseling Service, 1965. Hoffman, Jeffrey L. "Personality Relationships Between Supervising Teachers and Student Teachers as Determined by the Myers-Briggs Type Indicator." Unpublished doctoral dissertation, University of Florida, 1974. Hulbert, James C. "A Brief Study Showing the Frequency of Types in Art and Creativity Classes." Paper read at the First National Conference on the Uses of the Myers—Briggs Type Indicator, October, 1975, Gainesville, Florida. Jung, Carl G. Psychological Types, trans. H. G. Baynes. New York: Harcourt, Brace & Co., 192 . Jung, Carl G. (ed.). Man and his Symbols. New York: Dell Publishing Co., 1968. 171 Laney, Arthur R. ”Occupational Implications of the Jungian Person— ality Function—Types as Identified by the Myers—Briggs Type Indicator." Unpublished master's thesis, George Washington University, 1949. Lawrence, Gordon D., and Richard DeNovellis. "Personality Variables and the Middle School Teaching Learning Process." Paper read at the American Educational Research Association convention, April, 1974, Chicago, Illinois. Levell, James P. "Secondary School Counselors: A Study of Differ- entiating Characteristics." Unpublished doctoral dissertation, University of Oregon, 1965. Levy, Nissim, Glennie Murphy, Jr., and Rae Carlson. "Personality Types Among Negro College Students," Educational and Ps cholo ical Measurement, Vol. 32 (Autumn, 1972), pp. 41-653. Lieberman, Myron. Education as a Profession. Englewood Cliffs, N.J.: Prentice-Hall, 1956. Likert, Rensis. "A Technique for the Measurement of Attitudes," Archives of Psychology, Vol. 22 (June, 1932), pp. 5—55. MacKinnon, Donald W. "The Nature and Nurture of Creative Talent." Personality Research, ed. D. Byrne and M. L. Hamilton. Englewood Cliffs, N.J.: Prentice—Hall, 1966. Madison, Peter, David H. Wilder, and Walter B. Suddiford, "The Myers—Briggs Type Indicator in Academic Counseling." Statistical Research Report No. 13. Princeton, N.J.: Princeton University Counseling Service, 1963. McCaulley, Mary H. "Type and Education." Gainesville, Florida: University of Florida Typology Laboratory, 1971. McCaulley, Mary H. "Presenting MBTI Information on Individuals and Groups." Paper read at the First National Conference on the Uses of the Myers—Briggs Type Indicator, October, 1975, Gainesville, Florida. McCaulley, Mary H., et al. "Myers—Briggs Type Indicator Applications." Gainesville, Florida: University of Florida Typology Laboratory, 1973- McCaulley, Mary H., and Frank L. Natter. "Psychological Type Differ— ences in Education." The Governor's Task Force on Disruptive Youth, ed. F. L. Natter and S. A. Rollin. Tallahassee, Florida: Office of the Governor, 1974. 172 McNamara, Thomas C. ”A Study of Philosophical Identities in a Counseling Practicum." Unpublished doctoral dissertation, University of Illinois, 1967. Mendelsohn, Gerald A. "Review of the Myers—Briggs Type Indicator." Sixth Mental Measurements Yearbook, ed. 0. K. Buros. 3d ed. Highland Park, N.J.: Gryphon Press, 1965. Mendelsohn, Gerald A., and M. H. Geller. "Similarity, Missed Ses— sions, and Early Termination," Journal of Counselipg Psychology, Vol. 14 (May, 1967), pp. 210—215. Miller, Paul Van Reed, Jr. "The Contribution of Non—Cognitive Var- iables to the Prediction of Student Performance in Law School." Unpublished doctoral dissertation, University of Pennsylvania, 19650 Morales, Carmen A. "Teacher Personality Types and Teacher Morale in Selected Open Space and Self—Contained Elementary Schools." Unpublished doctoral dissertation, University of Florida, 1975. Myers, Isabel Briggs. "Type as the Index to Personality." Privately published, Swarthmore, Pennsylvania, 1945. Myers, Isabel Briggs. The M ers—Bri s T e Indicator Manual. Princeton, N.J.: Educational Testing Service, 1962. Myers, Isabel Briggs. "Introduction to Type." Privately published, Swarthmore, Pennsylvania, 1970. » Myers, Isabel Briggs. Personal correspondence to the writer, December 1, 1975- Myers, Isabel Briggs, and Junius A. Davis. ”Relation of Medical Students' Psychological Type to Their Specialties Twelve Years Later." Research Memorandum 64-15. Princeton, N.J.: Educational Testing Service, 1964. Nunnally, Jum C. Psychometric Theory. New York: McGraw—Hill Book Co., 1967. Paraskevopoulos, Ioannis. ”How Students Rate Their Teachers," Journal of Educational Research, Vol. 62 (September, 1968), pp. 25-29. Richek, Herbert G. "Note on Intercorrelations of Scales of the Myers- Briggs Type Indicator," Psychological Reports, V01° 25 (August, 1969), pp- 28—30. 5:?! I ' ‘ L'I-L- .- ' I ' (l-i' .'. : f3"... 10_ ”if.“ 1 1 . u _ ' .. ‘ 31.1. o :1; o ' .