ABSTRACT CLIENT—THERAPIST ATTRACTION AS PERCEIVED ON THE PICTURE IMPRESSIONS TEST by John Andrew Mullen The concept of attraction in interpersonal relation— ships is guite important as a determining factor in whether or not an interpersonal relationship is continued. Attrac- tion is defined in this study as statements made by §fs on the Picture Impressions Test which reflect positive or negative feelings about a therapist after an initial inter— view. This study was concerned with the ability of the Picture Impressions Technique to provide relevant information which would aid in communication of attraction in client— therapist relationships. A total N of 137 students was used, split into experimental and control groups. The experimental group consisted of students coming for help at a university counseling center. The control group was defined as those not seeking help. The predictive validity of this instrument was measured in a different setting than that in which it was originally measured. In a previous study, the instrument demonstrated predictive validity in a more medically oriented setting. The trend was in the direction of the original study but results were not John Andrew Mullen significant. The test apparently did predict well on the basis of what transpired in an interview with a client, i.e., whether or not the client was attracted after one personal contact. Factors other than attraction seemed to play a large part in the return rate of this population. A second goal of this study was to develop a base rate of attraction scores for people not having any personal contact. The control group, consisting of advanced under— graduates in a psychology class, was administered the test. Their results were contrasted with the experimental group, i.e., people who came to the university counseling center seeking therapeutic contact. The latter were administered the test after an initial interview. The experimental group obtained a significantly higher mean of attraction scores. A qualitative analysis of the data indicated material which was important for communication between client and therapist. ; { ~ "' J' t w \ V {I “‘. Ir ' '_ .P w l ; Approved: Committee Chairman \( EQI‘ Date: "\.\ "‘5 "’ Thesis Committee: Norman Abeles, Chairman John R. Hurley Josephine Morse CLIENT—THERAPIST ATTRACTION AS PERCEIVED ON THE PICTURE IMPRESSIONS TEST by John Andrew Mullen A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1966 ACKNOWLEDGMENTS I would like to give special thanks to Dr. Norman Abeles, chairman of this thesis committee. His ready availability, guidance and understanding helped make the undertaking of this study possible, and a very meaningful experience to me. I also would like to extend my appreciation to Dr. Josephine Morse and Dr. John Hurley for serving as members of the thesis committee. Their support and observations were very beneficial in furthering my understanding. Finally, a special thanks to my wife, Nance, whose patience and many hours of typing I found invaluable. ii Abstract Introduction Procedure . Results . Discussion Summary . Appendix . Bibliography TABLE OF CONTENTS iii Page 12 19 27 34 37 39 Table II. III. IV. VI. VII. LIST OF TABLES Scoring Criteria . . . . . . . . . Summary of Agreement Between Two Independent Scorers on the PIT Using Three Measures of Agreement . . . . . . . . . . . Summary Table of T—Ratios for Experimental versus Control Groups . . . . . . . Relation Between Test Prediction and Actual Attendance O O O O O O O O O O 0 Relation Between Test Prediction and Actual Attendance for Males . . . . . . . Relation Between Test Prediction and Actual Attendance for Females . . . . . . . Relation Between Test Prediction and Continuing Attendance . . . . . . . Agreement on Three Scoring Criteria by Two Independent Scorers . . . . . . Computation of Pearson Product-Moment Correlation Coefficient . . . . . . iv Page 16 2O 21 22 23 24 24 37 38 INTRODUCTION This investigation is concerned with the expectations people have of others. The communication between individuals is a critical issue here in terms of what people do with their expectations of others. This is extremely important in therapeutic relationships, and is, indeed, an important area in all social interaction also. Goldstein, in an article on client dependency and therapist expectancy as relationship—maintaining variables (19499, quotes Borden on the meaning of the client-therapist relationship: The key to the influence of psychotherapy on the patient is in his relationship with the therapist. Wherever psychotherapy is accepted as a significant enterprise, this statement is so widely subscribed to as to become trite. Virtually all efforts to theorize about psycho- therapy are intended to describe and explain what attributes of the interactions between the therapist and the patient will account for whatever behavior changes result. This relationship includes variables such as the therapist's expectation of patient improvement, the client's expectation of what therapy is like, the client's perception of his role, the magnitude of client attraction toward the therapist, and whether or not the expectations of both therapist and client are congruent. 2 In more recent years, much more attention has been given to the client-therapist variables. Snyder (1961) spoke of the value of knowledge of this relationship and how it might be studied and measured: About five years ago we became very much con— vinced that the relationship that develops between client and therapist is the essential core of therapy. If the relationship can be measured, then psychotherapy can be improved. (p. 127) He was especially interested in the attitudes and feelings of both, believing it would be very desirable to measure those fairly subtle attitudes of client and therapist toward each other. One way of learning about interpersonal relationships and perceptions is through the use of projective techniques. The basis of all projective techniques rests on the fact that a shift takes place in the importance of internal versus external factors in perception, depending on various stimulus fields that a person might be exposed to. With regard to studying the internal factors behind an individual's percep- tion, Bellak (1950) noted that: The basic assumption in the use of projective tests (e.g., ROR, TAT, Szondi, etc.) is that the subject is presented with a number of ambiguous stimuli and responds to them. It is assumed that the subject projects his own needs and press and that these will appear as responses to the ambigu— ous stimuli. (p. 9) In recent years the projective hypothesis and assumptions underlying it have been subjected to close study. Abt and Bellak (1950) speak of the need to understand the projective hypothesis: 3 Behind the projective hypothesis itself stands a whole matrix of assumptions which probably differ from one projective psychologist to another and which have largely been kept implicit. If projective psychology is to grow in acceptance and validity it is essential that these assumptions be made fully explicit and it is necessary that they be tested to ascertain whether they have established validity and generality within the area of inquiry in