. EFFECTS OF INDMDUAL SUPERViSION 0N SELECTED: 33,, - w AFFECTWh AND coemnvr CHAAAATEAmcs :_~oF COUNQELORS m TRAINING iii A PILOT 3mm " ' Theszs for the Degree of Ph D M!CHICAN STATE umvmsm/ _ ’ JOAN NANCY HAMALHEK AAA/AAA A _ 0242 Michigan State University "3 3 1' . ' i This is to certify that the thesis entitled A Effects of Individual Supervision 0n Selected l Affective and Cognitive Characteristics ' 0f Counselors-ln-Training: A Pilot Study presented by Joan NancyHamachek has been accepted towards fulfillment of the requirements for Ph-nl degree 1n_D.epaLt.me-.n1; of Counseling, .. Personnel Services and Educational Psychology ///‘\ 5&6va Major professor Date Maj ‘9. DH (f 0-7639 ..\ MSU LIBRARIES “ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. _FINES will be charged if book is returned after the date stamped below. g .— ___—.— —_-— _— - -— ‘- impact of affe ness an problem COURSEIl the natl affecti\ these t“ SenSitiv t0 eithe. facilita‘ hYPOthese Supervis,c T ViSors Se ABSTRACT EFFECTS OF INDIVIDUAL SUPERVISION ON SELECTED AFFECTIVE AND COGNITIVE CHARACTERISTICS OF COUNSELORS-IN-TRAINING: A PILOT STUDY BY Joan Nancy Hamachek A major purpose of this study was to measure the impact of the supervising counselor on the deve10pment of affective variables such as empathy, respect, genuine- ness and concreteness and cognitive variables such as problem sensitivity and problem solving ability in the counselor-in-training. A second purpose was to determine the nature and degree of the relationships between these affective and cognitive variables. In order to achieve these two major purposes, an instrument to measure problem sensitivity and problem solving ability was developed. This pilot study was not designed as an attempt to either support or refute any formal interactive- faCilitative process theory, but rather to generate hypotheses for further investigation of this process in supervisory experiences. The sample consisted of twelve counseling super- visors selected from the faculty of a state university and seventeen counselors-in-training being supervised by these Kit (P Functi visors of the perform into hi done by mention median ; median l Scored k Next, th affectiv the ariti ratings . in order Supervisc in the hi median We T Cognitive each of t}. On the p55 anine~mon Joan Nancy Hamachek these twelve supervisors. Two tests, the Problem Solving and Sensitivity Kit (PSSK) and the Scales of Measurement for Facilitative Functioning (SMFF), were administered to the twelve super- visors. Problem solving and problem sensitivity were two of the cognitive measures and these were derived from performance on the PSSK. The supervisors were divided into high and low cognitive groups. The division was done by adding the T-scores of each of the two afore- mentioned cognitive variables in order to determine the median score. Those supervisors who scored above the median were in the high cognitive group and those who scored below the median were in the low cognitive group. Next, the supervisors were divided into high and low affective groups. This division was done by computing the arithmetic averages of their SMFF affective variable ratings on empathy, respect, genuineness and concreteness in order to determine the median. Once again, those supervisors with scores above the median were considered in the high affective group and those who scored below the median were in the low affective group. To note possible changes in affective and/or cognitive functioning as a consequence of supervision, each of the seventeen counselors-in-training were tested on the PSSK and SMFF at the beginning and at the end of a nine-month supervisory period. _ _———'-—- superi superx the St the su measur affect solving level I Pearsor order t affecti changes the beg. a trend R “u\1ts and an 6 provides Problem 5 ‘- l superViSe Joan Nancy Hamachek Analysis of covariance was used to compare the supervisees of the high cognitive supervisors with the supervisees of the low cognitive supervisors and also the supervisees of the high affective supervisors with the supervisees of the low affective supervisors on the measures of empathy, respect, genuineness, concreteness, affective functioning level, problem sensitivity, problem solving I, problem solving II, cognitive functioning _ level I and cognitive functioning level II. In addition, Pearson product-moment correlations were computed in order to examine the relationships between the selected affective and cognitive variables. In order to examine changes in the supervisee's overall counseling style from~ the beginning to the end of his supervisory experience, a trend analysis was employed. Results 229 Conclusions Based on the statistical treatment of the data and an analysis of data trends, six conclusions appear to be worth noting: l. The Problem Solving and Sensitivity Kit (PSSK) provides reliable measures of problem solving ability and problem sensitivity of counselors, as defined by the PSSK. 2. There are significant differences between supervisees trained by high cognitive supervisors and supervisees trained by low cognitive supervisors on measures of TIE of hig respec low cc superv superv; measure lem sol tioning Visees . Visees < empathy, fUnCtior. PIOblem function of the 81 other, in dynamics 6 patterns, functiOniJ affective Joan Nancy Hamachek of respect and affective functioning. The supervisees of high cognitive supervisors gain more on the measures of respect and affective functioning than the supervisees of low cognitive supervisors. 3. There are no significant differences between supervisees trained by high cognitive supervisors and supervisees trained by low cognitive supervisors on measures of problem sensitivity, problem solving I, prob- lem solving II, cognitive functioning I, cognitive func- tioning II, empathy, genuineness and concreteneSS. 4. There are no differences between the super- visees of the high affective supervisors and the super- visees of the low affective supervisors on measures of empathy, respect, genuineness, concreteness, affective functioning, problem sensitivity, problem solving 1, problem solving II, cognitive functioning I and cognitive functioning II. 5. The cognitive scores and affective scores of the supervisors are not significantly related to each other, implying that these dimensions may be supervisor dynamics which function independently of each other. 6. From examination of overall data trends and patterns, the data suggest that unless a supervisor is functioning at high levels on both the cognitive and affective dimensions, his supervisees will tend to either decrease or remain the same in their affective functioning. _.—.-_—-——_ If, h< on Egg visees levels facili cognit high 0 the co: more 1; cognit: Joan Nancy Hamachek If, however, the supervisor is functioning at high levels on bg£§_the cognitive and affective dimensions, his super- visees will tend to increase on their affective functioning levels in the direction of becoming "fully functioning facilitative counselors." In addition, the data suggest that supervisees' cognitive scores are not consistently related to either high or low affective/cognitive supervisors. That is, the consistent and predictable changes in supervisees are more likely to be in their affective rather than their cognitive functioning. EFFECTS OF INDIVIDUAL SUPERVISION ON SELECTED AFFECTIVE AND COGNITIVE CHARACTERISTICS OF COUNSELORS—IN-TRAINING: A PILOT STUDY BY Joan Nancy Hamachek A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services and Educational Psychology 1971 To Our Children Deb, Greg and Dan ii ‘Illl I'll)! l' [II‘ II II:' II ACKNOWLEDGMENTS This thesis was possible through the assistance of Dr. Norman Kagan. I am especially grateful for his intellectual curiousity and his continuous assistance throughout my doctoral program. I also wish to express my appreciation to the other members of my committee, Dr. Arthur Elstein, Dr. William Farquhar, Dr. Sue Jennings and Dr. William Mehrens. I owe thanks to Dr. Maryellen McSweeney for her advice concerning the analysis of the data, and to Dr. Bill Kell, Dr. William Mueller and Dr. Richard Pierce for their constant encouragement and support. The thirty-seven participants in this study were counseling center staff members, interns and practicum students. Special thanks go to each individual for the time and energy he devoted to my study. I am particularly thankful to my husband, Don, who helped me in so many Ways, but, I think, the greatest help was that he expected me to remain "human" during this research endeavor. iii __‘ .— - —.—— _—— LIST OF LIST OF Chapter I II III TABLE OF CONTENTS LIST OF TABLES O O O O O O O O O O O O O O O O O O 0 LIST OF FIGIIRES O O O O O O O O O O O O O O O O O 0 Chapter I THE PROBLEM . . . . . . . . . . . . . . . . Purpose of the Study . . . . . . . } . . Definition of Terms . . . . . . . . . . . Major Research Hypotheses . . . . . . . . Theory . . . . . . . . . . . . . . . . Need for the Study . . . . . . . . . . . Delimitations of the Study . . . . . . . Organization of the Study . . . . . . . . II REVIEW OF THE LITERATURE . . . . . . . . . Evaluation of Counselor Education Programs . . . . . . . . . . . . . . . Impact of Training on Counselor Behavior Counselors and Supervisors as Models . . The Counselor as a Model for Clients The Supervisor as a Model for Counselors-in-Training . . . . . . . Variables Involved in Counseling and Counselor Training . . . . . . . . . . Empathy, Respect, Genuineness and Concreteness as Variables in Individual and Group Therapy . . . Empathy, Respect, Genuineness and Concreteness as Variables in Counselor Training Programs . . . . Cognitive Variables in Therapy and Training . . . . . . . . . . . . . III INSTRUMENT DEVELOPMENT . . . . . . . . . . Theory and Selection of Instrument . . . Description of the Problem Solving and Sensitivity Kit (PSSK) . . . . . . . . Administration and Scoring of the Problem Solving and Sensitivity Kit (PSSK) . . iv vii xi 13 15 20 21 22 24 25 31 34 36 36 39 42 __.__ _- ChaptE IV VI VII Chapter IV V VI VII Initial Pilot Study . . . . . . . . . summary 0 O O O O O O O O O O O O O 0 DESIGN AND METHODOLOGY . . . . . . . . Sample Selection . . . . . . . . . . Supervisor Selection . . . . . . Supervisee Selection . . . . . . Instrumentation . . . . . . . . . . . Scales of Measurement for Facilitative Functioning (SMFF) Scoring . . . . . . . . . . . Standardization Information on Scales of Measurement for Facilitative Functioning (SMFF) Problem Solving and Sensitivity Kit (PSSK) O O I O O O O O O 0 Administration and Scoring . Standardization Information on the Problem Solving and Sensitivity Kit (PSSK) . . Procedure . . . . . . . . . . . . . . Design . . . . . . . . . . . . . . . 'Statistical Hypotheses . . . . . . . Analysis . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . ANALYSIS OF DATA: Hypothesis One . . . . . . . . . . . Hypothesis Two . . . . . . . . . . . Hypothesis Three . . . . . . . . . . Hypothesis Four . . . . . . . . . . . Hypothesis Five . . . . . . . . . . . Summary . . . . . . . . . . . . . . . ANALYSIS OF DATA: Supervisor Groups . . . . . . . . . . High Cognitive--High Affective Group High Cognitive--Low Affective Group . Low Cognitive--High Affective Group . Low Cognitive--Low Affective Group . Summary of the Findings . . . . . . . s WMARY O O O O O O O O O O O O O O O 0 Results . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . Discussion of Results and Conclusions Limitations of the Study . . . . . . STATISTICAL TREATMENT DESCRIPTIVE TREATMENT 47 53 55 55 55 S6 57 57 58 59 59 60 63 63 65 66 68 69 71 71 79 87 88 88 89 92 92 98 101 104 106 110 112 114 116 118 Lb!»- M, ' .M—w—J Chapte BIBLIOA APPEXDL APPENDI APPENDI APPENDI. APPENDIZ Chapter Implications for Further Research . . . . . 119 Implications for Supervision . . . . . . . . 121 BIBLIOGRAPHY O O O O O O O O O O O O O O O O O O O O O 124 APPENDIX A INSTRUCTIONS FOR PSSK, FORM A . . . . . . 133 APPENDIX B POTENTIALLY PROBLEMATIC SITUATIONS EMBEDDED IN PSSK, FORM A . . . . . . . . 135 APPENDIX C SCORING CODE FOR PROBLEM SOLVING ABILITY ON PSSK, FORM A . . . . . . . . . 145 APPENDIX D LETTER TO PARTICIPANTS IN STUDY . . . . . 147 APPENDIX E SCALES OF MEASUREMENT FOR FACILITATIVE FUNCTIONING . O C O O O O O O O O O O O O 148 vi Table FCLVICLCL 3.4 FCLAFI S o 3 AAR l . 4 AFPP 2 O 4 D.Anhd Adam mm 1 2 . . 5 s Table 3.1 5.1 LIST OF TABLES ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SENSITIVITY IN THE INITIAL PILOT STLIDY O O O O O O C O O O O O C O O 0 ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SOLVING I IN THE INITIAL PILOT ST [IDY O O O O O C O C O O O O O O O 0 ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SOLVING II IN THE INITIAL PILOT STUDY 0 O D O O O O O O O O C O C O O RELIABILITY OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II SCORES, WITH AND WITHOUT EFFECTS OF SYSTEMATIC DIFFERENCES REMOVED, BASED ON THE AVERAGE OF SUBJECT'S FOUR SCORES IN THE INITIAL PILOT STUDY . . . . . . . . . . . . . . . . . . . RELIABILITY OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II SCORES, WITH AND WITHOUT EFFECTS OF SYSTEMATIC DIFFERENCES REMOVED, BASED ON AN INDIVIDUAL SCORE IN THE INITIAL PILOT STUDY . . . . . . . . . . . . AVERAGE RATINGS AND RANGE OF SCORES FOR HIGH AND LOW SUPERVISORS ON DIMENSIONS OF EMPATHY, RESPECT, GENUINENESS AND CONCRETENESS . . . AVERAGE RAW SCORE RATINGS AND RANGE OF SCORES FOR HIGH AND LOW SUPERVISORS ON DIMENSIONS OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II . . . . . . . . . . . . . PRE AND POST MEAN SCORES, STANDARD DEVIATIONS AND ADJUSTED MEAN SCORES ON TEN VARIABLES FOR HIGH AND LOW COGNITIVE TREATMENT GROUPS . . ANALYSIS OF COVARIANCE FOR PROBLEM SENSITIVITY SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROIJP S C O O C O O O O O O O C O O O C C C 0 vii Page 49 50 50 52 52 59 62 72 73 Table 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 H1 nrn'n Arr-N Yb! 5:3, 3:.» mm: 533’ Table 5.3 5.8 5.10 5.11 5.14 5.15 5.16 ANALYSIS OF COVARIANCE FOR PROBLEM SOLVING I SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS 0 e e o o o o o o o o o o o o o o o o 0 ANALYSIS OF COVARIANCE FOR PROBLEM SOLVING II SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS 0 o o o o o o o o o o o o 0 ANALYSIS OF COVARIANCE FOR COGNITIVE I SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS. ANALYSIS OF COVARIANCE FOR COGNITIVE II SCORES WITH HIGH AND LOW COGNITIVE SUPER- VISOR GROUPS 0 O O O O O O O O O O O C O O O 0 ANALYSIS OF COVARIANCE FOR EMPATHY SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS 0 O O O O I O O O O O O C O O C C O O 0 ANALYSIS OF COVARIANCE FOR RESPECT SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUP S O C O O O O O O O O O O O O O O O O O 0 ANALYSIS OF COVARIANCE FOR GENUINENESS SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUP S O O O O O O O O O O O O O O O O O O O 0 ANALYSIS OF COVARIANCE FOR CONCRETENESS SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS 0 O O O O O O O O O O O O O O O O C O 0 ANALYSIS OF COVARIANCE FOR AFFECTIVE FUNCTIONING LEVEL SCORES WITH HIGH AND LOW COGNITIVE SUPERVISOR GROUPS . . . . . . . PRE AND POST MEAN SCORES, STANDARD DEVIATIONS AND ADJUSTED MEAN SCORES ON TEN VARIABLES FOR HIGH AND LOW AFFECTIVE TREATMENT GROUPS . . . ANALYSIS OF COVARIANCE FOR EMPATHY SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS . . . ANALYSIS OF COVARIANCE FOR RESPECT SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS . . . ANALYSIS OF COVARIANCE FOR GENUINENESS SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS. ANALYSIS OF COVARIANCE FOR CONCRETENESS SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS. viii Page 73 73 74 74 74 75 75 75 76 80 81 81 81 82 Table 5.17 I‘I} 5.18 2. 