Tl-E DEVELOPMENT AND VALEQATION OF A SCALE TO mwm EMPATHY Thesis for ”as Degree of DB. D. mcmm STATE; UNEVERSETY James Lee Chapman 1966 F". m M LIFER/4"" Michigan Sun University u gum; 1%"ij Lm fl Mil 1m "me W2! I This is to oerttfg that the , . } thesis entitled g THE DEVELOPMENT AND VALIDATION OF 1 A SCALE TO MEASURE EMPATHY presented by James Lee Chapman has been accepted towards fulfillment of the requirements for Ph.D. degree in Guidance and Counseling XMW/t’fitzo’ Malnr vat-afoul:- Date May 20, 1966 0-169 ”day-«H I 1" i. I ABSTRACT THE DEVELOPMENT AND VALIDATION OF A SCALE TO MEASURE EMPATHY by James Lee Chapman The purpose of this study was to develop and validate an instrument that tested a subject's ability to identify the emotion or emotions ex- pressed by another in a series of videotaped excerpts from actual coun- seling interviews. The instrument was composed of selected scenes of videotaped recordings of counseling interviews. Each scene was followed by four to seven descriptive adjectives. Subjects were to respond to each ad- jective on a continuum indicating how strongly they possessed the feeling described by the adjective at the end of each scene. Two criterion groups, judged high and low in empathic regard, responded to the instru- ment and generated items from a list of 57 adjectives expressing various feelings. The final list of four to seven descriptive adjectives was produced from the list of 57 adjectives. The instrument was administered to 53 subjects in two universities for the purpose of validation. In order to cross-validate the results of the item analysis of the instrument, it was administered to another 88 persons enrolled in three National Defense Education Act Summer Institutes (1964). The instrument was scored for "right" and "wrong" answers on 172 scorable items from the original 280. Correct and incorrect answers were determined on the basis of the responses given by the criterion James Lee Chapman groups of high and low empathizers. The total sample was increased from 88 to 148 by adding the data from two additional National Defense Education Act Summer Institutes (1964). The purpose of scoring the in- strument and increasing the sample was to determine the predictive validity and reliability of the instrument. In addition, a t-test for significance among the sub-groups of the sample and a factor analysis to determine commonality of meaning in adjectives used were made possible by scoring the instrument. The desired data from the subject's responses were placed on IBM cards for statistical analysis. Chi-square was used in the statistical treatment of the data for validation and cross-validation purposes with the test of significance being established at .20 or higher. The two- by-two chi-square contingency was employed on the response scale. The re8ponse scale was split at the point at which the median occurred in the response pattern of each item. Analysis of variance to ascertain predictive validity; t-test to determine significant differences among means of the sample sub-groups; Kuder-Richardson formula 20 to establish reliability and factor analysis for commonality of meaning were also used in working with the data. A total of 65 items of the original 280 included met the test of significance for the purpose of validation, and nine items were found to be significant in cross-validating the instrument. These nine items which held up on cross-validation included the adjectives "bitter" (three of the items), "disgusted," "scared," "help- less, confused," "mixed-up," "lonely," and "satisfied." The large number of negative adjectives results from the fact that the adjectives James Lee Chapman in general were descriptive of negative affect. There was no major concentration on any one client or episode although nearly all of the items were toward the end of the scale. Because the nine items repre- sent about a fifth of the 65, and one would expect about a fifth of the items to show significance by chance at the 20% level, it would appear that there is nothing more than chance relationship between the item re5ponses and the criterion. The predictive validity of the instrument was nonexistent. The analysis of variance values for three effects (sex grouping, high-low grouping and interaction) were low enough to indicate that should the sample be increased to infinity, significance would not be produced. The correlation of reliability was .441. The t-test did not find a significant difference between the highest and lowest means of the two sub-groups involved. There were five factors operative in a ten-factor matrix. A total of 29 scenes and 62 items were found in the five factors. THE DEVELOPMENT AND VALIDATION OF A SCALE TO MEASURE EMPATHY By James Lee Chapman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education Department of Guidance, Personnel Services and Educational Psychology 1966 ACKNOWLEDGMENTS I wish to express sincere appreciation to my adviser, Dr. Norman Kagan, for his kind assistance and helpful suggestions in the preparation and presentation of this thesis. I am also grateful to Drs. Marion Kinget, James Costar and Floyd Parker for their assistance as members of my guidance committee. In addition, I express gratitude to Drs. Walter Johnson and David Krathwohl who were members of the guidance committee earlier in my program. I wish to thank my wife, Mary, and son, Ronald, who were willing to be without a husband and father on many occasions throughout my entire doctoral program. ii TABLE OF CONTENTS CHAPTER I. INTRODUCTION. Background of the Study . Need for the Study. The Problem . Definition of Terms Used. Limitations Assumptions Hypotheses. Organization of the Thesis. II. PREVIOUS APPROACHES TO THE DEFINITION AND MEASUREMENT OF EMPATHY. Definitions: Empathy and Affective Sensitivity. The Measurement of Empathy. Predictive Studies of Empathy . Situational Tests of Empathy. Summary of Literature III. DESIGN OF THE INSTRUMENT AND METHODOLOGY. Construction of the Instrument. Selection of Videotaped Scenes. Selection of Criterion Groups for Generating Items iii Page 10 13 15 18 21 23 23 23 24 CHAPTER Generating Items. Selection of Items for the Affective Sensitivity Scale Methodology for Cross-Validation. Selection of the Sample Procedures for Obtaining the Data . Validity. Reliability . Factor Analysis Summary . IV. ANALYSIS OF DATA. Results of Validating the Instrument. Results of Cross-Validating the Instrument. Predictive Validity Results Reliability Results Factor Analysis Results Analysis. Validating and Cross-Validating the Instrument. Predictive Validity Reliability . Factor Analysis Summary . V. SUMMARY, CONCLUSIONS AND DISCUSSION . Summary Findings. iv Page 25 26 29 29 3O 31 32 32 32 34 34 4O 42 46 47 48 48 48 50 50 51 52 52 53 CHAPTER Conclusions Discussion. Implications for Further Research BIBLIOGRAPHY. Page 54 55 62 115 TABLE LIST OF TABLES Composition of the Criterion Groups. Categorization of Subjects from which Validation Groups were Obtained. Categorization of Enrollees from which Cross-Validation Groups were Obtained . Frequency Distribution and Significance of Items on the Affective Sensitivity Scale which Failed to Meet the .20 Level of Significance in Validating the Instrument. Frequency Distribution and Significance of Items on the Affective Sensitivity Scale which Failed to Meet the .20 Level of Significance in Cross-Validating the Instrument Summary Data of Scored Instrument. Analysis of Variance of Instrument Scores with Means Adjusted for Disproportionality. NDEA Institute Means Adjectives in Operative Factors. vi Page 25 27 3O 35 41 43 45 47 49 FIGURE 1. 2. 3. LIST OF FIGURES Illustrations on Determining the Median. Table of Symbols Illustration on Determining Factors. vii Page 28 44 47 LIST OF APPENDICES APPENDIX A. List of Fifty-seven Adjectives. B. Affective Sensitivity Scale C. Item Analysis of Items Found Nonsignificant in Validating the Instrument. D. Rating Scale for Counseling Ability E. Rating Scale for Empathy. F. Item Analysis of Items Found Nonsignificant in Cross- Validating the Instrument G. Items Scored and Correct Response for Obtaining Total Score viii Page 63 68 81 99 101 103 109 CHAPTER I INTRODUCTION What would be desirable preparation for the person who is to receive training as a therapist? . . . If the student is to become a thera- pist, the more he has been able to achieve of empathic experiencing with other individuals, the better will his preparation be. (Rogers, 1951, p. 437) Theorists of both the psychoanalytic and client-centered schools of counseling have emphasized that one important therapist variable, independent of technique, is the therapist's ability to understand the client empathically (Cartwright and Lerner, 1963). Although, therorists and practitioners of different positions agree upon the importance of the therapist's empathic ability, studies investigating this variable are few when compared to those relating to intelligence, interests and personality development. The confusion which surrounds the definition of empathy could be one explanation for the dearth of studies in this area. The inability to replicate human behavior prior to the advent of motion picture and television facilities in university and research settings may be another. »Since one important therapist's variable in counseling is the therapist's ability to understand the client empathically, the extent to which a counselor possesses empathic regard could be one major factor leading to success in counseling. The instrument developed in the pre- sent study is an attempt to provide an aid to assist counselor educators in predicting counseling success by potential counselors. Background of the Study Studies relating to the concept of empathy, prior to 1956, have been confusing since theoretical and operation definitions do not agree. Cron- back (1955) indicates that studies of "social perception" have been chiefly concerned with differences among perceivers either in terms of their accur- acy or in terms of their tendency to view others as similar to themselves. These studies have usually been built around a particular operation in "predicts" how another person will respond in a given situa- which a judge tion or to a test. The judge's response is then compared with character- istics of the judge himself. Often, for example, both persons describe themselves on a personality inventory, and the judge is then asked to fill out the inventory as he thinks the other person did. The extent to which the prediction agrees with the other person's actual response is taken as a measure of the judge's accuracy of social perception (or C O I C C O ' " "soc1al sensrtiVity," "diagnostic competence'). Scores ob- "empathy, tained in this manner are difficult to interpret, and several investigators have reported low consistency for them (Crow, 1954; Gage, 1953; Stone and Leavit, 1954). Recently attempts have been made by Astin (1957), O'Hern (1964), and Stefflre (1962) to study empathy by confronting the subject with a simu- lated counseling situation which has been recorded on audio tape. The recording is made from a "staged" counseling interview and is designed to be stopped at the appropriate time. The subject is required to choose a proper response from a number of alternatives given. The situation is structured in such a way that the behavioral variable or emotion in ques- tion can be revealed and responded to by the subject. Kelly and Friske (1951) have employed this approach in assessing behavioral variables of clinical psychology trainees in Veteran Administration hospitals. Buchheimer (1964) is currently conducting research on empathy by using an actual counseling situation. The situation is recorded on film and the subject is requested to respond verbally at selected points in the interview as though he were the counselor. The subject's response is recorded on an audio tape recorder. The results of most studies of empathy continue to indicate that the definition of this variable, as espoused by the researcher, is not operationally consistent with the design. The confusion between theoretical definitions of empathy and operational definitions have contributed to a lack of insight into this important variable. It becomes necessary that the nature of this vari- able be investigated due to the important part it plays in counseling and other interpersonal relationships. The possibility of uniting theoretical and Operational definitions of empathy has been enhanced with the availability and use of closed circuit television. Researchers are able to record human behavior as it occurs in the counseling interview. A new research tool has been developed in the College of Education at Michigan State University through use of closed circuit television which is designed to assist in the study of empathy (Kagan, Krathwohl and Miller, 1963). The technique which is known as Interpersonal Process Recall (IPR) can be classified in the general area of stimulated recall methodology and permits identification of the nature of emotions expressed by clients in recorded behavioral episodes. Need for the Study In an extensive review of studies relating to empathy, Taft (1955) describes the different methods of measuring empathy, classifications of the tests used, the reliability of tests and concludes the article by listing the characteristics (i.e., age, sex, intelligence, training in psychology, esthetic ability, sensitivity and emotional stability) of those who should be good judges of others. Although these character- istics are given, one is still faced with the question as to whether em- pathic ability is being tested or simply the ability to predict basic personality traits of individuals or groups from a reservoir of stereo- typic concepts and attitudes. The subject seems to be required in most of these tests of "empathi' to calculate and analyze the facts presented and come to a decision that requires a great amount of cognitive ability but little, if any, affective sensitivity ability. The point is made by Allport (1937) that, "the theory of empathy is a peculiar blend, and must be regarded both as a theory of inference and a theory of intuition depending upon the coloring given it by different authors.” If one looks at empathy as having at least two dimensions, that of inference and intuition, the need continues for an instrument to be de- veloped which measures a subject's affective sensitivity ability by ascertaining the momentary psychological state of another person. In order to ascertain the affective state of another, one must be capable of determining the feelings at the moment they are expressed. At the present time there is not a dependable test for empathy for as Buchheimer (1963) indicates, the empathic process has been studied from only a limited point of View. The Problem The purpose of this study is to develop and validate an instrument that tests a subject's ability to identify the emotion or emotions that are being expressed by another in a series of videotaped scenes from actual counseling interviews. The specific problem is to determine whether or not such an affective sensitivity instrument will signifi- cantly correlate with counseling effectiveness. Definition of Terms Used Instrument. In this study, the term "instrument" shall refer to the device used to obtain an objective score of affective sensitivity. The instrument itself consists of selected scenes of videotaped record- ings of actual counseling interviews. Each scene is followed by four to seven descriptive adjectives, from which subjects are to respond to each word on a continuum from feeling strongly to not possessing the feeling at all. Affective sensitivity. For the purpose of this study, affective sensitivity shall refer to the subject's ability to identify the momen- tary feelings of a client as revealed through the videotaped recording of the client. Affective sensitivity score. The raw score a subject earns on the instrument shall be referred to as the affective sensitivity score. The higher scores shall indicate a greater degree of agreement with criterion groups of known empathizers. Sample. -Students enrolled in five summer institutes established by the National Defense Education Act in different areas of the United States comprised the sample. Known empathizers. Those persons who by reputation were judged by their instructors or colleagues to be high in empathic regard are referred to as known empathizers. Interpersonal Process Recall. The IPR technique Operates in the following manner: A counselor and client conduct a counseling interview within the studios of the closed circuit television installation. The studio is curtained, the cameras are pre-set and unmanned so that a minimum of distraction exists. The interview is enacted and the two participants are videotape recorded on a split screen with head and torso head-on views enlarged as much as the screen permits. Immediately after the interview is concluded, the client witnesses a playback of the inter- view in the presence of another trained counselor. The trained counselor encourages the client to describe his feelings, interpret statements, and translate body movements at various times during the replayed interview. The trained counselor or subject may stop the playback and discuss re- called feelings and elaborate on meanings. Limitations The study is limited by the following: 1) There was no follow-up or replication of the study. 