II_II‘.\I4II '13 :‘Iuguv ‘IJ‘?{ It “will“! . “ I III- - III?" A; J2}? " a» II ') 'I-g. I: .'2$:I-'3I W; , ”'5?"- ' 3‘. II. If“. 2 H: ‘ fun“! .'\ u “I I ;J"'€'I “be. 'fw U" ‘ 3". ;' T.‘ '1 lard. ' . “£12m I7. ‘5' "In” If!“ hm r',~!J- 11"? «.. "Vl'h XI... . . -I. HI!" 'I‘ Jr‘s-WI fis'N in I “:5 I4I1IIIIIIIIIII‘5I WIN .II (53%: I‘m IIIIMII I K' I J w' , ' ;' ’J-It'. WI ‘ "' ’ . (II . ". I.“l I... I: . ,I '..f .~* ' h';‘]“'1'I‘~':\1E:“”-I" I‘. Q-JI‘ ‘ f C I I . , .l‘ D {‘4 b I, I ‘ l I " ;' ' ; my - .. :I I -. I. I -F II ‘A -- I ' . I IN ‘ .9. ‘.,". I. .~ 'i.‘ ' ' . I. . I J1" a) J I IIn \ n V 'I “I! : 1|"; I I. . I I I I l ,1 I . II . . | I I ‘ v OI I ‘ I ‘ I A ”,3 fi I; I. ‘I J ' I I ' I I 1 " I r I.I . I_ . I- I} ; . .. .. u . .p I. I I . .. K ¥.;I I' )I. .J ' r. '.' ‘ II. '1' ‘ ' .I l';. I. I ' ’ . “ '8 ‘ '; I ' ~ ' I If. I VI I u I. l fl" ,. 'V‘. .‘ w I w I": - [II I I - i ’0 H S . ‘ / 1' J = ' I} .n" '. ' ' lI " , .I' v." I |I J I I .. w I. I . I I I} F. ‘r I II I, I .. I II , I .I . O.‘ ,u . . I , I \ r v . . I . I I .I L. I. I ~ I I ,I I‘. ‘ "NV . U .' l . u ‘ J . .. ‘ “. I "' q n l I ’ fig 2% I I I" I II I I I a' . I “ . I Is . J’I ‘ J .v. I I u" ' I II .II . '_ . u . I. .J '. II .I I I II nl‘.‘ I I III I) fi ‘1. . I ". \'. I II' “ . 1. I ' I " I .' ' ' ‘ ‘ I 1-I ‘nf , 'o‘ I w I I" I ‘.I'| II. I. II I I A ’ "-'. . "I , II I I I . III I . I ‘9 I I . I II I I I I I. I I If. (._ “ ‘ H ‘ 4 ‘ 'l I III I “J ~—. I'I “'s‘II‘“ I I J '3 J. " .I ‘1’ II “I: -I .I a I I I I. I” . . I I. I ,I , I I. I- I _ I I HI.” I I1” III II > .. ‘I I I I n I l ”I". _. - I -l . I n - _ I ' I " }:\ '. II . 4%: o‘ 'I‘ .II n, - , 'l \ I “‘4' ' :,-,:I 5 r II ’ .I ‘ IL I. I . I l I W: I, . $$ an I. III, I' I ' a I , ' " - ‘ ' ' 1 . . .I I I‘ II:- o . . I q '1 I I . IIII I .‘l I I I ‘. I . II I -. 'Il’. '{ II ‘- II I H I .- \ I t; . . I .I , I . - . -I I - E. I" :1. I ' - .I ‘ 'J . I ‘4 h ‘ ; ; v -I‘ " I If “It“ | I & I I ~ I < a I I W" I . II“ I n | ' ‘ . I I I . ‘ IIIV'II l ’ n | I‘ ‘I' I \ . I l I; .0 I ' | . ' 5 ‘ n . a I\ I Q . , ' . -- I ‘ ' - I I ‘ - I t ' ' IV _ - I I'.' I O I - I I. -. I ' . l j ' .‘ O - h . x ‘ ., I I I l o - .0 ' We ' uaéARY'm ' Michigan State University This is to certify that the thesis entitled A Study of Differences in Therapists' Perceptions of Clients Between Therapists' With Different Levels of Adherence to Community Mental Health Ideology presented by Robert Dean Smith has been accepted towards fulfillment of the requirements for __Bh.._fl._ degree in W9 Psychology Major professor Dr. William w. Farquhar Date 37’/o'/% 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution 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. A STUDY OF DIFFERENCES IN THERAPISTS' PERCEPTIONS OF CLIENTS BETWEEN THERAPISTS WITH DIFFERENT LEVELS OF ADHERENCE TO COMMUNITY MENTAL HEALTH IDEOLOGY BY Robert Dean Smith A DISSERTATION Submitted to Michigan State University ' in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling and Educational Psychology 1984 ABSTRACT A STUDY OF DIFFERENCES IN THERAPISTS' PERCEPTIONS OF CLIENTS BETWEEN THERAPISTS' WITH DIFFERENT LEVELS OF ADHERENCE TO COMMUNITY MENTAL HEALTH IDEOLOGY BY Robert Dean Smith The primary purpose of the present study was to examine differences in therapists' perceptions of clients. A sec- ondary purpose was to examine whether therapists with different degrees of adherence to CMH ideology perceived clients with different socioeconomic backgrounds differently. The dependent variables were therapist ratings of (1) the Motivation for Treatment Scale, (2) the Global Assessment Scale (GAS) at intake, (3) the GAS at a subse- quent date, (4) Residual Gain Scores, (5) the Client's Reported Degree of Satisfaction Scale, and (6) the number of therapy sessions clients recieved. Therapists' degree of belief in assumptions underlying the community mental health movement was assessed with the Community Mental Health Ide- ology Scale (CMHI). Twenty-four community mental health outpatient thera- pists volunteered for the study and took the CMHI. Ten clients were randomly selected from each therapist's caseload. Robert Dean Smith A set of six hypotheses was postulated to examine the prediction that low CMHI therapists would rate clients as less motivated, more disturbed, perceive less improvement, perceive less client satisfaction, and have fewer therapy sessions than therapists with high adherence to CMH ideology. A second set of six hypotheses was postulated to exa- mine if there was an interaction on therapists' perceptions of clients, between client socioeconomic background and therapists' level of adherence to CMH ideology. A combination of analysis of variance and multiple regression was used for each of the dependent variables to test its associated hypothesis. Low CMHI therapists were found to rate clients as less motivated for treatment and less improved over time than high CMHI therapists. No differences in ratings of initial severity of disturbance, client satisfaction and the number of therapy sessions were found between high and low CMHI therapists. None of the null hypotheses pertaining to an interaction between therapist CMHI and client socioeconomic background were rejected. The results supported the general hypothesis that low CMHI therapists were inclined to make personalistic attribu- tions and high CMHI therapists were inclined to make situa- tional attributions about the causes of clients'ldisturb- ance. Implications for future research were discussed. To my mother, Thelma Smith, who made things possible. ii ACKNOWLEDGMENTS The author wishes to express appreciation to Dr. William Farquhar, chairman, whose adept comments and edito- rial guidancewmade it possible to transform vague notions into an actual dissertation. Appreciation is also expressed to Dr. William Hinds, Dr. Gilbert De Rath and Dr. Judith Taylor for serving on the committee. A special note of gratitude goes to Dr. Douglas Miller who provided narcissis- tic supplies for the journey. A very personal thanks goes to Fran Field who was understanding when I was not. iii TABLE OF CONTENTS LIST OF TABLES O O O O O O O O O O O O O 0 O O O O O 0 CHAPTER I. II. III. IV. TH E PROBLEM O O O O O O 0 O O O O O O O. O O 0 Purpose . . . Need . . . . Hypotheses . Theory . . . Overview . . REVIEW OF THE LITERATURE . . . . . . . . . . . CMH Ideology and Treatment . . . . . . . . . Therapists' Personality and Community Mental Health IdeOJ-Ogy O O O O 0 O O O O O O Therapists' Socio-Political Beliefs and Perceptions O O O O O O O O O O O O O O O Therapists' Clinical Judgments About Clients Therapist & Client Variables Related to Length of Therapy . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . DESIGN OF THE STUDY 0 O O O O O O O O O O O 0 Sample . . Procedure . Measures . Hypotheses Analysis . Summary . . ANALYSIS OF RESULTS 0 O O O C C O O O O O O 0 Motivation for Treatment Scale Ratings . . . Initial Global Assessment Scale (GAS) Ratings Subsequent Global Assessment Scale (GAS) Ratings . . . . . . . . . . . . . . . . . . Residual Gain Scores . . . . . . . . . . . . Number of Therapy Sessions . . . . Client's Reported Degree of Satisfaction Scale Ratings . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . iv Page vi 0 \O come-uh) H H w PJH an» FJH mpg N N hUWNNN N‘Dh‘lU’lw .5 U1 Ul-b “KO mmm SQUID \IQ CID-h CHAPTER page V. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . 85 Summary . . . . . . . . . . . . . . . . . . . . 85 COHCIUSionS O O O O O O O O O O O O O O O O O O 95 Discussion . . . . . . . . 96 Implications for Community Mental mHealth .. . 102 Implications for Future Research . . . . . . . 103 APPENDICES A. Baker- -Schu1berg CHMI Scale . . . . . . . . . . 105 B. Global Assessment Scale (GAS) . . . . . . . 110 C. Distribution of Clients for High and Low CMHI Therapists . . . . . . . . . . . . . . . 112 D. Consent Form . . . . . . . . . . 115 E. Cover Letter and Instructions for Therapists . 116 F. Therapist Demographic Questionnaire . . . . . . 118 G. Case Histories for Reliability Study of Motivation for Treatment Scale and Global Assessment Scale . . . . . . . . . . . . . . 119 REFERENCES 0 O O O O O O O O O O O O O O O O O O O O O O 124 TABLE 3.1 4.1 4.2 4.3 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 LIST OF TABLES Therapist-Subjects Characteristics for High CMHI and Low CMHI Correlations Between Client Education, Client Income, CMHI and Two Quasi Variables . . . . . ANOVA Summary of Groups . . . . . Two-Way ANOVA of tion of CMHI and Two-Way ANOVA of tion of CMHI and ANOVA Summary of Two-Way ANOVA of Function of CMHI Two-Way ANOVA of Function of CMHI ANOVA Summary of CMHI Groups . . Two-Way ANOVA of Function of CMHI Two-Way ANOVA of Function of CMHI ANOVA Summary of Groups . . . . . Two-Way ANOVA of Function of CMHI Two-Way ANOVA of Function of CMHI ANOVA Summary of Groups . . . . . Motivation Ratings for CMHI Motivation Ratings as a Func- Client Education . . . . . . Motivation Ratings as a Func- Client Income . . . . . . . . Initial GAS for CMHI groups . Initial GAS Ratings as a and Client Education . . . . Initial GAS Ratings as a and Client Income . . . . . . Subsequent GAS Ratings for Subsequent GAS Ratings as a and Client Education . . . . Subsequent GAS Ratings as a and Client Income . . . . . . Residual Gain Scores and CMHI Residual Gain Scores as a and Client Education. . . . . Residual Gain Scores as a and Client Income . . . . . . Number of Sessions for CMHI vi Page 24 47 50 54 54 55 59 59 61 64 64 66 68 69 70 TABLE Page 4.15 Two-Way ANOVA of the Number of Therapy Sessions as a Function of CMHI and Client Education . . . 73 4.16 Two—Way ANOVA of the Number of Therapy Sessions as a Function of CMHI and Client Income . . . . 73 4.17 ANOVA Summary of Client Satisfaction for CMHI Groups 0 O O O O O O I O O I O O O O O I O O I O 75 4.18 Two-Way ANOVA of Client Satisfaction Ratings as a Function of CMHI and Client Education . . . . 77 4.19 Two-Way ANOVA of Client Satisfaction Ratings as a Function of CMHI and Client Income . . . . . . 78 4.20 Results of Statistical Tests of Main Effect Hypotheses O O O O O O I O O O O O O O O O O O O 80 4.21 Results of Statistical Tests of Main Effect Hypotheses O O O O O O O C O O O O O O O O O O O 83 A1 Client Distribution by Income for High and Low CMHI Therapists . . . . . . . . . . . . . . . . 112 A2 Distribution of Clients by Education for High and Low CMHI Therapists . . . . . . . . . . . . 113 A3 Client Distribution by Sex for High and Low CMHI Therapists . . . . . . . . . . . . . . . . 114 vii CHAPTER I THE PROBLEM The community mental health movement is a socio- political phenomena that has introduced many services rapid- ly through the country during the past twenty years. There is fairly wide spread agreement on the important concepts and goals that have been introduced (Schulberg & Baker, 1975). Baker and Schulberg (1967) identify the beliefs underlying the community mental health movement as follows: (1) the practitioner is responsible for the population of identified and unidentified potentially sick members of the community, (2) harmful forces can be counteracted before they produce mental illness, (3) the primary goal of treat- ment is to help clients achieve social adjustment as soon as possible, not personality reconstruction, (4) there should be a continuity of professional responsibility as the client moves from one program to another in an integrated network of services, and (5) the mental health specialist is only one member of a group of community agents and that his effectiveness can be extended by working with and through other peOple. Jeger & Slotnick (1982) point out that the community mental health movement contains a fundamental contradiction 2 at its core. While espousing a philosophy that emphasizes the role of environmental and social factors as causes of mental illness, the movement is mandated by law to emphasize direct services to individuals that are based on the medical-disease model. Implementation of services has been inconsistent with regard to the concepts and goals of the community mental health movement. Many of the services have, for practical reasons, hired professionals who do not believe in the guiding principles of the movement and who practice therapies based on traditional models (Henry, 1977). In studies of the nature of the community mental health center as an institution, staffs have been found to be heterogeneous in their endorsement of community'mental health ideology (Hoffnung, Taylor, Tylim & Koretsky, 1982: Robin & Wagenfeld, 1976). Therapists have preferences about the kinds of problems with.which they want to work and have assumption that relate client characteristics to prognosis for their preferred brand of therapy (Rabkin, 1977; Berzins, 1977). Therapist perceptions and clinical judgments about clients are influ- enced by these beliefs (Fiske, 1977; Wills, 1978) and, are important factors that govern interaction with clients (Rabkin, 1977: Fiske, 1977; Wills, 1978). RRIEQSS The purpose of this study is to examine the difference in therapists' clinical judgments, perceptions, and length 3 of treatment clients receive, between therapists who have different degrees of adherence to community mental health ideology. The therapists' clinical judgments are the ini- tial severity of disturbance, and severity of disturbance after prolonged contact. Therapists' perceptions are client motivation and satisfaction. The outcome variable is ad- justed change scores of severity of disturbance. Need Community mental health is a primary source of direct mental health services to the general public and especially to the poor and working classes. 