which they are being employed. (p. 5) An important consideration in studying the personality of an individual with projective techniques is that any given projective method will produce information about only one expression of the personality. Abt (1950) refers to this, stating that: The configurational nature of personality is the justification for a multidimensional approach to analysis which the several projective proce- dures represent. Each aspect of the configuration called personality which any given projective method attempts to throw light on must be regarded as only one expression of the total personality as process and must be considered in the light of the other behavioral expressions of the individual. (p. 61) The part of the configuration of the total personality that this study is concerned with is the eXpression of the apper— ception of the present relationship between the client and the therapist, as it is measured by a projective technique. Miller (1953), using the Rorschach technique as an example of an interpersonal relationship, studied Rorschach protocols to determine conditions under which the most valid results were obtained. Five variables were found which seem to be pertinent to any interpersonal situation, whether it is therapy, diagnosis, or attitude measurement. These variables are: 4 1) setting: The characteristic of the test situation. 2) task: Nature of the test, how it is introduced and how responses are required from the subject. 3) The examiner's social stimulus value. (Readily observable characteristics such as height or certain mannerisms.) 4) The subject's character structure. 5) The relationship between the examiner and the subject. Not all of these five variables are usually evaluated in projective testing, yet they are important sources of vari_ ance. For example, a protocol of those actively seeking help for personal difficulties will be very different from prisoners told by authority to take tests. Kimble (1945) found that subjects tested under standard conditions produced significantly fewer responses than others tested in an "intimate and friendly" atmosphere in a cafeteria. It also was found that peOple respond differently for individual versus group administration of some tests, thus indicating that the specificity of the situation affects what they do. Of the five above mentioned variables, the most critical one, and the one this study is specifically interested in, is the fifth one, concerning the relationship between the examiner and the subject. Miller states that: Jung represents all schools of therapy in his statement that the therapist does not 'analyse' an object at a distance, but is as much a part of the relationship as the client. Both are interacting and mutually influenced in the specific situation between the therapist now and the client now. (1953, p. 370) Thus, whether or not the examiner and the subject have the same goals, whether they communicate, their appearance, etc., 5 all influence a specific projective test protocol. All projective test protocols appear to be, at least to some extent, measures of the interpersonal situations and reactions between the specific client and therapist or examiner in a specific situation. It can be seen, then, that there are several important variables which will affect the impact of the relationship I on the client. This study is concerned with the role of a client attraction for the therapist within the theory of social preference and the perception of such, and the ' concepts of role, role expectation, and the mutuality of ' role expectations in determining the effect of interpersonal interaction. Tagiuri (1958) talks about the significance of "social preference and its perception" in relation to person per— ception and interpersonal behavior; he assesses how people perceive their environment, and emphasizes the variables of like and dislike between persons, which has consequences concerning interpersonal behavior. This variable of attraction is one way for a person to evaluate his human environment within the context of active interaction with others. Tagiuri speaks of the like-dislike variable being so basic that where efficient operation depends upon specifically non—affective relationships, as in a hierarchy, special safeguards have to be built into the system to avoid as much as possible the interference of "favoritism". He states that person perception is interwoven with interaction in such a way that they cannot be separated, and that choice 6 or rejection has meaning in terms of perception. He points out it is difficult practically to distinguish between what a person feels and what he says he feels. Positive feelings are displayed relatively openly, especially if one feels the choice is reciprocated by the one chosen. Negative feelings are more difficult to identify because of inhibiting the display of negative attitudes. One may also fail to correctly identify negative feelings of another because of defense of your own self—image. Tagiuri speaks of the importance of confidence, i.e., if a person feels confident that a choice will be reciprocated, such a choice is usually more visible. Concerning dyadic relationships, he feels the most crucial social situation is the two person group, and perhaps it is the most crucial of all human situations. The variable of like—dislike, then, is seen as all important in social preference, in choice resulting from it, and in expression of that choice, at least partially through role play and expectation. In this study, clients entering therapy who remain are seen as "liking" and making the choice related to it, leading to the eXpected role play. The concept of role has had several definitions in recent years. It has received importance since 1920 when social psychologists first began to emphasize the concept of self as the basic element in a process of symbolic inter- action in the develOping personality. Emphasis was placed on the person's attitude toward himself in relation to its determination by the attitudes and expectations of others toward him. 7 Role has been defined (Goldstein, 1962) as the appro- priate behavior expected of people in certain positions, and also as the common content of the role expectations for different members of a social group. Role expectation involves two classes, usually rights and obligations. With rights, the person assuming the role eXpects certain behavior of the individual reciprocating. Obligations are eXpected things the role player must do regarding the reciprocator. Heavy importance is placed on mutuality of role expectations. According to Goldstein (1962), client expectations in psycho- therapy are: (1) An end in itself rather than a means to an end, i.e., psychotherapy is another of the "good things". (2) A way to produce a fixed or rigid "healthy state of mind". (3) A virtuous act, one which will be rewarded by relief of difficulties. (4) A means of altering circumstances. ) A confirmation of one's illness, i.e., if he takes me as a patient, I must be sick. (6) Proof of the objective difficulty of one's life circumstances. (7) An environment in which already imminent changes may take place. Apfelbaum claimed to find three clusters of