1J Cu 5.19 IJCL 5.20 . AAC FCLC .anhC NACC NA 5 21 S 22 5.23 524 6.1 RHA 6.2 punt me 6.3 bmhu mu 6.4 Table 5.17 5.20 5.23 ANALYSIS OF COVARIANCE FOR AFFECTIVE FUNCTIONING LEVEL SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS . . . . . . . ANALYSIS OF COVARIANCE FOR PROBLEM SENSITIVITY SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS . . . . . . . . . ANALYSIS OF COVARIANCE FOR PROBLEM SOLVING I SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROUPS 0 O O O O O O O C O O O O 0 ANALYSIS OF COVARIANCE FOR PROBLEM SOLVING II SCORES WITH HIGH AND LOW AFFECTIVE SIIPERVISOR GROIIPS O O C O O O C I O C O O C 0 ANALYSIS OF COVARIANCE FOR COGNITIVE I SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROIIPS O O O O O O O O O O O O O O O O C C 0 ANALYSIS OF COVARIANCE FOR COGNITIVE II SCORES WITH HIGH AND LOW AFFECTIVE SUPERVISOR GROI-IPS O O I O O C O O O O O O C C C O O O C MATRIX OF INTERCORRELATIONS AMONG BASIC COGNITIVE VARIABLES FOR PRE AND POST SCORES OF SEVENTEEN COUNSELORS-IN—TRAINING . . . . . MATRIX OF INTERCORRELATIONS AMONG BASIC AFFECTIVE VARIABLES FOR PRE AND POST SCORES OF SEVENTEEN COUNSELORS-IN-TRAINING . . . . . MEANS AND STANDARD DEVIATION ON THE BASIC VARIABLES FOR SUPERVISORS AND SUPERVISEES . . RAW SCORES FOR TEN BASIC VARIABLES FOR HIGH COGNITIVE--HIGH AFFECTIVE SUPERVISORS AND THEIR SUPERVISEES O O O O O O O O O O O 0 RAW SCORES FOR TEN BASIC VARIABLES FOR HIGH COGNITIVE--LOW AFFECTIVE SUPERVISORS AND TPIEIR SUPERVISEES O O O O C O O O O O C O O 0 RAW SCORES FOR TEN BASIC VARIABLES FOR LOW COGNITIVE-~HIGH AFFECTIVE SUPERVISORS AND THEIR SUPERVISEES O O O O O O O I O O O O O 0 ix Page 82 82 83 83 83 84 87 88 94 96 99 102 ‘7 Table 6.5 6.6 6.7 Table 6.5 Page RAW SCORES FOR TEN BASIC VARIABLES FOR LOW COGNITIVE--LOW AFFECTIVE SUPERVISORS AND THEIR SUPERVISEES . . . . . . . . . . . . . 105 FREQUENCY OF PRE-TO-POST MEAN SCORES FOR AFFECTIVE AND COGNITIVE VARIABLES . . . . . . 107 FREQUENCY OF PRE—TO-POST RAW SCORE CHANGES FOR AFFECTIVE AND COGNITIVE VARIABLES . . . . 108 Figure 3.1 3.2 3.3 LIST OF FIGURES CONTENTS OF PSSK IN-BASKET, FORM A . . . . SCHEMATIC DIAGRAM OF THE RESEARCH SETTING INITIAL PILOT STUDY TESTING SEQUENCE DESIGN xi Page 41 44 49 F Bergin (l professic in produc levels. and Kelln their res does inde tional St 0f the 'q] some fOrm resUlts u that We 1 therapy a aCCUIatel therapeut by Carkhu patients Peutic en that faCi Carkhuff explicate CHAPTER I THE PROBLEM Research evidence presented by Eysenck (1952), Bergin (1963) and Levitt (1957) suggest that the helping professions are not always as effective asthey could be in producing improvement in clients at all developmental levels. Rogers and Dymond (1954), Carkhuff and Truax (1966) and Kellner (1967), on the other hand, have shown from their research efforts that psychotherapy or counseling does indeed make a difference toward improving the emo- tional status of clients. In any case, the consequences of the "help" received by clients who have experienced some form of therapy is not always clear nor are the results universally agreed upon. It seems imperative that we increase our knowledge about both the process of therapy and the training of therapists in order to more accurately assess the consequences and prognosis of therapeutic endeavors. For example, the review of research by Carkhuff and Truax suggests that students, clients and patients may be hindered as well as helped in the thera- peutic encounter. While the literature on the conditions that facilitate psychological change has been growing, Carkhuff and Truax note that, "there is still a need to explicate those process variables that facilitate positive 1 movemel 725). a crit: the im; a nine- Purpose the imp the dev acteris the int solving HESS an Study, is the lying t' ability fOr 900. that tr where t: modifiai meaSUre movement and those that inhibit this goal" (1966, p. 725). In an effort to examine what may very well be a critical "process variable," this study will investigate the impact of supervision on counselors-in-training over a nine-month period. Purpose 2; the Study The purpose of this study is twofold: 1) examine the impact of the supervisor's counseling "style" on the development of specific counseling-related char- acteristics in the counselor-in-training and 2) examine the interrelationships of problem sensitivity, problem solving ability, empathic understanding, respect, genuine- ness and concreteness and their change over time. In addition to the experimental aspects of this study, there are several descriptive aspects. One aspect is the hypotheses generating nature of the design. Under- lying this study is a search for knowledge of the modifi- ability of various characteristics considered important for good counseling skills. Implicit is the assumption that training programs can be developed more effectively where there is an awareness of what characteristics are modifiable. Developing a valid and reliable instrument to measure problem sensitivity and problem solving ability ___._—.___‘ of COUflSl search. measure ‘ administ: Saunders Definiti< consistex reacting 1 minute 5. 0f empat] "Accurat. sensitiv to C0mmu: to the c minute 3 °f reSpe Regers ( Dist's w hepes 0r minUte S of counselors is itself one of the purposes of this re- search. Several scales have already been developed to measure these variables for use with engineer managers, administrators and teachers (Joyce, 1970; Frederikson, Saunders and Wand, 1957; Shulman, Loupe and Piper, 1968). Definition of Terms Style is the term used to denote a subject's consistent mode of initiating, conducting, relating and reacting to specific problematic situations. Empathy level is the average rating of 3-three minute segments of a counseling interview on the measure of empathic understanding in interpersonal processes. "Accurate empathy" is judged on both the therapist's sensitivity to current feelings and his verbal facility to communicate this understanding in a language attuned to the client's current feelings (Shapiro, 1969, p. 350). Respect_level is the average rating of 3-three minute segments of a counseling interview on the measure of respect or positive regard in interpersonal processes. Rogers (1962, pp. 420-422) describes respect as the thera- pist's willingness to share equally the client's fears and hOpes or achievements and failures, without placing condi- tions upon the warm acceptance of the client's inner self. Genuineness level is the average rating of 3-three minute segments of a counseling interview on the measure of facilit The thera; ness verbs response 1' (1962, pp. therapist going on i ship. 3 3-three mj measure 01 0f EXpres: and complq the theta; of facilitative genuineness in interpersonal processes. The therapist functioning at a high level of genuine- ness verbalizes positive cues that indicate a genuine response in a nondestructive manner to the client. Rogers (1962, pp. 417-419) has observed that to be genuine the therapist must allow his client to see everything that is going on in the therapist which is relevant to the relation- ship. Concreteness level is the average rating of 3-three minute segments of a counseling interview on the measure of personally relevant concreteness or specificity of expression. High concreteness is the fluent, direct and complete expression of specific feelings and experi- ences regardless of emotional content expressed by either the therapist or the client (Carkhuff and Berenson, 1967, p. 7). Affective functioning level is the arithmetic average of the ratings for empathy, respect, genuineness and concreteness. Problem refers to a psychological state of dis- comfort or disequilbrium sensed by an individual. This discomfort might be caused by: l) discrepancy between anticipated and an encountered event, 2) imbalance gen- erated by the gap between desired and actual conditions, that is, between intended goal and a current status or 3) the ambiguity resulting from contradictory sources of information in a situation. F the envir thus havi problemat potential to feelin been sens 230 poten g 0f potent to (sense 3 resolutio to three . PIOblem a. that are . degree of SubjeCt b: were not a for that l how Well a seleCted 13 scores in the deepes respect t materials Pr< \ Potential problem refers to a configuration of the environment that is intrinsiCally indeterminate, thus having a high likelihood of being perceived as problematic by an individual encountering it. When the potentially problematic situation is encountered and leads to feelings of disequilibrium, we say that a problem has been sensed. Each form of the PSSK has approximately 230 potential problems. Problem sensitivity is a measure of the number of potentially problematic elements in the PSSK reacted to (sensed) as problems by the subject. Problem solving I is a measure of overall problem resolution. The subject is rated on a scale from zero to three on each of ten major problem areas. Each major problem area encompasses a number of the potential problems that are imbedded in the PSSK. The rating depends on the degree of completion or comprehensiveness to which the subject brings his problem resolutions. If the problem were not attempted, the subject is given a score of zero for that problem. The total score is assumed to indicate how well a subject understands the nature of the ten selected problem areas in the situation. Those with high scores in problem solving ability I are assumed to reach the deepest and most complete level of understanding with respect to the general problems embedded in the test materials.° Problem solving ability II is a measure of specific problem divided ‘ the prob fleets ti problems with big to come standing ject sel average . sensitiv 1 average . sensitiv problem resolution. The problem solving I total score is divided by the number of problems attempted to arrive at the problem solving ability II score. This score re- flects the depth of the problem resolution in only the problems that the subject attempts to resolve. Those with high scores in problem solving ability II are assumed to come to the deepest and most complete level of under- standing with respect to specific problems that the sub- ject selects to resolve. Eggnitive functioning level I is the arithmetic average of the T scores of problem solving I and problem sensitivity measures. Cognitive functioning level II is the arithmetic average of the T scores of problem solving II and problem sensitivity measures. Major Research Hypotheses The hypotheses tested in this study are state- ments of how the behavior of a counselor-in-training changes as a consequence of his supervision experience. The basic hypotheses are as follows:* 1. The supervisors who scored high on measures of empathy, respect, genuineness and concreteness will have supervisees who score higher on these variables than will the supervisees working with supervisors who scored * Research hypotheses are restated in testable form in Chapter IV. -—d-——--——i——b low on t1 of proble supervise the supei on these between t variables 4 the empat ables. 5 between t nitive Sc 6 Will beCc Tiesrx T Carkhuff, 1967‘ Car study Car lsnsOlVi four dime Condition low on these measures. 2. The supervisors who scored high on measures of problem solving and problem sensitivity will have supervisees who score higher on these variables than will the supervisees working with supervisors who scored low on these measures. 3. There will be positive intercorrelations between the problem solving and problem sensitivity variables. 4. There will be positive intercorrelations among the empathy, respect, genuineness and concreteness vari- ables. 5. There will be no significant correlation between the supervisors' affective scores and their cog- nitive scores. 6. The supervisee's overall counseling "style" will become more like his supervisor's "style" over a nine- month training period. Theory The theory undergirding this study grows out of Carkhuff's model for predicting facilitative growth (Carkhuff, 1967: Carkhuff and Berenson, 1967, pp. 44-60). For this study Carkhuff's theory has been modified to include prob- lem solving and problem sensitivity in addition to the four dimensions described by Carkhuff as the major core conditions for effective interpersonal communication, which he concreter 1 positive movement the a in re signi to th patho couns ditio tion C condition all learn Secondary action of Situation to facili T by Carkhu trainer f dimension dimensiOn (Carkhuff which he described as empathy, respect, genuineness and concreteness. This modified theory can be used to predict positive movement and gain as well as to predict negative movement or deterioration. Carkhuff and Truax write, the absence or low levels of facilitative conditions in relationships with parents, teachers and other significant figures in all likelihood contributes to the development of the difficulty or psycho- pathology in the first place. It makes sense that counselors offering a continuation of these same con- ditions will continue to produce further deteriora- tion (1966, p. 726). Carkhuff's model suggests that a primary core condition such as empathic understanding is critical to all learning and relearning processes. In addition, secondary dimensions peculiar to a particular inter- action of lst person (therapist), 2nd person (client) and situational variables (environmental settings) may Operate to facilitate or retard the outcomes of the primary pro- cess variables. Three critical classes of variables are encompassed by Carkhuff's model of counselor training: 1) level of trainer functioning on facilitative and action-oriented dimensions, 2) level of trainee functioning on relevant dimensions and 3) type of training programs operationalized (Carkhuff, 1968; Carkhuff, 1969). According to Carkhuff, the most critical variable in effective counselor training is the level at which the counselor-trainer is functioning on the facilitative and action-oriented dimensions related to constructive client change (1969). The present study focuses but 815! sensiti' that box part be is func1 Differer functior accordir 0f both Persons at lowex Variable experier