2) The criterion groups of high and low empathizers were selected by "reputatiod' among their peers or instructors rather than by an objective measure. 3) No systematic or equated stratification of samples was attempted. Assumptions The following assumptions are implicit in the study: 1) That client affective behavior in a counseling interview is measurable. 2) That client behavior observed and recorded at Michigan State University is not markedly different from that observed in other areas of the country. 3) That counselor educators can judge between effective and less effective counselors based upon their conduct during the coun- selor's professional education. 4) That colleagues can judge between counselors who are high and low in affective sensitivity. Hypotheses Counselor education has as one major goal the education of effective counselors. There seems to be general agreement in the field that many personal traits including one's attitudes toward others are included in counselor effectiveness (Weitz 1957 and O'Hern 1964). The present study investigates only one of these traits, that of affective sensitivity, under the assumption that this trait can be measured. The hypotheses to be tested in this investigation are: l) A scale based upon actual counseling interviews will differen- tiate between persons high and low in empathic regard. 2) A scale based upon actual counseling interviews will produce a significant correlation between scores received on the affective sensitivity scale and counselor effectiveness. Organization of the Thesis Studies related to the concept of empathy and their findings will be summarized in Chapter II. The basic design and statistical procedures that were used in the construction of the instrument will be presented in Chapter III. A description of the treatment of the data and results of the analysis will be reported in Chapter IV. Chapter V will contain the summary, conclusions and recommendations. CHAPTER II PREVIOUS APPROACHES TO THE DEFINITION AND MEASUREMENT OF EMPATHY The literature on empathy is found in various areas of psychology, from industrial to social and clinical psychology. Empathy is a respected concept in the German school of Verstehende Psychologie, which holds that "psychic processes are not to be described but to be intuitively under- stood' (English and English, 1958). Empathy or Einffihlung, in this con- text, is regarded as a faculty of immediate knowledge. This connotation has little acceptance in American psychology, in which empathy is partly synonymous with clinical inference. Sarbin (1960) pointed out the apparent distinction between clinical inference, which follows a general inference model, and intuition or immediate knowledge as a nonmediated process. He found that careful consideration of the elements of the intuitive process proved that it conformed to the logical model of inference, the under- lying postulates being immediately available from past experience. The absence of a sound theory which would explain the nature of empathic interpersonal communication either as an intuitive, cognitive model or as an inferential, learning model has permitted the operational meanings or usages of the term 'empathf' to become increasingly vague and contra- dictory. 10 Definitions: Empathy and Affective Sensitivity The problem of defining empathy is an epistemological one--how do we "knod' others in a counseling relationship? Attempts to define empathy have been as diverse and as inconclusive as the efforts of philosophers to describe the nature of knowledge. To establish an operational basis for the development of the present scale, various definitions and synonyms for empathy were compared to generate the testable concept of affective sensitivity. The term, Einfuhlung, later translated by Titchener as empathy, was introduced in 1897 by Theodore Lipps to describe the process of aesthetic appreciation as a form of inner imitation. This imitative characteristic is distinctive of those usages of the term which postulate an identity between the empathizer and his object. Sensitivity to the affect of another, according to such a definition, depends on analogy, identifica- tion, or real similarity. This connotation is illustrated in the common sense idea of "putting yourself in the other fellow's place” which Allport (1954) found as a point of agreement in many definitions of empathy. An important theoretical disparity exists between this group of definitions and those which stress the detachment and objectivity of the empathizer. Typical of the second class of definitions for empathy is the one given by English and English (1958): Apprehension of the state of mind of another person without feeling (as in sympathy) what the other feels. While the empathic process is primarily intellectual, emotion is not precluded, but it is not the same emotion as that of the person with whom one empathizes. The parent may empathize with the child's puny rage, feeling pity or amusement, whereas in sympathy he would feel rage along with the child. The attitude in empathy is one of acceptance and under- standing; of an implicit ”I see how you feel." 11 Carl Rogers (1959) emphasized the importance in counseling of the empathic understanding of the other person's "internal frame of reference." He described the state of being empathic as one of perceiving: the internal frame of reference of another with accuracy, and with the emotional components and meanings which pertain thereto, as if one were the other person, but without ever losing the"as if' condition. Thus it means to sense the hurt or pleasure of another as he senses it, and to perceive the causes thereof as he perceives them, but without ever losing the recognition that it is as if I were hurt or pleased, etc. If this "as if' quality is lost, then the state is one of identification (1959, pp. 210-211). The element of detachment, or as Rogers calls it, the ”as if' condition, is present in the psychoanalytic definition of empathy, as given by Fenichel (1945, p. 511): empathy consists of two acts: (a) an identification with the other person, and (b) an awareness of one's own feelings after the identification, and in this way an awareness of the object's feelings. Katz (1963) advocated a four phase model for empathy, based on the formula given by Reik in Listening with the Third Ear. This finer division of the continuum from identification to objectivity suggests that the more closely one inspects the elements of an empathic act, the further one can amplify the definition. The semantic controversy as to whether empathy is immediate, unmediated knowledge gained through a process of identifi- cation, or a logical process of inference culminating in objective awareness, may depend on the duration of the act or sequence of acts denoted as being empathic. An immediate focusing of attention on the affective state of another simultaneous with his behavioral expression of that state may be only one brief facet of interpersonal sensitivity. The nature of this particular operation is essentially passive and receptive, and if it is of extended duration and constancy it becomes sympathetic rather than 12 empathic behavior. Buchheimer (1963) points out that a sympathetic person does not need to interact with the other person. In sympathy, he says, there is a paralleling of thought and feeling between the two or more individuals, whereas empathy connotes a converging of behavior. Thus an empathic process may begin with a receptive state of sensitivity to the affect being communicated by another, but the state rapidly shifts into covert inferential activity and into overt operations intended to elicit meaningful or needed communication from the other; and this would be merely one phase of a cyclical process of interaction leading to increas- ing congruence. The plethora of different usages for the term empathy in the litera- ture points to the need for a clear analysis of the nature and duration of the empathic act. Some authors, to side-step the semantic confusion, have contrived fresh paraphrasings of a more generic type, such as "inter- personal perception" (Gage and Cronbach, 1955) or "ability to judge people” (Taft, 1955). With a similar intent, though hopefully taking finer aim in conceptualizing, this study has attempted to steer clear of the theore- tical shoals inherent in the discussion of empathy by purporting to measure only the more circumscribed trait of affective sensitivity, This trait is defined as the ability to detect and describe the immediate affective state of another; or in terms of communication theory, the ability to receive and decode affective communication. Affective sensitivity would be a component of both sympathy and empathy, depending on the nature and duration of the interaction. The demonstration of affective sensitivity would depend upon an opportunity to observe--though not necessarily to interact with--the other. A high 13 degree of affective sensitivity would not suffice to ensure effectiveness as a counselor; a low score, however, on a measure of affective sensitivity would suggest unsuitability for counselor training, indicating that the person perceives the feelings of others inaccurately, perhaps on the basis of stereotyping, assumed similarity, projection, or other distorting pre- mises. The Measurement of Empathy Investigators attempting to measure empathy have used approaches as varied as the efforts to define the term empathy. The earliest ostensible study of empathy was conducted by Gordon in 1934. He instructed his sub- jects to tell whether the right or the left arm was raised in four photo- graphs of a Mexican male, and four reversals of these images. This procedure was an Operationalizing of the original Lippsian concept of aesthetic empathy, in its postulation of an identity of feeling in the musculature of the empathic perceiver. Subjects were rated as empathic if they overtly or covertly attempted to emulate the attitude of the photographed image. The current wave of interest in devising instruments for estimating and correlating empathic ability mounted in the 1950's. As yet, however, only one normative test has been reported in the Fifth Mental Measurements Yearbook (Buros, 1959), a group paper and pencil device entitled The Empathy‘Test (Kerr and Speroff, 1954). Designed for use in industrial selection, the test was intended to measure "the ability to put yourself in the other person's position, establish rapport, and anticipate his feelings, reactions and behaviors" (1954, p. 269). The researchers then translated this definition into operational terms the ability to rank as 14 the average person would rank them: (a) the popularity of musical forms, (b) reading and buying preferences for magazines, and (c) the annoyance magnitude of such experiences as hearing a person chew gum. Validation criteria were obtained from national survey results and related research findings. Robert L. Thorndike commented on The Empathy Test as follows: There appears to be no inherent validity in the operations called for in this test, and so its validity must be established empirically through its ability to predict socially important criteria, or its relationships to other variables that would make it a meaningful construct. The manual reports several studies presenting evidence on the validity of the test, and certain of these appear quite im- pressive. However, the relatively few studies by persons not assoc- iated with the author have tended to yield predominately negative results. Unless the positive results reported in the manual are verified in the findings of other workers, this test cannot be recommended as either a useful practical device or a contribution to the description and understanding of an individual (Buros, 1959, p. 120). ' Norman and Ainsworth In a similar endeavor to measure "massempathyfi (1954) administered two forms of the GAMIN Personality Inventory at an interval of two weeks apart. 0n the first administration, the subjects answered for themselves; on the second, they responded as they thought most other people of their own age and sex would respond. The criterion for scoring the responses on the second form as being "massempathic" was based on the scores of the total group on the first form. While the authors were careful to define their terms appropriately, the demonstra- tion of ability to conform to the average response pattern on a person- ality scale cannot be used to predict sensitivity in individual interaction. As in the case of Kerr and Speroff's test the subject is required to take the role of the generalized other, rather than that of a specific other. Gage and Cronbach (1955) delineated four components of experimental design in studies of interpersonal perception: 15 The Judge whom the experimenter is attempting to measure. The Othergs} whom the Judge is asked to interpret. c. The Input, or information concerning the Other which is available to the Judge. d. The Outtake, i.e., the statement or predictions about the Other obtained from the Judge. (p. 412) 0‘!» The two studies discussed above attempted to assign a score to the judges on the basis of their sensitivity to the generalized other. Let us now turn to those research studies in which empathy for a specific other has been estimated and attempt to distinguish among these on the components of input and outtake, which between them determine the nature and avail- ability of validating criteria. Predictive studies of empathy. One of the most popular approaches to the measurement of empathy was devised by Dymond (1949), who defined empathy as "the imaginative transposing of oneself into the thinking, feeling and acting of another and so structuring the world as he does." As a measure of this ability she found the correspondence between sub- jects' ratings of others known to them through group interaction and the actual self-ratings made by the others. This approach thus emphasized the outtake component by requiring judges to make global predictions about others. A refinement of the original test appeared in 1950; it required rankings on a 5-point scale on each of six traits: superior- inferior, friendly-unfriendly, leader-follower, shy-self-assured, sym- pathetic-unsympathetic and secure-insecure. Other studies employing a predictive criterion followed in the wake of Dymond's articles. Bender and Hastorf (1950) asked subjects to pre- dict the responses of acquaintances on three personality scales and found low positive correlations between obtained and predicted scores. Cowden (1955) had 29 married couples take the California Test of 16 Personality and then predict how his or her partner would answer the questions; he considered the accuracy of prediction to be a measure of empathy. Predictive studies have an inherent weakness in the tendency of subjects to avoid negative or unfavorable ratings of themselves and others. The results usually produce a high clustering of scores. The authors cited above have attempted to apply corrections for the biasing factors of projection and assumed similarity. Hastorf, Bender and Wein- traub (1955) found that subjects were cautious in predicting, and used the middle "often" and "seldonP alternatives. They remark: In the light of this finding, we learn why S's who themselves had definite preferences and close associates who reply with midscale responses tended to obtain high refined empathy scores. In view of these adventitious findings it would then become possible to predict an S's "empathic ability" without even scoring his predic- tions. We must therefore impugn the refined empathy score as a true measure of empathic ability. (p. 343) A similar conclusion was reached by Lindgren and Robinson (1953) who found in evaluating Dymond's empathy rating scale that predictions tended to conform to a cultural stereotype or norm. They found the scale too cumbersome for use as a normative test and, finally, questioned the validity of the Dymond scale in that a high score for predicting test responses might not denote actual empathy. In a similar indict- ment of tests stressing the predictive component, Cronbach (1955, p. 191) comments: Social perception research has been dominated by simple, operation- ally defined measures. Our analysis has shown that any such measure may combine and thereby conceal important variables, or may depend heavily on unwanted components. Only by careful subdivision of global measures can an investigator hope to know what he is dealing with. Our analysis makes especially clear that the investigator of social perception must develop more explicit theory regarding the constructs he intends to study, so that he can reduce his measures to the genuinely relevant components. 