'The effect on services, especially to the individual client, of degree of belief in community mental health is an area that has received scant attention in the literature. Only'now are questions such as whether the quality of treatment received in such a center is comparable~to that available to the more economically' privileged groups through private practice. However, con- trary to the private consumer's ability to shOp for or change therapists if dissatisfied, the public consumer has very limited options. Virtually nothing is known about what difference, if any, belief in community mental health might have on the practitioner‘s attitudes toward and interaction with clients. This investigation is a beginning exploration of that neglected area of research. This investigation is important with respect to two broad issues. The first issue is related to the composition 4 of staff needed at a community mental health center. Clini- cal judgments and perceptions are presumed to influence the quality of treatment clients receive. It would be useful to know if therapist treatment values make a difference in therapist clinical judgments and perceptions. The second issue concerns recent funding cutbacks for community mental health centers that have resulted in increased pressure to limit services. Proposals have been made to prioritize client need, which for reasons of practicality, are based on clinician judgment and ratings. Another approach has been to limit the amount or length of services by monitoring cases through the peer review system. Such reviews are typically based on the written evaluations and ratings of the treating clinicians. In either case, systematic differ- ences in rating tendencies of clinicians might result in inequitable treatment decisions. Knowledge about this source of error in clinical judgments would allow for ad- justment of ratings prior to making decisions. We The basic assumption underlying this study was that therapists with different degrees of adherence to community mental health ideology react to clients differently. It was further assumed that this different reaction tendency would be reflected in clinical judgments, perceptions and behaviors. 5 The following hypotheses tested in this study are stated in general form. mail: Therapists with different degrees of adherence to CMH ideology perceive and evaluate clients differently. W: The perceptions and evaluations of therapists with different degreeslof adherence to CMH ideology are influ- enced to different degrees by client socioeconomic factors. The above hypotheses will be restated in testable form in chapter III. Theory Wills (1978) reviewed literature pertaining to profes- sional helpers' perceptions of clients. He concluded that therapists' perceptions are influenced along three dimen- sions: client manageability, treatability and likability. The dimension of treatability is the most relevant to the prOposed study because treatability is associated with de- gree of client disturbance and motivation. ‘Wills identified four processes that influence the therapist's perception of the client's treatability. The four processes are (a) the tendency of therapists to take a biased sample of negative aspects of the client's behavior and personality, (b) the principle of similarity and attraction, (c) the personalis- tic tendency in making attributions about behavior, and 6 k” the perceptual consequences of clients'resistance to influence. The last two processes are the most relevant to the present study and will be explained below. The personal istic tendency refers to a bias to attri— bute cause of behavior to the personality of the client and minimize the influence of situational factors. In attribu- tion theory, this is referred to as the fundamental attribu- tion error. It is defined as the tendency for observers to underestimate the impact of situational factors and overes- timate dispositional factors in controlling behavior. It has been noted that actors and observers differ in their susceptibility to this bias. Observers are more inclined to make dispositional attributions while actors are more in- clined to make situational attributions of their own be- havior (Ross, 1977). In the therapy relationship, thera- pists are the observers and clients are the actors. Adherence to community mental health ideology repre- sents the view that problems result from some interaction between the current environment and the individual. Lack of adherence represents the view that problems primarily reside with the individual personality. It is expected that thera- pists with lower degrees of adherence to community mental health ideology are more susceptible to making dispositional or personalistic causal attributions. Therapists with higher degrees of adherence to community mental health ide- ology are less susceptibleeto dispositional bias and more (receptive to attributing cause to situational factors. 7 Therapists who attribute cause to dispositional factors are expected to perceive more pathology and disturbance in clients than therapists who attribute cause to situational factors. The perceptual consequences of clients' resistance to influence are influenced by the therapist's degree of per- sonalistic interpretation of client motive for resistance. Clients tend to persist in the belief that their problems are essentially environmentally induced because of a need to protect self-esteem (Fisher, Nadler, & Whitcher-Alagna, 1982). Therapists with stronger personalistic tendencies may view such attitudes as resistance and evidence of poor motivation. Fisher et a1. (1982) pointed out that according to their threat of self-esteem model, personalistic interpreta- tions of a client's problems are threatening and prompt the client to terminate therapy as soon as possible. Therapists with lower degrees of adherence to community mental health ideology are expected to be more threatening to clients that therapists with higher degrees of adherence to community mental health ideology because of their personalistic bias. It is, therefore, expected that therapists with less adher- ence will have fewer therapy sessions with clients than therapists with more adherence~to community mental health ideology. 8 9191:1211 In Chapter II, research on community mental health ideology and related areas is reviewed. In Chapter III, the design of the study is described. The results of the analy- sis are presented in Chapter IV} Chapter V consists of a summary of the study, conclusions, discussion of the results and a statement of implications for future research. CHAPTER II REVIEW OF LITERATURE The focus of this section is on community mental health ideology of therapists. This research project is intended to investigate differences in the perceptions of clients, between therapists with different levels of adherence to community mental health ideology. The review will be divid- ed into six parts. The first part will examine the research on therapists' beliefs in community mental health ideology and its effect on treatment. The second part will briefly review literature on community mental health ideology and therapist personality. The third part will review litera- ture pertaining to the influence of therapists' socio- political beliefs on perceptions of clients. The fourth part will review general research on therapists' perceptions of clients. The fifth part will review research on vari- ables related to the number of sessions clients receive. Finally, the major conclusions of the literature review will be summarized. SEW Schulberg and Baker (1975) defined ideology as a belief system that justifies an individual's behavior and serves 10 the purpose of self-definition. Baker and Schulberg (1967) developed the Community Mental Health Ideology Scale (CMHI) to measure the degree of adherence to beliefs and values underlying the community mental health movement. In the sample used to develop the CMHI, low scoring therapists spent more of their time in direct service to clients than high scoring therapists. Low scores on the CMHI were also associated with a greater interest in new developments in individual psychotherapy, biochemistry, genetics, neurology, and neuropharmacology. High CMHI scores were associated with interest in new developments in culture and personali- ty, social psychiatry, epidemiology» group psychotherapy, and milieu therapy. Persons with high scores identified more with a sociotherapeutic orientation, while low scoring persons identified with a psychotherapeutic orientation. The sociotherapeutic and the psychotherapeutic orienta- tions have been considered as opposite poles of the same therapy continuum (Sharaf & Levinson, 1957). The psycho- therapeutic orientation is characterized by a view that intensive psychotherapy is the sole curative process. Psy- choanalytically oriented private practice is the role model for this orientation. Sociotherapeutic adherents do not view individual psychotherapy as the sole curative process and believe that various kinds of interactions with person— nel are potentially therapeutic. By definition, adherence to community mental health ideology represents the view that problems result from some 11 interaction between the current environment and the indivi— dual. Lack of adherence represents the view that problems primarily reside within the clientls personality. Clinical perceptions and behaviors are formed with regard to these criteria. A review of the literature resulted in only two studies of adherence to community mental health ideology and therapy processes. The first is a vignette study of clinical per- ception. The second is a field study of the length of therapy. Del Gaudio, Stein, Ansley, and Carpenter (1976) studied the relationships between therapists' scores on two attitu- dinal measures, clients'lcharacteristics and therapists' ratings of clients on likability, comfort, interest in treating, interest in friendship, and prognosis. Clients' social class, diagnosis, and insight level were systemati— cally varied within.case histories. Thirty-three therapists with 5 or less years of experience were included. Case history ratings were analyzed by a four factor analysis of variance with repeated measures. High CMHI therapists gave significantly higher ratings on liking and prognosis than low CMHI therapists. Low CMHI therapists gave high ratings mainly to less disturbed and middle-class clients. No significant interactions between CMH ideology and client social class on any of the ratings were found. 12 A primary limitation of this study is the generaliz- ability of the results to the clinical setting. The present study will examine the interaction.oflclient social class variables, and therapist ideology with therapist clinical judgment and perception of clients. Carpenter and Range (1982) used a step-wise multiple regression analysis to investigate relationships between the predicted variable (number of therapy sessions) and the predictor variables of therapists' sex, professional affili- ation, Democratic Values scores, and Community Mental Health Ideology scores. These variables accounted for only 3% of the variance>of the number of sessions and the CMHI alone accounted for most of that figure. A four-way analysis of variance revealed a significant main effect for the CMHI Scale. Low scoring CMHI therapists had significantly more sessions than high scoring CMHI therapists. However, even though significant, the strength of that relationship was not large. The general conclusion was that therapist vari- ables alone are not very important and that therapist-client interaction may have more impact on the duration of therapy. A weakness of the study was that it was done at a facility conducted as a community mental health center (Carpenter, Del Gaudio, Morrow & Ritzler, 1981), but one that served predominately middle-class clients. This was a university operated facility and may not have duplicate dthe client population or staff of a typical community mental health center. An important aspect not investigated in this 13 study was the possible interaction of client socioeconomic factors and therapist ideology with the number of therapy visits. In the proposed study, this interaction will be investigated. W W Penn, Baker & Schulberg (1976) found high degree of endorsement of community mental health ideology to be asso- ciated with lower needs for deference and order. Baker and Schulberg (1969) studied members of citizen mental health boards and found dogmatism (negatively) and liberalism (positively) correlated with endorsement of community mental health ideology. Del Gaudio, Stein, Ansley, and Carpenter (19760 found endorsement of CMH ideology to be positively correlated with democratic values. W W Several studies have been conducted to investigate clinician socio-political beliefs as a variable that might result in bias in clinical judgments. Lerner (1972) developed the Democratic Values Scale to measure degree of therapist authoritarianism and found out- come with lower class clients.to be correlated with demo- cratic values. She defined democratic values as consisting of l4 '. .. respect for individual autonomy, for one- self and for others, which transcends differences in status and ideology and which takes precedence over other, potentially conflicting values. Lerner (1973) found that therapists with high democratic values were more likely to indicate a preference for lower class clients. SHe also found that therapists with high democratic values obtained better outcome as measured by Rorschach raw change scores and therapists ratings. Del Gaudio et a1. (1976) found democratic values to be positively'correlated with CMH ideology. They found no main effect between the democratic values and ratings of case histories on likability, interest in friendship» comfort, interest in treating or prognosis. A significant interac- tion between democratic values and client social class on ratings of therapist interest in treating was found. Theraé pists grouped on the basis of high and low democratic values expressed the same degree of interest in treating lower class clients. However, therapists high in democratic values assigned higher ratings of interest in treating to middle class clients than therapists low in democratic values. Abramowitz and Abramowitz (1973) assessed political liberalism of counselors on the basis of two Likert scales that pertained to foreign affairs and social issues and found no significant effect for the politics of counselors 1B.Lerner, 'Democratic Values and Therapeutic Effi- cacy: A Construct Validity Study." W W, 82, 1973, p. 491. 