client role expectations: nurturant, model and critic (Goldstein). Goldstein sums up research in role expectation: On a gross level, it would seem clear that the studies of Apfelbaum, Ruesch, Deskins, and Goldstein and Heller amply demonstrate the use- fulness of examining role expectations in psycho- therapy in terms of the nurturant, model and critic expectancy dimensions. The ease with which DeHaan's and Chance's findings could be conceptualized in terms of this expectational triad tends to further enhance its usefulness. (p. 57) 8 The significance of mutuality of participant expec- tancies in the therapist—client interaction is heavily emphasized with regard to role expectations, and, in fact, with regard to most of the literature. An all important factor in making clear what the participant expectancies in the therapist—client inter— action are is communication between the two. It is known that emotional states affect communication and the desire for it. Festinger (1960) states: The force to communicate about 'item X' to a particular person will decrease to the extent that he is perceived as not a member of the group or to the extent that he is not wanted as a member of the group. Viewing the client—therapist relationship as a unique two-person group, it seems that communication between them should of necessity, be clear, frequent, and not withheld by either member. It also is apparent that many emotional and other factors as mentioned previously may hinder com- munication. Thus, a projective which will specifically measure the positive and/or negative aspects of their inter- action and yield indicators of the patients feelings toward the therapist can be an important aid to effective communi— cation. This investigation is concerned with the ability of the Picture Impressions Test to effectively communicate in an accurate way information about the specific client—therapist relationship. The purposes of this study are both empirical and methodological. One purpose is to investigate the concurrent validity (Anastasi, 1961) of the Picture 9 Impressions Test (PIT) as a measure of the client's expressed attraction to the interpersonal situation. Libo (1957) pre— sented evidence supporting the idea of attraction to the therapist as a prediction of whether or not the client will return for his second interview. In reference to measure— ment of attraction by the PIT and the rationale behind it, he stated: Several recent studies have demonstrated that the responses to projective techniques seem to be determined in part by interpersonal aspects of the examination situation. This sensitivity of projectives to the social situation in which they are administered is a serious problem for the clinician. Since the test is designed to measure enduring characteristics of the patient, its sensitivity to characteristics of his specific relationship to the examiner makes the test to that extent an unreliable instrument. However, it may be possible to harness the inter—examiner unreliability of projective techniques to serve a useful measurement function, namely that of characterizing specific relationships. (p. 33) L Wilfltfilfiiur r In? Libo goes on to explain the validity of the projective method in the new kind of use. It was shown that the sensitivity of projectives to immediate situations can be used to measure the cohesiveness of an interpersonal relationship. The PIT, which was based on application of the projective method to measurement of group cohesiveness, specifically measures the strength (in attractiveness) of the client—therapist inter— personal relationship. In Libo's study the instrument pre- dicted client behavior much more accurately than one could predict from knowledge of the return rate of their clinic alone. Another purpose of the study is to investigate the baseline of attraction scores between individuals requesting therapy and continuing in therapy, and a sample of college 10 students who are asked to imagine a person coming for therapy and then asked to respond to the PIT. It would seem possible that interest in psychological help, combined with the impact of one person on another, should produce high attraction scores by itself. From a methodological point of view, a comparison will be made between Libo's findings and the findings in a different setting, i.e., a college counseling center offering psychotherapeutic services. The qualitative aspect of this study will be to consider and make some tenta~ tive statements about individuals with low attraction scores. Here we are interested in the relationship between both groups (i.e., the students seen for an initial interview at the MSU Counseling Center and the students not connected with therapy or desiring any contact at this time) as to whether or not the material producing a low attraction score is similar for both. While we are interested in why a particular client has a low attraction score, we are also interested in whether this low score is indeed reflective of the specific interpersonal situation experienced, or if it reflects more general feelings toward therapists or therapy, or simply negative feelings toward interpersonal relationships. Because of these and similar questions that might be raised, it seems important to examine the qualitative aspects here so as to develop further hypotheses for study. This study has three specific hypotheses, two of which are quantitative and one qualitative. Of the two quantita- tive, one is methodological to determine the predictive validity of the instrument, while the other retests Libo's 11 original hypothesis to determine whether or not the same results will be obtained in a different setting. The general hypotheses of this investigation are as follows: A. Quantitative: 1. In order to determine the predictive validity of the Picture Impressions Test, this study predicts greater client-therapist attraction scores for students continuing in therapy as Opposed to students not returning, as measured by the projective technique after the initial interview. In order to determine the base rate of client- therapist attractions, it is predicted that students coming to one counseling interview will have higher client~therapist attraction scores than a group of students who have not had counseling contact. Qualitative: 1. A qualitative study of low attraction scores will be conducted for both experimental and control groups. PROCEDURE The subjects in this investigation were obtained from the student population at Michigan State University. The experimental group consists of students who came to the Counseling Center at Michigan State University requesting therapeutic counseling, and have been seen during an initial interview. The control group consists of students enrolled in advanced undergraduate courses in psychology, who have no previous contact with therapy and have not indicated any desire for therapeutic contact. Students in the control group who have previous contact with therapy and students over twenty-five years of age were omitted from this study. The total N used was 137. The instrument used in this study consisted of the Picture Impressions Test which includes a packet consisting of four pictures, an outline form, and an instruction page. There is also a manual for the examiner. This projective test consists of four cards similar to TAT cards. The four pictures can be used for white and Negro subjects and for adolescents and adults of both sexes (Libo, 1956). The first picture is of a man (or woman) standing with chin in hand, and above the person and to the right is a balloon— like structure such as is seen in cartoons. In this balloon is pictured two people seated in chairs facing each other. 