Kell anc Change , process when th. increas' Cogniti. ”tents Hosford focuses primarily on these dimensions of counselor growth but also includes problem solving ability and problem sensitivity as dimensions of interest. An implication of Carkhuff's modified theory is that how we interact with a given individual may in large part be determined by the level at which each participant is functioning on the affective and cognitive variables. Differential predictions concerning gains in interpersonal functioning by the counselor-in-training may be generated according to discrepancies in initial level of functioning of both the counselor and the counselor—in-training. Persons at higher levels of functioning can help persons at lower levels to acheive higher levels on these and other variables (Carkhuff, 1967). A goal of psychotherapy is change and change is experienced through feeling, thinking and interacting. Kell and Burrow (1970), for example, have noted that change occurs when feelings are associated with the thought process, but that this change does not necessarily occur when thinking or feeling alone is involved. There is increasing research attention being given to evaluating cognitive and affective factors which may affect a student's potential as a facilitative counselor (Carkhuff, 1966; Hosford and Briskin, 1969; Pierce and Schauble, 1970; Truax and Carkhuff, 1967; Anderson, 1968; Martin and Carkhuff, 1968). It is conceivable that important changes might take pla but that Need for source 0 counselo that the developi counseli further function Visees. imPortan Process, Should d and wi U) W Study is was take in—train trainers and C011: at Michi differeh 10 take place as a function of graduate training experiences but that these changes will not be uniform for all students. E229 m the _xStud While research has not identified the primary source of influence on the counseling behavior of counselors-in-training, the contention of this study is that the individual supervisor has a major impact on developing those characteristics relevant to effective counseling behavior of counselors-in—training. It is further argued that the manner in which the supervisor functions can influence the development of his super- visees. If in fact, the individual supervisor plays an I important and significant role in the counselor-training process, then this may have direct implications for who should do individual supervision, how it can best be done and with which counselors-in-training. Delimitations 9f_the Study Any generalizations which can be made from the study is limited to the population from which the sample was taken, namely, supervising counselors and counselors- in-training at Michigan State University. Counselor trainers and counselors-in-training from other universities and counselors not engaged in supervisory duties, either at Michigan State or other universities, may have had different academic experiences. Without a replication study t1 conclus design, to resu, partic11 and Sen; While 1' availabl come-va; and pro] the lmp< Cretene: FaCilit. t0 sugg, ample, Rat. COn. the the Und. Therefo‘ SolVing 11 study the conclusions must be considered only as tentative conclusions and results must be accepted with caution. In addition, because of the pretest and posttest design, findings of this investigation cannot be compared to results of subsequent studies unless the subjects also participate in a similar experimental design. It should also be noted that the Problem Solving and Sensitivity Kit (PSSK) is an unresearched instrument. While it appears to have face validity there is no data available relating the cognitive measures to other out- come-variables or to other measures of problem sensitivity and problem solving. Although there has been increasing support for the importance of empathy, respect, genuineness and con- creteness, as measured by the Scales of Measurement for Facilitative Functioning, there has been little evidence to suggest what the scales "actually" measure. For ex- ample, Shapiro states that, Raters' judgments may be based largely on behavioral concomitants of the types of utterance specified by the ”definitions" rather than on the extent to which the therapists' utterances are formally subsumable under the "definition" (1969, p. 352). Therefore, similar findings would not be expected if instruments other than the PSSK and SMFF were used to measure the concepts of problem sensitivity, problem solving, empathy, respect, genuineness and concreteness. In summary, this study is delimited by the nature of the S Organiza literatu‘ areas: grams, 2 counselo: affective Chapter 1 descriptj The desig Chapter 1 comparisc counselo; Variables counseloI groUps. Variables and affec analysis VII Cont; from the 12 of the sample studied and to the instruments used. Qgggnization g: the Study . The following chapter includes a review of the literature related to the study. The review is of four areas: 1) evaluation of existing counselor education pro- grams, 2) impact of training on counselor behavior, 3) counselors and supervisors as models and 4) cognitive and affective variables involved in counselor training. Chapter III contains a report on the development and description of the Problem Solving and Sensitivity Kit. The design and methodology of the study is examined in Chapter IV. The analysis begins in Chapter V with a comparison of counselors-in-training being supervised by counselors who score high and low on the affective variables. This is followed by a comparison of the counselors-in-training in the high and low cognitive groups. Finally, the correlations between cognitive variables, affective variables and supervisors' cognitive and affective scores are examined. The descriptive analysis of the data is reported in Chapter VI. Chapter VII contains a summary of the study and conclusions drawn from the results. evaluate a studen huff, 19 1970; Tr vestigat counselo counselo focuses . counselo CounSelo and 4) c CounSelo clarify prESent duc ' w nOtable 0Ver the was °nly CHAPTER II REVIEW OF LITERATURE Increasingly, research is being undertaken to evaluate cognitive and affective factors which may affect a student's potential as a facilitative counselor (Cark- huff, 1966; Hosford and Briskin, 1969; Pierce and Schauble, 1970; Truax and Carkhuff, 1967). Since this study in- vestigates the impact of counseling supervisors on counselors-in-training during a nine-month period in a counselor training program, the review of literature focuses on the following areas: 1) evaluation of existing counselor education programs, 2) impact of training on counselor behavior, 3) counselors and supervisors as models and 4) cognitive and affective variables involved in counselor training. In addition, an attempt is made to clarify the specific issues of direct relevance to the present study emerging from the review. Evaluation g£_Counselor Education Programs According to Whiteley (1969), there had been notable accomplishment in the area of counselor education over the past three years. But he also stated that there was only the barest knowledge about some of the central 13 issues 1 educatio l966; Wh had been For exam 3 cross related deterior of train training function 0f empatl ing to C the high. tended t and Supe use in c [“0 SPEC have bee SuperVis institut the mini recently currentl eXpreSSe 14 issues in counseling. Evaluating the effects of counselor education programs had received scant attention (Carkhuff, 1966; Whiteley, 1969) and, in fact, when such evaluation had been carried out, the results were not encouraging. For example, Carkhuff, Piaget and Pierce (1968) reported a cross sectional study which found that, on dimensions related to constructive client change, psychology trainees deteriorated in functioning from the beginning to the end of training. An analysis of a rehabilitation-counselor training program by Anthony (1968) showed a low level of functioning among graduate students across the dimensions of empathy, respect, genuineness and concreteness. Accord- ing to Carkhuff (1969), counselor trainees functioning at the highest levels of empathy, respect and genuineness tended to deteriorate over the course of training. In 1967 the Association for Counselor Education and Supervision adOpted a revised set of Standards for use in counselor education. However, Stripling noted that, "no specific criteria for accrediting counselor education have been developed; and, in many cases, no qualified supervisor, counselor or counselor educator is on an institutional visiting committee" (1968, p. 201). Because the minimal standards in counselor training had only recently been established and accreditation as it was currently practiced lacked substance, Whiteley (1969) expressed a need as recently as 1969 for a careful study of what enhance similar tionshi underst grams w counsel and, in the reS' gating ' that th. the beg dimensi. alga: \m that on graduat MiSChel be have about p graduat counsel attitUd signifil 15 of what counselor education should and could be doing to enhance skill development in counselors-in-training. A similar plea that studies of behavior in counselor rela- tionships become a more prominent aspect of efforts to understand, evaluate and improve counselor education pro— grams was made by Schoch (1966). In essence, the evaluation of the effects of counselor education programs has received scant attention, and, in fact, when such evaluation has been carried out, the results are discouraging. Previous research investi— gating the impact of counselor training programs suggests that the participants in training programs deteriorate from the beginning to the end of the training programs on dimensions that are thought to be critical in counseling. Impact 2: Training 22 Counselor Behavior A review of the literature on consistency reported that only minimal and superficial changes occurred during graduate counselor training programs (Mischel, 1969). Mischel also suggested in his review that counselors-to- be have a "style" of approaching, dealing with and feeling about people that was quite stable by the time they entered graduate school. In addition, Kassera (1968) found that counselor education had only a modest impact in changing attitudinal orientations of students. After one semester significant changes were found only on a measure of acceptar and role empathy 118 sub; or cont} there we the beg: (1969) 1 client 1 client 1 fluence verbal : SUbjects sistanc, ourSEIVE probl em conSistG disorgar evidence the hum; 9ff8cti\ the Perc petualii Consists 16 acceptance of others. Dell (1967) reported a study in which lecturing and role-playing were used as strategies to increase empathy in a group of college sophomores. A total of 118 subjects were placed in either role-playing, lecture or control groups. Dell found that after seven sessions, there were no group changes in their empathic levels from the beginning to the end of the training program. In addition, a study by Hountras and Redding (1969) reported no significant differences on client initiated indirect talk ratio or ' - CO ns '11.. client response direct u elor 1 fluence ratio when fifteen subjects were trained in verbal interaction analysis and compared with fifteen subjects who received no training. Mischel (1969) examined the stability and re- sistance to change of "cognitive constructions about ourselves and the world" and of cognitive styles in problem solving. He observed that the loss of behavioral consistency may be a chief characteristic of personality disorganization. His interpretation of the overall evidence from reviewing the topic of consistency described the human mind as functioning like an "extraordinarily effective reducing valve" that created and maintained the perception of continuity even in the face of per- petually observed changes in actual behavior. Because consistency in the cognitive and intellective domain was easier persona to than by Kell Reighar ever, h as the direct counsel an inst trainin. measure 0f frie: and stai measuIe, any of 1 (1968), Characte PrOgram' determin ifiCally training HIOre 11nd 17 easier to establish than in the personality and inter- personal behavior domain, he hypothesized more resistance to change in the cognitive realm than the affective domain. No change in cognitive complexity, as measured by Kelly's Role Construct Repertory Test, was found by Reighard (1968) as a result of a practicum course. How- ever, he did report an increase in verbal behaviors, such as the use of negators, qualifiers, retractors, evaluators, direct questions and expressions of feelings. Gallagher (1968) investigated the changes in counseling and mathematics trainees that occurred during an institute training program. At the end of the institute training program the counseling trainees scored higher on measures of pasSivity and moodiness and lower on measures of friendliness, persistency, sensitivity, predictability and stability than they had scored on pre-institute measures. The mathematics trainees showed no changes on any of the measures. On the other hand, an investigation by Ashmore (1968), related to changes in attitudes and personality characteristics among counselors in a counselor education program, offered evidence to refute the notion of early determinism of personality characteristics. More spec- ifically, Ashmore reported that counselors in an institute training program changed in several ways. They became more understanding, less probing and less supportive, as measurec addithl pretive Personai needs fl measure: decreas. less de more by; reflecti 18 measured by the Helping Relationship Inventory. In addition, the female counselors also became more inter- pretive as measured by this instrument. On the Edwards Personal Preference Schedule the males decreased in their needs for endurance and abasement, and increased on measures of heterosexuality and dominance. Females decreased in measures of endurance. Males reflected less defensive, depressed and psychosthenic behaviors and more hypomanic behavior of the MMPI while the females reflected less defensive and paranoid behaviors. Over the past twenty years many divergent points of view have developed in relation to the teaching and learning of psychotherapy under supervision. Several studies have been reported that suggested specific kinds of training programs in order to modify counseling be- haviors. For example, Reddy (1969) concluded that the group receiving immediate feedback became more empathic than the delayed-feedback or no-feedback groups.when thirty- six subjects were divided into three equal groups and each group was given a different type of feedback on their empathic behavior. The delayed-feedback group was not significantly different from the no-feedback group on the measures of empathy. In addition, the immediate-feedback and delayed-feedback groups increased in the number of reflective responses made and also gave fewer supporting and advice giving statements at the end of the training .