15 on Likert ratings of psychological adjustment for client case histories. Amira, Abramowitz and Gomes-Schwartz (1977) sent hypo- thetical case histories to 600 members of the APA division of school psychology. The assessed school psychologists on the Traditional Moral ism factor of the New Left Scale and found no significant effect on diagnosis or remediation recommendations. iHowever, they conjectured that this may have resulted from the low reliability of the Traditional Moralism factor for this sample. Authoritarianism has been considered especially rele- vant in reference to social prejudice and refers to a per- sonality trend toward making oversimplified, rigidly crys- tallized judgments (lack of tolerance for ambiguity) and social prejudice in general (Adorno, Frenkel-Brunswik, Levinson and Sanford, 1950). Authoritarianism has been an important consideration in studies of less educated mental health workers. However, the sephistication of more highly educated professionals has apparently prevented the assess- ment of authoritarian tendencies in professionals (Rabkin, 1972). Trachtman (1971) concluded that the failure to detec authoritarianism in psychologists was due to the unsophisti- cated nature of the F scale and constructed a measure of authoritarianism with an associated coefficient alpha of .77. Clinicians, divided into high and low authoritarian groups, rated Rorschach protocols with attached lower and 16 middle class histories. Designated lower-class protocols were judged to be more disturbed than middle-class proto- cols. Lower-class protocols were more frequently assigned a diagnosis of character disorder while middle class protocols were more frequently assigned a diagnosis of neurosis. Authoritarianism was not found to be related to the results, although the author speculated that the authoritarianism instrument may have been inadequate. In a study of attitudes toward different social classes, Kurtz, Kurtz & Hoffnung (1970) divided social work students and psychiatric residents into high and low author- itarian groups based on the F scale. Case histories that represented lower and middle class clients were rated on three 7-point semantic differential scales (good-bad, awkward-graceful, and beautiful-ugly). Higher authoritarian psychiatric residents judged lower social class case histo- ries more negatively than lower authoritarian residents. A similar analysis for social work students failed to reach significance (p<.06) . WWW Clinical judgments frequently reflect beliefs that particular kinds of problems and client characteristics are appropriate for the brand of therapy and particular thera- pist practices (Garfield, 1978). These initial judgments and perceptions frequently have a lasting effect on the quality of later interaction with the client (Fiske, 1977), l7 and often act as a self-fulfilling pr0phecy (Rabkin, 1977). The therapist's clinical judgments have been shown to be influenced by the therapist's attitudes, personal beliefs, and values (Strupp, 1973; Rabkin, 1977). Abramowitz and Dokecki (1977) reviewed the literature on the influence of value biases of therapists on clinical inferences. The methodology used has been either correla- tional or vignette studies. Vignette studies have usually had therapists rate a case study or audiotape using a one item Likert scale on variables such as motivation, progno- sis, likability, severity of disturbance and acceptability for treatment. They have been more rigorously controlled for internal validity, but external validity was questioned. Abramowitz and Dokecki suggest the need to supplement vig- nette studies with correlational studies. Client social class factors have been associated with less acceptability for treatment and fewer therapy sessions. They concluded that client social class has more influence on therapist judgments and attitudes toward the client than any other factor. W W Baekeland and Lundwall (1975) reviewed results and methodological considerations pertaining to the length of treatment. They recommended that the number of sessions be counted, rather than number of months. The variables iden- tified from previous research as significantly related to 18 dropping out of treatment were: age, client and therapist gender, socioeconomic status, symptom levels and symptom relief, aggressive and passive-aggressive behavior, socio- pathic features, motivation, psychological mindedness, be- havioral and or perceptual dependence, therapist attitudes and'behavior, drug and alcohol dependence, and discrepant treatment expectations. Socioeconomic status was associated with drop-out in most of the studies conducted in clinics that emphasized psychoanalytical 1y oriented psychotherapy. However, no such relationship has been found in clinics that do not emphasize psychoanalytically oriented treatment. Garfield (1978) concluded that social class is the strongest predictor of duration of treatment, with a lesser inverse relationship to educational level. A clear rela- tionship between treatment duration and age, or sex, or psychiatric diagnosis has not been established. Baekeland and Lundwall (1975) supported the conclusion that psychiatric diagnosis was generally found to be unre- lated to drOp out. However, low levels of anxiety or de- pression, paranoid symptoms, sociopathic features such as aggressive behavior, legal trouble, or hostility to authori- ties, and alcoholism have been related to drapping out. Therapist gender, particularly being male, was identified as an important influence in dropping out. Another important factor identified was a positive relationship between thera- pists' years of experience and length of therapy. 19 Three main influences on dropping out have been de- scribed: (1) client intrapsychic factors, (2) therapist's personality, attitudes or style, and (3) environmental fac- tors. Therapists'lattitudes toward their client were ob- served to be more important than socioeconomic status or client motivation as a cause of drapping out (Baekeland & Lundwall, 1975). Sum Therapists high in adherence to community mental health ideology have been found to be more interested in environ- mental influences on psychiatric disorders than therapists low in adherence to community mental health ideology. Therapists with low adherence to CMH ideology have been found to be primarily interested in individual psychother- apy, biochemistry and other subjects that suggest a focus on the individual personality as the repository of pathology. This difference in interests is consistent with the hypothe- sized personalistic bias by low adherents of CMH ideology when making causal attributions about psychopathology. Results have been found that suggest a negative rela- tionship between CMH ideology and related personality attri- butes often associated with authoritarianism, such as dogma- tism, deference and political conservatism. In general, authoritarianism related measures have not been found re- lated to therapist attitudes or clinical judgments, although scattered evidence of a therapist values by client social 20 class interaction has been found. This limited result may be due to the inadequacy of the devices used to measure therapist authoritarianism and socio-political values. In one study of CMH ideology, it was found that thera- pists high in CMH ideology adherence gave higher ratings of clients likability and prognosis than therapists low in CMH ideology. However, no significant interaction was found between CMHI and client social class. Information about client socioeconomic status may be used by therapists to make inferences about insight level and diagnosis. Inde- pendent variation of insight and diagnosis may have reduced the information value of socioeconomic information about the clients, which may have prevented detection of bias. It appears that CMH ideology'isjpositively'correlated with a general optimism about helping people in distress. In the other study of CMH ideology, a relationship between thera- pist adherence to CMH ideology and the number of therapy sessions was found. Although statistically significant, this result appears to have little practical significance. The possibility of an interaction between CMH ideology with client social class was not investigated. Such an interac- tion might enhance the magnitude of the relationship between CMH ideology and the number of therapy sessions. Studies of the interaction of clinician values with client socioeconomic factors have generally yielded negative results. However, client socioeconomic factors alone have consistently been associated with less acceptability for 21 treatment and fewer therapy sessions and has had more influ- ence on therapist judgments and attitudes toward clients than any other factor. Consequently, potential bias toward socially marginal clients cannot be easily ruled out and warrants consideration in the design of research on thera- pist beliefs and values. CHAPTER III DESIGN OF THE STUDY This study was designed to meet two objectives. One objective was to investigate whether adherence to community mental health ideology is associated with different response dispositions to clients from different socioeconomic back- grounds. The independent variables were therapists' scores on the Community Mental Health Ideology Scale, client years of education and client family income. The dependent vari- ables were therapists' ratings of clients on the Motivation for Treatment Scale, Global Assessment Scale (GAS) at in- take, GAS at a subsequent date, Residual Gain Scores (ad- justed GAS change scores), Client Satisfaction Scale, and the Number of therapy sessions. The main effects of CMHI on the dependent variables were examined with six separate one- way analyses of variance. The interactions of CMHI with client education and income on the dependent variables were examined with multiple regression. A series of two-way analyses of variance were performed to provide a comparison between the significance of the interactions tested with multiple regression and significance tested with traditional analysis of variance. 22 *‘H 23 Therapist completion of the Motivation for Treatment Scale, Client Satisfaction Scale, and Global Assessment Scale was required by the agency for every client as a part of the client's file. A progress note was required for every therapy session, and was part of the client's file. The Community Mental Health Ideology Scale was the only scale that needed to be administered to therapists. 53ml: The subjects of this study were 24 adult outpatient therapists employed by the Clinton-Eaton-Ingham Community Mental Health Board of Lansing, Michigan. There were nine female and fifteen male therapists. Nine therapists had degrees in social work, five had degrees in clinical psy- chology and ten had degrees from various areas that were grouped as educational psychology. The mean years of paid experience was 10.6 and ranged from 1 year to 21 years. (See Table 3.1.) The Clinton-Eaton-Ingham Community Mental Health Board has four outpatient treatment centers serving the adult population of the tri-county area. Sixteen of the subjects worked at the Ingham center, which is the largest center of the four and is located in an urban setting. The other eight subjects worked at two of the satellite centers serv- ing a small town, rural population. 24 Table 3.1. Therapist-Subjects Characteristics for High CMHI and Low CMHI. LQW_§MHI Educational Psychology (M.Ed.) Clinical Psychology (M.A.) Counseling Psychology (M.A.) Social Work (MSW) Social Work (MSW) 5-Female Clinical Psychology (M.A.) 7-Ma1e Counseling Psychology (Ph.D.) Counseling Psychology (Ph.D.) Rehabilitation Counseling (M.A.) Clinical Psychology (M.A.) Social Work (MSW) Social Work (MSW) Years of experience: 1, 4, 6, 6, 9, 12, 12, l2, l4, 16, 16, 18 (x a 10.5) CMHI scores: 94, 106, 109, 141, 144, 152, 157, 164, 167, 174, 180, 193 (x = 148) EI§B_§MHI Clinical Psychology (Ph.D.) Social Work (MSW) Counseling Psychology (M.A.) Educational Psychology (M.S.) Community Mental Health (M.S.) 4-Female Social Work (MSW) 8-Male Social Work (MSW) Counseling Psychology (M.A.) Counseling Psychology (M.A.) Social Work (MSW) Social Work (MSW) Clinical Psychology (Ph.D.) Years of experience: 1, 1, 6, 7, 7, 9, 9, 10, 13, 15, 19, 21 (x = 9.8) CMHI scores: 198, 199, 202, 207, 210, 213, 214, 226, 240, 240, 266 (x a 218) The order of therapist-subjects by degree, years of experi- ence and CMHI scores do not correspond in order to protect confidentiality of participants. 