12 13 One (the man) has a note pad on his knee. The second pictures a man (or woman) sitting in a chair at a distance from another individual seated behind a desk. The third pictures a man (or woman) leaving an office, and the door to the office has "Dr. Jones" lettered on it. The person is leaving through the doorway and a man is seated behind the desk in the office, watching the person. The last picture consists of a man (or woman) sitting in a waiting room alone. Also in the picture is an office door with a sign that reads "the doctor is in” hanging at an angle from the door. Administration time for the Picture Impressions Test is twenty to thirty minutes. The manual's instructions to the examiner are as follows: This is a set of four pictures designed to elicit stories from clients. This is a set designed for use with male patients and another for use with female patients. You should have the set marked M if you have a man, F if a woman. The procedure is introduced to the client at the end of the initial interview. Each picture is handed to the client, and the client writes a story on the outline provided. One picture at a time is to be shown; other pictures should be placed face down or otherwise kept from view. Questions from the client should be answered by repeating the relevant parts of the instructions. If the client asks what he is supposed to write or say, he should be answered simply: "It's up to you". No time limit is set up. The following instructions should be read to the client, slowly and distinctly: "I am going to give you some pictures to look at. There are four pictures in all. As I give you these pictures, one at a time, look at the picture, form your own ideas about it, then make up a short story about what you see in the picture. Use your imagination. There is no right or wrong way. Your story should be about what you see in the picture. 14 ‘"To help you make up a story I am giving you this outline to follow. (Hand the four-page outline to the client.) By answering these questions, all the parts of your story's plot will be covered. Look at the picture briefly, then write your story by answering the questions on the outline sheet. Write a definite, interesting story, rather than a simple description of the picture. "Please work fast. Each story should take about five minutes to write. "All right? Here is the first picture." Using the above instructions, the test was administered to the experimental and control groups. The experimental group consisted of the students seeking therapeutic counseling, who were given the test after the initial interview. There was deviation from standard procedure of administration here, in that the PIT was administered by a psychometrist in the testing room of the Counseling Center rather than the therapist who had conducted the interview. The client was given the complete set of pictures and read the instructions, then sent into the testing room with the instructions and pictures, and he wrote out a response to each picture by himself, one at a time. The control group consisted of students in advanced undergraduate courses in psychology, who were administered the test during a class period. The administration for the control group followed the usual pro- cedure except that it was administered to the whole group rather than to each person on an individual basis. This entailed projecting the four pictures on a screen with a slide projector. Data from students in the control group 15 who were found to have previously been in therapy, and all subjects over the age of twenty—five, were eliminated from the study. The stories were scored using the attraction concept.) This concept views attraction as an immediate contemporaneous ("here and now") rather than a genetic ("central need") concept. This projective technique is designed to make use of the interpersonal, situational influences on projective expression. Thus the test is designed to elicit stories that will reflect immediate, situation—determined themes. According to Libo (1956): The basic approach is one which treats story themes as a direct reflection of the subject's present need-state with respect to approach toward or avoidance of a specific object. For the present purpose, the way in which a client depicts client— therapist relationships in the stories is treated as a direct reflection of his own attraction to his therapist (i.e., the client projects his present need to "belong with" his particular therapist). Thus an individual who is strongly attracted to his doctor should write stories depicting a client's actual or desired movement toward a therapist and actual or eXpected benefit received through contact with him. On the other hand an individual who is only minimally attracted to or repelled by his doctor should write stories describing a client's actual or desired movement away from a therapist and actual or expected dissatisfaction with him. Feelings of frustration or deprivation by the client should be reflected in low attraction scores. Though frus- tration of needs might increase their intensity, the increase should show up in regard to other interpersonal relationships, instead of the present unsatisfactory one. Libo (1956) explains this: 16 This line of reasoning follows from the present view of attraction as a contemporaneous situational phenomenon rather than as a general need. The question must always be: attracted to what? The need for affiliation with whom? In short, the present purpose is to characterize the relationship only of a specific client to a specific therapist. Scoring of stories is set upon the basis of certain criteria. The first of these criteria is that it must include a doctor, therapist or examiner (e.g., medical man, dentist, psychiatrist, psychologist, etc., as Opposed to friend, lawyer, employee, etc.), and an individual client or patient, relating to the therapist for diagnostic or therapeutic functions for the self. When a story does not meet the above it is scored zero. In the actual scoring, a (+1) is given for any indicator of clientmtherapist attraction and a (—l) is given for every indicator of repulsion. There is a total score consisting of an algebraic sum of all "+" and "~" scores. The number of stories which receive a score (0, l, 2, 3, or 4) are considered as the involvement in the task. Libo (1956) feels this in itself may be an indicator of attraction. Both total score and number of stories receiving scores are recorded. The following table summarizes the scoring criteria. Table I.