‘ — . —.- W. . program no chang ments, 0 treatmen warmth a on measu pre-to-p Jordan f treatmer data the change c treatmer with rec aSpects Carkhuf: 1966? T: Goldberc trainee learnin, 19 program than was the case during pretesting. There were no changes in the no-feedback group. Jordan (1968) compared two training group treat- ments, one experiential and one didactic, with a no— treatment group on measures of empathy, non-possessive warmth and genuineness. Both treatment groups gained more on measures of empathy and non-possessive warmth from pre-to-posttesting than did the no—treatment group. While Jordan found no significant differences between the treatment groups, it was tentatively concluded from the data that the didactic treatment had produced a greater change on all the measures than did the experiential treatment. Several research investigations have concluded with recommendations to combine didactic and experiential aspects into a counselor training program (Dahmen, 1967; Carkhuff and Truax, 1965b; Berenson, Carkhuff and Myrus, 1966; Truax, Carkhuff and Douds, 1964; Grzegorek, 1970; Goldberg, 1967) with one assumption being that the trainee's growing awareness of intellectual content and learnings in psychotherapy in the context of a relation- ship which nurtures his own self-exploration would lead to his growth as a therapist. In summary, little is known about the kinds of systematic changes that can occur in counselor training programs and the research that is available reports conflicti: counselor suggest t1 counselor. On the ot‘.‘ that repo acteristi I suggest t has is de have reco perientia \ p (I) when Sions att that 0's aSpeCtS Heller (: be impom COlll'lselOJ of resea; CEdureS 20 conflicting evidence on the impact that training has on counselor behavior. On the one hand, some investigators suggest that counselor education has minimal impact on counselors-in-training, especially in the cognitive realm. On the other hand, more optimisitc data are presented that report changes in attitude and personality char- acteristics as the result of counselor educator programs. Increasingly, research is becoming available to suggest that the impact that a counselor training program has is dependent on the program. Several investigators have recommended programs that combine didactic and ex- periential aspects. Counselors and Supervisors g§_Models An observer (0) is said to model an individual (I) when observation of the behavior of I, or of expres- sions attributing certain behavior to I, affects 0 so that O's subsequent behavior becomes more similar to the observed behavior of I (Flanders, 1968). In his review of the literature covering various aspects and applications of modeling and identification Heller (1969) concluded that while modeling processes can be important facilitators of change, application to counselor-client relationships was impeded by a paucity of research investigating the operation of modeling pro- cedures in clinic-like settings. the clil therapi: his the: more ad; this prc attemptc therapy "Wheneve likely 1 They als the cor: SPeCific For exam beCBME n liter-atL Came mo; while SC this was dUring c modifiec‘ sistent 21 The Counselor as a Model for Clients Bandura (1961) suggested that during psychotherapy the client learned to incorporate certain aspects of the therapist's intact ego. That is, he became more like his therapist with regard to values and mannerisms and more adaptive in his own behavior. According to Bandura, this process occurred whether or not the therapist directly attempted to transmit his values and attitudes. Concluding from data drawn from eight, year—long therapy cases, Lennard and Bernstein (1960) observed that, "whenever persons freely interact with each other, it is likely that they will become more and more alike" (p. 249). They also found that over time there were increases in the correlations of counselors and their counselees on specific kinds of patient and counselor communications. For example, verbal behavior and affective communications became more alike. Bednar (1970) concluded from his review of the literature on persuasion that some improved clients be- came more like their counselors in the counseling process while some unimproved clients did not. He noted that this was, "support for the concept of a convergence effect during counseling" (p. 651). According to Rosenthal (1955), improved clients imodified their system of moral values in a direction con- sistent with the beliefs of their counselor. Sixty statemer by the < therapy. improven moral va visor's by Under student Sensitiv ship bet empathy influenc in Each Hm main with the reSPECt Sistentl Supervis theraDis 1969). 22 statements about sex, aggression and authority were sorted by the counselors and their counselees before and after therapy. Rosenthal reported that the rho correlation of improvement and change in the direction of the therapist's moral values equaled .68. The Supervisor as a Model for Counselors-in-Training A study that examined the influence of the super- visor's personality on trainee's perceptions was reported (by Underhill (1968). Using videotape exerpts to rate student teachers and their supervisors on the Affective Sensitivity Scale, Underhill found a positive relation- ship between a student teacher and her supervisor's empathy at the end of the student teaching experience. Hansen and Barker (1964) also investigated the influence of the supervisor's personality and found that in each of three groups being studied, the supervisor was the main affecting variable. The trainees in the group with the counselor rated highest on levels of empathy, respect and genuineness by his trainees were rated con- sistently higher than the trainees of the other two supervisors by judges on the Experiencing Scale. A study was conducted at the University of Kansas to investigate the relationship between the student- therapist and his supervisor (McAllister and Neuringer, 1969). Fifteen counselors-in-training and two supervisors were the minimum measures of speci The ider larity, correlat gators c mechanis personal 0f funct ness and reported Studies availabl mOre lik meaSUreS from the ChaDQES to the C thereDis area is some of Very Spe situatio 23 were the subjects in the study and each student had a minimum of three months of continuous supervision. Five measures of identification were correlated with measures ‘of specificity, initiative, relevance and warmth-acceptance. The identification measures referring to specific simi- larity, either real or assumed, showed significant correlations with the criterion-measures. The investi- gators concluded that identification was a modeling mechanism which appeared to operate on very specific personality variables in very circumscribed situations. The differential effects of supervisor's level of functioning on measures of empathy, respect, genuine- ness and concreteness upon counselors-in-training was reported by Carkhuff (1969). He reviewed the available studies and concluded that in cases where data were available, the trainees moved in the direction of becoming more like their trainers on levels of functioning on measures of empathy, respect, genuineness and concreteness from the beginning to the end of training. In general, the literature would suggest that changes do occur and these changes are oftentimes related to the client's or supervisee's interactions with his therapist or supervisor. Much of the research in this area is still in the exploratory stages of investigation. Some of the studies on modeling appear to Operate on very specific personality variables in very circumscribed situations. Variable and Coun standing selor an that beh aspects. striving: referred involved affectiv< to under: understax the clier I "free flc necesSar) 0f underE E h‘COUnse Viewed th that empa PartiCula was more C . fOur "lees 24 Variables Involved ig_Counseling and Counselor Training Both cognitive and affective aspects of under- standing are manifest in the interactions between coun- selor and client. Bordin (1968, pp. 166-182) suggested that behavior should be looked at as possessing these two aspects. The affective aspect referred to people's strivings, feelings and emotions and the cognitive aspect referred to the conceptual, perceptual and motor processes involved in response to the pressures inherent in the affective aspects of behavior. Bordin further stated that to understand the client completely the counselor must understand both the cognitive and affective aspects of the client's communications. Fenichel (1941) argued that some balance between "free floating" and compulsive intellectualism was necessary for a therapist to provide the greatest degree of understanding of his client. Empathy has been the focus of numerous studies in counseling and psychotherapy. Gladstein (1970) re- viewed the studies on empathy in counseling and concluded that empathic type responses without regard for the particular client was often inappropriate; that is, there was more to counseling than empathy. Cooper (1967) investigated the relationship between four measures of empathy and nine measures of cognitive control . rigid, m« than ind The othe. to the mo [71 Is» [23. sufficie: change t. reGard (. lieved t1 sufficie: tion to 1 a major ; therapeu Whitehor BetZ, 19 Their ma Psychiat ment rat SEVen 0t imprOVem 10wer th phrenics SUCCeSS 25 control and found that individuals high on empathy were less rigid, more field analytic and better at cognitive shifting than individuals who scored low on the measures of empathy. The other measures of cognitive control were not related to the measures of empathy. Empathy, Respect, Genuineness and Concreteness as Variables in Individual and Group Therapy Rogers (1957, 1962) recognized three "necessary and sufficient" conditions in order for constructive personality change to occur. These conditions were genuineness, positive regard (respect) and empathy. While few researchers be- lieved that any three therapist characteristics would be sufficient to account for the therapist's total contribu- tion to patient outcome, Rogers' theoretical formulation was a major stimulus for research in this area. Research related to the role of the therapist in therapeutic outcome grew out of the pioneering work of Whitehorn and Betz at John Hopkins Hospital (Betz, 1963a; Betz, 1963b; Whitehorn, 1964; Whitehorn and Betz, 1954). Their major contribution was a retrospective study of seven psychiatrists whose schizophrenic patients had an improve- ment rate of seventy-five per cent, as contrasted with seven other psychiatrists of similar training who had an improvement rate of only twenty-seven per cent, a percentage lower than the improvement rate found in groups of schizo- phrenics with no treatment. Analysis showed that the success of the patients appeared to be dependent on the in was be1 sct lee of the of pat In Con low to 1 Were thei OUtc fro“ SUcc WEre 26 the "type" of counselor seen for help, but the differences in the two "types" were not systematically explored. It was concluded that psychotherapy could indeed by "for better or for worse" (Truax, 1963, p. 256). A study of psychotherapy with sixteen hospitalized schizOphrenic patients was started in 1958 under Rogers' leadership at the University of Wisconsin. The purpose of the research program was to study the effects of the therapist's levels of: l) accurate empathic understanding of the patient, 2) unconditional positive warmth for the patient and 3) therapist self-congruence or genuineness. In 1963 Truax reviewed the findings of the study to date. Comparisons of the levels of therapist conditions offered during therapy with measures of constructive personality change in the patient, using a matched control group, suggested that: 1) high levels of therapist-offered conditions were related to patient improvement, but 2) low levels of therapist-offered conditions were related to patient deterioration. These and additional findings were reported by Truax and Carkhuff in 1967 (pp. 80-143). One of the first studies attempting to relate the therapist's level on the therapeutic triad to patient outcome was by Halkides (1958), who selected brief samples from early and late therapy interviews from ten most sucCessful and ten least successful therapy cases. Ratings were made on the therapist's levels of empathic understanding 27 unconditional positive regard and genuineness. She reported that the therapists of the most successful cases showed significantly higher levels on the three conditions than did the therapists of the least successful cases. The work of Barrett-Lennard (1962) supported the relevance of the counselor characteristics of empathy, respect and genuineness for success with counseling center cases. However, it should be noted that a replication of these studies by Hart (1960), using the Halkides' data and similar procedures but different judges, failed to confirm Halkides' and Barrett-Lennard's findings. Truax, Carkhuff and Kodman (1965) attempted to evaluate the generality of empathy by studying its relation- ship to outcomes in group psychotherapy. The study in- volved forty hospitalized mental patients, all relatively chronic, who were given group therapy sessions twice weekly over a three-month, time-limited period. The patients were divided into equal groups receiving high or low levels of empathy and those receiving high levels of empathy showed improvement on all the MMPI subscales equal to, or greater than, that of the patients receiving low levels of empathy. In still other studies (Rogers, 1962; Truax, 1963; Truax and Carkhuff, 1964b; Truax, 1961), rating scales were used to divide hospitalized schizophrenic patients according to the level of empathy, respect and genuineness 28 provided by their therapists and it was found that patients receiving high levels of empathy, respect and genuineness demonstrated significant process movement and constructive personality and behavioral change. Patients who received low therapeutic conditions did not become engaged in positive therapeutic process movement and actually deteriorated on the outcome criteria. Studies reporting positive counseling outcomes experienced by chronic hospitalized mental patients when they interacted with hospital attendants trained only in operationalizing the dimensions of empathy, respect and genuineness in the counseling encounter (Carkhuff and Truax, 1965a, Carkhuff and Truax, 1965b) provided an additional source of supportive evidence to the notion of modeling certain facilitative conditions. A study of forty outpatients treated by resident psychiatrists at the Phipps Psychiatric Clinic at John Hopkins (Truax, t al., 1966) was an attempt to cross- validate the findings of the studies on individual psycho- therapy with hospitalized schizophrenics with data from a very different patient and therapist