25 Procedure , A packet of materials containing a cover letter, in- structions, the Community Mental Health Ideology Scale, and background information sheet was distributed to 28 thera- pists during the first week in June, 1983. Therapists were requested to complete the CMHI Scale plus the background information sheet and to return them in an enclosed stamped, addressed envelope within two weeks. Twenty-seven thera- pists voluntarily returned the materials. These were clients seen for initial interviews from January 1, 1983 through April 1983. Three therapists had either an insufficient number of cases or case records lacking most of the required informa- tion and were not included in the study. Ten patient cases were randomly sampled from each ther- apist's caseload. Certain patient characteristics have been noted as identifiers of patient populations that respond to treatment in unique ways. Consequently, the client sample was restricted to clients who were not court-ordered, and who received an initial diagnosis of a situational, neurotic or personality disorder. These patients make up the bulk of referrals who come to the clinic. In addition, patients whose presenting problem is identified in the initial eval- uation as primarily drug-dependent or alcohol-dependent will not be included. It is clinic policy to not accept such referrals and only'a small number escape detection by the referral worker prior to the initial interview. This 26 restriction was imposed to maintain a comparable sample of clients for each therapist. Therapists are employed on both a full time and a part time basis. Consequently, they accumulate new patients at different rates. Control for possible systematic assignment and selection of patients was done by obtaining a list of all new patients seen by each participating therapist during January through April 1983. These patients were listed by date in the order they were seen. Clients were eliminated from the list according to the criteria described above. Then, the total number of clients seen by each therapist during this period was divided by 10, the number of required clients for each therapist. That number was then used as an interval to select every nth case until the required ten were obtained. Necessary information about patient variables and out- come variables were obtained from the intake evaluation forms, termination of treatment forms, and quarterly status update forms kept in individual case files and transferred to a separate form containing the therapist code. A time limit of 32 weeks was allowed for treatment between the initial appointment and final collection of data for each case. Client satisfaction and GAS ratings were obtained from the termination of treatment form or the quarterly status update form if the case were still open as of the thirty-fourth week. Therapists are required as part of clinic procedure to complete termination of treatment and 27 quarterly status update forms case. Completion of Quarterly status update forms are required every three months for each case. Frequently, these are not completed until two to three weeks past the due date. This resulted in the most current quarterly update having been completed about two weeks before or after the 32-week time limit. In some cases, the required paper work was not completed with this four-week cushion. In these cases, therapists were briefly interviewed by the thirty—fourth week to obtain the client satisfaction and GAS ratings. Measures Four instruments and two measures were used in the study. The Community Mental Health Ideology Scale (CMHI) was used to measure therapists'ldegrees of belief in assump- tions underlying the community mental health movement. The Motivation for Treatment Scale was used to assess thera- pists' perceptions of clients' motivation for treatment. The Global Assessment Scale (GAS) was employed to measure therapists' perceptions of the severity of clients'ldisturb- ance. An instrument called the Client's Reported Degree of Satisfaction Scale and used to assess therapists' percep- tions of client satisfaction. The number of therapy ses- sions was used to measure length of therapy. Finally, Residual Gain Scores were employed to adjust for statistical regression in GAS change scores. 28 W W Baker and Schulbert (1967) develOped the Community Mental Health Ideology Scale (CMHI) to measure the degree of endorsement.of basic underlying concepts of the community mental health movement. These concepts were defined as: (a) a population focus, (b) primary prevention, (c) social treatment goals, (d) comprehensive continuity of care, and (e) total community involvement. The CMHI consists of 38 positive and negative statements. (Answers are recorded on a 6-point Likert Scale anchored by Strongly Agree and Strongly Disagree. Content Validity was derived by sending an 88-item preliminary questionnaire to 16 judges prominent in the community mental health.movement. These judges rated the degree to which both an 'agree' and Ndisagree' response to each item would reflect a subject's endorsement of the community mental health orientation. This was done on a 5- point continuum. Sixty-four items were selected from this pool on the basis of (a) degree of agreement on the meaning of an 'agree' and I'disagree" response, and U” a.distribu- tion of response extremes that would allow for a wide range of opinions. These 64 items were arranged in the Likert format described above with each item scored from 1 to 7. High total scores represent a high degree of endorsement to a community mental health orientation. Low scores represent (a low endorsement. 29 The final version was based on the following groups. Graduates of the Harvard School of Public Health and the Harvard Medical School Community Mental Health Training Program. Graduates of the Columbia University School of Public Health and Administrative Medicine, Division of Com- munity Psychiatry Post-Doctoral Training Program. Members of the Harvard Laboratory of Community Psychia- try Visiting Faculty Seminar. The participants in the 1965 Swampscott Conference on Training in Community Psychology. The American Psychological Association (Division of Clinical Psychology). The American Psychiatric Association. The American Occupational Therapy Association (Psychiat- ric Occupational Therapists) The American Psychoanalytic Association. The Society for Biological Psychiatry. Four hundred eighty-four questionnaires were returned out of 784. Return rates ranged from 46% for the American Psychoanalytic Association to 93.8% for the Harvard Visiting Faculty. The 38 items selected had corrected item total correlations of .44 or higher. These 38 items received a factor loading of .44 or higher on the first factor of the principal-components analysis. 30 The Kuder-Richardson formula 20 for the total 484 sub- jects on the 38-item scale was .94 and the odd-even split- half reliability was .95. A test-retest reliability of .92 was computed on the basis of 34 graduates of the Harvard Community Mental Health Program. The 484 individuals used for scale development com- pleted a general background questionnaire. One of the fac- tors that differentiated high and low scorers was their type of work setting. High scorers worked in universities, medi- cal schools, general hospitals, community clinics, and school systems. Low scorers spent more of their time in private practice than did high scorers. Item selection for the final version resulted in rela- tively few items from category of comprehensive continuity of care, and the category of social treatment goals. A high score thus represents subscription to only 3 of the original 5 conceptual categories. A high score, then, represents a belief in the need to focus on the total population, preven- tion through environmental intervention, and the involvement of community resources in treatment. A coefficient alpha of .96 was calculated for the sample in the present study. W The Motivation for Treatment Scale was a one-item, 5- point Likert Scale (high being 1 and low being 5) that was printed on the agency intake evaluation form. The statement 'Motivation for Treatment“ was printed above the scale and 31 therapist were required to circle the number that they perceived as representing client motivation. The meaning of the term “motivation" was not defined for Tri-County theta? pists. There have been no reliability or validity studies of this scale. For the present study, an inter-rater reli- ability study was done. Six therapists from the sample participated. Each therapist rated five written case histo- ries on the motivation for treatment scale. A coefficient alpha of .64 was obtained. WWW Endicott, Spitzer, Fleiss, and Cohen (1976) developed a 100¥point rating scale intended to measure the overall se- verity of psychiatric disturbance of a patient. The GAS consists of ten intervals each containing 10 one-point in- tervals (see Appendix B). Each of the ten broad intervals includes a corresponding description of characteristics, psychiatric symptoms, and behaviors. In five studies, in— terjudge reliabilities have ranged from .61 (and standard error of 6.0) to a reliability of .91 (and standard error of measure of 5.0 units). Tri-County therapists were required to rate every pa- tient with this scale on both the intake evaluation form and the termination of treatment form. The intake and termina- tion scores on the GAS will be two of the outcome measures of interest in this study. A reliability study of the GAS was conducted in the Southeastern Region of the Michigan Department of Mental 32 Health (Herman, 1981). Data for this study were collected at the Ypsilanti Regional Psychiatric Hospital, the Hillcrest Regional Center for Developmental Disabilities, the Lenawee Community Mengal Health Board and the Clinton- Eaton-Ingham Community Mental Health Board. The inter—rater reliability coefficients for all the samples, except Ypsilanti, were above .80. The Clinton-Eaton-Ingham sample consisted of 7 clients rated once by two therapists. The inter-rater reliability for this sample was .95. For the present study, an inter-rater reliability study was done. Six therapists from the sample participated. Each therapist rated five written case histories on the GAS scale. A coefficient alpha of .95 was obtained. 91W Satisfaction The Client's Reported Degree of Satisfaction Scale was a one-item, six-point scale that was printed on the agency termination of treatment form. The statement, "Client's Reported Degree of Satisfaction,” was printed above the scale. Therapists were instructed to check one of the options: very positive, positive, neutral, negative, very negative, and unknown. For scoring purposes of the study, the instrument was scaled from one (for very positive) to five (for very negative). The option unknown was treated as missing data. Therapists were required to complete the scale at the time the termination summary was written. There was no reliability or validity information on the 33 instrument. Only face validity can be claimed. Sometimes therapists asked about the patient's View» but most often the therapists made their own best guess. Such a judgment task would seem to involvera projective process. Therefore, this measure may be as much a reflection of therapist satis- faction as patient satisfaction. No reliability estimate was attempted for the scale in the present study. WWW Therapists were required to write a progress note for every therapy appointment. Completion of progress notes was monitored by clerical staff. The number of therapy sessions per client was easily and accurately counted. WW Residual gain scores were the deviation of final scores from the score predicted from the regression of final scores on initial scores. They have been recommended over raw change scores as a measure of improvement (Fiske, Hunt, Luborsky, Orne, Parloff, Reiser, & Tuma, 1970). In this study, residual gain scores will be calculated using initial and termination GAS scores to correct for ceiling effects in the GAS. It has been observed that, the post-treatment adjust- ment level appears to be more important to therapists than the degree of change when making judgments about the general success of treatment of any particular patient (Mintz, 1972). Mintz (1977) concluded that when making global 34 judgments about the relative success of a treatment, thera- pists seem to take into account the ceiling effects of any scale used to measure pre- and post-therapy change. Mintz (1977) re-analyzed correlations, from previous research, between therapist global outcome ratings and pre- and post- therapy adjustment. In two of the studies, adjustment rat- ings were based on therapist ratings of various scales. In the third study, adjustment ratings were composites of ther- apist, patient, and observer's ratings. Global outcome ratings were predicted from the adjust- ment measures used in each study. Squared multiple Rfls ranged from .439 to .670. Residual gain scores were also calculated as predictors of global outcome ratings and were almost identical to the multiple R Correlations. However, Mintz cautioned that the variance not accounted for by a linear function of pre-and post-treatment adjustment ranged from .33 to .56, and noted that therapists of different theoretical orientations seemed to have different standards for evaluating pre-treatment adjustment. Hypotheses A major interest of the study was to examine differ- ences between levels of adherence to community mental health ideology and clinical perceptions of clients. The Community Mental health Ideology Scale was compared to therapists' ratings of client motivation, initial GAS, subsequent GAS, 35 residual gain scores, number of sessions and ratings of client satisfaction. It was hypothesized that therapists with varying de- grees of community mental health ideology would respond differently to clients from different educational or econom- ic backgrounds. The potential interaction of CMHI with client education and CMHI with client family income on each of six dependent variables was tested with multiple regres- sion procedures. Therapists' CMHI scores, client education and client family income were used to construct 2 quasi predictor variables according to a method suggested by Cohen (1978). The correlation between a quasi variable and a dependent variable contains the potential interaction more frequently investigated in a factorial design. The statistical significance of the interactions was tested by inspection of the partial correlation of the quasi ‘variable, once the separate effects of the original inde- pendent variables were removed. The quasi variable EDCM was constructed by multiplying the client's years of education and the CMHI score of clientfis therapist. The quasi variable INCM was constructed by multiplying the clientfis family income and the CMHI score of the clientfs therapist. 