17 Thus, the ability to predict global attributes of another is a complex and fortuitous variable to investigate, and studies of this type have not contributed valid or reliable methods for assessing interpersonal sensitivity. The authors of two further studies attempted to avoid the problems of using a predictive criterion for empathy. Lifton (1958) asked the members of a class in music education to rank their fellow members as "someone--sensitive to the feelings others may have about the problems that face them.” Rosalind Dymond Cartwright appended to the article the comment that Lifton was measuring only the reputation for empathy. Lesser (1958) devised an Empathic Understanding Scale of 16 items such as the following: "I understand the client's feelings well"; or "I seem to see the client as he sees himself." A corresponding version of the scale allowed the client to rate the counselor. The scales were com- pleted by counselor-client pairs, and client progress in therapy was estimated by the amount of change of the self-concept toward the ideal self, measured by Q-technique. The correlations between client progress and Empathic Understanding as measured by the counselor self-rating scale or by the client rating scale were not significant and were, in fact, negative. However, as in the case of the previously described rating scales, the mean ratings by both clients and counselors were signifi- cantly beyond the midpoint of the scale. Thus, the validity of the Empathic Understanding Scale must be held in question. Neither Lifton's nor Lesser's approach to the rating of empathic ability can be adapted to yield normative data. 18 Situational tests of empathy. The most felicitous approaches for the generation of measures of empathy or sensitivity are those which stress the input component in their experimental design. These studies provide a situation or situations involving combinations of visual auditory or kinesthetic stimuli so as to provide a standard experience to elicit empathic responses from the subject. Reid and Snyder (1947) played phonographic recordings of passages from actual interviews, asking the subjects to name the feelings ex- pressed in each of 40 client statements within a 15-second time limit. They stated that approximately 80 percent of the 15 subjects agreed 50 percent of the time on the designation of feelings. The authors did not attempt, however, to define the "correct" answers or to develop an evalu- ative instrument. Astin (1947) devised a situational test of empathy by having ten client statements recorded on audiotape by a professional actor and re- quiring subjects to respond to the statement as if they were the counselor. The performance on this task of eight subjects who were counselors was significantly superior to that of the eight non-counselor subjects. Weinstein and McCandless (1959) used a more restricted visual situa- tion to derive a measure of empathy in medical students: they showed a slide of a doctor talking to a mother in a sickroom. Subjects were re- quired to choose from multiple-choice items those which best described what was happening in the picture. The empathy score was determined by the correspondence between the student's designation of the mother's feeling and the doctor's concern. The validity of this procedure is suspect, as it seems to postulate that an empathic person will project his empathic needs onto others. l9 O'Hern (1964) constructed an instrument to measure the sensitivity of trainee counselors. The instrument consisted of 30 different client problems portrayed by actors and recorded on audiotape. Thirteen indivi- duals with varied backgrounds responded to the stimulus tape. This group included full-time counselors, graduate students in counseling, and under- graduates. The refined instrument included those responses that met modal jury agreement with variance no greater than one in either direction. The four answers for each reSponse were weighted according to degrees of rightness and degrees of wrongness as determined by a panel of seven judges who had had counseling experience, adhered to various counseling philosophies, and were involved in counselor education. The final in- strument contained 29 client problems and 114 responses. A total of 212 students, enrolled in seven different Guidance and Counseling Institutes, participated in the study. O'Hern found no significant relationship be- tween degree of sensitivity possessed by potential counselors and grades received in counseling practicum courses. However, the potential coun- selors judged most effective in counseling practice scored significantly higher on the O'Hern scale than those judged least effective. An in- strument such as this would lend itself to the development of standard- ized norms if further refinement and validation studies were carried out. It is questionable whether tests based solely on the verbal text of an interview or on the audio recording alone are sufficiently complete to give all the cues and clues needed for empathic understanding. Sullivan pointed out: A verbatim record of an interview, until it has been heavily annotated, is almost invariably remarkably misleading. I have had some recordings of interviews which I have regarded as as- tonishingly good teaching material, but when I have sprung these 20 on intelligent colleagues, I have often found them barking up trees I hadn't seen--if, indeed, such trees were ever there, and I came to realize that they weren't. In other words the complete meaning of a conversation is not to be found in the verbatim verbal context of the communication, but is reflected in all sorts of subtle inter- play. For example, very slight changes of tone suggesting the faintest hint of irritation on the part of the psychiatrist often switch the patient from an attempt at concealment to a very reason- able compromise between what he thinks it is safe to tell and what the facts may have been. Such things do not appear in the most perfect verbal record. Thus, to give a third person a notion of all that happened in an interview, one would have to annotate the written record by adding the impressions that went with different statements, explaining why things were put as they were, and so on; only in this way could the richness of the interchange . . . begin to be apparent. (Sullivan, 1954, p. 19) As it becomes easier to make visual as well as auditory records, the possibility of using the total stimuli in a situation becomes more possible, and Sullivan's interpreter becomes unnecessary and undesirable. Two studies by Buchheimer (1964) and Rank (1964), as well as the present one, have moved in this direction. Buchheimer used a filmed counseling interview in which the subject free responded verbally at selected points in the film as if he were the counselor, as one phase of his total study of empathy. Buchheimer's results are not yet available. His free response situation presents obvious scoring problems, and the test requires a language laboratory or similar situation to group administer. The approach of having the subject play the role of the counselor has obvious and de- sirable face validity. It should, perhaps, be considered more a "work sample" type test in the prediction of counselor success than a measure- ment of empathy per se. This has its advantage in that work sample tests have proven to have greater predictive validity in a number of comparisons than aptitude variables. Thus measures of predictor variables 21 such as intelligence tests or special ability and aptitude tests have been repeatedly outshone by work sample tests or measures of previous related work in fields other than counseling. On the other hand, however, a predictor variable may add cases which might be missed by a work sample test. Subjects that have not yet matured or who have had sample test but may do quite well in a predictor test of related variables. A test of empathy would appear to fall in this category and thus be of value. Still closer to the approach proposed in this study, and indicative of the fact that the approach may prove fruitful, is the work of Rank (1964). Rank explored the relationship between trainee perceptions of film-presented counseling and post-training counseling competence. Test subjects responded from "strongly agree” to "strongly disagree” concerning statements about the client and the counselor presented on movie film. His "Film Test of Counselor Perceptions" produced cross-validation pre- dictive correlations of .41 on the criterion of trainee competence after a practicum. The test also measured significant increase in scores during the practicum. The test was scored with a complex weighting system. Rank plans to revalidate and improve this scale. At present it is limited to one counselor and client in a variety of scenes. Summary of Literature From this survey of the status of research in empathy it appears that there is a dearth of valid tests for the assessment of sensitivity to affective communication. Of all the possibilities, Rank's comes closest to being practical. Still, there is a need for a test which will be convenient to administer, provide normative scores, and present a sufficient number and variety of situations for the elicitation of 22 responses which demonstrate sensitivity to the affect being expressed. The use of videotaped sequences from actual counseling interviews appears to offer a highly realistic yet standardized mode of presentation of affective stimuli in items which would discriminate between those per- sons sensitive and those not sensitive to the expression of feeling in others. The variety of clients and counselors recorded in the IPR pro- cess would appear to be a rich source of material from which to create such test items. The availability of the IPR interpretation by the client seems to add a dimension beyond that found in previous work to date. CHAPTER III DESIGN OF THE INSTRUMENT AND METHODOLOGY The counseling interview protocols used in constructing the instrument were taken from a larger research project conducted in the College of Education at Michigan State University (N.D.E.A. Grant # 7-32-0410-216). The affective sensitivity scale was developed from the counseling interviews obtained by employing the IPR technique. Construction of the Instrument Selection of videotaped scenes. The scale was constructed by reviewing videotapes of counselor trainees involved in the counselor training part of the IPR study. A team of two of the principal in- vestigators and four of the graduate assistants reviewed the tapes and selected scenes which, in view of the interrogation and the con- tent itself, indicated that some describable client emotion was present. Typically these were changes in mood depicting the onset of client anger, fear, embarrassment, or avoidance behaviors. In most scenes selected there was a movement on the part of the client from one feeling state to another. The criteria were (a) that some emotion was displayed in the content and (b) that the nature of the emotion had been revealed during interrogation. In some instances the display of emotion was very subtle, and in other instances quite blatant. There was an attempt to choose scenes which ranged along the continuum of obviousness. The 23 24 purpose in selecting scenes in this manner was to differentiate between high and low empathizers. The final form contained 41 such scenes in- volving 11 different clients and counselors. The number of scenes for each client varied from two to six so that there was a variable exposure to the different clients and counselors. Both male and female clients were included. In general they were clients experiencing normal prob- lems of interpersonal conflicts, social maturity and educational planning. The scenes varied in the emotional depth of the content and the kinds of problems covered. The counselors were both male and female and varied considerably in their skill. The use of male and female coun- selors was done so that differences of affective sensitivity between both groups could be determined. Although most were beginning coun- selors, some experienced counselors were included. Several scenes of two married women as clients were included, but all other scenes were of high-school age students. Selection of criterion gropps for generatipg items. The two criterion groups were composed of subjects selected from several dif- ferent populations. The first group of subjects consisted of seven professional counselors at Michigan State University, all of whom held a doctorate in the area of guidance and counseling, had professional counseling experience and were judged to possess a high degree of empathy by other faculty members of the Guidance and Counseling Department of Michigan State University. The second group of subjects consisted of five doctoral candidates judged to possess a high degree of empathy by members of the Department of Guidance and Counseling of Michigan State University. The third group of subjects consisted of nine 25 beginning counselors judged to be high in empathy by supervisors who assisted in training each during a term of counseling practicum. These three groups of subjects (N=21) composed the criterion group of high empathizers. The criterion group of low empathizers (N=l6), judged so by pre- sent or former instructors who were asked to make such judgment due to reputation in evaluating people, was composed of six beginning counse- lors, four superintendents of schools, four high school counselors, and two undergraduates majoring in education. The composition of the two criterion groups is consolidated in the following table: Table 1 Composition of the criterion groups High Empathizers Low Empathizers Professional Counselors 7 Superintendent of Schools 4 Doctoral Students 5 Beginning Counselors 6 Beginning Counselors _2__ Secondary School Counselors 4 Undergraduates _2_. Total 21 Total 16 Generating items. The two criterion groups were shown the selected excerpts in a studio of the closed circuit television facility of Michigan State University. The videotape recording was stopped after each scene. The subjects were instructed to try to feel as the client last felt, and responses were obtained from a list of 57 adjectives expressing various feelings (Appendix A). The list of adjectives was chosen from those compiled by Osgood (1953) as representing the major affective dimensions expressed in the English language. The subjects were in- structed to go down the list of adjectives checking all the words which 26 they thought applied to the client's feelings whether expressed or not. Further, they were instructed to circle one check mark which corresponded to the word which they thought described most accurately the real feeling of the client. The same list of 57 adjectives was used for each scene. Those words which received a minimum of three checks and one circle were selected from the responses of each criterion group. Those words which were chosen most frequently by the high criterion group and not by the low criterion group and vice versa were used to create items for the affective sensitivity scale. The method of selecting the words em- ployed was by inspection. The words chosen by the criterion groups were used to construct an affective sensitivity scale of 280 items. Selection of items for the Affective Sensitivity Scale. The affective sensitivity scale (Appendix B) was administered at two univ- ersities, one located in the midwest and one in the south, to a combined total of 53 persons who were either counselor educators or master's candidates. Instructions were given the subjects to try to feel as the client last felt in each scene. The videotape recording was stopped after each scene, and the subjects were asked to respond to each ad- jective from a list provided for every scene on the following scale: 1. I have this kind of feeling strongly. 2. I have this kind of feeling somewhat. 3. I have this kind of feeling only a little. 