population. Analysis of these data indicated greater improvement for the patients of therapists that offered high levels of empathy, respect and genuineness than for the patients that received relatively lower levels of these conditions. The dif- ferences were significant on the two measures of patient im1 le‘ ra‘ th. re. th. Si: su. as em; He of ab pr of ac 0f 0f Co] re. f0: 29 improvement. Counselors that offered one of three different levels of empathy, respect and genuineness (combined ratings) were investigated by Leitner (1969). He found that the clients of the high level counselors demonstrated a greater increase in self-exploration, a crucial variable related to therapy outcome (Truax and Carkhuff, 1964b), than did the clients of the low functioning counselors. Similar results were reported by Hountras and Anderson (1969). Stoffer (1968) investigated the therapeutic success of mother helpers with elementary age children as a function of genuineness, nonpossessive warmth, empathic understanding and dogmatism in the helping person. He reported that mother helpers exhibiting high levels of nonpossessive warmth were more successful bringing about gains in achievement and a reduction in behavior problems than were the mother helpers exhibiting low levels of nonpossessive warmth. Dickenson and Truax (1966) also found a gain in academic achievement as a result of a training program offered by a counselor trained in offering high levels of empathy, respect, genuineness and concreteness. When compared with no-therapy underachievers and underachievers receiving moderate levels of therapeutic conditions, they found that underachievers receiving high levels of the conditio average therapy Aspy (l9 regard a indexes which su expressi for beha effect 0 Truax an cantly r SUCh as and the COUHSelc accounte variatic the pati 30 conditions showed significantly greater grade-point average improvement than the no-therapy or moderate- therapy students. Similar findings were reported by Aspy (1966) in his conclusion that empathy, positive regard and genuineness were related to psychological indexes as well as intellectual achievement. Truax and Carkhuff (1964a) reported research which suggested that concreteness or specificity of 'expression in therapy was a critical element accounting for behavior change. In a study that investigated the effect of sixteen different therapist-influenced variables, Truax and Carkhuff found that concreteness was signifi- cantly related to criteria measures of therapeutic process such as the Insight Scale, the Personal Reference Scale and the Process Scale. In addition, they found that two counselor variables, genuineness and concreteness, accounted for approximately forty percent of the total variation on the Personal Reference Scale, a measure of the patients' personal reference statements. In another study, Pierce and Drasgow (1969) found that therapists who were trained to attend to conflict areas elicited more client self-exploration than did therapists using non- directive reflection. Shapiro (1969) presented a paper that related empathy, warmth and genuineness in psychotherapy to out— comes of psychotherapy. He reviewed several studies that 31 reported positive relationships between the three thera- peutic conditions and outcome in individual and group psychotherapy with a wide range of patients. These find- ings have also been verified and/or summarized by a number of other investigators (Carkhuff and Truax, 1966; Carkhuff and Berenson, 1967, pp. 44-60; Pagell, Carkhuff and Berenson, 1967; Rogers, §t_§l,, 1967; Truax and Carkhuff, 1967; Chessick, 1965). Empathy, Respect, Genuineness and Concreteness as Variables in Counselor Training Programs Desrosiers (1967) studied the personal growth of thirty-three counseling trainees and found that supervisors who scored high in genuineness, empathy and respect were more likely to affect positive attitude changes in their supervisees than was the case for low scoring supervisors. In another study, the therapist's levels of functioning on measures of empathy, respect, genuineness and concreteness were the treatment variables (Pierce, Carkhuff and Berenson, 1967). Seventeen volunteer counselors-in-training and two counselors, one functioning at a high level and one at a low level, were used to investi- gate whether or not the model developed by Carkhuff to explain the change in behaviors of clients would also apply to counselors-in-training; that is, the counselors-in- training under the supervision of the high level counselor would gain more on the outcome measures than the 32 counselors-in-training under the supervision of the low level counselor. Pierce, Carkhuff and Berenson reported that the supervisees of the high functioning supervisor changed significantly on eleven of the fifteen outcome measures while the trainees of the low level supervisor changed significantly on only one variable. In addition, more trainees drOpped out of the low level supervisor's group. They concluded that a supervisor could effect positive change in a counselor-in-training on measures of empathy, respect, genuineness, concreteness and self- exploration only if the supervisor was already functioning at a higher level on these dimensions than the counselor- in-training. In still another study, Pierce and Schauble (1970) investigated the effects of individual supervision on levels of empathy, respect, genuineness and concreteness in counselors-in-training throughout a nine-month training program. They found that the trainees of the high functioning supervisors reflected more growth on levels of empathy, respect, genuineness and concreteness than did the trainees of the low functioning supervisors from the beginning to the end of the training program. Carkhuff (1969) combined the data from sixteen studies that focused on the effects of training and found that different studies have reported somewhat similar results. According to Carkhuff, those trainees whose 33 trainers were functioning at or above the minimally facili- tative level and approximately one level above the trainees were more apt to encourage positive changes in their supervisees than were trainers functioning at lower levels. In addition, para-professional trainees who were functioning initially at low levels on measures of empathy, respect, genuineness and concreteness tended to demonstrate posi- tive gains over the course of training on these same four measures if they had been offered high levels during the training program but did not change or they deterioriated in functioning if they were offered low levels during the training program. In summary, there is some conflicting data re- ported on the importance of the affective variables of empathy, respect, genuineness and concreteness in counse- ling and counselor training but much of the literature that was reviewed reported that therapists functioning at high levels of empathy, respect, genuineness and concreteness had clients that improved on a variety of improvement criteria, while the clients of therapists that offered low levels of these dimensions deteriorated on indices of change or gain. Similar findings were re- ported when changes that occurred for supervisees were investigated as a function of the supervisor's level of functioning on these dimensions. 34 Cognitive Variables in Therapy and Training Bordin observed that, "the very fact that one cannot stop the movement of the world around him means that he cannot completely neglect major relationships to it" (1968, p. 175). Any analysis of a person's efforts to deal with his strivings inevitably must include the cognitive aspects of what he did or is doing. Thus, attending to specific "facts" related to a client's problem, such as vocational information or certain marital difficulties, may well contribute to a client's ego strength (Bordin, 1968, pp. 166-182). While the concepts of problem solving ability and problem sensitivity have been investigated in other areas, there has been no effort to relate the importance of problem solving ability and problem sensitivity to counseling. Shulman, Loupe and Piper (1968) developed a test using the "in-basket" technique to measure the inquiry process of teachers—in-training. Two of the variables investigated were problem solving and problem sensitivity. But since Shulman gt_§l,, did not have measures of problem solving and problem sensitivity on the supervising teachers, an analysis of change as a function of the supervising teacher scores on the same measures was not possible. Comparisons of "seeking style," as determined by ten paper and pencil tests, were made. From these data they tentatively concluded that if the 35 student teacher and her supervisor had different "seeking styles” at the beginning of student teaching, they became even more unlike each other as the term progressed. In summary, while several investigators discuss the importance of cognitive aspects in counselors, no research is available that investigates the impact that a supervisor's cognitive "style" has on his counselor trainees. CHAPTER III INSTRUMENT DEVELOPMENT The test to measure problem sensitivity and problem solving ability is the Problem Solving and Sensitivity Kit (PSSK),* a simulated counselor's office with an assortment of embedded potential counseling problems. The test is a modification of the Teacher In- Basket, a test develOped by Shulman to study the inquiry process of teachers-in-training (1965). The theory, description, administration and validation of the PSSK are presented in this chapter. Theory and Selection 2: Instrument The model of inquiry presented by Dewey (1938) was the framework around which the PSSK was developed. Based on Dewey's model, the process of inquiry was divided into four parts: 1) problem sensing, 2) problem formula- tion, 3) search and 4) resolution. The PSSK waa developed to investigate the stages of problem sensing and resolution. Problem sensing involved the recognition by the *Because of the length of the PSSK and the com- plexity of the scoring procedures it was not reproduced in its entirety in this dissertation. Copies of the com— plete instrument are available, on loan, from the investi- gator of this study. Write to: Joan Hamachek, Michigan State University Counseling Center, East Lansing, Michigan, 48823. 36 37 subject that a problem existed. First, problem sensitivity asked whether or not the subject perceived each embedded problem as problematic and, secondly, how many potential problems did he respond to in the situation. Resolution or problem solving occurred at the time when the subject's curiosity was satiated and his inquiry ceased. Problem solving ability was a measure of how well the subject understood the nature of selected embedded problems. Extensive research on problem solving has been conducted by psychologists (Shonksmith, 1969; Duncker, 1945; Maier, 1936; Wertheimer, 1945; Bloom and Broder, 1950). The traditional way of observing problem solving behavior was to pose a problem situation for subjects, individually or in groups, and observe their attempts to resolve that problem. In some studies it was presumed that the subject possessed the necessary information to solve the problem and it was the manner in which he brought the information to bear upon the problem which formed the focus for research. In other studies all necessary information to solve the problem was arrayed for the problem solver. Problem solving under field conditions differed in a number of major dimensions from the above experi- mental situations. As Shulman, Loupe and Piper stated, The real world does not consist of carefully constructed situations that are presented to 38 individuals as problems-for-solution. Instead, individuals move through an array of stimulus situa- tions which are potentially problematic in varying degrees, selectively reacting to some and not to others. Those situations that are problematic do not present themselves one at a time in predetermined numerical order but rather derive both their defini- tion and the order in which they are handled from the cognitive activity of the inquirer. . . . More often the inquirer is operating with his ideas and feelings focused upon matters in which he has an emotional investment, such as teacher with her students, a doctor with his patients, or a therapist with his client. This affect—invested inquiry may differ markedly from the same individual's problem-solving activities in relation to, say, the area of a parallel- ogram (1968, pp. 17-18). The range of experiences that were perceived and categorized as problematic varied from individual to individual. In addition, individuals reacted selectively to problems in the situation and dealt with them in an order and depth of his own choosing (Mercer, 1968). There- fore, there were different perceptions of what constituted a problem and a resolution of the problem. A basic assumption of this View of c0ping was that, In any situation, an individual will attempt to transorm the problematic and uncertain into a state that corresponds most closely to that picture of the universe with which that individual is most comfortable (Shulman, Loupe and Piper, 1968, p. 25). The major criteria for selecting an instrument were: 1.. Allow for observations of problem sensing and problem solving behaviors. 2. Maximize amount of observable search and problem solving behavior. 39 3. Simulate a real-life problem solving situation in order to elicit emotional involvement from the subject. A realistic setting which maximized the need for the subject to determine where he would begin and how he would proceed for himself was develOped by Frederiksen, Saunders and Ward (1957). This method was called the Administrator In-Basket. This technique was promising because it did not necessarily specify the problems to be handled, or their order. The in-basket technique left room for potential problems to be embedded, to which some subjects reacted and others did not. Shulman (1965) adapted the in-basket situation to study the inquiry processes of teachers-in-training and named his instrument the Teacher In-Basket. He found that through simulation of a complex problem situation and use of "thinking aloud" (Benjafield, 1969) techniques it was possible to conduct systematic studies of problem sensitivity and problem solving performance, among other things. Reliable and stable measures were obtained using this technique. The PSSK was modeled after the Teacher In-Basket. Description 2: the Problem Solving §n_d SensitivitJ Kit (PSSK) The kit contains three kinds of materials. These are: 1) contents of the in-basket, 2) written records concerning the clients and 3) the "human resource." The "human resource" consists of a secretary who can be 40 contacted by intercom. The contents of the in—basket include telephone messages, memoranda from various members of the staff and casenote information on selected clients. These materials presumably vary in their likelihood of being viewed as problems by the subject and in the ways in which they could be perceived as problems, if at all. Figure 3.1 contains the contents of the in-basket of PSSK, Form A.