36 W W H01: H 4: There is no difference in therapists' ratings of client motivation between therapists with high adherence to community mental health ideology and therapists with low'adherence to community mental health ideology as measured by the Com- munity Mental Health Ideology Scale. Motivation ratings made by high CMHI therapists are higher than motivation ratings made by low CMHI therapists. There is no linear relationship between the quasi variables and motivation ratings once the separate effects of the variables used in constructing the quasi variables have been removed. There is a linear relationship between the quasi variables and motivation ratings once the sepa— rate effects of the variables used in construct- ing the quasi variables have been removed. InitialfilooaLAment Whoa There is no difference in therapists' ratings of client in initial severity of disturbance as measured by the Global Assessment Scale, between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. Initial GAS ratings made by high CMHI therapists are higher than initial GAS ratings made by low CMHI therapists. There is no linear relationship between the quasi variables and the initial GAS ratings once the separate effects of the variables used in constructing the quasi variables have been removed. There is a linear relationship between the quasi variables and the initial GAS ratings once the separate effects of the variables used in con- structing the quasi variables have been removed. 37 W W H 6: H07: There is no difference in therapists' subsequent ratings of client severity of disturbance as measured by the Global Assessment Scale, between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. Subsequent GAS ratings made by high CMHI thera- pists are higher than subsequent GAS ratings made by low DMHI therapists. There is no linear relationship between the quasi variables and subsequent GAS ratings once the separate effects of the variables used in constructing the quasi variables have been removed. There is a linear relationship between the quasi variables and subsequent GAS ratings once the separate effects of the variables used in con- structing the quasi variables have been removed. Wm There is no difference in the residual gain scores between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. Residual gain scores for clients of high CMHI therapists are higher than residual gain scores for clients of low CMHI therapists. There is no linear relationship between the artificial variables and residual gain scores once the separate effects of the variables used in constructing the quasi variables have been removed. There is a linear relationship between the quasi variables and residual gain scores once the separate effects of the variables used in con- structing the quasi variables have been removed. N H 9: H 10: H011: Hall: H012: 38 S There is no difference in the number of therapy sessions received by clients of high CMHI thera- pists and the number of therapy sessions re- ceived by clients of low CMHI therapists. Clients of high CMHI therapists received more therapy sessions than clients of low CMHI therapists. There is no linear relationship between the quasi variables and the number of therapy ses- sions once the separate effects of the variables used in constructing the quasi variables have been removed. There is a linear relationship between the quasi variables and the number of therapy sessions once the separate effects of the variables used in constructing the quasi variables have been removed. W Whoa There is no difference in therapists' ratings of client satisfaction between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. Client satisfaction ratings made by high CMHI therapists are higher than 'client satisfaction ratings made by low CMHI therapists. There is no linear relationship between quasi variables and satisfaction ratings once the separate effects of variables used in construct- ing the quasi variables have been removed. There is a linear relationship between quasi variables and satisfaction ratings once the separate effects of variables used in construct- ing the quasi variables have been removed. 39 Anallaio The data were analyzed with a combination of analysis of variance and multiple regression procedures. The indivi- dual client was the unit of statistical analysis. A tradi- tional approach with three independent variables would be to use a factorial design to investigate main effects and interaction effects. The distribution of clients with vary~ ing education and family income was expected to be unequal among the therapist-subjects. A factorial design analysis of variance that would allow for examination of interaction of CMHI with client education and family income therefore, was expected to result in unequal and disproportionate cells and would decrease the robustness of the ANOVA procedure in regard to violation of the homogeneity of variances assumption. The potential main effects of CMHI on the six dependent variables were tested with six one-way analyses of variance. Equal numbers of clients were selected from each therapistls caseload and therapists were divided equally into high and low CMHI groups. Therapists' CMHI scores were split at the median. Twelve therapists were assigned to the high CMHI group and twelve therapists were assigned to the low CMHI group. The potential interaction effects of CMHI with client education and CMHI with client family income on each of the six dependent variables were investigated with multiple regression procedures. Therapists' CMHI scores, client 40 education and client family income were used to construct two quasi independent variables according to a method sug- gested by Cohen (1978). The correlation between the quasi variable and the dependent variable contains the potential interaction more frequently investigated in a factorial design. According to the method described by Cohen, the subject's score on one independent variable is multiplied with the subjectfls score on a second independent variable to form a score on an artificial third quasi independent vari- able. Such a procedure is performed for all the subjects in a study; The presence of an interaction between the inde- pendent variables is statistically tested by entering the original independent variables in the first step of a hier- archical multiple regression equation. The score for the quasi independent variable is entered in the second step of the equation. Cohen emphasized that once the quasi inde- pendent variable was entered in the above hierarchal order the cross products of the original independent variables will no longer be operating. Instead, it is the partialed product of the quasi independent variable that is operating. This partialed product is the interaction of the original independent variables and its significance can be tested with an F statistic. The partialed product of the quasi independent variable is correlationally invariant over lin— ear transformation of its components. Cohen emphasized that it is important to remember that the interaction is not the cross products of the independent‘variables. Rather, the 41 cross product of the independent variables contains the interaction plus the variance of the dependent variables accounted for by the original independent variables. When the variance of the dependent variable accounted for by the original independent variables is removed from their cross product by partialing, the remaining variance of the depend- ent variable that is accounted for is the interaction and is not correlated with the original independent variables. The quasi independent variables in the present study were: (1) therapist CMHI score x client years of education, and (2) therapist CMHI score x family income. Family income was broken down into fifteen levels on the client informa— tion form, used by the agency (see Appendix C). The dependent variables were (1) the client's motiva- tion rating, (2) the client's GAS score at intake, (3) the number of therapy visits, (4) the client's GAS score at termination or 6 months from intake (whichever occurred first), H” the client's reported degree of satisfaction, and (6) the residual gain scores. A series of two-way analysis of variance were performed to provide a comparison between the significance of the interactions tested with multiple regression and signifi- cance tested with traditional analysis of variance. The independent variables of therapist CMHI scores and client years of education were used in a two-way analysis of vari- ance for each of the six dependent variables. (A second two- say analysis of variance was performed for each of the six 42 dependent variables using therapist CMHI scores and client family income for the independent variables. 'Therapists' scores on the Community Mental Health Ideology Scale were split at the median into a high and low group. Clients' years of education were trichotomized into high 914 years or more), medium (12-13 years), and low (1-11 years) based on the Hollingshead and Redlich (1958) socioeconomic status levels summarized by Lorion (1973). The fifteen categories of client family income were trichotomized into high, medium and low categories. A Client was assigned to the low income category if their family income was at or below the federal median income level and above the poverty line. All other clients were assigned to the high income category. It was assumed that there would be a linear relation- ship between the criterion variables and the predictor vari- ables. Second, it was assumed that the residuals would be normally distributed. Third, it was assumed that the dis- tribution of residuals along the best-fitting straight line had approximately the same variance at all levels. Proba- bility statements about the correlations may be incorrect when any one of the assumptions are not met. However, accuracy is not markedly affected unless the violation is extreme. Snmmarx Four psychometric instruments were used in the present study; The instruments were the Motivation for Treatment 43 Scale, the Global Assessment Scale (GAS - severity of dis- turbance), the Client's Reported Degree of Satisfaction Scale, and the Community Mental Health Ideology Scale (CMHI) . The independent variables in the study were the thera- pists! scores on the Community Mental Health Ideology Scale (CMHI), client years of education and client family income. The dependent variables were therapists' ratings of clients on the Motivation for Treatment Scale, GAS ratings at in- take, GAS ratings at a subsequent date, Residual Gain Scores (adjusted GAS change scores), the Client's Reported Degree of Satisfaction Scale, and the number of their sessions. Twenty-four therapists employed by the Tri-County Men- tal Health Board volunteered to participate in the study. Ten clients were randomly selected from each therapist's caseload. Twelve hypotheses were stated. Five hypotheses postu- lated that therapists high in community mental health ide- ology would give higher ratings on the Motivation for Treat- ment Scale, GAS at intake, GAS at a subsequent date, Residue Gain Scores, and the Client's Reported Degree of Satisfac- tion Scale than therapists low in community mental health ideology. A sixth hypothesis was that therapists high in community mental health ideology would have more therapy sessions with clients than therapists low in community men- tal health ideology. Six one-way analyses of variance were conducted to test the hypotheses. 44 The six other hypotheses concerned the potential inter- action of CMHI scores with client education, and CMHI scores with client family income on each of the six dependent variables. Two quasi variables were constructed by multi- plying the client's years of education with his or her therapist’s CMHI score to construct the variable EDCM. The clientfls family income was multiplied with his or her thera- pist's CMHI score to produce the variable INCM. It was hypothesized that there would be a linear relationship be- tween the quasi variable and each of the dependent variables once the relationship of the original independent variables to the dependent variable was removed. Multiple regression was used to investigate the potential relationships between the dependent variables and the quasi variables. A series of two-way analyses of variance were performed to provide a comparison between the significance of the interactions tested with multiple regression and signifi- cance tested with traditional analysis of variance. CHAPTER IV ANALYSIS OF RESULTS The analysis of data is presented in this chapter. The research hypotheses associated with each dependent variable is presented followed by the analysis of the data. The primary interest of the study was to explore dif- ferences between therapists' with varying degrees of adher- ence to community mental health ideology (CMHIL. Therapists were split at the median of CMHI scores into two groups. High and low CMHI groups were the independent variable and a one-way analysis of variance was performed for each of the six dependent variables. It was expected that therapists would react to clients from varying socioeconomic backgrounds differently depending on therapists' CMHI scores. The variables reflecting socio- economic background were client education and income. Traditionally, such differences would be examined with an analysis of variance approach. This approach was feasible for examining the potential main effect of CMHI. However, examination of potential interaction effects of therapists' CMHI and client education, or CMHI and client income would result in unequal and disproportionate cell sizes. Robust— ness to violation of the anova assumption of equal variances 45 46 would be lowered. Consequently, the analysis of data used a combination of analysis of variance, and multiple regression procedures. First, the