——Scoring Criteria (See manual: Libo, 1956) No. of stories . with a score Total score Interpretation Prediction 2, 3, or 4 +1 or higher Attracted Return Any 0 or lower Not attracted Not return 0 or 1 Any Not attracted Not return 17 There are certain rules for scoring set up by Libo which were followed here. The scorable unit is any word, phrase, clause, or sentence (or combination) qualifying under one of several coding categories. Only one score can be credited to each scorable unit. In scoring, an attempt is made to score the shortest part of a story necessary to meet the criterion for a score, so that other parts of the story will be left available to receive other scores. With regard to series of words, a series in the same category with the same sign is scored as one unit. Each verb-adjective combination is treated as one unit, and when a series can be scored both plus and minus, it receives only one plus and one minus, regardless of the total number of positives and negatives in the series. Either—or statements are scored with regard to each scorable unit there. Conditional statements are not scored and moralistic statements are not scored. Chance events are not scored, and forced events (e.g., "They made him see a therapist") are not scored. Unless things meeting scoring criteria are explicitly mentioned, they cannot be assumed by the scorer. There are five coding categories which are: actual or attempted locomotion (AL), desired locomotion(DL), barriers to desired locomotion (BL), barriers overcome (BLO), and satisfaction (S) (Libo, 1956). Actual locomotion indicates an attempt on the part of the client to remain in contact with the therapist, or strengthen their relationship. It includes movement toward and involvement with the therapist. It is scored (AL) positive and the opposite of these desires 18 is scored (AL) negative. The desire for all of the above is scored as (DL) positive or negative (e.g., desire to remain in contact with the therapist is evidence of desired loco- motion). Barriers to desired locomotion (BL) are indicators of any kind of barrier to actual locomotion not imposed by the client or therapist. In order to score these positive, they must follow an indication of desired locomotion toward the therapist. If they follow an indication of desired locomotion away from the therapist (DL_) then they are scored as negative. Next is the category of barriers to desired locomotion overcome (BLO), which are scored when a barrier to locomotion earlier in the story is overcome. They are scored positive or negative depending on whether or not the barrier to locomotion earlier in the story was positive or negative. The last category is satisfaction (S), which is judged by an indication of benefit, whether actual or eXpected, caused by the therapist or the relationship with the therapist. Indications of benefit are scored as positive (5+) while indications of loss of benefit attributed to the relation_ ship with the therapist or caused by the therapist are scored as negative (5-). The Picture Impressions technique can also be analyzed for qualitative thematic content. The stories are studied to detect motivation for treatment, preferred role for the therapist, and the client's preferred role. Information can also be gathered concerning symptoms, defense problems, and perceptions. RESULTS The reliability of scoring the stories was determined by studying the results of two independent scorers for agreement. Each scorer scored twenty records1 in the eXperiu mental group and twenty records in the control group. The four criteria used were: 1) correlation coefficient for total score, using the Pearson Producthoment Correlation Coefficient; 2) agreement on interpretation of "attracted" or "not attracted" measured in percentage; 3) agreement on "total score" measured in percentage; and 4) agreement on "number of stories receiving a score" measured in percentage. Table A indicates the total breakdown of the above four criteria for the forty records (see Appendix). The Pearson Product—Moment Correlation Coefficient (see Table B in Appendix) was found to be r 69, which is signi- 122° ficant. This result indicates only moderate scoring con- sistency. However, the criterion here requires consistent agreement on the number in "total score" (e.g., the two scores must agree exactly on each protocol in terms of the numerical score), which is a more rigid reliability requirement than originally used by Libo (1957). This is a more stringent criterion than is commonly used in quantitative scoring of 1A constant of +25 was added to each score in this study to simplify arithmetic operations on the data. 19 20 any projective technique, and even moderate consistency is important in terms of this strict criterion. Table II summarizes the last three criteria for relia— bility. The four criteria used to measure reliability in this study would seem to indicate that the scoring system is con~ sistent when used by independent scorers. Table II.-—Summary of Agreement Between Two Independent Scorers on the PIT Using Three Measures of Agreement. INTERPRETATION TOTAL SCORE NUMBER OF STORIES Attrac— Not Attrac- 26 and 25 and RECEIVING A SCORE ted ted more less 2,3, or 4 0 or 1 Scorer A 25 15 33 7 26 14 Scorer B 26 14 33 7 27 13 Percent Agreement 97.5% 100% 97.5% In order to determine a base rate of attraction scores on this instrument, data from the control and experimental groups were compared in the following ways: 1) comparison of means for total experimental versus total control group; 2) comparim son of means for males in the experimental group versus means for males in the control group; 3) comparison of means for females in the experimental group versus means for females in the control group; 4) comparison of means for males in the eXperimental group with means for females in the experimental group; and 5) comparison of means for males in the control group versus means for females in the control group (see Table III). Results indicate that the mean for the total experimental group is significantly greater (P (.01) than the mean for the 21 Table III.—-Summary Table of T-Ratios for Experimental Versus Control Groups. EXPERIMENTAL CONTROL ._ 2 -— 2 N x1 51 N x2 52 t Males 34 29.15 18.87 42 28.83 10.24 2.89a Females 35 28.71 52.98 28 27.88 20.27 0.77 Total 89 28.90 18.24 88 27.24 14.09 2.50a N I" s2 N Y 52 t Experi- mental 34 29.15 18.87 35 28.71 52.98 0.45 Control 42 28.83 10.24 28 27.88 20.27 1.12 = significant at .01 level. total control group. This result supports our hypothesis #2 that students coming to one counseling interview will have greater attraction scores than a group of students who have not had counseling contact. There is also a significantly (F’<.01) greater mean for experimental males as opposed to control males. A significantly greater mean for experimental females versus control females was not obtained although the trend was in that direction. Differences in means between males and females within both experimental and control groups were examined, and it was found that the means were not significantly different. Finally, data from this study was analyzed to determine the predictive validity of the instrument in a college 22 counseling center setting. Libo (1957), in a study of 40 patients coming to a clinic which deals extensively with patients presenting psychosomatic problems, states: A significant relationship between the Picture Impressions judgement (“attracted" versus "not attracted") and the patient's behavior ("return" versus "not return") was found. He later goes on to say: The projective technique seems to have pre- dicted behavior of both those who returned and those who did not return more accurately than one would eXpect from knowledge of the clinic's total return rate alone. (p. 36) From the total experimental group a check was made six months later to determine who had returned for one or more interviews. The Picture Impressions technique classified accurately seven out of twenty—three clients who did not return and thirty—four out of forty—one who did return. This relationship between predicted and actual attendance is not significant (P;>.05) (see Table IV). From the total experi- mental group, the mean of the students not returning Table IV.