* Subjects could approach these stimuli in a potentially infinite number of ways. For example, a phone call from Tom (A-4) asking whether or not they would meet on Thursday could trigger search behavior in a number of ways. The subject may set the call aside and wait for another call from Tom. The call may be sensed as prob- lematic but deferred for future inquiry or stored for future reference. Or the subject may look in the schedule book and find that in the past there were two Toms that met with Dr. Binaca on Thursday. Another memo from Dr. Bailey (A-9) may be sensed as problematic and set off search behavior on the part of the subject. The materials on Stuart Strong have em- bedded in them a series of potentially problematic elements *Because of the pre—post design, two forms of the PSSK were developed. A mental health clinic was the setting for Form A; a factory was the setting for the other form, Form B. A-10 A-ll A-12 A-13 41 Brief description of Madison Mental Health Clinic and its location. Calendar indicating the date as October 29, 1969. Memorandum from the secretary indicating that she will need information on clients so she can start scheduling appointments. Phone call from Tom. Wants to know if you would be meeting Thursday. Last name of caller was not noted. Phone call from Patricia Conwell. Wants to know what time her appointment will be. Phone call from Jeff Murray's mother. Jeff left home and she wants to talk to you about him. Phone call from ex-client, Jill Asher. Has a son with a problem and wants to talk to you about him. Memorandum from Dr. Bailey indicating that the con- sulting psychiatrist will be available on Friday A.M. Memorandum from Dr. Bailey indicating that three parent permission forms have been sent home with Stu and asks if psychologist will take care of it. Form is attached. Memorandum from Florence Carter indicating her dissatisfaction with a group therapy sesSion. Asks if they could meet on the thirtieth to discuss plans for future meetings and the sociogram. Socio- gram made during last meeting is attached. Memorandum from Dr. Bailey to comply with request in attached letter. The letter to Bailey indicates that Claire Powers is moving to Indianapolis and will enroll in school there on October 27. The letter, from her new principal, requests informa- tion about Claire from the Madison Mental Health Clinic. Casefile on Cynthia Boring, a referral client from Harriet Brown. Casefile on Mark Garrison. No indication why folder is in the in-basket is noted. FIGURE 3.1 CONTENTS OF PSSK IN—BASKET, FORM A 42 which, if sensed and followed up, would lead the subject to the conclusion that Stuart is a high school drop-out who is currently having difficulties in his family with his step-sister and step-father. In addition, his step- sister is in the same therapy group with him which also presents a problem to Stu. The written records include information in the client's case file, schedule book and current notes. Some of the case files include medical information, application forms, work evaluation sheets, case notes and educational and family history data on the client. Other case files are not as complete. The schedule book lists appointment time and dates. The current notes are recent reports on individual clients and/or group members. The in-basket materials and the written records have over two hundred potential problems embedded in them. It is assumed that the problems vary along an obvious— obscure dimension. Administration and Scoring 2: the Problem Solving and Sensitivity Kit (PSSK)7 During the administration of the PSSK the coun- selor is placed at a desk and informed that he is to role-play a new counselor working in a clinic or factory.* * There are two forms of the PSSK. A mental health clinic was the setting for Form A; a factory was the setting for Form B. _.______, _ ——_—'_ -u ._—_——_‘ w —- . -1 M This been verba He ca No ti; a tra arran: subje. the 51 subje< 5) gex scheme the ol answei ing it 10g is sessit functj that ~ C \ data 43 This is to be his first day at work and materials have been piling up on his desk. The subject is instructed to verbally express his thoughts during the entire period. He can begin where he wants and do whatever he pleases. No time limit is suggested.* The PSSK is administered, observed and scored by a trained observer. Via the use of a one-way mirror arrangement the observer records: 1) the materials the subject was looking at, 2) embedded problems observed by the subject, 3) conclusions or decisions reached by the subject, 4) questions the subject asked the secretary and 5) general observations and comments. Figure 3.2 is a schematic drawing of the research setting. In addition, the observer plays the role of the subject's secretary, answering questions about policy and personnel and bring- ing in any folders requested by the subject.** The written log is coded and rated by the observer after the testing session. Three basic scores and two overall cognitive functioning scores constitutes the cognitive variables that are used for the analysis of cognitive performance. * For complete instructions given to subjects taking Form A of the PSSK see Appendix A. **This dual role did not seem to interfere with data gathering procedures. The ideal conditions, of course, would be to have another person serve in the secretarial role. ONE—WAY MIRROR 44 OBSERVER ROOM A — Microphone-output B — Observer C - Observer's Record Book D - Intercom unit E - Office Records FIGURE 3.2 I J SCHEMATIC DIAGRAM OF THE SUBJECT ROOM Subject Files and Schedule Book Microphone-input In-basket Intercom Unit RESEARCH SETTING 45 Problem sensitivity is the number of potentially problematic elements reacted to (sensed) as problems by the subject. Each form of the PSSK has approximately 230 potential problems embedded in the materials. Scoring for problem sensitivity is based upon a list of the embedded problems!r For example, in the cumulative folder for Claire Powers, a sixth grade student who saw the previous psy- chologist at the clinic, one cardex contains achievement scores of a very capable student while the scores on an- other cardex indicate that Claire is working below grade level. Each of these elements may be reacted to as a prob- lem. In addition, the subject might combine the two prob- lematic elements and note the variability of the scores from One year to the next. Adding more problematic elements concerning Claire, the subject may note that Claire has changed schools every year and when her father is home her grades are excellent, but when her father is working out of town her grades drop considerably. What and how the elements are reacted to and combined depends on the subject. Problem solving I is a measure of overall problem resolution. The ten "major problems” each made up of a number of the potentially problematic elements, form the basis for the scoring. On each of the ten problems the subject is given a rating from zero to three, depending * The list of potential problems for PSSK, Form A is reproduced in Appendix B. 46 on the degree of completion or comprehensiveness to which the subject's problem resolutions are brought. Each point from zero to three for the ten problems is specifically defined for the rater to use in determining the level of resolution adequacy.* Those who score high in problem solving ability come to what is defined as the deepest and most complete level of understanding with respect to the problems embedded in the test materials. The total of a subject's ratings on the ten problems constitutes the problem solving I score. Problem solving II is purported to be a measure of specific problem resolution. The purpose of this measure is to have a measure of depth of problem resolu- tion on specific problems but also to eliminate any effect of fatique. Each subject's problem solving I score is divided by the number of problems he attempts, that is, that he has a rating of one or above, to determine the problem solving II score. Because a time limit is not suggested, subjects may continue working until they are tired and/or feel they have completed the task. To reduce the effect of fatigue on the score, the researcher uses only those problems that the subject attempts. Thus, an individual could have a problem solving I score of nine by resolving the problems * The scoring manual used to rate the ten major problems for PSSK, Form A is reproduced in Appendix C. 47 in a variety of ways. For example, he could have received a rating of one on nine of the ten problems, or he could have received a rating of three on three of the ten problems. By dividing the total score by the number of problems attempted there would be a difference of 1.0 and 3.0. The score is assumed to reflect how in- volved the individual is in problems that he attempts without being penalized for not attempting all of the problems. Cognitive functioning level I is the average score of the standardized problem solving I and problem sensi- tivity measures. Because the scoring system varies for each measure, T scores are calculated. The T scores are averaged for the cognitive functioning level I score. Cognitive functioningilevel II is the average score of the standardized problem solving II and problem sensitivity measures. Initial Pilot Study Two forms of the PSSK were developed because of the necessity of pre and post measures. The two forms needed to be sufficiently similar so that the same under- lying processes could be manifested and observed in the two settings, yet not so similar that there would be a transfer of learning between the two forms. Therefore, it was necessary to ascertain the degree of reliability 48 between the two forms of the test as well as to make an initial attempt at identifying the patterns of relation- ships among the problem sensitivity and problem solving variables. To do this a pilot study was conducted. The questions to be answered by the pilot study were: Are there differences in the forms of the PSSK on the vari- ables in question? Are there differences in the ratings of the observers on the variables in question? Are there differences between raters as a result of one judge ob- serving the subject's behavior as opposed to the other judge only hearing the audio-recording of the subject's behavior? Eight counselors were administered both forms Of the PSSK, at one week intervals. Each testing session was tape recorded. One rater observed a test session and the audio-recordings were rated by the other rater. Forms and raters were counterbalanced to reduce administration effects. Figure 3.3 contains the testing sequence design. A two stage analysis was designed to compare the forms of the PSSK and the raters on measures of problem sensitivity, problem solving I and problem solving II. A two factorial ANOVA for repeated measures (Kirk, 1968, pp. 237-242) provided information on the comparisons of the raw scores on measures of problem sensitivity and the two measures of problem solving. Tables 3.1, 3.2 and 3.3 contain the ANOVA tables for the three measures. 49 Administration Administration Form Rater . . Time 1 Time 2 A Observes n1 n3 B listens . n2 n4 Form A audio-tape B observes n3 n1 A listens n4 n2 audio-tape A observes n5 n7 B listens . n6 n8 Form B audio-tape B observes n7 n5 A listens n8 n6 audio-tape FIGURE 3.3 INITIAL PILOT STUDY TESTING SEQUENCE DESIGN TABLE 3.1 ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SENSITIVITY IN THE INITIAL PILOT STUDYa Source 85 df MS F Between Subjects 4369.95 7 624.28 Within Subjects 80.08 3 26.69 Between Raters .80 l .80 .000 Between Forms 63.26 1 63.26 1.229 Form X Rater 16.02 1 16.02 .311 Residual 1080.18 21 51.44 Total 5530.20 31 aFor 1, 21 d.f., the probability of a F value of __ the probability of a F value of 8.025 occurring by chance is S .01. 4.325 occurring by chance is <..05; 50 TABLE 3.2 ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SOLVING I IN THE INITIAL PILOT STUDYa Source SS df MS F Between Subjects 509.38 7 72.77 Within Subjects 81.38 3 27.13 Between Raters 3.13 1 3.13 .565 Between Forms 78.13 1 78.13 14.127* Form X rater .13 1 .13 .023 Residual 116.13 21 5.53 Total 705.875 31 * p 5..01 aFor l, 21 d.f., the probability of a F value of 4.325 occurring by chance is 5..05; the probability of a F value of 8.025 occurring by chance is :5.01. TABLE 3.3 ANOVA SUMMARY TABLE FOR THE MEASURE OF PROBLEM SOLVING II IN THE INITIAL PILOT STUDYa Source SS df MS F Between Subjects 4.3175 7 .6168 Within Subjects .7938 3 .2646 Between Raters .0709 l .0709 .808 \ Between Forms .6331 1 .6331 7.211* Form X Rater .0898 1 .0898 1.022 Residual 1.8438 21 .0878 Total 6.9550 31 'k p :7.05 aFor 1, 21 d.f., the probability of a F value of 4.325 occurring by chance is 53.05; the probability of a F value of 8.025 occurring by chance is :3.01. 51 Significant differences were found between forms on measures of problem solving I and II but not on the measure of problem sensitivity. No significant differences were found between the raters on any of the measures. The differences between forms may have been systematic. If the differences were systematic, the analysis of co— variance used in the study would take these differences into account and the differences in forms of the test would not confound the Obtained results. A further analysis was conducted to check the reliability of the ratings and to investigate whether the differences might be systematic or not (Winer, 1962, pp. 124-132). The lower limit on the reliability for a score based on the average of each subjects' four scores was computed in two ways. One estimate treated all differences within subjects (whether systematic or random) as com- ponents of the errors of measurement. The other estimate removed systematic differences associated with raters, forms and rater and form interaction from the errors of measurement. If the differences were systematic, and thus would not effect the methods of analysis in a negative way, it could be concluded that the two forms of the PSSK were measuring the same dimensions. Table 3.4 contains these findings. The reliability of a single score on the measures of problem sensitivity, problem solving I and problem solving II is reported in Table 3.5. 