potential main effect of therapist CMHI was investigated with six separate one-way analysis of variance. Second, the potential interaction effects of therapist CMHI with client education and income were inves- tigated through multiple regression procedures. The possibility that therapists with varying degrees of adherence to community mental health ideology were differen- tially affected by client socioeconomic factors was investi- gated. This was done by constructing two quasi variables that represented potential interaction effects between client education, income and therapists' CMHI scores. Quasi variables were constructed by multiplying the numerical figures representing the independent variables that were to make up the quasi variable. For each client case, the associated therapistls CMHI score was multiplied with the client's years of education, which resulted in a score representing the quasi variable EDCM. The therapist's asso— ciated CMHI score was multiplied with the client's income, which resulted in a score representing the quasi variable INCM. The two quasi variables were (1) INCM (client income x therapists' CMHI scores), and (2) EDCM (client education x therapists' CMHI scores). Client education, income, and therapists' CMHI scores were entered in a forced entry hierarchal multiple regression procedure. The partial cor- relation coefficient for each interaction variable, and its 47 associated significance level, was then examined. The lin- ear relationship between an interaction variable and a cri- terion variable contains the main effects of the separate independent variables plus the interaction effects of these independent variables. The significance of the interaction effects can be tested by first, entering the separate inde- pendent variables, then second, entering the interaction variables and examining the significance of the partial correlation for the quasi variable of interest. Correla- tions between the independent and quasi variables are pre- sented in Table 4.1. Table 4.1. Correlations Between Client Education, Client Income, CMHI and Two Quasi Variables. Client Client CMHI Education Income EDCM Client Education .08 Client Income .07 .02 INCM .35** .03 .93** .21** N = 239 CMHI = Community Mental Health Ideology Scale INCM = interaction variable-CMHI x Client Income EDCM = interaction variable-CMHI x Client Education ** p = .01 48 The multiple regression method used to test the signif- icance of the interactions was a more powerful procedure than analysis of variance because the numerical values of the independent variables remained in continuous form. How— ever, the appropriateness of the multiple regression method has been questioned in the literature. Consequently, the significance of the therapist CMHI x client education inter- action and the therapist CMHI x client family income inter- action was also tested with traditional factorial analyses of variance. A series of two-way analysis of variance were performed to provide a comparison between the significance of the interaction tested with multiple regression and sig- nificance tested with traditional analysis of variance. The independent variables of therapist CMHI scores and client years of education were used in a two-way analysis of variance for each of the six dependent variables. A second two-way analysis of variance was performed for each of the six dependent variables using therapist CMHI scores and client family income for the independent variables. Thera- pists' scores on the Community Mental Health Ideology Scale were split at the median into a high and low group. Clients' years of education were trichotomized into high (14 years or more), medium (12-13 years), and low (1-11 years) based on the Hollingshead and Redlich (1958) socioeconomic status levels summarized by Lorion (1973). The fifteen categories of client family income were trichotomized into high, medium and low categories. A client was assigned to 49 the low income category if their family income was at or below the federal poverty line. A client was assigned to the medium income category if their family income was at or below the federal median income level and above the poverty line. All other clients were assigned to the high income category. HofixationJoLJroafmenLSoolLBatingo Three types of analyses were performed to test hypoth- eses about the dependent variable of therapists'jperceptions of client motivation as measured by the Motivation for Treatment Scale. An analysis of variance was used to test a main effect hypothesis and multiple regression was used to test an interaction hypothesis. Two factorial analyses of variance were performed to provide a comparison between the significance of the interactions tested with multiple re- gression and significance tested with traditional analysis of variance. The independent variables of therapist CMHI scores and client years of education were used in a two—way analysis of variance for motivation ratings. A second two- way analysis of variance of motivation ratings was performed with therapist CMHI scores and client family income for the independent variables. Win A one-way analysis of variance was performed on motiva- tion ratings between high and low CMHI therapists (see Table 442%. The research hypothesis tested using therapist' CMHI 50 scores as the independent variable and therapists' ratings and client motivation was as follows: H01: There is no difference in therapists' ratings of client motivation between therapists with high adher- ence to community mental health ideology and thera- pists with low adherence to«community mental health ideology as measured by the Community Mental Health Ideology Scale. H 1: Motivation ratings made by high CMHI therapists are higher than motivation ratings made by low CMHI therapists. Table 4.2. ANOVA Summary of Motivation Ratings for CMHI Groups Source DF SS MS F P Between Groups 1 17.07 17.07 13.12 .0004 Within Groups 191 248.49 1.30 Total 192 265.56 The motivation for treatment scale is a one-item, 5- point Likert scale with a “1" being the highest motivation and a '5' being the lowest motivation. The high CMHI thera- pists obtained a mean of 2.12 and the low CMHI therapists obtained a mean of 2:72. The F was 13.12 with an associated p = .0004 significance level. Therefore, the null hypothe- sis was rejected and the alternate hypothesis was accepted. 51 The therapists was high adherence to community mental health ideology gave significantly higher ratings of client motiva- tion than therapists with low adherence to community mental health ideology. W In the regression model used for analysis of the data, the predictor variables were the clients' years of educa- tion, income, therapists' CMHI scores, and two quasi vari- ables representing potential interaction effects usually associated with analysis of variance. The dependent vari- able was the therapists' ratings of client motivation. The correlation between the quasi variable INCM and therapists' ratings of client motivation contained.the main effect of therapist CMHI scores, client family income and the interac- tion of therapist CMHI along client family income. The effects of therapist CMHI and client income were statisti- cally removed from the correlation by multiple regression. The remaining partial correlation represented the effects of a therapist CMHI with client family income interaction on ratings of client motivation. The correlation between the quasi variable EDCM and therapists' ratings of client moti- vation contained the main effect of therapist CMHI scores, client education and the interaction of therapist CMHI with client education. The quasi variable EDCM was analyzed with the same procedure as the quasi variable INCM. The remain- ing partial correlation represented the effects of a thera- pist CMHI with client education interaction on ratings of 52 client motivation. The research hypothesis tested using the two interaction variables INCM and EDCM against therapists' ratings of clients motivation was as follows: H02: There is no linear relationship between the quasi variables and motivation ratings once the separate effects of the variables used in constructing the quasi variables are removed. H 2: There is a linear relationship between the quasi vari- ables and motivation ratings once the separate effects of the variables used in constructing the quasi vari— ables are removed. The overall regression equation using client education, income and therapists' CMHI scores resulted in a multiple regression coefficient of .29 (N - 193) with F = 3.75 (df = 3) and a p =.01 significance level. The addition of the quasi variable INCM as the fourth variable entered resulted in a partial correlation coefficient of .13 with F a 2.19 (df = 4) and a p 3’.114 significance level. Inclusion of the quasi variable EDCM as the fourth variable entered resulted in a partial correlation coefficient of .13 with F =- 2.20 (df = 4) and a p a .14 significance level. The partial correlation coefficients for quasi variables INCM and EDCM were not significant at the p = .05 level. Therefore, null hypothesis two was not rejected. 53 1Mn:flo¥_flmihflflfiLJfiLflbxion£§ Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of motivation ratings (see Table 4M”. A second analysis of variance of motivation ratings was performed with CMHI scores and client income as the independent vari- ables (see Table 4.4). CMHI scores were split at the median and therapists were categorized into high and low CMHI groups. Clients' years of education were divided into high, medium and low education groups. Clients' family income was divided into high, medium and low income groups. High CMHI therapists obtained a mean of 2.12 and low CMHI therapists had a mean of 2.72. The F value for the main effect of therapist's CMHI resulted in a p = .001 significance level. Therefore, null hypothesis one was rejected and the alternative hypothesis was accepted. 'The F value for the therapist CMHI x client education interaction was 2.89 with an associated p = {79 significance level. The therapist CMHI x client income interaction had an F of 1.74 with a p a .18 significance level. The CMHI x client educa- tion interaction and the CMHI x client income interaction were not significant at the p==.05 levell Consequently, null hypothesis two was not rejected. W The hypotheses about the dependent variable of initial GAS ratings were analyzed with three procedures. An S4 Table 4.3. Two-Way ANOVA of Motivation Ratings as a Func- tion of CMHI and Client Education. Source DF SS MS F P Client Ed. (A) 2 7.44 3.72 2.89 .06 Therapist CMHI (B) 1 15.33 15.33 11.92 .001 A x B 2 .62 .31 .24 .79 Residual 187 240.43 1.29 Total 192 265.56 1.38 Table 4.4. Two-Way ANOVA of Motivation Ratings as a Function of CMHI and Client Income. Source DF SS MS F P Client Inc. (A) 2 4.43 2.21 1.73 .18 Therapist CMHI (B) l 17.23 17.23 13.45 .001 A x B 2 4.47 2.23 1.74 .18 Residual 187 238.60 1.28 Total 192 265.56 1.38 55 analysis of variance was used to test a main effect hypothe- sis and multiple regression was used to test an interaction hypothesis. Two factorial analyses of variance were per- formed to provide a comparison between the significance of the interactions tested with multiple regression and signif- icance tested with traditional analysis of variance. The independent variables of therapist CMHI scores and client years of education were used in a two-way analysis of vari- ance for initial GAS ratings. A second two—way analysis of variance of initial GAS ratings was conducted with therapist CMHI scores and client family income for the independent variables. W A one-way analysis of variance was performed on initial GAS ratings between high and low CMHI therapists (see Table 4u5). The research hypothesis tested using therapists' CMHI scores as the independent variable and therapists' ratings of client initial GAS was as follows: Table 4.5. ANOVA Summary of Initial GAS for CMHI groups Source DF SS MS F P Between Groups 1 400.30 400.30 3.86 .05 Within Groups 237 24,568.40 103.66 Total 238 24,968.70 56 03: There is no difference in therapist' ratings of client initial severity of disturbance as measured by the Global assessment scale, between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology'as measured by the Community Mental Health Ideology Scale. H 3: Initial GAS ratings made by high CMHI therapists are higher than initial GAS ratings made by low CMHI therapists. The GAS is a one-hundred point rating scale with a rating of one being the highest severity of client disturb- ance and one-hundred being the least disturbance. The high CMHI therapists obtained a mean of 55.60 and the low CMHI therapists obtained a mean of 53.02. The F was 3.86 with an associated p = .05 significance level. Therefore, the null hypothesis was rejected and the alternate hypothesis was accepted. Therapists with high adherence to community men- tal health ideology rated clients as being less disturbed than therapists with low adherence to community mental health ideology. Ween In the regression model used for analysis of the data, the predictor variables were therapist CMHI scores, client education, client income and two quasi variables, INCM and EDCM . 