-~Relation Between Test Prediction and Actual Atten— dance (N=64). Did Not Return Returned Predicted Return 16 34 2 Predicted Not X = 1.54 Return 7 7 (XbNR=29.08) equalled the mean (Xfi:29.08) of the students returning. The total eXperimental group was broken down into 23 two groups in order to investigate sex differences. One of these was thirty-five females and the other was twenty-nine males. In the group of twenty—nine males, the Picture Impressions technique classified accurately four of four— teen male clients who did not return and thirteen out of fifteen male clients who did return. This relationship between predicted and actual attendance is not significant (F’>.05) (see Table V). In the group of thirty—five females, the Picture Impressions technique classified accurately three out of nine female clients who did not return and twenty—one out of twenty~six female clients who did return. This relationship between predicted and actual attendance is not significant (P:>.05) (see Table VI). These results do not support hypothesis #1 of this study on predictive validity. Using a measure of "continuing Table V.——Re1ation Between Test Prediction and Actual Atten— dance for Males. Did Not Return Returned Predicted Return 10 13 2 Predicted Not X = 1-02 Return 4 2 in therapy" (the criterion used was more than two returns necessary to be considered "continuing”), the Picture Impressions technique accurately classified twelve out of thirty-eight students not continuing and twenty-three out of 24 twenty—six students continuing. This relationship (see Table VII) between predicted continuing in therapy versus actual continuing is not significant (Pj>.05). These Table VI.-—Relation Between Test Prediction and Actual Atten- dance for Females Did Not Return Returned Predicted Return 6 21 2 Predicted Not X = 0-75 Return 3 5 Table VII.--Relation Between Test Prediction and Continuing Attendance Did Not Return Returned Predicted Return 26 23 2 X = 3.45 Predicted Not Return 12 3 results do not concur with Libo's (1957) expressed above in his statement on the PIT's judgment and client's behavior, at a significant level, but one can see the trend is in the same direction. If one looks at Libo's (1957) statement that the PIT more accurately predicted behavior than one could find from knowledge of the return rate alone, we find the results here showing the same direction, although the indication is not as great. This study found that: 1) for predicting return for a second interview, the PIT accurately classified 68% of 25 those associated with predicted return and 50% associated with predicted not return, compared to a base rate of 64% returning and 36% not returning; 2) for predicting return for a second interview, the PIT accurately classified 57% of those males (from the total group) associated with predicted return and 66.7% of those males associated with predicted not return compared to a base rate of 52% returning and 48% not returning; 3) for predicting return for a second inter~ view, the PIT accurately classified 78% of those females (from the total group) associated with predicted return and 37.5% of those females associated with a predicted not return compared to a base rate of 74% returning and 26% not return“ ing; and 4) for predicting continuing in therapy (three or more interviews), the PIT accurately classified 47% associated with a prediction of continuing and 80% associated with a prediction of not continuing, compared to a base rate of 40% continuing and 60% not continuing. Low attraction scores for both experimental and control groups contained similar qualitative material. Characteris— tically, stories in these protocols centered around job interviews and problems related to working situations for both males and females. There seemed to be more stories dealing with seeking medical aid from the females. In the control group protocol, the aid was usually being sought'for someone else. Many of these females talked of problems concerning pregnancy but not always in the light of seeking medical or counseling help about them. There was a tendency for those males and females in the experimental group who 26 obtained negative scores to talk about lack of satisfaction, high cost, etc., as a result of an attempt to get self help. The control group also talked of lack of satisfaction from doctors but in a tone more indicative of a general negative attitude to these professionals rather than a specific situation. These people in both groups tended in general to appear to take a somewhat dimmer, more pessimistic view of life. DISCUSSION Of the two quantitative hypotheses of this study, results were significantly in a direction of agreement with one, and were in the predicted direction of the other, although not at a statistically significant level. Let us look at some possible inferences that can be made about the meaning of these results in relation to the theory discussed earlier and purposes of this study. Our first quantitative hypothesis predicted greater scores for students continuing in therapy as opposed to those not returning. Results indicate that the means of the two groups of students (i.e., those who actually returned and those who didn't) were equal. Using Libo's (1957) cri- terion of actual return measured by the client coming for a second interview, we obtained the direction expected but not at a significant level. This was also true when a criterion of "more than two return visits necessary to be classed as continuing in therapy" was used. When one looks at the tables (see Tables IV through VII in Results section), it can be seen that while the PIT classified correctly 34 out of 41 students that returned and 23 out of 26 respectively for the two criteria above, it only classified correctly 7 out of 23 and 12 out of 38 not returning. When these were separated into males and females, results were approximately the same. 27 28 The difference here in our results and Libo's lies mainly in the fact that we had a higher number of people for whom a "predicted return" was the judgment made on the basis of the PIT, who, in fact, did not return. The reason for the fact that many more of the "predicted returns" of the study were in actuality "not returns", using a waiting period of six months, might lie in the nature of our experimental group. We have here in our experimental group