52 TABLE 3.4 RELIABILITY OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II SCORES, WITH AND WITHOUT EFFECTS OF SYSTEMATIC DIFFERENCES REMOVED, BASED ON THE AVERAGE OF SUBJECT'S FOUR SCORES IN THE INITIAL PILOT STUDY Variable Di¥f22e5Z25eIizlfided Digfzgeizzzeggmiged gzggiiivity .957 .918 22:51:: I iiiSiSE II '571 '858 TABLE 3.5 RELIABILITY OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II SCORES, WITH AND WITHOUT EFFECTS OF SYSTEMATIC DIFFERENCES REMOVED, BASED ON AN INDIVIDUAL SCORE IN THE INITIAL PILOT STUDY L :— Variable .With Systematic .With Systematic Differences Included Differences Removed $232522“), -848 .812 giisiig I -296 .752 EEISifig II '249 .601 The reliability of the scores on the problem solving I and problem solving II variables increased substantially when the effects of systematic differences were removed from the error of meaSurement. Therefore, it was concluded that the differences between forms and raters 53 were primarily systematic and that both forms of the PSSK measure essentially the same processes and characteristics. It should be noted that the two forms of the PSSK represent parallel rather than equivalent forms of the same character- istics. In addition, correlation coefficients were computed on each of the three cognitive measures for the score ob- tained by observing and rating the subject's behavior and the score from rating the audio-recording of the subject's behavior. The high positive correlations of .98, .94 and .61 for problem sensitivity, problem solving I and problem solving II, respectively, would suggest that rating by either the observation or the audio-recording technique allows for reliable scoring of the data. Summary The Problem Solving and Sensitivity Kit (PSSK) was based on a model of inquiry presented by Dewey (1938) and on research by Shulman (1965) and Shulman, Loupe and Piper (1968). The PSSK was develOped so as to allow for observa- tions of problem sensing and problem solving behavior and to simulate a real-life counseling situation in order to elicit emotional involvement from the subject. The contents of the testwere described as were the administration and scoring procedures used to score the PSSK. A pilot study was conducted to investigate several em... . ...____._-~ ~ m. -._' "m .1 54 questions. They were: 1) Are there differences in the ratings of the observers on the problem solving and problem sensitivity variables? 2) Are there differences in the forms of the PSSK on the variables in question? 3) Are there differences between raters as a result of one judge Observing the subject's behavior as Opposed to the other judge only hearing the audio-recording of the subject's behavior? From a two factorial ANOVA for repeated measures and an estimate of reliabilities based on the components of variance, it was concluded that differences in the ratings by two judges on the same test were not significant and that the two forms of the PSSK represented parallel rather than equivalent forms to measure problem sensitivity and problem solving ability of counselors. High correla- tions between measures obtained from observing the subject and listening to an audio-recording of the subject's be- havior suggested that rating by either procedure allowed for reliable scoring of the data. CHAPTER IV DESIGN AND METHODOLOGY In this chapter the 1) sample selection, 2) instrumentation, 3) procedure for collecting data, 4) the design of the study, 5) hypotheses to be tested and 6) the type of analysis used to test these hypotheses are examined. Sample Selection The sample consisted of twelve counselor training supervisors and seventeen of their supervisees. Super- visors were faculty members at Michigan State University and the supervisees were advanced graduate students at the same university. Supervisor Selection As soon as the list of faculty members eligible to do supervision was released in the fall of 1969, a letter was sent to each potential supervisor asking whether he would be willing to participate in this investigation.* Twenty-two letters were sent. Two supervisors declined to participate. An additional five supervisors were * See Appendix D for c0py of this letter. 55 56 eliminated from the selection because they were not assigned to supervise first and second term practicum students and/or interns, the supervisees that were to participate in this study. Twelve supervisors of the fifteen that were eligible and interested were randomly chosen to participate in this study. The supervisors were ten male and two female counseling center faculty members employed at Michigan State University during the 1969-70 academic year. The theoretical orientation of the supervisors ranged from neo-psychoanalytic to interpersonal or eclectic, and they had a range of one to twenty-three years of counseling experience. Each had a doctoral degree in either clinical psychology or counseling psychology. Supervisee Selection The supervisees were ten male and seven female first or second term practicum students or interns that were to be supervised by the twelve supervisors selected to participate in this study. The supervisees were assigned to supervisors by either the Assistant Director for Training or by the practicum instructor. The procedure that was used to make the pairings was described as "essen- tially no systematic process." The supervisees were twelve interns, three first term practicum students and two second term practicum students. Each had completed seventy-five _57 per cent or more of the course work for his Ph.D. require- ments. Eight supervisors were assigned only one counselor- in-training, three supervisors were assigned two counselors- in-training and one supervisor was assigned to three counselors—in-training. Instrumentation The two instruments used in this study were Cark- huff's Scales of Measurement for Facilitative Functioning (SMFF) and the Problem Solving and Sensitivity Kit (PSSK). Scales of Measurement for Facilitative Functioning (SMFF) The SMFF consists of four scales: "Empathic Understanding in Interpersonal Processes;" "Respect or Positive Regard in Interpersonal Processes;" "Facilitative Genuineness in Interpersonal Processes;" ”Personally Relevant Concreteness or Specificity of Expression."* Three excerpts of three minutes each are selected randomly from the beginning, middle and end of a taped counseling interview and ranked on a nine-point scale (1.0, 1.5, 2.0, . . ., 5.0), with the use of the four scales of measurement, to determine the subject's level of functioning on dimen- sions of empathy, respect, genuineness and concreteness. The average ratings of the 3—three minute segments on a * See Appendix E for copies of the four scales. 58 given dimension constitutes the measure of therapeutic functioning on that dimension. Scoring Each tape segment was scored by two judges independantly of the other segments. All four dimensions were rated at the conclusion of each tape segment. The two raters for the Carkhuff's Facilitative Functioning Scales were trained by an experienced re- searcher that had been shown by previous research to be functioning above level 3.0 across all the dimensions of empathy, respect, genuineness and concreteness (Pierce and Schauble, 1970). The interrater reliabilities in this study were E. = .890, R. = .899, G. = .874 and C. = .893. The raters were neither aware of the supervisor-supervisee pairings nor did they have knowledge of which tapes were interviews by the supervisors and which interviews were i by the supervisees. The averaged ratings across all four dimensions were used to divide the supervisors into high or low affective functioning groups. The average ratings on the high and low supervisors on the four dimensions are reported in Table 4.1. The high supervisors had average ratings of E. = 3.29, R. = 3.33, G. = 3.54 and C. = 3.46. Across all dimensions the high supervisors' scores ranged from a low of 2.75 to a high of 4.00. The average ratings 59 TABLE 4.1 AVERAGE RATINGS AND RANGE OF SCORES FOR HIGH AND LOW SUPERVISORS ON DIMENSIONS OF EMPATHY, RESPECT, GENUINENESS AND CONCRETENESS guperVisor Range Empathy Respect Genuine— Concrete- roup ness ness of the low supervisors were E. = 2.29, R. = 2.29, G. = 2.25 and C. = 2.17. Across all conditions the low super- visors' scores ranged from a low of 1.00 to a high of 2.75. Standardization Information on the Scales of Measurement for Facili- tative Functioning (SMFF) Validation of the scales, apart from consideration of their face validity, depends almost entirely on the research evidence relating them to outcome and to other therapy variables. The data suggest that empathy, respect, genuineness and concreteness are related to such outcome variables as self-exploration, the MMPI, achievement score gains and constructive personality and behavioral change.* Problem Solving and Sensitivity Kit (PSSK) The PSSK** is used to measure the subject on dimen- sions of problem solving and problem sensitivity. The * For a fuller discussion of the validity information on the SMFF, see Chapter II. * at ' For a detailed description of the PSSK and its development, see Chapter III. 60 PSSK is used during a structured role playing session and requires the subject to come to grips with problematic situations which requires the discovery of available techniques or the invention of new means for a resolution of the imbedded problems. Each subject's performance is rated on problem sensitivity from 0 to 233, depending on the number of embedded problems he "senses." The subject is also rated from 0 to 3 on ten complex problems, de- pending on the problem resolution that occurs. The total of the ratings on the ten problems yields data for two measures of the subject's problem solving ability. Administration and Scoring A room with a one-way mirror and wired with an intercom was made available by the Reading Clinic in the College of Education for the PSSK testing. The in- structions* required the subject to role play a counselor working in a mental health clinic (Form A) or a factory (Form B). This was to be his first day at work and materials had piled up on his desk and he was to respond to the materials as if it was, in fact, his job. The subject was instructed to give verbal expression to his thoughts during the entire period. The subject was to react to the array of materials before him, any of which * See Appendix A for a copy of the instructions used when administering Form A of the PSSK. 61 could have been perceived as either possessing discrepant characteristics or as having been in equilibrium and non- problematical. Each subject was to work until the task was completed to his satisfaction. The observer, who was on the other side of the one-way mirror during the entire period, recorded, in writing: 1) the materials the subject was looking at, 2) embedded problems observed by the subject, 3) conclusions or decisions reached by the subject, 4) questions asked of secretary and 5) general observations and comments. The written record was used to rate the participants on their problem solving ability and problem sensitivity. The observer (A) for the PSSK was an experienced researcher and had been closely involved for three years in the development and rating of the Teacher In-Basket, (Shul- man gt, 21,, 1968) the instrument on which the PSSK was . modeled.* While it was not possible for two raters to be used throughout the ratings of the PSSK tests, a second rater (B) was trained to score the audio-recordings of the PSSK for problem sensitivity and problem solving ability. Rater B rated the audio-recordings of three randomly se- lected administrations of the PSSK as a check on the reliability of Rater A's scores. Using the Pearson *Rater A was the investigator of this study. While it would have been preferable to have the data collected by an unbiased judge, the large amount of time and money this would involve and the complexity of the instrument prohibited this. 1 ‘ I .1 , I ' . imme- Wuwafl 62 product-moment formula, the interrater reliabilities for problem sensitivity, problem solving I and problem solving II were .96, .99 and.8Q respectively. The sums of the supervisors' scores on problem solving and problem sensitivity, converted to T scores, were used to divide the supervisors into high and low cognitive functioning groups, the treatment groups. The division into groups was the same whether problem solving I or problem solving II was used as the problem solving dimension to divide the groups. As indicated in Table 4.2, the average raw score ratings of the high supervisors were P. Sen. = 50.00, P. Sol. I= 13.17 and P. Sol. II = 1.77. Their range of scores was 39 to 71 for problem senSitivity, 10 to 20 for problem solving I and 1.3 to 2.5 for problem solving II. The average raw score ratings of the low super- visors were P. Sen. : 28.83, P. Sol. I = 5.33 and P. Sol. II = 1.50. Their range of scores was 16 to 40 for problem sensitivity, 3 to 10 for problem solving I and 1.0 to 2.0 for problem solving II. TABLE 4.2 AVERAGE RAW SCORE RATINGS AND RANGE OF SCORES FOR HIGH AND LOW SUPERVISORS ON DIMENSIONS OF PROBLEM SENSITIVITY, PROBLEM SOLVING I AND PROBLEM SOLVING II m 1 S v' or Problem Problem Problem Guper ls Sensitivity Solving I Solving II roup _ _ _ X Range X Range X Range High (n = 6) 50.00 39-71 13.17 10-20 1.77 1.3-2.5 Low (n 6) 28.83 16-40 5.33 3-10 1.50 1.0-2.0 63 Standardization Information on the Problem Solving and Sensitivity Kit (PSSK) The PSSK was an empirical yet unreasearched, instrument to measure problem solving ability and problem sensitivity in counselors. Since it was a modification of a test developed and validated by Shulman (1965) and Shulman, Loupe and Piper (1968), it was assumed that the concepts of problem solving and problem sensitivity would also be valid concepts in the PSSK. In addition, a pilot study was conducted in which eight counselors were administered both forms of the PSSK at one-week intervals. The ANOVA for the two-factor experiment with repeated measures yielded no significant F values for the rater or form and rater interaction effects. While there were significant main effects for form on both of the problem solving variables, it was concluded that the differences were primarily systematic, and thus, would not interfere with the planned analysis.