4. I have this kind of feeling not at all. There were four counselor educators of the total 53 subjects who took the affective sensitivity scale. They were judged high in empathy by reputation. The 49 master's candidates were rated on a normal 27 distribution curve from high to low in empathy. The 14 master's candi- dates, in addition to the four counselor educators, that were rated the highest and the 18 that were rated the lowest comprised the high and low groups. These two groups were the validation groups. Table 2 gives a categorization of the 53 subjects from which the validation groups were taken: Table 2 Categorization of subjects from which validation groups were obtained University High Middle Low* Group Group Group A 9 9 9 B 9 8 9 Totals l8 l7 l8 *High and low groups refer to the top and bottom 33% of the sample based on ratings obtained from instructors. Only the high and low groups were used for validation purposes. Items that significantly differentiated between the high and low groups were being sought in the investigation. The response pattern for each item was obtained by simple addition and the median determined in the following manner: 28 Figure 1 Illustrations on determining the median An Example: Response Categories Item #15 1 2 3 4 High Group (N=18) 13 I, I 3 Low Group (N=18) 6 8 l 3 Totals 19 9 2 6 (In this example the median occurs between re3ponses l and 2.) An Example: ReSponse Categories Item #30 1 2 3 4 High Group (N=18) '6 7' E. '3 Low Group (N=18) 2 3 8 5 Totals 8 10 10 8 (In this example the median occurs between responses 2 and 3.) An Example: Response Categories Item #45 l 2 3 4 High Group (N=18) 3 8 I 6 Low Group (N=18) 2 l 1 14 Totals 5 9 2 20 (In this example the median occurs between responses 3 and 4.) The study included all items which might discriminate so long as they were consistent in keying. Thus a dual criterion was used in selecting items as validated: (a) they had a chi-square significance at the .20 level or better and (b) the direction of difference between 29 high and low empathizers was the same as in the data for the group used in selecting adjectives for the instrument. Analysis was determined by a two-by-two chi-square contingency after the median had been established. The formula used was(Siegel, 1956): 2 x = N(|AD-BC' -y_)2 2 (A+B) (C+D) (A+C) (B+D) A total of 65 items of the original 280 included in the scale met the test of significance at .20 or higher. This low level of signifi- cance was employed because of the small number of subjects and the exploratory nature of the study. Those items found to be significant are listed in Table 4 in Chapter IV (p. 35) and those which are not significant at .20 are given in Appendix C (p. 81). The frequency and significance of each item are also listed. Methodology for Cross-Validation Selection of the sample. Students enrolled in three National Defense Education Act Summer Institutes (1964) comprised the sample of 88 persons who responded to the instrument for the purpose of cross- validation. The 88 Institute enrollees were rated on a normal distri- bution curve from high to low in empathy by their instructors and practicum supervisors. The 28 enrollees rated the highest and the 28 enrollees rated the lowest comprised the high and low groups. These two groups were the cross-validation groups. Table 3 gives a categoriza- tion of the 88 enrollees from which the cross-validation groups were taken: 30 Table 3 Categorization of enrollees from which cross-validation groups were obtained University High Middle Low* Group Group Group A 10 10 10 B 10 10 10 C 8 12 8 Totals 28 32 28 *High and low groups refer to the top and bottom 33% of the sample based on ratings obtained from instructors. Only the high and low groups were used for cross-validation purposes. Procedures for obtaining the data. A kinescope film of the original videotape was made and the film was shown to the Institute enrollees. An exception was made in the case of one group which viewed the video- tape recording. A member of the IPR research project staff went to each Institute and gave instructions for taking the instrument and answered questions relating to the entire procedure. The director of each Institute was requested to have his staff rate enrollees for counseling effectiveness (Appendix D) and empathy (Appendix E). Data obtained from the 88 Institute enrollees and the ratings from the directors were analyzed in the following manner: 1. To cross-validate the results obtained on the 65 significant items, listed in Table 4, data received from the top and bottom 33% (N=56) of the sample (N=88) in the three Institutes were selected for item analysis purposes. The results are contained in Chapter IV in Table 5 (p. 41). 31 2. To determine the correlation between scores received on the affective sensitivity scale and counseling effectiveness, it was planned to obtain a Pearson Product-Moment correlation from data received from all enrollees of the three Institutes. However, this was inappropriate since too few items were sig- nificant under the procedures of step 1. Validity An item analysis was conducted to determine those items that sig- nificantly differentiated between those persons judged high and low in empathy. The item analysis was conducted on those items that signifi- cantly differentiated in order to determine the validity of the instrument. However, an item analysis that employs a two-by-two chi-square contingency considers each item independently of all other items. There is a possi- bility that items of a test considered separately will not produce enough significant items to prove the test to be valid, but a combination of a part or all of the items will. A total score was obtained for the instrument as each item was scored for ”right" and "wrong” answers. Thus the total scale could be assigned correct and incorrect answers on the basis of the responses given by the criterion groups of high and low empathizers. A score of l_was given for the correct response and Q for an incorrect response. The sample was increased from 88 to 148 by adding data obtained from two additional National Defense Education Act Summer Institutes (1964). The increased sample and the scoring of the 172 items permitted an analysis of variance to determine the predictive validity of the in- S trument . 32 Reliability The data from the scored instrument was used to determine a correlation of reliability by employing the Kuder-Richardson formula 20. Further, a t-test was utilized to test for significance among the sub-groups of the five Institutes. Factor Analysis A ten-factor matrix was employed to ascertain commonality of meaning in adjectives used in the 172 items. Summary The videotape scenes used in the instrument were selected by a panel of six counselors. Two criterion groups, judged high and low in empathic regard, responded to the instrument from a list of 57 adjectives representing major affective dimensions. Their choices of adjectives were used to comprise the affective sensitivity scale of 280 items. The instrument was administered to 53 subjects in two universities. A total of 65 items of the original 280 included, met the test of signifi- cance at .20 or higher. In order to cross-validate the results of the item analysis of the instrument, another 88 persons responded to the instrument. These sub- jects were enrolled in three National Defense Education Act Summer Institutes (1964). The instrument was scored for "right" and "wrong" answers. A value of l was given for the correct response and Q for an incorrect response on 172 scoreable items. The total number of subjects taking the test 33 was increased from 88 to 148 by administering the test to two additional Institutes. The purpose of scoring the instrument and increasing the sample was to determine the predictive validity and reliability of the instrument. In addition, a t-test for significance among the sub-groups L/ of the sample and a factor analysis to determine commonality of meaning in adjectives used in the 172 items were made possible by scoring the instrument. The desired data from the subject's responses were placed on IBM cards for statistical analysis. Analysis of variance, t-test, chi-square, Kuder-Richardson formula 20 and factor analysis were used in the statistical treatment of the data. CHAPTER IV ANALYSIS OF DATA This chapter contains the analysis of the results obtained in the steps to validate, cross-validate, factor analyze and determine the predictive validity and reliability of the affective sensitivity scale. Results of Validating the Instrument The median response for each item on the four-point scale response was determined, and the scale dichotomized at the combined median for both high and low empathy groups so that a two-by-two chi-square analysis could be performed. The median was at the middle of the scale for 39 items, between the first and second responses for 15 items and between the third and fourth responses for 11 items. The following table contains the results of the 65 items which were significant at .20 or higher by employing a two-by-two chi-square contingency after determining the point at which the median occurred in the response pattern (chi-squares of 215 items which were not significant are given in Appendix C (p. 81): 34 35 Table 4 Frequency distribution and significance of items on the affective sensitivity scale which failed to meet the .20 level of signifi- cance in validating the instrument Part A (Median occurs between respgnses 2 and 3.) Client Item Selected Selected and and Group2 ReSponse Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % I 3 High 33.333 16.67 4.000 1 Lonely Low 16.67 33.33 I 18 High 27.78 22.22 1.800 2 Ambivalent Low 16.67 33.33 I 19 High 36.11 13.89 4.050 3 Hopeful Low 19.44 30.56 I 21 High 25.00 25.00 1.870 3 Refreshed Low 13.89 36.11 I 33 High 11.11 38.89 2.000 4 Ambivalent Low 22.22 27.78 I 35 High 22.22 27.78 2.000 4 Embarrassed Low 11.11 38.89 I 38 High 38.89 11.11 4.028 5 Resistant Low 22.22 27.78 II 52 High 38.89 11.11 2.000 2 Surprised Low 27.78 22.22 II 54 High 30.56 19.44 2.786 2 Anxious Low 16.67 33.33 1A chi-square value of 1.642 or higher was needed for the item to be significant at .20. 2Group rating refers to the top or bottom 33% of the sample based on ratings obtained from instructors and supervisors. 3The figure given in each cell is the percent of total response (N=36). Table 4 (Continued) 36 Part A Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % II 59 High 30.56 19.44 1.778 3 Anxious Low 19.44 30.56 II 60 High 33.33 16.67 1.800 3 Guilty Low 22.22 27.78 III 69 High 47.22 2.78 2.090 1 Controlled Low 38.89 11.11 III 77 High 38.89 11.11 4.208 2 Bored Low 22.22 27.78 IV 90 High 16.67 33.33 1.800 2 Capable Low 27.78 22.22 IV 94 High 38.89 11.11 4.208 2 Annoyed-Angry Low 22.22 27.78 IV 95 High 38.89 11.11 5.600 2 Abused Low 19.44 30.56 IV 100 High 27.78 22.22 4.985 3 Talkative Low 44.44 5.56 V 111 High 20.59 32.35 4.480 1 Disgusted- Low 35.29 11.76 Fed-up V 113 High 20.59 32.35 4.480 1 Upset- Low 35.29 11.76 Disturbed V 116 High 23.53 29.41 3.265 1 Bitter Low 35.29 11.76 V 117 High 47.06 5.88 3.278 1 Cautious Low 29.41 17.65 V 135 High 5.88 47.06 2.101 4 Erratic- Low 14.71 32.35 Disorganized 37 Table 4 (Continued) Part A Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % VI 167 High 48.48 6.06 2.418 4 Happy Low 30.30 15.15 VI 177 High 36.36 18.18 5.241 5 Disgusted- Low 12.12 33.33 Fed-up VII 179 High 27.27 27.27 3.182 1 Comfortable- Low 36.36 9.09 At ease VII 180 High 6.06 48.48 5.215 1 Regretful Low 21.21 24.24 VII 185 High 27.27 27.27 1.866 2 Abused Low 12.12 33.33 VII 191 High 15.15 39.39 2.238 2 Scared Low 24.24 21.21 VIII 194 High 48.48 6.06 2.418 1 Embarrassed Low 30.30 15.15 VIII 199 High 27.27 27.27 1.866 2 Frustrated Low 33.33 12.12 VIII 203 High 42.42 12.12 2.200 2 Foolish- Low 24.24 21.21 Riciculous IX 223 High 39.39 15.15 4.911 4 Defeated Low 45.45 .00 X 243 High 42.42 12.12 4.891 3 Worried Low 18.18 27.27 X 247 High 27.27 27.27 1.866 3 Bitter Low 12.12 33.33 X1 .265 High 18.18 36.36 1.782 2 Helpless Low 6.06 39.39 38 Table 4 (Continued) Part A Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % XI 271 High 45.45 9.09 3.487 3 Courageous- Low 24.24 21.21 Daring X1 272 High 18.18 36.36 2.347 3 Stubborn- Low 27.27 18.18 Resistant X1 278 High 39.39 15.15 2.451 4 Sensible- Low 42.42 3.03 Reasonable Part B (Median occurs between responses 1 and 2;) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square % % I 26 High 47.22 2.78 3.200 3 Controlled Low 36.11 13.89 I 30 High 38.89 11.11 2.090 4 Controlled Low 47.22 2.78 I 40 High 47.22 2.78 3.200 5 Reluctant Low 36.11 13.89 11 46 High 47.22 2.78 2.090 1 Anxious Low 38.89 11.11 III 71 High 50.00 0.00 3.273 1 Confused- Low 41.67 8.33 Mixed-up IV 82 High 38.89 11.11 2.000 1 Cautious Low 27.78 22.22 IV 99 High 44.44 5.56 2.118 3 Discouraged Low 50.00 0.00 IV 108 High 50.00 0.00 2.118 4 Controlled Low 44.44 5.56 39 Table 4 (Continued) Part B Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square % % V 123 High 47.06 5.88 1.889 2 Anxious Low 47.06 0.00 V 133 High 29.41 23.53 2.555 4 Proud Low 38.24 8.82 VI 173 High 39.39 15.15 2.451 5 Disappointed Low 42.42 3.03 VIII 192 High 54.55 0.00 2.555 1 Confused- Low 39.39 6.06 Mixed-up IX 214 High 54.55 0.00 2.555 2 Optimistic Low 39.39 6.06 IX 221 High 42.42 12.12 3.793 3 Bitter Low 45.45 0.00 X 232 High 42.42 12.12 3.793 1 Discouraged Low 45.45 0.00 Part C (Median occurs between responses 3 and 4.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square _1 % % I 23 High 0.00 50.00 2.118 3 Trapped Low 5.56 44.44 I 24 High 5.56 44.44 2.118 3 Annoyed-Angry Low 0.00 50.00 II 49 High 2.78 47.22 2.090 2 Bitter Low 11.11 38.89 III 76 High 2.78 47.22 2.090 2 Confused- Low 11.11 38.89 Mixed-up 40 Table 4 (Continued) Part C Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square % % IV 86 High 8.33 41.67 3.273 1 Forced Low 0.00 50.00 IV 88 High 5.56 44.44 2.118 1 Embarrassed Low 0.00 50.00 IV 106 High 19.44 30.56 2.215 4 Careful Low 8.33 41.67 VII 183 High 6.06 48.48 1.774 1 Lonely Low 0.00 45.45 X 251 High 0.00 54.55 2.555 4 Bitter Low 6.06 39.39 X1 268 High 6.06 , 48.48 1.774 2 Satisfied Low 0.00 45.45 XI 276 High 15.15 39.39 2.451 4, Ambivalent Low 3.03 42.42 Results of Cross-Validating the Instrument The median was at the middle of the scale for five items, between the first and second responses for one item and between the third and fourth responses for three items. The following table contains the results of the cross-validation analysis of the 65 items. Of the 65, nine items were found to be significant at .20 or higher by employing a two-by-two chi-square contingency (chi-squares of 56 items which were not significant are given in Appendix F (p.101 ): ’l 41 Table 5 Frequency distribution and significance of items of the affective sensitivity scale which were significant at .20 or higher obtained in cross-validating the instrument Part A ,(Median occurs between responses 2 and 3.); Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % v 116 High 17.863 32.14 5.793 1 Bitter Low 33.93 16.07 VI 177 High 41.07 8.93 4.139 5 Disgusted- Low 28.57 21.43 Fed-up VII 191 High 5.36 44.64 10.220 2 Scared Low 25.00 25.00 X 247 High 32.14 17.86 2.585 3 Bitter Low 21.43 28.57 X1 265 High 32.14 17.86 12.469 2 Helpless Low 8.93 41.07 Part B (Median occurs between responses 1 and 2.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square ' % % III 71 High 10.71 39.29 2.947 1 Confused- Low 21.43 28.57 Mixed-up 1A chi-square value of 1.642 or be significant at .20. 2Group rating refers to the tOp on ratings obtained from instructors 3The figure given in each cell (N=56). higher was needed for the item to or bottom 33% of the sample based and supervisors. is the percent of total responses 42 Table 5 (Continued) Part C (Median occurs between responses 3 and 4.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square % % VII 183 High 35.71 14.29 1.697 1 Lonely Low 42.86 7.14 X 251 High 28.57 21.43 7.376 Bitter Low 44.64 5.36 XI 268 High 46.43 3.57 4.383 Satisfied Low 35.71 14.29 Predictive Validity Results An examination of Tables 4 and 5 indicates the instrument does not discriminate consistently across samples of high and low empathic coun- selors when each item is considered independently of all other items. In order to consider the data from a different perspective a total score was obtained for the instrument. Those items that discriminated between high and low counselor criterion groups were scored on a scale 0-1. A score of l was given for the correct response and Q for an incorrect re- sponse. A total of 172 items from the original 280 were scoreable, in that the high counselor criterion group selected the item and the low criterion group did not or vice versa. The items that were not scored were equally attractive to both criterion groups and would not discrimi- nate between the two. The scored instrument is given in Appendix G (p. 109). Once a total score was obtained, an analysis of variance was employed to determine the predictive validity of the instrument. If the instru- ment discriminated between persons judged high and low in empathy, then one could state that it possessed predictive validity. 43 The summary data on the 172 scored items not only considered high and low groupings but male and female as well. This was done to determine whether or not there were significant differences between sexes. These data are contained in Table 6: Table 6 Summary data of scored instrument Type or Male (3) Female (4) Totals Group K 'X K 'X K 'X High (1) 35 92.571 9 96.778 44 93.432 Low (2) 33 90.575 11 91.818 44 90.886 Totals 68 91.603 20 94.050 88 92.159 The differences in means between high-low and male-female groupings were: D1,2 = XH ' xL Dl,2 = 2.555 D3,4 = XM ' XF D3,4 — 2.457 where: D = difference 1 = high; 2 = low; 3 = male; 4 = female The analysis of variance operates on the assumption of equal N's among the different sub-groups. This assumption is not met in the data presented in Table 6. Therefore, an adjustment term (Wert, Neidt, and Ahmann, 1954) was employed to correct for disproportionality among the 44 sub-groups. The formula for the adjustment term is: )2 - 2(01’2) (03,4) (ad-bc) (ad-bc)2 [E§;) (K2) (D1 2)2 + (K3) (K4) (D3,4 (K1K2K3K4> ' N 1 - (ad-bc)2 (K1K2K3K4) The alphabetical symbols of Figure 2 correspond to the numerical values of Table 6: Figure 2 Table of symbols Type Male (3) Female (4) Totals or _ _ _ Group K X K X K X High (1) 3 p K1 Low (2) lg d K2 TOtals K3 K4 The following sum of squares and their formulae (Wert, Neidt and Ahmann, 1954) were employed in computing the analysis of variance: 1) Sum of squares for total ssT = 8x2 - ggxzz N 2) Sum of squares for sex 353ex= (Exrp2 + (Exf)2 _ (EX)2 Km “1§?