57 The dependent variable was the initial GAS rating. The partial correlations of INCM with initial GAS and EDCM with initial GAS were derived in the same manner as with the motivation ratings. The partial correlation of each quasi variable with initial GAS ratings represented potential interaction of client family income with therapist CMHI (INCM) and client education with therapist CMHI on the dependent variable initial GAS ratings. The interaction hypotheses were statistically tested by examination of the statistical significance of the partial correlation of INCM with initial GAS ratings and the statistical significance of the partial correlation of EDCM with initial GAS ratings. The research.hypothesis tested using.the two interac- tion variables INCM and EDCM against therapists' initial GAS ratings of clients was as follows: IBM: There is no linear relationship between the quasi variables and initial GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. H 4: There is a linear relationship between the quasi vari- ables and initial GAS ratings once the separate ef- fects of the variables used in constructing the quasi variables are removed. The regression equation using client education, income and therapists' CMHI scores resulted in a multiple regres- sion coefficient of .17 with F = 1.31 (df = 3) and a p = .27 58 significance level. Inclusion of the quasi variable INCM as the fourth variable entered resulted in a partial correla- tion coefficient of .13 with F = 2.06 (df = 4) and a p = .15 significance level. The quasi variable EDCM entered as the fourth step resulted in a partial correlation coefficient of .01 with F = .28 (df = 4) and a p = .87 significance level. The partial correlation coefficients of INCM and EDCM did not reach the minimum significance level (p a .05) for inclusion in the regression equation. Therefore, Infill hy- pothesis four was not rejected. WWW Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of initial GAS ratings (see Table 4A”. Thera- pists' CMHI scores and client family income were the inde- pendent variables in a two-way analysis of variance of initial GAS ratings (see Table 4.7). CMHI scores were split at the median and therapists were categorized into high and low CMHI groups. Clients' years of education were divided into high, medium and low education groups. Clients' family income were divided into high, medium and low income groups. High CMHI therapists had a mean of 55.60 and low CMHI therapists had a mean of 53.02. The F value for the main effect of CMHI on initial GAS ratings was 3.05 with and associated p = .08 significance level. Therefore, null hypothesis three was not rejected. The F value for the client education x Therapist CMHI interaction was 1.69 with 59 Table 4.6. Two-Way ANOVA of Initial GAS Ratings as a Func- tion of CMHI and Client Education. Source DF SS MS F P Client ED. (A) 2 649.16 324.58 3.21 .04 Therapist CMHI (B) 1 308.48 308.48 3.05 .08 A x B 2 341.23 170.61 1.69 .19 Residual 233 23,578.01 101.19 Total 238 24,968.70 104.91 Table 4.7. Two-Way ANOVA of Initial GAS Ratings as a Func- tion of CMHI and Client Income. Source DF SS MS F P Client INC. (A) 2 607.44 303.72 2.96 .05 Therapist CMHI (B) 1 348.57 348.57 3.40 .07 A x B 2 78.00 39.00 .38 .68 Residual 233 23,882.96 102.50 Total 238 24,968.70 104.91 60 a p==.19 significance level. The F value for the client income x therapist CMHI interaction was .38 with a p = .68 significance level. Therefore, null hypothesis four was not rejected. WWW 1W An analysis of variance was used to test a main effect hypothesis and an interaction hypothesis was tested with multiple regression. Two factorial analyses of variance were performed to provide a comparison between the significance of the inter- actions tested with multiple regression and significance tested with tradition analysis of variance» The independent variables of therapist CMHI scores and client years of education were used in a two-way analysis of variance for subsequent GAS ratings. A second two-way analysis of vari- ance of subsequent GAS ratings was performed with therapist CMHI scores and client family income for the independent variables. W13. A one-way analysis of variance was performed on subse- quent GAS ratings between high and low CMHI therapists (see Table 4A”. The research hypothesis tested using thera- pists' CMHI scores as the independent variable and thera- pists' subsequent ratings of client GAS was as follows: 61 Table 4.8. ANOVA Summary of Subsequent GAS ratings for CMHI groups. Source DF SS MS F P Between Groups 1 1,570.81 1,570.81 14.01 .0002 Within Groups 235 26,348.25 112.12 Total 236 27,919.06 5: There is no difference in therapists' subsequent rat- ings of client severity of disturbance as measured by the Global assessment scale, between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. H 5: Subsequent GAS ratings made by high CMHI therapists are higher than subsequent GAS ratings made by low CMHI therapists. The high CMHI therapists obtained a mean of 62.97 and the low CMHI therapists obtained a mean of 57.82. The F was 14.01 with an associated p = .0002 significance level. Therefore, the null hypothesis was rejected and the alter- nate hypothesis was accepted. Therapists with low adherence to community mental health ideology rated clients as being more disturbed than therapists with high adherence to com- munity mental health ideology. 62 WWI: In the regression model used for analysis of the data, the predictor variables were the client years of education, income, therapists' CMHI scores, and two quasi variables, INCM and EDCM. A client income by therapist CMHI interac- tion on subsequent GAS ratings was represented by the quasi variable INCM. A client education by therapist CMHI on subsequent GAS ratings was represented by the quasi variable EDCM. The research hypothesis tested using the two interac- tion variables INCM and EDCM against therapists' GAS ratings of clients was as follows: HOG: There is no linear relationship between the quasi variables and subsequent GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. H 6: There is a linear relationship between the quasi vari- ables and subsequent GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. The overall regression equation using client education, income and therapists' CMHI scores resulted in multiple regression coefficient of .28 with F = 3.57 (df = 3) and a p a .02 significance level. The quasi variable INCM entered as the fourth variable resulted in a partial correlation coefficient of .14 with F = 2.39 (df = 4) and a p = .12 significance level. Addition of the quasi variable EDCM as 63 the fourth step resulted in a partial correlation coeffi- cient would be .03 with F = .98 (df = 4) and a p = .75 significance level. The partial correlation coefficients of INCM and EDCM did not reach the minimum significance level (p = .05) for inclusion in the regression equation. Conse- quently, null hypothesis six was not rejected. WW Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of subsequent GAS ratings (see Table 4E9). Thera- pists' CMHI scores and client family income were the inde- pendent variables in a two-way analysis of variance of subsequent GAS ratings (see Table 4.10). CMHI scores were split at the median and therapists were categorized into high and low CMHI groups. Clients' years of education were divided into high, medium and low education groups. Clients' family income were divided into high, medium and low income groups. High CMHI therapists obtained a mean of 62.97 and low CMHI therapists obtained a mean of 57.82. The F value of 12.83 was significant at the p = .001 level. Consequently, null hypothesis five was rejected and the alternative hy- pothesis was accepted. The F value for the client education x therapist CMHI was 1.59 with an associated p a .21 signif- icance level. The F value for the client income x therapist CMHI was .54 with a p = .58 significance level. The inter- actions of client education x therapist CMHI and client 64 Table 4.9. Two—Way ANOVA of Subsequent GAS Rates as a Func- tion of CMHI and Client Education. Source DF SS MS F P Client ED. (A) 2 780.59 390.29 3.57 .03 Therapist CMHI (B) 1 1,400.95 1,400.95 12.83 .001 A x B 2 346.26 173.13 1.59 .21 Residual 231 25,221.40 109.18 Total 236 27,919.06 118.30 Table 4.10. Two-Way ANOVA of Subsequent GAS Ratings as a Function of CMHI and Client Income. Source DF SS MS F P Client INC. (A) 2 559.40 279.70 2.52 .08 Therapist CMHI (B) 1 1,464.40 1,464.40 13.18 .001 A x B 2 120.18 60.09 .54 .58 Residual 231 25,668.67 111.12 Total 236 27,919.06 118.30 65 income x therapist CMHI failed to reach the p a .05 signifi- cance level and null hypothesis six was not rejected. BosidoalJiaimfioorea Residual gain scores were initial and subsequent GAS change scores statistically adjusted for initial severity of disturbance. The main effect hypothesis was tested with analysis of variance. The interaction hypothesis was tested with multiple regression. Two factorial analyses of variance were performed to provide a comparison between the significance of the inter- actions tested with multiple regression and significance tested with traditional analysis of variance. The independ- ent variables of therapist CMHI scores and client years of education were used in a two-way analysis of variance for residual gain scores. A second two-way analysis of variance of residual gain scores was performed with therapist CMHI scores and client family income for the independent variables. Wool: A one-way analysis of variance was performed on resi- dual gain scores between high and low CMHI therapists (see Tab1e24.11). The research hypothesis tested using thera- pists' CMHI scores as the independent variable and residual gain scores as the dependent variable was as follows: 66 Table 4.11. ANOVA Summary of Residual Gain Scores and CMHI Groups. Source DF SS MS F P Between Groups 1 557.52 557.52 10.42 .001 Within Groups 235 12,572.12 53.50 Total 236 13,129.64 H07: There is no difference in residual gain scores between clients of therapists with high adherence to community mental health ideology and clients of therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. H 7: Residual gain scores for clients of high CMHI thera- pists are higher than residual gain scores for clients of low CMHI therapists. The high CMHI therapists obtained a mean of 1.55 and the low CMHI therapists obtained a mean of -1.51. The F was 10.42 with an associated p = .001 significance level. Therefore the null hypothesis was rejected and the alterna- tive hypothesis was accepted. Therapists with high adher- ence to community mental health ideology rated clients as improving more than therapists with low adherence to com- munity mental health ideology. 67 Infomotimflxpotheaio Multiple regression was used to predict residual gain scores from therapists' CMHI scores, client income, educa— tion and two quasi variables, INCM and EDCM. The variable INCM was used to examine the potential interaction of client family income by therapist CMHI on residual gain scores. The variable EDCM was used to examine the interaction of client education and therapist CMHI scores. The research hypothesis tested using the two interac- tion variables INCM and EDCM against residual gain scores was as follows: H08: There is no linear relationship between the quasi variables and residual gain scores once the separate effects of the variables used in constructing the quasi variables are removed. H 8: There is a linear relationship between the quasi vari- ables and residual gain scores once the separate ef- fects of the variables used in constructing the quasi variables are removed. The regression equation using client education, income and therapists' CMHI scores resulted in a multiple regres- sion coefficient of .22 with F = 2.17 (df = 3) and a p =- .09 significance level. The quasi variable INCM entered as the fourth variable resulted in a partial correlation coeffi- cient of .06 with F a .45 (df = 4) and a p = .50 signifi- cance level. The quasi variable EDCM entered as the fourth 68 step resulted in a partial correlation coefficient of .02 with F =- .63 (df =- 4) and a p = .80 significance level. The partial correlation coefficients of the quasi variables INCM and EDCM did not reach the p = .05 significance level. Therefore null hypothesis eight was not rejected. WW Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of residual gain scores (see Table 4.12). Thera- pists' CMHI scores and client family income were the inde- pendent variables in a two-way analysis of variance of residual gain scores (see Table 4.13). Table 4.12. Two-Way ANOVA of Residual Gain Scores as a Function of CMHI and Client Education. Source DF SS MS F P Client ED. (A) 2 126.08 63.04 1.18 .31 Therapist CMHI (B) 1 545.66 545.66 10.18 .002 A x B 2 63.49 31.75 .59 .55 Residual 231 12,382.54 53.60 Total 236 13 ,129.64 55.63 69 Table 4.13. Two-Way ANOVA of Residual Gain Scores as a Function of CMHI and Client Income. Source DF SS MS F P Client INC. (A) 2 30.49 15.24 .28 .75 Therapist CMHI (B) 1 538.03 538.03 9.95 .002 A x B 2 47.27 23.63 .44 .65 Residual 231 12,494.36 54.09 Total 236 13,129.64 55.63 The clients of high CMHI therapists obtained a mean of 1.55 and the clients of low CMHI therapists obtained a mean <3f -l.51. The F value was significant at the p a .002 level. Consequently, null hypothesis seven was rejected and iihe alternative hypothesis was accepted. The F values for the client education x therapist CMHI interaction and client income x therapist CMHI interaction respectively were .59 (p =- .55) and .44 (p = .65). Consequently, null hypothesis eight was not rejected. W A main effect hypothesis was tested with analysis of Variance and an interaction hypothesis was tested with multiple regression. Two factorial analyses of variance were performed to provide a comparison between the significance of the inter- actions tested with multiple regression and significance 70 tested with traditional analysis of variance. The independ- ent variables of therapist CMHI scores and client years of education were used in a two-way analysis of variance for the number of therapy sessions. A second two-way analysis of variance of the number of therapy sessions was performed with therapist CMHI scores and client family income for the independent variables. Wheels A one-way analysis of variance was performed on the number of therapy sessions between high and low CMHI thera- pists (see Table 4.14). The hypothesis tested using thera- pists' CMHI scores as the independent variable and the number of sessions per client as the dependent variable was as follows: Table 4.14. lANOVA Summary of Number of Sessions for CMHI groups. Source DF SS MS F P Between Groups 1 13.81 13.81 .324 .57 Within Groups 235 10,015.97 42.62 Total 236 10,029.78 9: There is no difference in the number of therapy sessions received by clients of high CMHI therapists and the number of therapy sessions received by clients of low CMHI therapists. 