* Procedure The supervisors were tested on two instruments, the Problem Solving and Sensitivity Kit, Form A (PSSK) and Carkhuff's Scales of Measurement for Facilitative Functioning (SMFF). The supervisors were divided into two * For a more complete description of this pilot study, see Chapter III. 64 groups, twice, for the analysis. First, the supervisors' scores on the PSSK were used to divide the supervisors into high and low cognitive groups, and, secondly, the supervisors' average scores on the SMFF were used to divide the supervisors into high and low affective groups. Each participating supervisee submitted tape recordings of two of his counseling interviews within the first three weeks of the academic year. One of the tapes was randomly selected and rated by two judges on the dimensions of empathy, respect, genuineness and con- creteness, as defined by the SMFF. In addition, the supervisees submitted a tape of a counseling interview during the final three weeks of the academic year, with the exception of the two second term practicum students who submitted tapes during the final three weeks of winter term. These tapes were rated on empathy, respect, genuine- ness and concreteness by the same two judges who rated the pretapes and were used as post measures of affective therapeutic functioning. During the first three weeks of fall term each participating supervisee was scheduled to take the PSSK, Form A. In addition, the supervisees participated in taking the second form of the PSSK during the last three weeks of spring term, with the exception of the two second term practicum students. The second term practicum students were tested at the beginning of spring term. The ratings 65 on the PSSK were used as pre and post measures of cogni- tive functioning. It should be noted at this point that neither the supervisor nor the supervisees had any information about the design of the study. Design The design of this study is both predictive and descriptive in nature. The predictive aspects of the study follows a pretest-posttest control group design, as described by Campbell and Stanley (1966, pp. 13-24). Rather than one treatment and one control group this study uses two treatment groups, one of high supervisors and one of low supervisors. The basic design has the following form: R 01 THi 02 R 03 TLO 04 where R is randomized placement of the supervisees into the two treatment (T) groups,* 01 and 03 are pretreatment observations and O2 and 04 are posttreatment observations. *While pairing of the supervisees to super- visors was described as a relatively random process by the instructors that make the assignments, it should be noted that on a few occasions pairing of a particular supervisee to a supervisor was made for specific reasons. If the assignments were not all random, confounding variables may have entered into the supervisory process and these were not taken into account in this study. 66 While the sources of internal invalidity are accounted for in this design, an external source of in- validity, interaction of the testing and the treatment, may influence the final observations. The descriptive aspects of the study follows a trend analysis design. This design can not provide em- perical validation nor will it test any hypotheses. None- theless, id: may help to generate some hypotheses that might be investigated in future studies. Statistical Hypotheses The following are the principle testable hypo- theses of the experiment as stated in their null form: 1. No difference will be found in the supervisees of the high cognitive supervisors and the supervisees of the low cognitive supervisors on the measure of: A. Problem sensitivity. Symbolically: HO: PSen Hi cog = or51in 1 . , n. .n 93 of three supervisors and their respective supervisees.* The supervisors in group one each scored above the median scores of the twelve supervisors on both the affective and cognitive functioning dimensions. Two of the supervisors in this group supervised one counselor-in-training each while one supervisor supervised two counselors-in- training. In the second group the supervisors scored above the median score on cognitive functioning level and below the median score on the affective functioning level. There were six supervisees in this group with one super— visor working with one counselor-in—training, one super- visor supervising two counselors-in-training and one supervisor supervising three counselors-in-training. In the third group each supervisor trained one counselor- in-training. These supervisors were functioning below the median score on the cognitive functioning dimension but above the median score on the affective functioning dimension. The fourth group was composed of three super— visors functioning below the median on the affective and cognitive dimensions and their four supervisees. One supervisor in this group trained two counselors-in- training and the other two supervisors trained one counselor-in-training each. *The fact that there were three supervisors in each group was by chance rather than by design. The classifications of high and low on the cognitive and affective dimensions were based on each supervisor's score in relation to the other supervisors' scores on the same dimension. 94 The theoretical orientation of the ten male and two female supervisors ranged from neo-analytic to existential, and they had a range of one to twenty—three years of post Ph.D. counseling experience. Each had a doctors degree in either clinical psychology or counseling psychology. The ten male and seven female supervisees were second term practicum first term practicum students, students or interns. There were twelve interns, three first term practiCum students and two second term practicum stu- dents. Each had completed seventy—five percent or more of the course work for his Ph.D. requirements. Table 6.1 contains the means and standard devia- tions of the ten basic variables for the supervisors and supervisees participating in this study. TABLE 6-1 MEANS AND STANDARD DEVIATION ON TEN BASIC VARIABLES FOR SUPERVISORS AND SUPERVISEES Supervisors Supervisees Variables 2' S.D. Pre R Pre S.D. Post 2 Post S.D. Prob. Sen. 39.42 14.80 30.56 18.67 23.31 12.63 Prob. Sol. 9.25 5.05 7.56 4.73 7.31 5.21 Prob. Sol. LI 1.72 .45 1.57 .07 1.69 .78 Cog. I 100.88 17.43 99.00 18.69 97.63 17.48 Cog. II 101.04 15.25 98.94 15.12 98.28 17.01 Empathy 2.79 .62 2.42 .47 2.22 .55 Respect 2.81 .66 2.42 .47 2.22 .54 Genuineness 2.90 .73 2.48 .53 2.25 .55 Concretness 2.81 .84 2.39 .66 2.27 .65 Aff. Funct. 2.81 .73 2.42 .51 2.27 .56 95 High Cognitive-—High Affective Group The raw scores on the ten basic variables and an indication if the score was above (+) or below (-) the mean for the high cognitive-~high affective supervisors and their supervisees are presented in Table 6.2. The scores of supervisor 103 were all above the mean except for problem sensitivity and cognitive I scores, which were just slightly below the mean scores for the twelve supervisors. The supervisee of supervisor 103 was subject 208. The scores of subject 208 increased on the five cognitive variables and the affective variable of concreteness from below to above the mean from pre-to- posttesting. The four additional affective scores were above the mean during both testing periods. In other words, subject 208 scored above the mean on all the cognitive and affective variables during posttesting. Supervisor 104 had scores above the mean on all measures except problem solving II, empathy and respect. The supervisees of supervisor 104 were subjects 216 and 217. The scores of supervisee 216 dropped to below the mean or remained below the mean on the five cognitive variables from pre-to-posttesting. On the other hand, the scores below the mean on the affective dimensions increased to above the mean at the time of posttesting. While super- visor 104 had scores on empathy and respect below the mean of the twelve supervisors, the scores were still above the 96 .m.m w4nme SODOHSH N.m mw4nm9 ou 04Qmow4mmm ma wuocuoow m4se 4.0 OHQmB . mcmme wmwfiu. WCHMUCOU .mmsoum m>4uuommmu 024 Mo some on» 30409 m4 wuoom 3mm umgu mmmeHUca I ”mmsouo m>4uomammu may mo some was m>onm m4 muoom 3m4 umsv mwumUHUCH +m um~.m +oo.m +om.~ +oo.m oooe os4so4oossm uoo.~ +om.m +om.~ -m~.~ +oo.m +om.m +oo.e who o>4oooem< -oo.~ +oo.m +om.~ +om.m onoo noosooonoooo -oo.~ +om.m +om.~ -m~.~ -me.~ um~.~ +oo.o one -oo.~ +oo.m +om.~ +oo.m Doom nnoeosasooo -m~.~ +me.m +om.~ -m~.~ +oo.m +me.~ +me.m one . +mm.m +oo.m +om.~ +me.~ umoo Doodnom -oo.~ +m~.m +om.~ -m~.~ (me.m +om.~ +me.m one +m~.~ +oo.m +om.~ +om.~ shoe ssomosm uoo.~ +oo.m +om.~ +om.~ -me.m +om.~ +me.m one -o.om +o.444 um.em +o.~44 soon 44 o>auasooo no.4e +m.~o4 +o.m44 no.mo +o.~o4 -m.mo +o.oo4 one . . ao.em +o.e04 um.4m +o.4~4 once 4 o>444sooo no.4e +o.444 +o.e44 +m.mO4 +o.e44 :m.mo -o.oo one . . no.4 +m.~ uo.4 +m.4 Doom 44 os4>4om no -m.4 +0.4 14.4 -m.4 -o +0.4 one 804nonn 1m +04 .4 +64 once 4 os4>4om Io +44 +44 +m +44 no +o4 who 564Eouo uom +e~ ue4 +48 unoo >u4>4u4ncmm .6 +46 +44 +mm +mm um muom one SO4nonm bow e04 e4~ G44 eO4 mom 404 wmmfl> H064> mmm4> mmm4> Hom4> TTM4> Hom4> mTHQm4Hm> lummsm luomsm lawman luoasm lumasm thwasm tummsm mummH>mmmDm mHmmB 924 mmOmH>mmmDm m>H90mhh¢ EUHm Ilm>HBHZUOU EOHm mom mmqdem<> UHmdm ZmB mom mmmoum Sém N.o mdm¢9 97 post—scores of supervisee 216. Supervisee 217 had scores above the mean on all ten variables during pre and post4 testing. In general, supervisees 216and 217 remained the same or increased their scores on the affective measures and stayed the same or decreased in their scores on the cognitive measures from pre-to—posttesting. The scores of supervisor 107 were above the mean on all the variables with the exception of problem solving II. Subject 206 was the supervisee of supervisor 107. While the scores of subject 206 remained below the mean on the five cognitive variables from pre-to-posttesting, his raw scores did increase markedly. In fact, during post- testing his scores were only slightly below the means of the seventeen supervisees' scores. Scores on two of the five affective variables increased from below the mean on pretesting to above the mean during posttesting. The other three affective variables remained below the mean. In summary, for the students of supervisors termed high cognitive--high affective the scores on the five affective variables remained the same or increased from pre-to-posttesting. In fact, only three of the twenty post- test affective scores (four subjects, five scores each) fell below the mean scores for these variables. Those which did fall below had been low on pretesting also. Seven of the scores increased from below to above the mean from pre-to-posttesting. There was no consistent direction of 98 change for the supervisees on the cognitive variables. High Cognitive-~Low Affective Group Table 6.3 contains the raw scores on the ten basic variables and an indication if the score was above (+) or below (-) the mean for the high cognitive-~1ow affective supervisors and their supervisees. Supervisor 106 had scores above the mean on the cognitive variables and below the mean on the affective variables. The supervisees of supervisor 106 were sub- jects 213, 214 and 215. All of the affective scores of subject 213 dropped considerably from pre-to-posttesting but only two of the five scores dropped below the mean of the supervisees' scores. The cognitive scores remained essentially the same from pre-to-posttesting with only slight variations in the raw scores obtained during both testing periods. Subject 214 had scores above the mean on the cognitive variables and below the mean for the affective variables during pretesting. The affective scores dropped considerably during posttesting and all remained below the mean of the supervisees' scores. In addition, the cognitive scores also dropped and were all below the mean during posttesting. All of the raw scores of subject 215 dropped to below the mean from pre-to- posttesting. In general, the supervisees of high cognitive--1ow affective supervisor 106 tended to decrease on measures of both the cognitive and affective variables. 99 +0e.0 +00.0 +00.0 100.4 100.4 100.0 smog 004so400000 100.4 100.4 +00.0 +00.0 100.0 100.0 100.0 +00.0 100.0 040 0>4Dom440 +0e.0 +00.0 +00.0 100.0 100.4 100.0 . omoO mmmsmomhosoo 100.4 100.4 +00.0 +00.0 100.0 100.0 100.0 +00.0 100 0 000 +00.0 +00.0 +00.0 100.0 100.4 +00H0 . omoO 00000040000 100.0 100.0 +00.0 +00.0 100.0 100.0 100.4 +00 0 100 0 040 +00.0 +00.0 +00.0 100.4 100.4 +00.0 Enos 0000000 100.0 100.0 +00.0 +00.0 100.0 100.0 100.0 +00.0 100.0 040 +00.0 +00.0 +00.0 100.4 100.4 +00.0 onoo >0000E0 100.0 100.0 +00.0 +00.0 100.0 100.0 100.0 +00.0 100.0 000 +0.404 10.40 +0.004 10.00 10.00 +0.044 Enos Hfi4m>4u4s0oo +0.044 +0.044 10.00 10.00 +0.444 10.00 +0.444 +0.004 +0.004 040 +0.004 10.00 10.00 10.00 10.00 +0.044 onom 4 0>4u4s0oo +0.00 +0.004 10.00 10.00 +0.044 10.00 +0.004 +0.404 +0.004 040 +0.4 +0.0 +0.0 10.4 10.4 +0.0 Dnom 44 004>4o0 +0.0 +0.0 +0.0 10.4 +0.4 +0.4 +0.0 +0.4 +0.0 000 5040000 10 10 10 1e 10 +04 smog 4 004>4o0 1e +04 10 10 +04 +0 +44 +04 +00 040 0040o40 +00 10 104 104 104 +00 Enos >44>4o4nsmm +40 +00 104 100 +00 100 +40 +00 +40 040 sm4nonm 000 444 040 440 004 040 040 040 004 wwm4> Hom4> 0004> wwm4> Hom4> wwm4> mmm4> momH> 4004> 00430440> lummsm 140Q5m 14mmsm lummzm 1ummsm lummsm 140Q3m IHTmsm lummsm IIm>HBHZOOU EOHE mmmmH>mmmDm MHWIB QZ< mmOmH>mmmDm W>HBUmmm< 304 mom mmddem¢> UHmHDUmmm¢ 100.0 100.4 100.4 smog 000s0o0nosoo 100.0 +00.0 +00.0 +00.0 100.0 +00.0 040 100.N 10m.4 Im>.4 . smog 000:0:45000 +00.0 +00.0 +m~.m +00.0 +00.0 +00 m 044 100.0 100.4 100.4 smog 0000000 100.0 +00.0 +mm.m +m>.m +00.0 100.0 044 100.0 100.4 100.4 unom 0:00080 100.0 +00.0 +00.0 +00.0 100.0 +00.0 040 +0.0N4 +m.mm4 Im.4> smog HH m>444cmoo +0.044 10.00 10.00 10.00 +0.004 10.00 049 +0.004 +0.004 10.00 smog 4 0>40400oo +0.044 10.00 10.00 10.00 +0.004 10.00 040 +0.0 +0.0 10.4 0000 44 004>4O0 +0.0 10.4 +0.0 +0.0 +0.4 10.4 04m 5049044 +0N +04 IN smog H os4>4om +04 10 10 10 +04 10 mum 5044044. +00 +00 10 smog >u4>4040000 +mm 104 1v4 10m +0m 10m mum 5043044 400 044 000 004 N00 404 mmm4>ummsm How4>ummsm mom4>uwmsm Hom4>ummsm mmm4>ummsm 4004>Hmmsm mw4Q0440> mmmH>mmmDm mHmmB 02¢ mmOmH>mmmDm m>HBUMhm< EOHS IIW>HBHZUOU 204 mom mmddem<> UHm400000< 100.0 100.4 +00.0 100.0 0000 000000000000 100.0 100.0 100.0 +00.0 100.0 100.0 100.0 000 100.0 100.0 +00.0 100.0 0000 00000040000 100.0 100.0 100.0 +00.0 100.0 100.0 100.0 000 . 100.0. 100.0 +00.0 100.0 0000 000000,0 100.0 100.0 100.0 +00.0 100.0 100.0 100.0 000 100.0 100.0 +00.0 100.0 0000 0000080 100.0 100.0 100.0 +00.0 100.0 100.0 100.0 000 10.00 10.00 +0.004 +0.044 0000 -44 m>404coou 10.00 +0.004 10.00 +0.044 10.00 +0.004 10.00 000 . . 10.00 +0.404 +0.044 +0.444 0000 4 0>40400oo 10.00 +0.044 10.00 +0.004 10.00 10.00 10.00 040 . . 10 10.4 +0.4 +0.0 0000 44 004>400 10.4 +0.4 10.4 +0.4 10.4 +0.4 10.4 000 0040000 10 +0 +04 +04 0000 4 004>400 10 +04 10 +04 10 10 +04 000 0040040 10 100 +00 +00 0000 004>4040000 100 +00 100 +40 +00 100 100 000 0040040 040 000 044 000 004 000 004 mwm4> 0004> Hom4> mwm4> Mom4> 0004> 0004> mm4Qm4Hm> 14003m 10005m 14003m lawman 14005m 14009m 14003m WHMMH>mmmDm MHmmB 92¢ mmOmH>mmmDm m>HBUWmh¢ BOA IIH>HBHZOOU 304 mom mmqdem¢> UHm 4om4> . . 0:040 00440440> 00440440> 14005m 14005m 0>4000wm< 400 000020 0002 0>404cmou 404 000050 0002 mm4m<4m<> m>HEHZOOU Q24 m>HBUmmm< mom mmmOUm dez EmOQIOEImmm mo >UZmDOmmh 0.0 m4m<9 r—l H (I) H O ‘1‘ Fl 0 108 m 4 4 11 m 11 O4N m N 11 11 m 11 moN N44 11 m 11 11 v 4 mON 004 00<104 m 11 11 11 11 m 000 m00 000100 v 44 o o m 04 v. 4 11 11 11 m 4ON 044 11 m 11 11 11 m VON N04 000140 11 m 11 11 m 11 New 404 000100 N 04 04 N 4N 0 11 11 m 11 0 4 mON 444. 11 11 m N m 11 N4N 11 m 11 11 N m 44N 004 N m 11 11 m 11 m4N 11 m 11 11 m 11 v4N 004104 11 m 11 11 N m m0m 000 000100 m 4 04 o 0 44 4 4 m 11 11 m ooN 004 11 11 m 11 v 4 04N 4 11 0 11 m 11 04N 004 000140 4 11 v 11 11 m mON mo4 000140 0000 00000000 00000004 0000 00000000 00000004 0000> 0004> a 00400000> 0>400000¢ 004D0400> 0>4000000 100030 100050 3000 000 000000 00000 300 000 000000 00000 300 mmdmfiHm<> m>HBHZOOU QZ< m>H80mmm¢ mom MMOZ