-. I I . - . ' ' . ’ =1 . . n ' u u I , I t 173 Richek, Herbert G. "Personality Characteristics of Male and Female Prospective Teachers: A Multivariate Analysis," Alberta Journal of Educational Research, Vol. 19 (September, 1973i, pp. 208—215. Richek, Herbert G., and Oliver H. Bown. "Phenomenological Correlates of Jung's Typology," Journal of Analytical Psychology, Vol. 13 (January, 1968), pp. 57—65. Ross, John. "Progress Report on the College Student Characteristics Study." Research Memorandum 61—11. Princeton, N.J.: Educational Testing Service, 1961. Ross, John. "The Relationship Between a Jungian Personality Inventory and Tests of Ability, Personality, and Interest," Australian Journal of Psychology, Vol. 18 (April, 1966), pp. 1—17. Rudisill, Edwin M. ”An Investigation of the Relationship Between Mathematics Teachers' Personality Characteristics, as Measured by the Myers—Briggs Type Indicator, and Their Preferences for Certain Teaching Strategies." Unpublished doctoral dissertation, University of Florida, 1972. Saunders, D. R. "Evidence Bearing on Use of the Myers—Briggs Type Indicator to Select Persons for Advanced Religious Training." Research Bulletin 57—8. Princeton, N.J.: Educational Testing Service, 1957. Saunders, Do R. "Evidence Bearing on the Existence of a Rational Correspondence Between the Personality Typologies of Spranger and Jung." Research Bulletin 60—6. Princeton, N.J.: Educational Testing Service, 1960. Schmidt, Lyle D., and Bruce R. Fretz, ”The Effects of Teacher— Student Similarity in an Educational Skills Course.” Cooperative Research Project No. S-217. Washington, D.C.: U. S. Office of Education, 1965. Stalcup, Donna L. K. "An Investigation of Personality Character- istics of College Students Who Do Participate and Those Who Do Not Participate in Campus Activities." Unpublished doctoral dissertation, Auburn University, 1967. Story, Garth E. "An Analysis of the Relationships Between Person— ality Types of Mathematics Teachers 7—12), as Measured by the Myers—Briggs Type Indicator, and Selected Factors Related to Teaching." Unpublished doctoral dissertation, University of Florida, 1972. Stricker, Lawrence J., and John Ross. "A Description and Evaluation of the Myers—Briggs Type Indicator." Research Bulletin 62—6o Princeton, N.J.: Educational Testing Service, 1962. . :3:- 5 yet": all : g 'r:-" . .-;.':'_-l:a." .. 174 Stricker, Lawrence J., and John Ross. "Intercorrelations and Relia— bility of the Myers-Briggs Type Indicator Scales," Psychological ‘ Reports, Vbl. 12 (February, 1963), pp. 287—293. ‘ Stricker, Lawrence J., and John Ross. "An Assessment of Some Struc- tural Properties of the Jungian Personality Typology," Journal of Abnormal and Social Psychology, Vol. 68 (January, 1964), pp. 72—71. Stricker, Lawrence J., and John Ross. "Some Correlates of a Jungian Personality Inventory," Psychological Reports, Vol. 14 (April, 1964), pp. 623—643. Stricker, Lawrence J., Harold Schiffman, and John Ross. "Predictions of College Performance with the Myers-Briggs Type Indicator," Educational and Psychological Measurement, Vol. 25 (Winter, 1965), pp. 1081—1095. Taylor, Robert E. ”An Investigation of the Relationship Between Psychological Types in the College Classroom and the Student Perception of the Teacher and Preferred Teaching Practices." Unpublished doctoral dissertation, University of Maryland, 1968. Van der Hoop, J. H. Character and the Unconscious, trans. E. Trevelyan. London: Routledge & Kegan Paul, 1923. ‘ Van der Hoop, J. H. Conscious Orientation, trans. Laura Hutton. London: Kegan Paul, Trench, Trubner & Co., 1939. Von Fange, Erich A. "Implications for School Administration of the Personality Structure of Educational Personnel.” Unpublished doctoral dissertation, University of Alberta, 1961. Von Franz, Marie-Louise. "The Inferior Function." J 's 010 . New York: Spring Publications, 1971. Webb, Sam C. ”An Analysis of the Scoring System of the Myers—Briggs Type Indicator,” Educational and Ps cholo ical Measurement, Vol. 24 (Winter, 1964), pp° 765-781. Williams, David L. "An Analysis of the Interrelationships Among Elementary School Teachers' Personality Types, Beliefs, Observed Classroom Practices, and Reports of How Broadcast Instructional Television Should Be Used." Unpublished doctoral dissertation, University of Florida, 1972. Wright, Judith Anne. "The Relationship of Rated Administrator and Teacher Effectiveness to Personality as Measured by the Myers— Briggs Type Indicator." Unpublished doctoral dissertation, Claremont Graduate School, 1966. MICHIGAN STATE UNIV. LIBRARIES 1|”thth WI ”WI W W W “WWI WI 3129310576531