“‘ “Tf“‘ 3) Sum of squares for high-low 2 SSH-L = (exH) + (Exp2 _ (€X)2 KH T N with 45 4) Sum of squares for interaction ssI = (EXHM)2 + (éxHF)2 + (€XLM)2 + (Emu?)2 _ (8)02 KHM KHF KLM KLF N 5) Sum of squares for within ssw = $3 (ssSex + ssHL + 331) T The material in Table 7 contains the data for the analysis of variance means adjusted for disproportionality by sex, high-low groupings and interaction of all possible combinations: Table 7 Analysis of variance of instrument scores with means adjusted for disproportionality Source Degrees of of Sum of Squares Mean Variation Freedom Unadjusted Adjusted Square (N-l) Sex 1 92.543 80.640 80.640 High-low 1 142.533 130.730 130.730 Interaction 1 171,475.435 171,487.238 171,487.238 Within 87 -l64,910.738 - 1,895.526 Totals 90 6,799.773 Adjustment Term = 11.803* For Sex: F1 87 = 80.640 = - .0043** ’ -1,895.526 For High-Low: F1 87 = 130.730 = - .0690** ’ -1,895.526 For Interaction: F1 87 = 171,487.238 = 90.469** ’ - 1,895.526 *The adjustment term is added algebraically to main effects (sex and high-low) and subtracted from minor effect (interaction). **An analysis of variance value of F(.05) = 253 was needed for the in- strument to discriminate on a source of variation. 46 Reliability Results The scoring of the instrument permitted a correlation of reliability to be determined. The Kuder-Richardson formula 20 for reliability was .441. The correlation of reliability obtained (.441) is high enough to indicate the instrument possesses average consistency of measurement when compared with other tests of personality variables. (This would not be the case where tests of aptitude and achievement are involved since correlations of reliability are usually higher in these two areas.) Although the reliability of the instrument is acceptable, the instrument does not possess the ability to differentiate between persons judged high and low in empathy other than that produced by chance. A score obtained in using the instrument could not be relied upon since it is not known whether or not affective sensitivity is measured. A t-test was employed to determine any significant difference be- tween the highest mean (104.7) and the lowest mean (82.) of all the NDEA Institutes sub-groups. These data were the same used for purposes of determining the validity and reliability of the instrument. This was done in order to answer the question, "are we dealing with any combina- tion of effects that is significant." The data in Table 8 contains the means of each NDEA Institute according to high-low and male-female groupings. There was no significant difference between the highest mean of 104.7 (Institute D, high-female) and the lowest of 82 (Institute B, low-female).1 Since the difference between the highest and lowest means did not produce significance, it could be concluded that all other possible combinations of means would also fail to produce significance. It (°{= .05) = 2.447 had to be reached to obtain significance. The difference produced a t score of only .013. 47 Table 8 NDEA Institute means High Low Institute Male Female Male Female K X K X K X K X A 9 87.6 1 99 8 85.3 2 90.5 B 10 92.9 0 -- 9 92 l 82 C 6 98 2 86 5 5 89 3 90 D 5 96.8 3 104.7 5 93.8 3 96 E 5 90.2 3 95 6 94.2 2 94.5 Totals 35 92.6 9 96.3 33 90.6 11 91.8 Factor Analysis Results The data for the 172 items, which were scoreable from the original 280, were factor analyzed. The purpose of employing factor analysis to the data was to determine commonality of meaning in adjectives used in the items for a maximum of ten factors. A factor had to possess a value of .450 or higher to meet a level of significance of .05. The selecting of factors for each item was as follows: Figure 3 Illustration on determining factors An Example: Item Number 1 2 3 4 5 6 4 .649 .115 .106 .006 .029 .065 9 .033 .071 .023 .144 .187 .083 .0 7 8 9 10 65 .004 .049 .005 .344 .312 .209 .174 (In the example given, factor 1 of item 4 was selected as the factor for that item. There was no factor that was significant in item 9.) 48 A particular factor was considered operative when at least eight items had values of .450 or higher from the 172 items. The cut-off point of eight items was determined by N—2, where N equals total number of factors or a total of ten in this factor analysis. There were five factors operative in the data analyzed, and these are given in Table 9. As a final step in factor analysis, all scenes that contained an item or items that were operative in a particular factor were considered for commonality of meaning. This was accomplished by viewing those scenes containing operative items and determining the common thread of feeling or thought contained in the combined group of scores. The descriptive statements of feeling or thought are found in the next section under "Analysis" (p.51 ). Analysis Validating and cross-validatingkthe instrument. The two-by-two chi-square analysis was performed on each item independently of all other items. This was done to determine those items that were empiri- cally significant. Then, all significant items would have been used to produce a valid instrument. However, an analysis of Tables 4 and 5 re- veals the instrument cannot be relied upon as being valid since the number of significant items obtained in cross-validation was no better than chance. Predictive validity. The analysis of variance values obtained for sex grouping, high-low grouping and interaction were not significant. These three values were obtained to discover whether one or more of these effects may be significant. If only one had proven to be significant, Table 9 Adjectives in operative factors FACTOR 1 FACTOR 2 FACTOR 4 FACTOR 6 FACTOR 7 Client Item Client Item Client Item Client Item Client Item and and and and and and and and and and Scene Adjective Scene Adjective Scene Adjective Scene Adjective Sceng, Adjective I 4 II 46 I 19 I 23 II 43 1 Depressed l *Anxious 3 Hopeful 3 *Trapped l *Baffled I 6 II 54 I 20 I 32 II 56 1 Discouraged 2 *Anxious 3 Optimistic 4 *Baffled 2 *Frustrated III 67 IV 82 I 22 I 33 III 72 1 Anxious l *Cautious 3 *Relieved 4 *Ambivalent l *Listless- III 68 IV 84 II 58 VI 169 Indifferent 1 Worried 1 *Reluctant 3 *Daydreaming 4 Baffled III 74 III 71 IV 85 V 121 VI 170 2 *Listless- 1 *Confused- 1 Anxious 2 Annoyed- 4 *Confused- Indifferent Mixed-up V 139 Angry Mixed-up III 77 IV 99 5 *Bitter VI 157 VIII 194 2 *Bored 3 *Discouraged IX 226 2 *Satisfied l *Embarrassed III 79 IV 100 4 *Annoyed- VI 158 VIII 201 2 Reluctant 3 Talkative Angry 2 Appreciated 2 *Embarrassed VII 180 IV 105 X 242 VI 160 XI 258 l *Anxious 4 Defeated 3 Depressed 2 *Amused 1 *Embarrassed IX 219 V 113 VI 171 3 Amused 1 Upset- 4 *Helpful IX 221 Disturbed VIII 197 3 Bitter V 124 l *Discouraged IX 224 2 Lonely X 242 4 Bitter V 126 3 *Depressed IX 225 3 Depressed 4 Hopeless V 127 3 *Confused- Mixed-up V 135 4 Erratic- Disorganized V 136 4 Gailty V 140 5 *Discouraged V 142 5 Depressed V 150 6 Disgusted VIII 200 2 Impatient IX 207 1 Cautious IX 227 4 Baffled X 233 1 Baffled X 241 2 Confused- Mixed-up X 246 3 Confused- Mixed-up XI 259 1 Forced *The Opposite in meaning should be given since the value was negative. 50 this would have been evidence enough to continue the investigation in the direction of the significant effect. The values for the three effects were far enough below significance to indicate that had the sample been increased to infinity the values needed for significance could not have been reached. Reliability. The correlation of reliability obtained (.441) has little significance since the instrument does not possess predictive validity. There would always be the question of exactly what variable was being measured by the instrument. The t-test score obtained, when comparison was made in the difference between the group possessing the highest mean and the group with the lowest, indicated the groups were homogeneously grouped in ability on the variable measured. The lowest mean was five points below chance and the highest mean 18 points above chance. A chance score was 86, which implies that the subjects re- sponding to the instrument were average in their ability on the variable measured by the scale. The logical interpretation to come from the t- test score is that it is difficult to know where the items chosen lie on an absolute scale of empathy. Factor analysis. The five factors that were operative went from a high of 24 words under factor 1 to 11 words in factors 4 and 7 to eight words in factors 2 and 6. There were 29 different scenes in the 62 items found under the five factors. Each operative factor was con- sidered for commonality of meaning by considering all scenes found in a particular factor. The following statements give the common idea for all scenes under each factor as observed by the researcher: 51 Factor 1 - There was a generalized distrust of the future. (The clients were anxious regarding their plans for the future.) Factor 2 - There was a need for relatedness to others and self- acceptance. (The clients felt a need to force themselves upon other people.) Factor 4 - There was a positive personal regard to the extent of being in agreement with one's self and self-potential. (The clients felt they were accepted and needed by others which enhanced positive feelings toward self.) Factor 6 - There was a concept of self-certainty. (The clients felt sure of themselves to the point of being decisive.) Factor 7 - There was a free-floating reaction that bordered on indifference. (The clients simply wanted ”to live and let live.") The statements describing each factor are independent of each other and in some cases opposite in meaning. This is to be expected since the descriptive adjectives used in the instrument were representative of the major affective dimensions expressed in the English language. Summary The results of the data obtained in validating and cross-validating failed to demonstrate that the affective sensitivity scale could reliably discriminate between persons judged high and low in empathy. The pre- dictive validity of the instrument thus could not be demonstrated and the reliability was average. There was no commonality of meaning in the five factors considered operative in the factor analysis. However, each factor did possess a common idea when considered independently of each other. CHAPTER V SUMMARY, CONCLUSIONS AND DISCUSSION Summary The purpose of this study was to develop and validate an instrument that tested a subject's ability to identify the emotion or emotions expressed by another in a series of videotaped excerpts from actual counseling interviews. The instrument was composed of selected scenes of videotaped re- cordings of counseling interviews. Each scene was followed by four to seven descriptive adjectives. Subjects were to respond to each adjec- tive on a continuum indicating how strongly they possessed the feeling described by the adjective at the end of each scene. Two criterion groups, judged high and low in empathic regard, responded to the instru- ment and generated items from a list of 57 adjectives expressing various feelings. The final list of four to seven descriptive adjectives was produced from the list of 57 adjectives. The instrument was administered to 53 subjects in two universities for the purpose of validation. In order to cross-validate the results of the item analysis of the instrument, it was administered to another 88 persons enrolled in three National Defense Education Act Summer Institutes (1964). The instrument was scored for "right" and "wrong" answers on 172 scoreable items from the original 280. Correct and incorrect answers 52 53 were determined on the basis of the responses given by the criterion groups of high and low empathizers. The total sample was increased from 88 to 148 by adding the data from two additional National Defense Education Act Summer Institutes (1964). The purpose of scoring the instrument and increasing the sample was to determine the predictive validity and reliability of the instrument. In addition, a t-test for significance among the sub-groups of the sample and a factor analysis to determine commonality of meaning in adjectives used were made possible by scoring the instrument. The desired data from the subject's responses were placed on IBM cards for statistical analysis. Chi-square was used in the statistical treatment of the data for validation and cross-validation purposes with the test of significance being established at .20 or higher. The two- by-two chi-square contingency was employed on the response scale. The response scale was split at the point at which the median occurred in the response pattern of each item. Analysis of variance to ascertain predictive validity; t-test to determine significant differences among means of the sample sub-groups; Kuder-Richardson formula 20 to establish reliability and factor analysis for commonality of meaning were also used in working with the data. Findings. A total of 65 items of the original 280 included met the test of significance for the purpose of validation, and nine items were found to be significant in cross-validating the instrument. These nine items which held up on cross-validation included the adjectives "bitter" (three of the items), "disgusted," ”scared," "helpless," "confused," "mixed-up,' ”lonely,' and "satisfied." The large number of 54 negative adjectives results from the fact that the adjectives in general were descriptive of negative affect. There was no major concentration on any one client or episode although nearly all of the items were toward the end of the scale. Because the nine items represent about a fifth of the 65, and one would expect about a fifth of the items to show signi- ficance by chance at the 20% level, it would appear that there is nothing more than chance relationship between the item responses and the criterion. The predictive validity of the instrument was nonexistent. The analysis of variance values for three effects (sex grouping, high-low grouping and interaction) were low enough to indicate that should the sample be increased to infinity significance would not be produced. The correlation of reliability was .441. The t-test did not find a significant difference between the highest and lowest means of the two sub-groups involved. There were five factors operative in a ten-factor matrix. A total of 29 scenes and 62 items were found in the five factors. Conclusions The hypotheses tested in this investigation were: 1. A scale based upon actual counseling interviews will differen- tiate between persons judged high and low in empathic regard. 2. A scale based upon actual counseling interviews will produce a significant correlation between scores received on the affec- tive sensitivity scale and counselor effectiveness. The affective sensitivity scale did not differentiate between persons judged high and low in empathic regard as stated in the first hypothesis; hence the first hypothesis is rejected. Since the affective sensitivity scale proved to be invalid, it was inappropriate to test the second hypothesis. 