college students who express an original interest in psychological help, and this combined with the impact of one person on another seems (when one scans the stories of the experimental group) to produce higher means of attraction scores by itself. This is backed by our findings on the second quantitative hypothesis concerned with the baseline of attraction scores wherein the people who come to the Counseling Center requesting therapeutic counseling and have an initial interview have a significantly (P"<.01) higher mean of their attraction scores than control students who have no contact. One can interpret this finding as indication that the PIT 3 \ 3. ‘2 \,-‘K I \\ / ‘l‘l material which is scored_is influenced by the impact of the relationship just experienced with another person. This is in keeping with Libo's (1957) desire for the test to be influenced by immediate interpersonal relationships. There are situational factors that may also effect why people with high scores "did not return" in actuality: 1) there is a waiting list at the Counseling Center, which can be longer than the interval considered in this study, and during which the students could not get return appointments; 29 and 2) the interval included a period of summer vacation which could have influenced returning in several ways other than interpersonal attraction. The interval used in this study was much longer than Libo's (1957) criterion of returning one week later, thus one would expect a higher possibility of things other than attraction to influence actual return rate over this longer period of time. The possibility should not be overlooked that due to the period of waiting time, students sought further therapeutic counseling elsewhere. In conclusion here, although the direction was the same, we did not obtain the significance of results obtained before, when Libo's original hypothesis was tested in this different setting. Looking at the qualitative content involved in the interviews for the people who were predicted to return but did not return in actuality, there are several notable trends. Two—thirds of the people who had a PIT prediction of return and did not return were in situations where: l) the client felt that he had obtained enough help from the one interview, and just had some "loose ends" he wanted to pull together. He felt he did not need further counseling and could handle things by himself on the basis of what he had obtained from the one interview; 2) the counselor told him there was no need for him to return because he had just needed some things clarified and could handle his own problems in a realistic manner; 3) the counselor felt the client was not yet ready for help from others; and 4) extenuating circumstances such as graduating and leaving the 30 school. If these were dropped from the study (as not appli- cable because of factors other than attraction influencing return), we would find the PIT correctly associated with a predicted return 81% of the time, while the return rate would only be 73% returning. In that situation the results would also be statistically significant. It should be noted that for all of the "predicted returns", the interview preceding the administration of the PIT had a positive flavor, with both client and therapist seeing it as generally beneficial. For all predicted non—returns the opposite seemed to be true. Many of the predicted non-returns who returned seemed to be encouraged to, either directly or indirectly by the therapist. This may have influenced the return rate, in a way not applicable to the use of the PIT, and the concepts it measures. Our second quantitative hypothesis, as already indicated above, predicted correctly the direction and results we obtained. The inference here was made that interest in m) psychological help combined with the impact of one person . on another was the underlying reason for the significantly ii) higher mean of attraction scores of the experimental group. :2 There are several possible implications of this: 1) the impact of one person on another can be the basis of the measure of attraction in the interpersonal relationship experienced through contact with a therapist (even for one initial interview); 2) the decision to obtain help and the interest in it may lend themselves to stories having positive connotations regardless of the immediate recent interpersonal I" r ‘ \ . 31 situation. The latter possibility does not seem as likely as the former when one views the specific qualitative content of the stories. The qualitative aspects of stories with low scores for both eXperimental and control groups show some similarities and some differences for the two groups. There appears to be a trend in the experimental group toward reflecting a specific unsatisfactory interpersonal situation, while the control stories with low scores reflect a more general feeling of ill will toward medicine and therapy of any kind. Content of interviews with peOple in the experimental group who obtained low scores indicated little or no motivation for therapy at the time. Both groups with low scores have a noticeable fraction of people who simply do not get involved in the story telling in a way that reflects their feelings about therapists and therapy. Not getting involved in this way is interpreted as a way of expressing lack of attraction. These are the people who tend to talk about job interviews, work on a job, and disciplinary situations (e.g., professor- student or employer-employee). This, again, seems to be in keeping with the tendency of people who have had a previous individual contact shortly before being administered the PIT to tell stories which seem to reflect a more immediate and specific interpersonal relationship. People with high scores in both groups tended to: l) talk more about medical advice and psychotherapy; 2) express positive feelings about doctors, psychologists, or psychiatrists; and 3) get more involved in general. The control group tended to talk more about medical examination and treatment than psychological help. 32 A general qualitative inspection of all protocols revealed: 1) females in both control and experimental groups told many stories concerned with pregnancy and seeking medical advice in terms of diagnosis of it, etc. This may have been a large contribution to the control group females' somewhat elevated mean, which, while lower than the mean of the experimental females, was not significantly lower; 2) there seemed to be a difference among the four pictures used in the ability to elicit stories of the type desired here. Picture one was noticeably lower in this ability than the other three on the material from this study; and 3) there seemed to be a wealth of material that one might use to speculate on in each individual case concerning present life problems viewed by the person as important and their modes of adaptation, defenses, etc. Thus, in this manner, also, this projective technique might serve as a useful tool for the clinician, perhaps in some ways as useful as the specific predictive aspects studied in this investigation. The findings of this study, on the whole, were in agree— ment with theoretical expectations. The stories told in response to