55 Discussion The affective sensitivity scale, as developed in this study, did not differentiate between persons judged high and low in empathic regard. This resulted in rejecting one major hypothesis which indicated that the scale would differentiate and as a result a second one could not be tested. It is true that variations in methodology (see section on Implications below) might yield different data; however, the methodology used appeared appropriate. Further, unlike so many past "empathy" studies, the use of videotape and insights provided through recall sessions, made possible the creation of an instrument and instructions for its use which seemed to come closer to a congruence of theoretical and operational definition than ever before. Operationally, the subject was asked only to "feel ' He was not asked to predict future client behavior or as the otherf to determine the counselor's next best statement--both typical of other studies and both subject to the influence of knowledge and training in the subject of counseling or personality theory. These characteristics of the scale place it in the category of situational tests of empathy which is described in Chapter II. These studies provide situations involving combinations of visual, auditory or kinesthetic stimuli so as to provide a standard experience to elicit empathic responses from the subject (Reid and Snyder, 1947; Astin, 1957; O'Hern 1964; and Rank, 1964). The contradictory meanings and vague usages of the term "empathy" were described in Chapter II by presenting definitions of this variable as found in the professional literature. The definition of affective sensitivity, as used in this study, has more compatibility with the 56 German term, Einfuhlung, and the definition of empathy stated by Katz (1963) than with those espoused by English and English (1948) and Fenichel (1945). The former definitions indicate that detecting the momentary affective state of another is essential rather than the detachment and objectivity of the empathizer, as held by the latter. The question must be asked: what contributed to negative results? One possibility is that adjectives, without qualifying phrases, have different meanings for different peOple. Thus individuals have similar feelings in empathizing but may be reSponding to the adjectives in dif- ferent ways. It is possible that the use of phrases might be more specific and therefore might give more consistency of interpretation than the ad- jectives alone. This point is supported by Katz (1963, p. 181): The empathizer is not concerned merely with feelings labeled "sibling rivalry," ”loneliness," "fear of authority,” or "ambiva- lence." He enters into a specific situation and imaginatively places himself in the moment, in time and space when his client responded spontaneously to an event and an object. He empathizes with the person in the situation and with the meanings, values and feelings spontaneously projected in that specific context. A second possibility is that the level of interpretation varied with the experience of the counselor. Thus, what was most indicative for coun- selors at one level might not have been at another. For instance, a sub- ject seeing a client overtly express fright might initially feel fright when he tried to feel as the person did. On closer examination, he might realize that actually the fright is fear of being angry and that under- lying this is anger at the counselor which is being suppressed. In a real sense then, at the deepest level of interpretation, the most appro- priate response might be anger. Thus the counselor who reads much deeper meanings into the scene may check different responses than the less 57 experienced counselors. If this is the case, one will have to define better the level of insight requested. This could enhance considerably the method of using videotaped recordings as compared to tests based solely on the verbal text of an interview or on audio recordings. The weakness of the latter methods is described by Sullivan (1954) in Chapter II. It is possible that the experience of the counselor can be a hin- drance to his ascertaining the feeling of the client. Some counselors indicate some of their best empathic endeavors occurred when they were just starting in the profession. The fixed routines and traditions of the profession may decrease the empathic ability of the counselor. The need to experiment or become involved with a person is not so great once a counselor is established. The counselor who possesses a high degree of empathy is willing to explore new and unprotected areas. It is also possible that as the individual checked down the list, the original feeling that the individual had at the time of the end of the scene may have been lost. Other feelings suggested by the adjectives may have modified the original feeling. If this were true, items at the beginning of the list should cross-validate more readily. An examination of the nine items which did cross-validate, however, indicated that the next to the last or lower half of the list in seven out of the nine in- stances were the items which held up. It would appear, then, that this is not a valid hypothesis. On the other hand, because the cross-validation indicates only chance results, a combination of several factors may have masked such results. 58 The point is made by Katz (1963) that social psychologists view empathy as adaptive rather than reactive and as a skill acquired in the process of general social interaction. The process is a rehearsal for realistic and appropriate action where it is not the other person one wishes to emulate but the other person's ideas and image of one's self that is desired to be understood. When empathy is viewed in this manner, the results obtained in validating and cross-validating are not surprising for two reasons. First, the subjects were instructed to assume a psy- chological "set" of viewing the client and respond or react as the client responded in the interview. The process of interaction possibly could have been accomplished by requiring the subjects to reply verbally to the client and record the reply on an audio tape recorder. Second the subjects focused their attention upon the client. Perhaps, if Katz is right, they should have been required to ascertain the ideas and the image of the counselor. There is a possibility that using most of the words many times in the instrument created an attitude of redundancy on the part of the sub- jects. This in turn could lead most of the subjects to become bored or flippant in the testing situation. A form containing 280 items is an extremely long test. It often took nearly two hours to administer the test (particularly when mechani- cal adjustments were needed). Fatigue certainly could have been a factor. The criterion problem is unavoidable. Although the counselor educa- tors ranked their students on a normal distribution and each had instruc- tions on rating procedures, it is not known whether the different universities were using similar criteria for ranking or how valid each 59 university's ranking was. Two of the professors pointed out that although they knew their students well, empathy was so elusive a quality for them they found themselves revising their rankings several times before sub- mitting them. "The only valid criterion," one professor stated, "will be their--the counselors--1ong term influence of the clients they counsel." This was beyond the scope of this study. The suggestion has been made that people do not differ significantly in their affective sensitivity, that this quality is distributed over a very narrow range. This idea stems from the apparent insight on re- call of individuals who might otherwise have been thought to be socially insensitive. A number of such individuals indicate an awareness of what is going on, but seem to be incapable of action based upon this awareness. Frequently they are able to verbalize their frustration at this gap. Such persons might be able to perform well on a non-threatening test of such behavior even though they could not act on it. Viewed in this fashion, effectiveness in counseling would result, not so much from a difference in perception, as in what the counselor does based on this perception. The effective counselor enters the phenomenological field of the client and then confronts the client or deals with the feelings in other ways appropriate to his theoretical base. The ineffective counselor, on the other hand, equally aware of the client's communication, avoids confrontation, changes the topic, or perhaps "uses" the client's feeling to satisfy his own needs. It is possible that the recognition of affective communication is basic to survival within a given society and is learned well quite early. The child suffers pain who is unable to recognize the meaning "behind" 60 father's words or unable to understand the meaning of the clenched teeth of his larger peers. Any teacher has experienced situations in which a classroom of 30 or more children will immediately come to order as he or she enters the room, thus making unnecessary the speech she was going to make about their poor conduct on previous days--the "speech" apparently having been already communicated by her posture, gait and expression on walking from the classroom door to her desk. Would the theater and motion picture industries have been so successful finan- cially if audiences differed significantly in affective sensitivity? But people do appear to differ in this quality. How could one "clumsiness" and apparent lack of sensitivity? explain the social boor's Where the behavior results in annoyance and irritation with the behavior, could one as readily attribute the cause to a need for this manner of attention rather than social insensitivity? It is not so much that the individual does not perceive accurately according to this point of view, as it is that he does not know what to do about it, or he does not behave in socially acceptable ways. It is interesting to Speculate that if research shows this view to be valid, then texts and training procedures in counselor as well as teacher education would require extensive revision. Students in these and other areas would need less to "learn" the meaning of the communication of others and more to overcome effectively their own need to avoid dealing with the communication. Studies conducted in the area of social suggestion may give another reason why negative results were obtained. In a given group each in- dividual receives and gives stimuli which condition the response pattern 61 for the entire group. Despite the instructions and frequent reminders, many individuals began rating before the end of the scene and so did not react to the last feeling in the scene. This at times meant they would laugh at their response or give a sign of boredom. This situation would have been corrected by having each subject respond in private to the instrument. However, the instrument was designed to be given in a group testing situation. To administer it individually was beyond the limits of this study. The criteria under which the enrollees of each institute were selected were fairly homogeneous since minimum requirements for accep- tance at an Institute were established by the U.S. Office of Education. Also, most counselor educators have some similar expectation for persons entering the field of counseling. This could have contributed to the means of the sub-groups being rather close to each other. The range of difference in empathic regard from institute to institute or within an institute may not be wide due to the factors of selectivity and ex- pectations held in common by counselor educators. The fact that "face validity" appeared to be operationally high and predictive validity empirically low or nonexistent is difficult to interpret. This may have been otherwise had a panel of judges been asked to view each scene and had the researcher used only those items where agreement was reached as to the correct response for the four- point scale. This would have meant: (a) each item used in the in- strument would have an agreed upon correct response; (b) the score obtained by each subject would have determined high and low empathic groups and (c) the item could have been weighed on a four-point scale. 62 The scoring procedure that was employed had to be on a two-point scale since only direction of response was given by the criterion groups in- stead of both direction and degree of response. Implications for further research. At this juncture the criterion problem, the similarity of the subject's reactions to the scale, and the necessity of getting subjects to react to the same portion of the scene are points of departure for additional study. The following recommen- dations are suggested: 1. The present scale or the proposed revisions of it should be administered to full-year rather than to summer Counseling Institutes so that counselor criterion ranking can be based on more extensive contact with the counselor educators making the ranking. The IPR interrogations responses themselves, along with alter- natives, could be used as items instead of simple adjectives. An accuracy score could thus be determined without the use of judges, and this score studied in relation to other criteria e.g. grades, ranking by practicum supervisors, client rating of counselor. Replicate the study but utilize professional counselors as the only criterion group to rate the feeling(s) being exhibited by the client and to indicate the correct response. Conduct another study where subjects respond to descriptive statements instead of a list of descriptive words or respond verbally to each episode. Conduct another study where subjects respond in a private setting instead of a group setting. APPENDIX A LIST OF FIFTY-SEVEN ADJECTIVES 64 INSTRUCTIONS In the next two hours you will be viewing some kinescopes of actual counseling sessions. We are trying to identify what feelings were being expressed by the clients in these interviews. We are concerned about what observers would judge to be the actual feeling. On the next page you will find a list of 57 adjectives expressing various feelings. Parallel to these words are spaces which you can check for episodes shown on the TV monitor. For example, all the spaces under Arabic numeral (1) will refer to the first episode of Client 1. After you have seen counseling episode (2) go down the adjective list checking all those words which you feel apply to the client's feelings whether expressed or not. Then, circle only pp; check mark which corres- ponds to pp; word which you think describes most accurately the feeling of the client. In other words we would like you to put yourself in the place of the client and as empathically as possible record what you think is the real feeling the client is experiencing. Name lr—I "0 *Episode 1. Abused 2. Ambivalent 3. Amused 4. Annoyed - Angry 5. Anxious 6. Appreciated 7. Baffled 8. Bitter 9. Bored 10. Capable 11. Careful 12. Cautious l3. Comfortable - at ease 14. Confused - mixed up 15. Controlled 16. Courageous - Daring l7. Daydreaming 18. Defeated l9. Depressed 20. Determined 21. Disappointed *This list of words was used for 65 IL») all episodes. ID IUI IO‘ Episode 22. Discouraged 23. Disgusted - Fed-up 24. Embarrassed 25. Enthusiastic 26. Erratic - 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. Disorganized Exhausted Flexible Foolish - Ridiculous Forced Frustrated Guilty Happy Helpful Helpless Hopeful Hopeless Impatient Listless - Indifferent Lonely Optimistic Patient Protected Proud |r—' IN [Co ID |U1 |O\ Episode 45. Refreshed 46. Regretful 47. Relieved 48. Reluctant 49. Satisfied 50. Scared 51. Sensible - Reasonable 52. Stubborn - Resistant 53. Surprised 54. Talkative 55. Trapped 56. Upset - Disturbed 57. Worried lv—I IN lb.) lb |U1 lox APPENDIX B AFFECTIVE SENSITIVITY SCALE 69 INSTRUCTIONS You will be viewing some kinescopes of actual counseling sessions. We are trying to identify what feelings are being experienced by the clients in these interviews. We want to know what you as an observer judge to be the actual feeling(s). On the following pages you will find lists of adjectives which describe various feelings. After viewing each episode, try to feel as the client last felt. Then, taking each adjective in turn rate it on the following scale: I have this kind of feeling strongly. I have this kind of feeling somewhat. I have this kind of feeling only a little. I have this kind of feeling not at all. war—I For example, if you view a specific episode in which the client states, "I am happy with my experiences at school," you are to try to feel as the client last felt. If you find yourself feeling very annoyed and defeated, you would rate the words annoyed and defeated #1 to indicate, "1 have this kind of feeling strongly." Assuming annoyed has arabic numeral 7 preceding it and defeated has arabic numeral 8 preceding it, your answer sheet would appear as follows: 7.1I2Z3:4:5: 8. 112:3:4js: If you feel as the client last felt, and you find yourself feeling some- what happy, you would rate the word happy #2 to indicate, "I have this kind of feeling somewhat." Assuming happy has arabic numeral 9 pre- ceding it, your answer sheet would appear as follows: 9. l __ 2 '[3 ::4 ::5': Please put yourself in the place of the client and, as honestly as possible, record the real feeling(§}* the client is last experiencing. Be sure to rate each word. CLIENT I Episode 1 l. 2. 10. Frustrated Anxious Lonely Depressed Ambivalent Discouraged Resistant Confused - Mixed-up Worried Pushed CLIENT I Episode 2 11. 12. 13. 14. 15. 16. 17. 18. Satisifed Abused Anxious Hopeful Frustrated Confused - Mixed-up Baffled Ambivalent 70 CLIENT I Episode 3 l9. Hopeful 20. Optimistic 21. Refreshed 22. Relieved 23. Trapped 24. Annoyed - Angry 25. Anxious 26. Controlled 27. Depressed CLIENT I Episode 4 28. Cautious 29. Appreciated 30. Controlled 31. Optimistic 32. Baffled 33. Ambivalent 34. Anxious 35. Embarrassed 71 CLIENT I Episode 5 36. Forced 37. Baffled 38. Resistant 39. Trapped 40. Reluctant 41. Upset - Disturbed 42. Annoyed - Angry CLIENT II Episode 1 43. Baffled 44. Talkative 45. Surprised 46. Anxious 47. Annoyed - Angry 48. Embarrassed 49. Bitter CLIENT II Episode 2 50. Forced 51. Embarrassed 52. Surprised 53. Trapped 54. Anxious 55. Scared 56. Frustrated CLIENT 11 Episode 3 57. Frustrated 58. Daydreaming 59. Anxious 60. Guilty 61. Embarrassed 62. Confused - Mixed-up 63. Appreciated 64. Scared 65. Ambivalent CLIENT III Episode 1 66. 67. 68. 69. 70. 71. 72. 73. Determined Anxious Worried Controlled Defeated Confused - Mixed-up Listless - Indifferent Annoyed CLIENT III Episode 2 74. 75. 76. 77. 78. 79. 80. Listless - Indifferent Controlled Confused - Mixed-up Bored Helpful Reluctant Surprised CLIENT IV Episode 1 81. 82. 83. 84. 85. 86. 87. 88. 89. Dependent Cautious Discouraged Reluctant Anxious Forced Worried Embarrassed Infantile CLIENT IV Episode 2 90. 91. 92. 93. 94. 95. Capable Bitter Disgusted - Fed-up Determined Annoyed - Angry Abused 73 CLIENT IV Episode 3 95. Frustrated 97. Confused - Mixed-up 98. Baffled 99. Discouraged 100. Talkative 101. Annoyed - Angry 102. Abused CLIENT IV Episode 4 103. Reluctant 104. Frustrated 105. Defeated 106. Careful 107. Depressed 108. Controlled 109. Hopeful CLIENT V Episode 1 110. Anxious lll. Disgusted - Fed-up 112. Baffled 113. Upset - Disturbed 114. Confused - Mixed-up 115. Pushed 116. Bitter 117. Cautious CLIENT V Episode 2 118. Abused 119. Disappointed 120. Frustrated 121. Annoyed - Angry 122. Comfortable - At ease 123. Anxious 124. Lonely CLIENT V Episode 3 125. 126. 127. 128. 129. 130. 131. Inflexible Depressed Confused - Mixed-up Stubborn - Resistant Abused Worried Resentful CLIENT V Episode 4 132. 133. 134. 135. 136. 137. 138. Lonely Proud Dependent Erratic - Disorganized Guilty Anxious Frustrated CLIENT V Episode 5 139. 140. 141. 142. 143. 144. 145. Bitter Discouraged Resentful Depressed Annoyed - Angry Abused Ambivalent CLIENT V Episode 6 146. 147. 148. 149. 150. 151. Trapped Bitter Resentful Annoyed - Angry Disgusted Abused 75 CLIENT VI Episode 1 152. 153. 154. 155. 156. Talkative Anxious Foolish - Ridiculous Guilty Embarrassed CLIENT VI Episode 2 157. 158. 159. 160. Satisfied Appreciated Guilty Amused CLIENT VI Episode 3 161. 162. 163. 164. 165. 166. Impatient Anxious Stubborn - Resistant Annoyed - Angry Disappointed Determined CLIENT VI Episode 4 167. 168. 169. 170. 171. Happy Relieved Baffled Confused - Mixed-up Helpful 76 CLIENT VI Episode 5 172. Bitter 173. Disappointed 174. Annoyed - Angry 175. Exhausted 176. Defeated 177. Disgusted - Fed-up 178. Baffled CLIENT VII Episode 1 179. Comfortable - At ease 180. Anxious 181. Regretful 182. Ambivalent 183. Lonely 184. Disappointed CLIENT VII Episode 2 185. Abused 186. Helpless 187. Ambivalent 188. Protected 189. Embarrassed 190. Anxious 191. Scared CLIENT VIII Episode 1 192. Confused - Mixed-up 193. Defeated 194. Embarrassed 195. Frustrated 196. Worried 197. Discouraged 198. Baffled 77 CLIENT VIII Episode 2 CLIENT IX Episode 2 199. Frustrated ll. Hopeful 200. Impatient 12. Anxious 1. Embarrassed1 l3. Refreshed 2. Confused - Mixed-up l4. Optimistic 3. Foolish - Ridiculous 15. Courageous - Daring 4. Bored l6. Enthusiastic CLIENT IXpEpisode l CLIENT IX Episode 3 5. Helpful l7. Talkative 6. Enthusiastic 18. Disgusted - Fed-up 7. Cautious 19. Amused 8. Surprised 20. Annoyed - Angry 9. Hopeful 21. Bitter 10. Relieved 22. Baffled 1 This item began on a second answer sheet. 78 CLIENT IX Episode 4 23. Defeated 24. Bitter 25. Hopeless 26. Annoyed - Angry 27. Baffled 28. Discouraged 29. Disgusted - Fed-up CLIENT X Episode 1 30. Confused - Mixed-up 31. Upset - Disturbed 32. Discouraged 33. Baffled 34. Anxious 35. Scared 36. Ambivalent CLIENT X Episode 2 37. 38. 39. 40. 41. Daydreaming Optimistic Hopeful Anxious Confused - Mixed-up CLIENT X Episode 3 42. 43. 44. 45. 46. 47. 48. Depressed Worried Helpless Hopeless Confused - Mixed-up Bitter Sensible - Reasonable 79 CLIENT X Episode 4 49. Defeated 50. Lonely 51. Bitter 52. Helpless 53. Upset - Disturbed 54. Anxious 55. Discouraged 56. Embarrassed CLIENT XI Episode 1 57. Anxious 58. Embarrassed 59. Forced 60. Discouraged 61. Impatient 62. Scared CLIENT XI Episode 2 63. Listless - Indifferent 64. Hopeful 65. Helpless 66. Impatient 67. Determined 68. Satisfied CLIENT XI Episode 3 69. Anxious 70. Bored 71. Courageous - Daring 72. Stubborn - Resistant 73. Impatient 74. Ambivalent 80 CLIENT XI Episode 4 75. 76. 77. 78. 79. 80. Protected Ambivalent Anxious Sensible - Reasonable Satisfied Determined APPENDIX C ITEM ANALYSIS OF ITEMS FOUND NONSIGNIFICANT IN VALIDATING THE INSTRUMENT 82 FREQUENCY DISTRIBUTION AND SIGNIFICANCE OF ITEMS ON THE AFFECTIVE SENSITIVITY SCALE WHICH FAILED TO MEET THE .20 LEVEL OF SIGNIFICANCE IN VALIDATING THE INSTRUMENT1 Part A (Median occurs between 2 and 3.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % I 5 High 30.563‘ 19.44 .450 l Ambivalent Low 25.00 25.00 I 13 High 16.67 33.00 2.7864 2 Anxious Low 30.56 19.44 I 14 High 16.67 33.33 .554 2 Hopeful Low 11.11 38.89 I 22 High 27.78 22.22 1.003 3 Relieved Low 19.44 30.56 I 25 High 16.67 33.33 .468 3 Anxious Low 22.22 27.78 I 29 High 22.22 27.78 .468 4 Appreciated Low 16.67 33.33 I 32 High 16.67 33.33 .000 4 Baffled Low 16.67 33.33 I 39 High 25.00 25.00 .000 5 Trapped Low 25.00 25.00 1A chi-square value of 1.642 or higher was needed for the item to be significant at .20. 2Group rating refers to the top or bottom 33% of the sample based on ratings obtained from instructors and supervisors. 3The figure given in each cell is the percent of total responses (N=36). 4Eighteen items were listed as having failed the validation test even though the chi-square was significant because the direction of discriminating high from low empathizers reversed from the group used in constructing the items. Part A4(Continued) 83 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % I 47 High 16.67 33.00 1.333 5 Annoyed- Low 8.33 41.67 Angry II 48 High 13.89 36.11 1.084 1 Embarrassed Low 22.22 27.78 II 50 High 36.11 13.89 2.857 2 Forced Low 22.22 27.78 II 53 High 19.44 30.56 1.003 2 Trapped Low 27.78 22.22 11 56 High 19.44 30.56 .120 2 Frustrated Low 16.67 33.33 II 57 High 30.56 19.44 .450 3 Frustrated Low 25.00 25.00 11 61 High 30.56 19.44 1.003 3 Embarrassed Low 22.22 27.78 II 62 High 33.33 16.67 .120 3 Confused- Low 30.56 19.44 Mixed-up II 63 High 13.89 36.11 .000 3 Appreciated Low 13 89 36.11 II 64 High 11.11 38.89 .800 3 Scared Low 5.56 44.44 III 67 High 19.44 30.56 .500 1 Anxious Low 13.89 36.11 III 68 High 38.89 11.11 .148 l Worried Low 36.11 13.89 III 72 High 13.89 36.11 1.870 1 Listless- Low 25.00 25.00 Indifferent Part A (Continued) 84 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % III 74 High 19.44 30.56 .120 2 Listless- Low 16.67 33.33 Indifferent III 77 High 38.89 11.11 4.208 2 Bored Low 22.22 27.78 111 78 High 13.89 36.11 .643 2 Helpful Low 8.33 41.67 III 79 High 25.00 25.00 .450 2 Reluctant Low 19.44 30.56 IV 85 High 25.00 25.00 4.500 1 Anxious Low 8.33 41.67 IV 87 High 19.44 30.56 2.215 1 Worried Low 8.33 41.67 IV 93 High 16.67 33.33 .000 2 Determined Low 16.67 33.33 IV 98 High 36.11 13.89 .500 3 Baffled Low 30.56 19.44 IV 101 High 27.78 22.22 .000 3 Annoyed- Low 27.78 22.22 Angry IV 103 High 22.22 27.78 .468 4 Reluctant Low 16.67 33.33 IV 104 High 30.56 19.44 1.003 4 Frustrated Low 22.22 27.78 IV 105 High 22.22 27.78 .114 4 Defeated Low 19.44 30.56 V 110 High 26.47 26.47 .537 1 Anxious Low 29.41 17.65 V 114 High 32.35 20.59 3.031 1 Confused- Low 41.18 5.88 Mixed-up Part A (Continued) 85 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % V 126 High 29.41 23.53 .002 3 Depressed Low 26.47 20.59 V 127 High 38.24 14.71 .049 3 Confused- Low 32.35 14.71 Mixed-up V 128 High 32.35 20.59 .216 3 Stubborn- Low 32.35 14.71 Resistant V 129 High 26.47 26.47 .000 3 Abused Low 23.53 23.53 V 131 High 32.35 20.59 .216 3 Resentful Low 32.35 14.71 V 136 High 32.35 20.59 3.032 4 Guilty Low 14.71 32.35 V 138 High 35.29 17.65 .926 4 Frustrated Low 38.24 8.82 V 139 High 44.12 8.82 1.001 5 Bitter Low 32.35 14.71 V 140 High 35.29 17.65 .926 5 Discouraged Low 38.24 8.82 V 142 High 32.35 20.59 .747 5 Depressed Low 35.29 11.76 V 145 High 17.65 35.29 .971 5 Ambivalent Low 23.53 23.53 V 149 High 41.18 11 76 .551 6 Annoyed- Low 41.18 5.88 Angry VI 155 High 42.42 12.12 1.224 1 Guilty Low 27.27 18.18 86 Part A (Continued)fi Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % VI 159 High 15.15 39.39 2.238 2 Guilty Low 24.24 21.21 VI 160 High 24.24 30.30 .259 2 Amused Low 24.24 21.21 VI 162 High 30.30 24.24 .259 3 Anxious Low 21.21 24.24 VI 163 High 27.27 27.27 .036 3 Stubborn- Low 24.24 21.21 Resistant VI 164 High 33.33 21.21 .550 3 Annoyed- Low 33.33 12.12 Angry VI 165 High 18.18 36.36 .609 3 Disappointed Low 21.21 24.24 VI 168 High 36.36 18.18 .157 4 Relieved Low 27.27 18.18 VI 172 High 30.30 24.24 .066 5 Bitter Low 27.27 18.18 VI 174 High 30.30 24.24 1.630 5 Annoyed- Low 15.15 30.30 Angry VI 178 High 12.12 42.42 3.422 5 Baffled Low 24.24 21.21 VII 181 High 39.39 15.15 .550 1 Regretful Low 27.27 18.18 VII 182 High 24.24 30.30 .793 l Ambivalent Low 27.27 18.18 VII 184 High 30.30 24.24 .259 1 Disappointed Low 21.21 24.24 87 Part A (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square % % VII 186 High 39.39 15.15 .550 2 Helpless Low 27.27 18.18 VII 189 High 39.39 15.15 .550 2 Embarrassed Low 27.27 18.18 VII 190 High 36.36 18.18 .000 2 Anxious Low 30.30 15.15 VIII 193 High 39.39 15.15 3.478 1 Defeated Low 18.18 27.27 VIII 195 High 39.39 15.15 .269 1 Frustrated Low 36.36 9.09 VIII 196 High 36.36 18.18 .000 1 Worried Low 30.30 15.15 VIII 197 High 33.33 21.21 1.460 1 Discouraged Low 18.18 27.27 VIII 198 High 24.24 30.30 .259 1 Baffled Low 24.24 21.21 VIII 200 High 18.18 36.36 .000 2 Impatient Low 15.15 30.30 VIII 204 High 21.21 33.33 .203 2 Bored Low 21.21 24.24 IX 212 High 27.27 27.27 .330 2 Anxious Low 18.18 27.27 IX 215 High 36.36 18.18 .000 2 Courageous- Low 30.30 15.15 Daring IX 218 High 27.27 27.27 .036 3 Disgusted- Low 21.21 24.24 Fed-up IX 219 High 21.21 33.33 .550 3 Amused Low 12.12 33.33 88 Part A (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square Z Z IX 220 High 39.39 15.15 1.021 3 Annoyed- Low 39.39 6.06 Angry IX 224 High 36.36 18.18 1.782 4 Bitter Low 39.39 6.06 IX 226 High 36.36 18.18 .733 4 Annoyed- Low 36.36 9.09 Angry IX 229 High 33.33 21.21 .004 4 Disgusted- Low 27.27 18.18 Fed-up X 233 High 36.36 18.18 .733 1 Baffled Low 36.36 9.09 X 241 High 18.18 36.36 .000 2 Confused- Low 15.15 30.30 Mixed-up X 242 High 33.33 21.21 .203 3 Depressed Low 24.24 21.21 X 246 High 18.18 36.36 .733 3 Confused- Low 9.09 36.36 Mixed-up X 248 High 39.39 15.15 .005 3 Sensible- Low 33.33 12.12 Reasonable X 256 High 18.18 36.36 .172 4 Embarrassed Low 12.12 33.33 XI 257 High 33.33 21.21 .203 1 Anxious Low 24.24 21.21 XI 258 High 36.36 18.18 .172 1 Embarrassed Low 33.33 12.12 89 Part A (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square Z Z XI 259 High 30.30 24.24 .016 l Forced Low 24.24 21.21 XI 260 High 15.15 39.39 .120 1 Discouraged Low 15.15 30.30 XI 261 High 24.24 30.30 .793 1 Impatient Low 27.27 18.18 XI 269 High 30.30 24.24 1.630 3 Anxious Low 15.15 30.30 XI 273 High 27.27 27.27 .930 3 Impatient Low 30.30 15.15 XI 277 High 21.21 33.33 .689 4 Anxious Low 24.24 21.21 XI 279 High 27.27 27.27 .036 4 Satisfied Low 24.24 21.21 90 Part B (Median occurs between 1 and 2.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square Z Z I 1 High 47.22 2.78 .364 l Frustrated Low 44.44 5.56 I 2 High 47.22 2.78 .000 1 Anxious Low 47.22 2.78 I 4 High 47.22 2.78 1.125 1 Depressed Low 41.67 8.33 I 6 High 47.22 2.78 .364 l Discouraged Low 44.44 5.56 I 8 High 47.22 2.78 .000 1 Confused- Low 47.22 2.78 Mixed-up I 9 High 44.44 5.56 .364 1 Worried Low 47.22 2.78 I 15 High 47.22 2.78 .000 2 Frustrated Low 47.22 2.78 I 16 High 47.22 2.78 .000 2 Confused- Low 47.22 2.78 Mixed-up I 17 High 44.44 5.56 .000 2 Baffled Low 44.44 5.56 I 20 High 44.44 5.56 .000 3 Optimistic Low 44.44 5.56 I 28 High 47.22 2.78 1.029 4 Cautious Low 50.00 0.00 I 36 High 38.89 11.11 .148 5 Forced Low 36.11 13.89 I 37 High 41.67 8.33 1.333 5 Baffled Low 33.33 16.67 Part B (Continued) 91 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square % Z II 43 High 38.89 11.11 1.178 1 Baffled Low 30.56 19.44 II 44 High 50.00 0.00 2.118 1 Talkative Low 44.44 5.56 II 45 High 47.22 2.78 .364 1 Surprised Low 44.44 5.56 II 51 High 47.22 2.78 1.029 2 Embarrassed Low 50.00 0.00 III 75 High 50.00 0.00 1.029 2 Controlled Low 47.27 2.78 IV 91 High 41.67 8.33 .232 2 Bitter Low 44.44 5.56 IV 92 High 44.44 5.56 .364 2 Disgusted- Low 47.22 2.78 Fed-up IV 96 High 47.22 ~2.78 1.029 3 Frustrated Low 50.00 0.00 IV 97 High 44.44 5756 .000 3 Confused- Low 44.44 5.56 Mixed-up . IV 102 High 47.22 ‘2.78 .364 3 Abused Low 44.44 5.56 x V 112 High 47.06 5'88 .016 1 Baffled Low 41.18 5.88 V 115 High 44.12 8.82 .117 1 Pushed Low 41.18 5.88 V 118 High 52.94 0.00 2.391 2 Abused Low 41.18 5.88 V 119 High 52.94 0.00 .000 2 Disappointed Low 47.06 0.00 Part B (Continued) 92 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square % Z V 120 High 52.94 0.00 .000 2 Frustrated Low 47.06 0.00 V 121 High 52.94 0.00 .000 2 Annoyed- Low 47.06 0.00 Angry V 124 High 47.06 5.88 1.889 2 Lonely Low 47.06 0.00 V 125 High 50.00 2.94 .916 3 Inflexible Low 47.06 0.00 V 130 High 52.94 0.00 .000 3 Worried Low 47.06 0.00 V 132 High 52.94 0.00 .000 4 Lonely Low 47.06 0.00 V 134 High 50.00 2.94 .007 4 Dependent Low 44.12 2.94 V 137 High 52.94 0.00 .000 4 Anxious Low 47.06 0.00 V 141 High 52.94 0.00 1.159 5 Resentful Low 44.12 2.94 V 143 High 50.00 2.94 .007 5 Annoyed- Low 44.12 2.94 Angry V 144 High 47.06 5.88 .294 5 Abused Low 44.12 2.94 V 146 High 44.12 8.82 .117 6 Trapped Low 41.18 5.88 V 147 High 52.94 0.00 .000 6 Bitter Low 47.06 0.00 V 148 High 52.94 0.00 .000 6 Resentful Low 47.06 0.00 93 Part B (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square Z Z V 150 High 50.00 2.94 .916 6 Disgusted Low 47.06 0.00 V 151 High 52.94 0.00 .000 6 Abused Low 47.06 0.00 VI 152 High 48.48 6.06 1.774 1 Talkative Low 45.45 0.00 VI 153 High 54.55 0.00 1.237 1 Anxious Low 42.42 3.03 VI 154 High 54.55 0.00 1.237 1 Foolish- Low 42.42 3.03 Ridiculous VI 156 High 54.55 0.00 .000 1 Embarrassed Low 45.45 0.00 VI 157 High 42.42 12.12 .024 2 Satisfied Low 36.36 9.09 VI 158 High 48.48 6.06 .196 2 Appreciated Low 42.42 3.03 VI 161 High 48.48 6.06 .038 3 Impatient Low 39.39 6.06 VI 166 High 48.48 6.06 .503 3 Determined Low 36.36 9.09 VI 171 High 45.45 9.09 .071 4 Helpful Low 39.39 6.06 VIII 201 High 48.48 6.06 1.774 2 Embarrassed Low 45.45 6.80 VIII 202 High 45.45 9.09 2.750 2 Confused- Low 45.15 0.00 Mixed-up IX 206 High 54.55 0.00 .000 1 Enthusiastic Low 45.45 0.00 Part B4(Continued) 94 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square Z Z IX 208 High 54.55 0.00 .000 1 Surprised Low 45.45 0.00 IX 209 High 54.55 0.00 1.237 1 Hopeful Low 42.42 3.03 IX 210 High 54.55 0.00 .000 1 Relieved Low 45.45 0.00 IX 211 High 51.52 3.03 .018 2 Hopeful Low 42.42 3.03 IX 213 High 48.48 6.06 .038 2 Refreshed Low 39.39 6.06 IX 216 High 54.55 0.00 .000 2 Enthusiastic Low 45.45 0.00 IX 217 High 42.42 12.12 1.540 3 Talkative Low 42.42 3.03 IX 225 High 51.52 3.03 .859 4 Hopeless Low 45.45 0.00 IX 228 High 51.52 3.03 .859 4 Discouraged Low 45.45 0.00 X 230 High 48.48 6.06 1.774 1 Confused- Low 45.45 0.00 Mixed-up X 231 High 48.48 6.06 1.774 1 Upset- Low 45.45 0.00 Disturbed X 234 High 54.55 0.00 .000 1 Anxious Low 45.45 0.00 X 235 High 48.48 6.06 .196 1 Scared Low 42.42 3.03 X 238 High 51.52 3.03 .018 2 Optimistic Low 42.42 3.03 Part B (Continued) 95 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square Z Z X 239 High 54.55 0.00 .000 2 HOpeful Low 45.45 0.00 X 240 High 45.45 9.09 .768 2 Anxious Low 42.42 3.03 X 244 High 54.55 0.00 .000 3 Helpless Low 45.45 0.00 X 245 High 51.52 3.03 .599 3 Hopeless Low 39.39 6.06 X 249 High 54.55 0.00 .000 4 Defeated Low 45.45 0.00 X 250 High 54.55 0.00 .000 4 Lonely Low 45.45 0.00 X 252 High 54.55 0.00 .000 4 Helpless Low 45.45 0.00 X 253 High 45.45 9.09 .071 4 Upset- Low 39.39 6.06 Disturbed X 254 High 54.55 0.00 1.237 4 Anxious Low 42.42 3.03 X 255 High 54.55 0.00 .000 4 Discouraged Low 45.45 0.00 XI 263 High 48.48 6.06 .196 2 Listless- Low 42.42 3.03 Indifferent XI 275 High 51.52 3.03 .018 4 Protected Low 42.42 3.03 96 Part C (Median occurs between 3 and 4.