the pictures of this technique seem to produce material which is influenced by the impact of a recent inter— personal relationship. They do indeed reflect preference or lack of it for the kind of experiences that this type of recent interpersonal relationship can provide. There was evidence in this study pointing toward the importance of like- dislike in preference and choice in continuing interpersonal relationships, and as a result of the preference, expected 33 role play of participants in future contacts. The material in these protocols would also be useful to any therapist, to determine if his expectations and his client's were in agreement which (Goldstein, 1962) views as critical for a meaningful relationship to be established. The Picture Impressions technique seems capable of being used as an important tool in effective communication needed between therapist and client. Further investigation might be useful to: 1) determine the relative reliability and validity of each of the four pictures used in relation to the total test; 2) obtain information from similar studies in populations differing from the one used in this one. Further work with the material of this study might shed more light on the actual attraction measuring ability of the PIT if the people who could not return for reasons other than attraction were discarded from statistical analysis. (These would be people such as were discussed earlier who had "predicted returns" on the PIT but for apparent reasons other than client-therapist attraction who did not return.) Libo's population consisted of over 70% women and 70% Negro subjects. Investigation of factors from the influence of differing content of these two popu- lations influencing results might be revealing. SUMMARY This study was concerned with expectations people have of others, and the communication of these expectancies. The specific implication here is for communication in client- therapist interpersonal relationships. An important factor in whether or not an interpersonal relationship is continued is the attraction it has for those involved. While attraction for another person in an interpersonal relationship may be openly displayed, the possibility is frequent that it will not be displayed because of fear of lack of reciprocation or various other things which influence communication of attraction. It has been noted in the past that all projectives tend to be sensitive to the specific interpersonal situation closely connected to the time of test administration. Libo (1957) devised a projective with which he hoped to capitalize on this aspect, and thus be able to measure the specific client" therapist interpersonal attraction from the client's view- point. His instrument, the Picture Impressions technique which provides an attraction score is investigated in this study to: 1) determine a base rate of attraction scores for people in a group administration of the test (i.e., without an interpersonal contact); 2) determine the predictive validity of this instrument in a somewhat different setting 34 35 than Libo used (e.g., a college counseling center); and 3) conduct a qualitative study of low attraction scores for people who had counseling contact and those who had none. Subjects were students at Michigan State University. A total of 137 students was used. There was an experimental group who had been seen at the counseling center for an initial interview before administration of the PIT to them, and a control group of students enrolled in advanced under_ graduate courses in psychology, who indicated they had no previous contact with a counselor and had no desire for any contact in the immediate future. Data from both groups were scored and analyzed. A reliability study for two independent scorers using protocols selected at random indicated that the utilized classification of scored PIT responses are consistent for independent scorers. Results indicated that the mean of attraction scores for peOple (the experimental group) who have had an inter~ personal contact just prior to administration of the PIT have a significantly higher mean in a positive direction (i.e., higher "attraction" score) than people who have had no contact (the control group). The results on predictive validity in this study were not significant although the trend was in the same direction as Libo's (1957). It was hypothesized that the difference in results may be accounted for by: 1) the difference in the populations in this study contrasted with Libo's; 2) a more rigid criterion of what constituted a clear "return" in this study in terms of length of time between interviews; 36 and 3) the effect of conclusion of counseling after only one interview as predetermined by the counselor or client. On the whole, findings were generally in agreement with those of Libo. Further research would seem indicated which might provide more meaningful comparisons than that provided by the two rather different populations (e.g., Libo's versus the one of this study) discussed here. APPENDIX Table A.——Agreement on 3 Scoring Criteria by 2 Independent Scorers a INTERPRETATION OF "ATTRACTED” OR "NUMBER OF STORIES "NOT ATTRACTED" "TOTAL SCORE" RECEIVING A SCORE" Scorer A Scorer B Scorer A Scorer B Scorer A Scorer B l. A A 33 35 4 4 2. NA NA 24 23 4 4 3. A A 28 28 4 4 4. A A 28 27 2 2 5. NA NA 26 26 l l 6. A A 26 26 3 3 7. A A 28 29 3 3 8. A A 29 28 2 2 9. A A 35 31 4 4 10. A A 27 3O 2 2 11. A A 28 34 3 3 12. A A 31 26 4 4 13. A A 32 3O 3 3 14. A A 30 30 2 2 15. A A 30 32 4 4 16. NA A 27 27 1 2 17. A A 31 27 3 3 18. A A 28 27 2 2 19. A A 29 3O 4 4 20. A A 26 27 2 2 a 21. NA NA 25 25 O 0 22. NA NA 26 26 l l 23. NA NA 25 25 0 0 24. NA NA 26 26 l 1 25. A A 26 26 2 2 26. A A 28 28 4 4 27. - NA NA 25 25 0 O 28. NA NA 26 26 l 1 29. A A 26 27 2 2 30. NA NA 28 27 1 l 31. NA NA 29 28 l 1 32. NA NA 28 26 1 1 33. A A 33 30 3 3 34 A A 35 31 4 4 35. NA NA 25 25 O O 36. A A 35 27 3 3 37. NA NA 25 25 0 0 38. A A 39 32 4 4 39. A A 31 30 4 4 40. NA NA 25 25 O 0 a: Items 1 through 20 represent the experimental group; items 21 through 40 represent the control group. 37 38 Table B.-—Computation of Pearson Product-Moment Correlation Coefficient SCORER A SCORER B 2 2 X Y X Y XY 1. 33 35 1089 1225 1155 2. 24 23 576. 529 552 3. 28 28 784 784 784 4. 28 27 784 729 756 5. 26 26 676 676 676 6. 26 26 676 676 676 7. 28 29 784 841 812 8. 29 28 841 784 812 9. 35 31 1225 961 1085 10. 27 30 729 900 810 11. 28 34 784 1156 952 12. 31 26 961 676 806 13. 32 30 1024 900 960 14. 30 30 900 900 900 15. 30 32 900 1024 960 16. 27 27 729 729 729 17. 31 27 961 729 837 18. 28 27 784 729 756 19. 29 30 841 900 870 20. 26 27 676 729 702 21. 25 25 625 625 625 22. 26 26 676 676 676 23. 25 25 625 625 625 24. 26 26 676 676 676 25. 26 26 676 676 676 26. 28 28 784 784 784 27. 25 25 625 625 625 28. 26 26 676 676 676 29. 26 27 676 729 702 30. 28 27 784 729 756 31. 29 28 841 784 812 32. 28 26 784 676 728 33. 33 30 1089 900 990 34. 35 31 1225 961 1085 35. 25 25 625 625 625 36. 35 27 1225 729 945 37. 25 25 625 625 625 38. 39 32 1521 1024 1248 39. 31 30 961 900 930 40. 25 25 625 625 625 Totals2X=ll42 ZY=lll3 ZX2=33,068 £Y2=31,247 iXY=32,024 M ='Y 28.55 27.83 NiXY - (2X) (2y) r = . =.69 l2 ( 11133842102][Ns.‘1(2—