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square Z Z I 7 High 2.78 47.22 1.029 1 Resistant Low 0.00 50.00 I 10 High 2.78 47.22 .000 1 Pushed Low 2.78 47.22 I 11 High 0.00 48.57 .972 2 Satisfied Low 2.86 48.57 I 12 High 2.78 47.22 .000 2 Anxious Low 2.78 47.22 I 27 High 0.00 50.00 1.029 3 Depressed Low 2.78 47.22 I 31 High 2.78 47.22 .000 4 Optimistic Low 2.78 47.22 I 34 High 0.00 50.00 1.029 4 Anxious Low 2.78 47.22 I 41 High 5.56 44.44 .364 5 Upset- Low 2.78 47.22 Disturbed I 42 High 8.33 41.67 1.125 5 Annoyed- Low 2.78 47.22 Angry II 55 High 2.78 47.22 .000 2 Scared Low 2.78 47.22 II 65 High 2.78 47.22 2.090 3 Ambivalent Low 11.11 38.89 III 66 High 0.00 50.00 .000 1 Determined Low 0.00 50.00 III 70 High 5.56 44.44 .232 1 Defeated Low 8.33 41.67 97 Part C (Continued) Client Item Selected Selected and and Group Reaponse Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square Z Z III 73 High 11.11 38.89 .177 1 Annoyed Low 8.33 41.67 III 80 High 0.00 50.00 .000 2 Surprised Low 0.00 50.00 IV 81 High 5.56 44.44 2.118 1 Dependent Low 0.00 50.00 IV 83 High 0.00 50.00 1.029 1 Discouraged Low 2.78 47.22 IV 84 High 8.33 41.67 .232 1 Reluctant Low 5.56 44.44 IV 89 High 2.78 47.22 1.125 1 Infantile Low 8.33 41.67 IV 107 High 0.00 50.00 1.029 4 Depressed Low 2.78 47.22 IV 109 High 2.78 47.22 1.029 4 Hopeful Low 0.00 50.00 V 122 High 2.94 50.00 .916 2 Comfortable- Low 0.00 47.06 At Ease VI 169 High 6.06 48.48 .038 4 Baffled Low 6.06 39.39 VI 170 High 6.06 48.48 .196 4 Confused- Low 3.03 42.42 Mixed-up VI 175 High 0.00 54.55 .000 5 Exhausted Low 0.00 45.45 VI 176 High 0.00 54.55 1.237 5 Defeated Low 3.03 42.42 Part C (Continued) 98 Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square Z Z VII 187 High 3.03 51.52 .018 2 Ambivalent Low 3.03 42.42 VII 188 High 0.00 54.55 .000 2 Protected Low 0.00 45.45 IX 205 High 0.00 54.55 .000 1 Helpful Low 0.00 45.45 IX 207 High 0.00 54.55 .000 1 Cautious Low 0.00 45.45 IX 222 High 0.00 54.55 .000 3 Baffled Low 0.00 45.45 IX 227 High 3.03 51.52 .859 4 Baffled Low 0.00 45.45 X 236 High 6.06 48.48 .038 1 Ambivalent Low 6.06 39.39 X 237 High 3.03 51.52 2.836 2 Daydreaming Low 12.12 33.33 XI 262 High 3.03 51.52 .599 1 Scared Low 6.06 39.39 XI 264 High 0.00 54.55 1.237 2 Hopeful Low 3.03 42.42 XI 266 High 3.03 51.52 .018 2 Impatient Low 3.03 42.42 XI 267 High 3.03 51.52 .018 2 Determined Low 3.03 42.42 XI 270 High 12.12 42.42 .024 3 Bored Low 9.09 36.36 XI 274 High 6.06 48.48 .196 3 Ambivalent Low 3.03 42.42 APPENDIX D RATING SCALE FOR COUNSELING ABILITY 100 Instructions for Rating Counselor Trainees on Counseling Ability Would you rate each person in the group in terms of his ability as a counselor. We are not concerned here with his ability as a general pupil personnel worker. We are interested only in counseling ability and skill. As you rate, would you keep the following question in mind: "If I were to refer someone dear to me to a counselor, and needed to find a highly sensitive, skilled person how likely would I be to choose each of the members of this group?" Least Ability Most Ability APPENDIX E RATING SCALE FOR EMPATHY 102 Instructions for Rating Counselor Trainees on Empathy We are asking that you rate a particular aspect of counseling ability, empathy. It is possible and likely that your ratings on this particular aspect will not correlate exactly with your overall ratings of the individuals as effective counselors. Because a number of de- finitions of empathy prevail, we are asking that you use this particular one: Empathy is the ability or capacity to feel as another person does. In other words, if the client feels depressed, the counselor feels the client's depression. You can easily see from the example above that it is possible that some counselors may pick up the client's feelings but be unable to deal effectively with the feelings and hence be rated low in counselor effectiveness but high in empathy. You have 28 clients to rate. We would like you to put their names into a normal distribution holding to the frequencies listed below. l 3 5 8 5 3 1 Least Empathic Most Empathic APPENDIX F ITEM ANALYSIS OF ITEMS FOUND NONSIGNIFICANT IN CROSS-VALIDATING THE INSTRUMENT 104 FREQUENCY DISTRIBUTION AND SIGNIFICANCE OF ITEMS ON THE AFFECTIVE SENSITIVITY SCALE WHICH FA LED TO MEET THE .20 LEVEL OF SIGNIFICANCE IN CROSS-VALIDATING THE INSTRUMENT Part A (Median occurs between 2 and 3.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square Z Z I 3 High 23.213 26.79 .0724 § 1 Lonely Low 25.00 25.00 I 18 High 23.21 26.79 .000 2 Ambivalent Low 23.21 26.79 I 19 High 30.36 19.64 1.149 3 Hopeful Low 23.21 26.79 I 21 High 12.50 37.50 .000 3 Refreshed Low 12.50 37.50 I 33 High 16.07 33.93 1.198 4 Ambivalent Low 23.21 26.79 I 35 High 14.29 35.71 .717 4 Embarrassed Low 19.64 30.36 I 38 High 25.00 25.00 .072 5 Resistant Low 26.79 23.21 II 52 High 16.07 33.93 .080 2 Surprised Low 17.86 32.14 II 54 High 30.36 19.64 .000 J 2 Anxious Low 30.36 19.64 1A chi-square value of 1.642 or higher was needed for the item to be significant at .20. 2 . a Group rating refers to the top or bottom 33A of the sample based on ratings obtained from instructors and supervisors. ‘9 3The figure given in each cell is the percent of total responses (N=56). 4Chi-square values were computed from frequency data, not from the percentage data reported here. 105 Part A (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square Z Z II 58 High 19.64 30.36 .000 3 Daydreaming Low 19.64 30.36 II 59 High 33.93 16.07 .350 3 Anxious Low 37.50 12.50 II 60 High 35.71 14.29 .381 3 Guilty Low 39.29 10.71 III 69 High 30.36 19.64 .074 1 Controlled Low 28.57 21.43 III 77 High 26.79 23.21 .286 2 Bored Low 23.21 26.79 IV 90 High 26.79 23.21 .072 2 Capable Low 28.57 21.43 IV 94 High 30.36 19.64 .000 2 Annoyed- Low 30.36 19.64 Angry IV 95 High 25.00 25.00 .000 2 Abused Low 25.00 25.00 IV 100 High 42.86 7.14 .487 3 Talkative Low 39.29 10.71 v 111 High 21.43 28.57 1.7881 1 Disgusted- Low 30.36 19.64 Fed-up V 113 High 33.93 16.07 1.524 1 Upset- Low 41.07 8.93 Disturbed V 117 High 25.00 25.00 .650 1 Cautious Low 19.64 30.36 “ 1Some items will appear in this table where the is higher than 1.642 because the response pattern of groups is Opposite that of the criterion groups. chi-square value the cross-validation 106 Part A (Continued) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 3 or 4 square Z Z V 135 High 16.07 33.93 1.198 4 Erratic- Low 23.21 26.79 Disorganized VI 167 High 30.36 19.64 .292 4 Happy Low 26.79 23.21 VII 179 High 46.43 3.57 4.383 1 Comfortable- Low 35.71 14.29 At ease VII 180 High 17.86 32.14 .000 1 Anxious Low 17.86 32.14 VII 185 High 8.93 41.07 .424 2 Abused Low 12.50 37.50 VIII 194 High 14.29 35.71 3.615 1 Embarrassed Low 26.79 23.21 VIII 199 High 32.73 18.18 .147 2 Frustrated Low 29.09 20.00 VIII 203 High 10.71 39.29 2.947 2 Foolish- Low 21.43 28.57 Ridiculous IX 223 High 71.43 28.57 .000 4 Defeated Low 71.43 28.57 X 243 High 35.71 14.29 2.828 3 Worried Low 44.64 5.36 XI 271 High 10.71 39.29 .100 3 Courageous- Low 12.50 37.50 Daring X1 272 High 48.21 1.79 29.944 3 Stubborn- Low 12.50 37.50 Resistant XI 278 High 23.21 26.79 .664 4 Sensible- Low 17.86 32.14 Reasonable 107 Part B 4_(Median occurs between 1 and 2.) Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 2 or 3 or 4 square Z Z I 26 High 3.57 46.43 1.469 3 Controlled Low 8.93 41.07 I 30 High 7.14 42.86 .487 4 Controlled Low 10.71 39.29 I 40 High 7.86 32.14 .076 5 Reluctant Low 9.64 30.36 II 46 High 8.93 41.07 .113 1 Anxious Low 10.71 39.29 IV 82 High 3.57 46.43 1.469 1 Cautious Low 8.93 41.07 IV 99 High 19.64 30.36 .074 3 Discouraged Low 21.43 28.57 IV 108 High 7.14 42.86 .163 4 Controlled Low 5.36 44.64 V 123 High 10.71 39.29 1.400 2 Anxious Low 17.86 32.14 V 133 High 10.71 39.29 .100 4 Proud Low 12.50 37.50 VI 173 High 16.07 33.93 .080 5 Disappointed Low 17.86 32.14 VIII 192 High 16.07 33.93 .000 l Confused- Low 16.07 33.93 Mixed-up IX 214 High 19.64 30.36 .074 2 Optimistic Low 21.43 28.57 IX 221 High 10.71 39.29 1.191 3 Bitter Low 5.36 44.64 X 232 High 17.86 32.14 .760 1 Discouraged Low 12.50 37.50 (Median occurs between 3 and 4.) 108 Part C Client Item Selected Selected and and Group Response Response Chi- Scene Adjective Rating 1 or 2 or 3 4 square Z Z I 23 High 26.79 23.21 1.198 3 Trapped Low 33.93 16.07 I 24 High 32.14 17.86 .000 3 Annoyed- Low 32.14 17.86 Angry II 49 High 21.43 28.57 .000 1 Bitter Low 21.43 28.57 III 76 High 37.50 12.50 1.018 2 Confused- Low 42.86 7.14 Mixed-up IV 86 High 16.07 33.93 1.198 1 Forced Low 23.21 26.79 IV 88 High 23.21 26.79 .286 1 Embarrassed Low 26.79 23.21 IV 106 High 41.07 8.93 1.469 4 Careful Low 46.43 3.57 XI 276 High 25.00 25.00 8.187 4 Ambivalent Low 42.86 7.14 APPENDIX G ITEMS SCORED AND CORRECT RESPONSE FOR OBTAINING TOTAL SCORE 110 ITEMS SCORED AND CORRECT RESPONSE FOR OBTAINING TOTAL SCORE Item Number Correct Response* Item Number Correct Response 1 3 or 4 31 Not scored 2 Not scored** 32 3 or 4 3 Not scored 33 3 or 4 4 3 or 4 34 Not scored 5 1 or 2 35 Not scored 6 3 or 4 36 Not scored 7 Not scored 37 3 or 4 8 1 or 2 38 Not scored 9 3 or 4 39 1 or 2 10 Not scored 40 Not scored 11 Not scored 41 Not scored 12 Not scored 42 3 or 4 13 1 or 2 43 3 or 4 14 Not scored 44 3 or 4 15 3 or 4 45 Not scored 16 1 or 2 46 l or 2 17 l or 2 47 l or 2 18 1 or 2 48 1 or 2 19 l or 2 49 Not scored 20 1 or 2 50 3 or 4 21 Not scored 51 1 or 2 22 3 or 4 52 Not scored 23 3 or 4 53 Not scored 24 l or 2 54 1 or 2 25 Not scored 55 Not scored 26 Not scored 56 3 or 4 27 Not scored 57 3 or 4 28 l or 2 58 3 or 4 29 Not scored 59 Not scored 30 1 or 2 60 Not scored *The data was dichotomized between responses 2 and 3 since direction of response was considered more important than degree of response. Thus, one of two responses was always correct. **These items did not present a clear pattern of discrimination be- tween high and low counselor criterion groups. They were used to con- stitute the 280 items as there was a slight pattern of discrimination. Item Number Correct Response 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 Not scored 1 or 2 l or 2 Not scored 1 or 2 Not scored 3 or 4 3 or 4 Not scored 3 or 4 1 or 2 1 or 2 Not scored 3 or 4 1 or 2 Not scored 1 or 2 3 or 4 3 or 4 Not scored Not scored 1 or 2 3 or 4 1 or 2 3 or 4 Not scored 3 or 4 Not scored Not scored 3 or 4 1 or 2 Not scored 3 or 4 l or 2 Not scored Not scored Not scored Not scored 1 or 2 111 Item Number Correct Response 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 3 or 4 Not scored 1 or 2 3 or 4 Not scored 3 or 4 Not scored 3 or 4 Not scored Not scored Not scored 1 or 2 Not scored 3 or 4 Not scored Not scored 3 or 4 3 or 4 1 or 2 Not scored or 2 or 2 or 4 or 4 or 4 ot scored ZUJUJL/Jr—‘H Not scored Not scored 1 or 2 3 or 4 Not scored 3 or 4 3 or 4 Not scored Not scored 112 Item Number Correct Responses Item Number Correct Responses 139 1 or 2 178 Not scored 140 l or 2 179 3 or 4 141 Not scored 180 3 or 4 142 3 or 4 181 Not scored 143 Not scored 182 Not scored 144 3 or 4 183 Not scored 145 1 or 2 184 3 or 4 146 3 or 4 185 Not scored 147 Not scored 186 3 or 4 148 Not scored 187 3 or 4 149 1 or 2 188 3 or 4 150 3 or 4 189 1 or 2 151 1 or 2 190 1 or 2 152 Not scored 191 3 or 4 153 Not scored 192 Not scored 154 3 or 4 193 3 or 4 155 l or 2 194 1 or 2 156 1 or 2 195 3 or 4 157 3 or 4 196 Not scored 158 1 or 2 197 3 or 4 159 l or 2 198 1 or 2 160 3 or 4 199 3 or 4 161 1 or 2 200 3 or 4 162 3 or 4 201 l or 2 163 Not scored 202 Not scored 164 Not scored 203 Not scored 165 3 or 4 204 3 or 4 166 1 or 2 205 Not scored 167 1 or 2 206 3 or 4 168 3 or 4 207 3 or 4 169 1 or 2 208 Not scored 170 3 or 4 209 l or 2 171 3 or 4 210 Not scored 172 Not scored 211 Not scored 173 3 or 4 212 3 or 4 174 1 or 2 213 3 or 4 175 3 or 4 214 Not scored 176 3 or 4 215 1 or 2 177 Not scored 216 1 or 2 Item Number 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 Correct Response Not scored Not scored or or or or or or or or 3 or Not scored Not scored Not scored Not scored 3 or 4 3 or 4 Not scored Not scored 1 or 2 3 or 4 1 or 2 1 or 2 Not scored 3 or 4 3 or 4 Not scored Not scored 3 or 4 3 or 4 Not scored Not scored I—va—dwwr-Jr—Iw bNDN-L‘J-‘NN-L‘ 113 Item Number 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 Correct Response Not scored 1 or 2 3 or 4 3 or 4 Not scored 1 or 2 1 or 2 1 or 2 Not scored 1 or 2 3 or 4 3 or 4 3 or 4 Not scored Not scored 3 or 4 1 or 2 3 or 4 3 or 4 Not scored 3 or 4 Not scored 1 or 2 3 or 4 Not scored Not scored 1 or 2 1 or 2 Not scored Not scored Not scored Not scored BIBLIOGRAPHY BIBLIOGRAPHY Allport, G. W. Personality: A Psychological Interpretation. New York: Holt, 1937. . "The Historical Background of Modern Social Psychology." In G. A. Lindzey (Ed.), Handbook of Social Psychology. Cambridge, Massachusetts: Addison-Wesley, 1954, 3-56. Astin, Helen S. "A Comparative Study of the Situational and Predictive Approaches to the Measurement of Empathy." Unpublished doctoral dissertation, University of Maryland, 1957. Bender, I. E. and A. H. Hastorf. "The Perceptions of Persons: Forecasting Another Person's Responses on Three Personality Scales.” Journal of Abnormal and Social Psycholpgy, 1950, 45, 556-561. . "On Measuring Generalized Empathic Ability (Social Sensitivity)." Journal of Abnormal and Social Psychology, 1953, 48, 503-506. Berne, E. Games Pepple Play. New York: Grove Press, Inc., 1964. Buchheimer, Arnold. "Progress Report--N.D.E.A. Grant # 7-42-0550-1670." Unpublished, Hunter College of the City University of New York, 1964. . "The Development of Ideas About Empathy." Journal of Counsel- ing Psychology, 1963, 10, 61-71. Buros, O. K. The Fifth Mental Measurements Yearbook. Highland Park, New Jersey: Gryphon, 1959. Cartwright, Rosalind D. and Barbara Lerner. "Empathy, Need to Change, and Improvement with Psychotherapy." Journal of Consulting Psychology, 1963, 27, 138-144. Cowden, R. C. "Empathy or Projection." Journal of Clinical Psychology, 1955, 11, 188-190. Cronbach, L. J. "Processes Affecting Scores on 'Understanding of Others' and 'Assumed Similarity'." Psychological Bulletin, 1955, 52, 177- 193. 115 116 Crow, W. J. "A Methodological Study of Social Perceptiveness." Unpublished doctoral dissertation, University of Colorado, 1954. Dymond, Rosalind F. "A Scale for the Measurement of Empathic Ability." Journal of Consulting Psychology, 1949, 13, 127-133. . I'Personality and Empathy." Journal of Consulting Psychology, 1950, 14, 343-350. English, N. B. and Ava C. English. A Comprehensive Dictionaryyof Psycho- logical and Psychoanalytical Terms. New York: Longmans, Green and Company, 1958. Fenichel, O. The Ppychoanalytic Theogy of Neurosis. New York: W. W. Norton, 1945. Fiedler, F. E. I'Quantitative Studies on the Role of Therapists' Feelings Toward Their Patients." Psychotherapy Theory and Research. New York: Ronald Press, 1953, 296-316. Gage, N. L. and L. J. Cronbach. "Conceptual and Methodological Problems in Interpersonal Perception." Psychological Review, 1955, 62, 411-421. . "Accuracy of Social Perception and Effectiveness in Interper- sonal Relationships." Journal of Personnel, 1953, 22, 128-141. Gordon, Kate. "A Device for Demonstrating Empathy." Journal of Experimental Psychology, 1934, 17, 892-893. Hastorf, A. H., D. E. Bender, and D. J. Weintraub. "The Influence of Response Patterns on the Refined Empathy Score." Journal of Abnor- mal and Social Ppychology, 1955, 51. Kagan, Norman, D. R. Krathwohl, and R. Miller. "Stimulated Recall in Therapy Using Videotape--A Case Study." Journal of Experimental Psychology, 1963, 10, 237-243. Katz, R. L. Empathy: Its Nature and Uses. London: Free Press of Glencoe, 1963. Kelly, E. L. and D. W. Fiske. The Prediction of Performance in Clinical Psychology. Ann Arbor: University of Michigan Press, 1951. Kerr, W. A. and B. J. Speroff. "Validation and Evaluation of the Empathy Test." Journal of General Psychology, 1954, 50, 269-276. Lesser, W. M. "The Relationship between Counseling Progress and Empathic Understanding." Unpublished doctoral dissertation, Michigan State University, 1958. 117 Lifton, W. M. "The Role of Empathy and Aesthetic Sensitivity in Counseling. Journal of Counseling_Psychology, 1958, 5, 267-275. Lindgren, H. E. and J. Robinson. "Evaluation of Dymond's Test of Insight and Empathy." Journal of Consulting Psychology, 1953, 17, 172-176. Norman, R. D. and Patricia Ainsworth. "The Relationships Among Projection, Empathy, Reality, and Adjustment, Operationally Defined." Journal of Consulting Psychology, 1954, 18. O'Hern, Jane S. and D. S. Arbuckle. "Sensitivity: A Measurable ConceptT' Personnel and Guidance Journal, 1964, 42, 572-576. Osgood, C. E. Method and Theory in Experimental Psychology. New York, Oxford University Press, 1953. Rank, R. C. "The Assessment of Counselor-Trainee Perceptions of Interview Protocols Before and After an Intensive Practicum Experience." Unpublished doctoral dissertation, University of Minnesota, 1964. Reid, Dorothy K. and W. V. Snyder. "Experiment on 'Recognition of Feeling' in Non-directive Psychotherapy.” Journal of Clinical Psychology, 1947, 3, 128-135. Rogers, C. R. "A Theory of Therapy, Personality, and Interpersonal Relationships, as Developed in the Client-Centered Framwork." In S. Koch (Ed.), Ppychology: A Study of a Science. Vol. III: Formulations of the person in the social context. New York: McGraw- Hill Book Company, 1959, 184-258. . Client-Centered Therapy, Boston: Houghton Mifflin Inc., 1951. Sarbin, T. R., R. Taft, and D. E. Bailey. Clinical Inference and Cognitive Theory. New York: Holt, Rinehart and Winston, 1960. Siegel, Sidney. Non-Parametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Book Company, 1956, 107. Stefflre, Buford. "Client Understanding Test (first version)." Unpublished, Michigan State University, 1962. Stone, G. C. and G. S. Leavitt. "Generality of Accuracy in Perceiving Standard Persons." Paper read at Midwest Psychological Association Conference; Columbus, Ohio, May, 1954. Sullivan, H. S. The Psychiatric Interview. New York: W. W. Norton, 1954. Taft, Ronald. "The Ability to Judge PeOple," Psychological Bulletin, 1955, 52, 1-23. 118 Thorndike, R. L. Review of Ken and Speroffs' "The Empathy Test" in O. K. Buros, The Fifth Mental Measurements Yearbook, Highland Park, New Jersey: Gryphon, 1959. Walz, G. R. and J. A. Johnson. "Counselors Look at Themselves on Video- tape," Journal of Counseling Psychology, 1963, 10, 232-236. Weinstein, M. and F. D. McCandless. "Empathic Communication and Anxiety in Medical Students," Mental Hygiene, 1959, 43, 526-531. Weitz, H. "Counseling as a Function of the Counselor's Personalityf' Personnel and Guidance Journal, 1957, 5, 277. Wert, J. E., C. O. Neidt, and J. S. Ahmann. Statistical Methods in Educational and Psychological Research. New York: Appleton-Century- Crofts, Inc., 1954.