71 Ha9: Clients of high CMHI therapists received more therapy sessions than clients of low CMHI therapists. Clients of high CMHI therapists obtained a mean of 6.98 therapy sessions and the clients of low CMHI therapists obtained a mean of 6.24. The F was .324 with an associated p = .57 significance level. Therefore, the null hypothesis was not rejected. There was no difference in the number of therapy sessions received by clients of therapists high in community mental health ideology and therapists low in ad- herence to community mental health ideology. Interactiomflyoofheaia CMHI scores, client education, income and two quasi variables, INCM and EDCM were used in a multiple regression equation to predict the number of sessions. The quasi variable INCM was used to examine the potential interaction of client family income and therapist CMHI on the number of therapy sessions. The quasi variable EDCM was used to examine the potential interaction.oflclient education and therapist CMHI scores. The research hypothesis tested using the two interac- tion variables INCM and EDCM against the number of sessions was as follows: IQQD: There is no linear relationship between the quasi variables and number of therapy sessions once the separate effects of the variables used in construct- ing the quasi variables are removed. 72 Halo: There is a linear relationship between the quasi variables and the number of therapy sessions once the separate effects of the variables used in construct- ing the quasi variables are removed. The overall regression equation using client education, income and therapists' CMHI scores resulted in a multiple regression coefficient of .19 with F a 1.63 (df = 3) and a p a .18 significance level. Inclusion of the quasi variable INCM as the fourth variable entered resulted in a partial correlation coefficient would be .11 with F a 1.41 (df = 4) and a p = .24 significance level. The quasi variable EDCM entered as the fourth variable resulted in a partial corre- lation coefficient of .01 with F a .11 Odf - 4) and a p = .92 significance level. The partial correlation coeffi- cients of INCM and EDCM did not reach the p =.05 signifi- cant level. Therefore, null hypothesis ten was not rejected. Tuo:Hax_Ana1¥ais_of_yarianoe Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of the number of therapy sessions (see Table 4.15). Therapists' CMHI scores and client family income were the independent variables in a two-way analysis of variance of the number of sessions (see Table 4.16). 73 Table 4.15. Two-Way ANOVA of the Number of Therapy Sessions as a Function of CMHI and Client Education. Source DF SS MS F P Client ED. (A) 2 194.55 97.28 2.29 .10 Therapist CMHI (B) l 14.69 14.69 .35 .56 A x B 2 3.39 1.69 .04 .96 Residual 231 9,818.02 42.50 Total 236 10,029.78 42.50 Table 4.16. Two-Way ANOVA of the Number of Therapy Sessions as a Function of CMHI and Client Income. Source DF SS MS F P Client INC. (A) 2 25.96 12.98 .30 .74 Therapist CMHI (B) l 12.96 12.96 .30 .59 A x B 2 9.10 4.55 .10 .90 Residual 231 9,980.91 43.21 Total 236 10,029.78 42.50 74 The F value for the main effect of therapist CMHI was .35 with a p = .56 significance level. Consequently null hypothesis nine was not rejected. The client education x therapist CMHI interaction obtained an F of .04 with a p .96 significance level. The F value for the client income x therapist CMHI interaction was .10 with a p =- .90 signifi— cance level. Consequently, null hypothesis ten was not rejected. WWW Soaluatinss The client's reported degree of satisfaction scale was a six point scale with a rating of '1' being very positive and a rating of "5" being very negative. The sixth point represented a rating option of unknown, which was treated as missing data in the analysis. Analysis of variance was used to test a main effect hypothesis and multiple regression was used to test an interaction hypothesis. Two factorial analyses of variance were performed to provide a comparison between the significance of the inter- actions tested with multiple regression and significance tested with traditional analysis of variance. The independ- ent variables of therapist CMHI scores and client years of education were used in.a two-way analysis of variance for satisfaction ratings. .A second two-way analysis of variance of satisfaction ratings was performed with therapist CMHI scores and client family income for the independent variables. 75 W A one-way analysis of variance was performed on client satisfaction ratings between high and low CMHI therapists (see Table 4.17). The research hypothesis tested using therapists' CMHI scores as the independent variable and therapists' ratings of client satisfaction as the dependent variable was as follows: Table 4.17. ANOVA Summary of Client Satisfaction for CMHI Groups. Source DF SS MS F P Between Groups 1 1.49 1.49 2.63 .11 Within Groups 147 83.34 .57 Total 148 84.83 Iagl: There is no difference in therapists' ratings of client satisfaction between therapists with high adherence to community mental health ideology and therapists with low adherence to community mental health ideology as measured by the Community Mental Health Ideology Scale. E111: Client satisfaction ratings made by high CMHI therapists are higher than client satisfaction ratings made by low CMHI therapists. The high CMHI therapists obtained a mean of 1.87 and the low CMHI therapists obtained a mean of 2.07. The F was 76 2.63 with an associated p = .11 significance level. There- fore, the null hypothesis was not rejected. 'There was no difference in ratings of client satisfaction between thera- pists high in adherence to community mental health ideology and therapists low in adherence to community mental health ideology. Interactionjmtheaio Therapists' CMHI scores, client education, income and two quasi variables, INCM and EDCM were used in a multiple regression equation to predict client satisfaction ratings. The research hypothesis tested using the two interac- tion variables INCM and EDCM against therapists' ratings of client satisfaction was as follows: H512: There is no linear relationship between the quasi variables and satisfaction ratings once the separate effects of the variables used in constructing the quasi variables are removed. Ha12: There is a linear relationship between the quasi variables and satisfaction ratings once the separate effects of the variables used in constructing the quasi variables are removed. The regression equation using client education, income and therapists' CMHI scores resulted in a multiple regres- sion coefficient of .09 with F = .32 (df = 3) and a p = .81 significance level. When the quasi variable INCM was added as the fourth variable its partial correlation coefficient 77 was -.14 with F = 2.63 (df = 4) and a p = .11 significance level. The partial correlation coefficient of the quasi variable EDCM entered as the fourth variable was -.01 with F = .58 (df = 4) and a p = .94 significance level. The partial correlation coefficients of the quasi variables INCM and EDCM did not reach the p =.05 significance level and null hypotheses twelve was not rejected. WW Therapists' CMHI scores and clients' education were used as independent variables in a two-way analysis of variance of satisfaction ratings (see Table 4.18). Thera- pists' CMHI scores and client family income were the inde- pendent variables in a two-way analysis of variance of satisfaction ratings (see Table 4.19). Table 4.18. Two-Way ANOVA of Client Satisfaction Ratings as a Function of CMHI and Client Education. Source DF SS MS F P Client ED. (A) 2 .21 .10 .18 .83 Therapist CMHI (B) 1 1.33 1.33 2.30 .13 A x B 2 2.22 .44 .77 .57 Residual 143 82.61 .58 Total 148 84.83 .57 78 Table 4.19. Two-Way ANOVA of Client Satisfaction Ratings as a Function of CMHI and Client Income. Source DF SS MS F P Client INC. (A) 2 1.26 .63 1.14 .32 Therapist CMHI (B) l 1.49 1.49 2.69 .10 A x B 2 2.80 1.40 2.52 .08 Residual 143 79.28 .55 Total 148 84.83 .57 The F value for the main effect of therapist CMHI was 2.30 with a p = .13 significance level. Therefore null hypothesis eleven was not rejected. The F values for the client education x therapist CMHI interaction and client income x therapist CMHI interaction was, respectively, .45 (p =- .64) and 2.52 (p = .08). Consequently, null hypothesis twelve was not rejected. Human: The results of the analysis of data were presented in Chapter IV. Six one-way analyses of variance were used to test a set of hypothesis about six dependent variables with thera- pists' CMHI scores as the independent variable. The de- pendent variables were therapists' ratings of clients on the Motivation for Treatment Scale, GAS at intake, GAS at a subsequent date, Residual Gain scores, the number of therapy 79 sessions and the Client Satisfaction Scale. These hypothe— sis were identified as main effect hypothesis (see Table 4.20 for a summary of the results of the tests of significance). The potential interaction of CMHI with client education and CMHI with client family income on each of the dependent variables was tested with multiple regression procedures. Two quasi variables were constructed that represented poten- tial interaction of CMHI with client education (EDCM) and CMHI with client family income (INCM). The hypothesis per- taining to the interaction variables were identified as interaction hypothesis (see Table 4.21 for a summary of the results of the tests of significance). A series of two-way analysis of variance were performed to provide a comparison between the significance of the interactions tested with multiple regression and signifi- cance tested with traditional analysis of variance. 80 pmumoooc a>oz¢ mmziosa ca emuomflom uoz emuoeflom mo. poumoooc emuomflmm «coo. .mumfim neuonu Hmzo 30H an acne manages mac Huauwca can» Loom“: mum mandamuonu Hmzu omen \3 some mmcHueu mam HeauficH .manom amoHoopH guano: House: auficsesoo way an codename no amoaoopa spasms any name muH:=EEoo ou oocouoacm 30H spas unmeasuonu can amoaooca guano: Houses muwczeaoo on cocouoscm cm“: spas mumaocuonu coosuoo .mamom unoEmmomm¢ Hmooao on» an caucuses mm oocmousumap mo muauo>om Hcfluwca ucofiao mo mosques .mum«&m~onu a“ mucouomufip on ma muons .mumfimcuoou Hmzo 30H mo once mmcauMu :ofium>wuoe can» swam“: mum mumflmnuosu Hmzo no“; so meme mmeaueu coaue>auo= .oacom amoHo umeH seamen Hence: suaaaaaou was an codename mm guano: Houses auacssEoo on mucouoocm 30a and; numammuosu can mmoaoopw guano: fiancee wuacsesoo ou mocouoccm smug spas mumaomuosu coozuoo newum>wuoe ucoaao mo mmcaumu .munaQMuonu a“ oosouommap on a“ muons pooped no coccowmacmam booflmm umoa .momonuoomm uoommm can: no mummy Hecaumapuum mo wueEEsm .om.v manna 81 ewuomnmm uoz pmummooc emuoeflmm hm. Nooo. .onom hmoaoocm nuamom amuse: muficsesoo on» an caucuses mm mmoaoopfi sudden an» Laos mousseeoo ou mucouoscm 30H suds unmeasuonu can mmoaoopa Sufism: fiancee auH:=EEOO ou mucouonpm no“: saws mumfiomuonu coosuoo mouoom camm Hosp names on» :a mucouummac o: m“ muons .mumammuonu Hmzu 30H an owns mmcfiumu mco newsman loan can» Loam“: cum mumfiomuonu Hmzo now: an come mmcaumu mfiooou acofimmon hmeuonu mo noose: «as can muaaaauoau Hmzu page no «new luau mo po>oaoou macammom mounds» mo cmuoommm uoz hm. m doossc on» a“ mosouommac on ma muons «mom .mumadeuweu Hmzo 30H mo mucofiao now mouoom swam Haspamou can» “mamas use muaaaeumeu Haze amen coumooo¢ no mucwwao uom mouoom swam Hcspamom «hum umoooc no cosmoamacmfim umoa uponom Aemaaauaooc .o~.e manna 83 .co>oeou coco o>mo noaoswus> “anew can mswuosuumcou cw poms noaonwun> on» no muoommo mono couponom uoz “zoom. om. lemon ecu coco mouoom swam Huscamou can moaoowum> Hedoauauuc on» sous» couponom uoz Azquv om. loo ofinmcofiunaou umoswa on ma muons ”mom .co>osou sumo u>sn moaocaum> “mono can mewuosuumcoo ca cums moaonaun> on» no nuoommo couponom uoz “zoomv ms. manuomon on» coco mmcfiumu m Hanna on» cums» wouooflom uoz Azosz NH. loo mfismsoHunHou unocHH on ma muons «mom .po>oeou coco obs: moaonausp House on» wcfiuoauuncoo cw cons moaocuucb on» no nuoommo cmuomfiem uoz Azuamv he. «baseman an» mono amassed mew Heauaaa on» can moflocwuo> “menu on» sous» couponom uoz Azosz ma. non manmsOuucHou unocqa on ma muons «com .co>osou coco o>mn moaoawus> «meow on» mcauosuuncoo ca pom: moHocauo> on» no nuoommo couponom uoz “sonny «a. uncommon on» coco mmcfiumu cow» Ic>fiuoe can noaosauc> House on» coosu caucuses uoz Azquv «a. sea aaamaoauoauu nausea on as means «mom uoonom no A.uuoo Haauummv uoooom ooccoumflsmam .momonuomwm upoumm can: no mummy Hmoaumaumum mo muasnom mo auceasm .HN.¢ wanna 84 cofiuomuoucu cofiumoscm ucoflao x Hmzo cofiuomuoucH oeoocm usoflau x Hmzo £00m ZUZH emuomflmm uoz emuomhwm uoz cmuomflwm uoz emuomnmm uoz Azonmv «a. Azosz ”H. “zonal um. Azosz em. .co>o&ou coco o>nn moaocfium> amaze on» mcauosuum Icoo ca pom: moaomaum> mo muommmo uncommon onu coco mangoes schoomw Imaumm can moaoeaum> fiasco on» some» too ofinmcoHumHou unocHH on ma muons .mo>o&ou coco o>cn meanness» “menu can mcauosuumcoo a“ poms moHomwum> on» no muoommo uncommon any coco aconmom ammuosu no noose: can moaomflum> Hmong on» some» loo magm:0aumaou nausea on ma muons «NH m noonom no nooooc A.uuou chuummv coccowmwcmam umoa Accusaucouv .HN.v oHnme CHAPTER V SUMMARY AND CONCLUSIONS A summary of the study with a restatement of the hy- potheses and associated results will be presented. The conclusions are discussed in the next section followed by a discussion of the implications for future research. Summary The primary purpose of the present study was to examine differences in therapists' clinical judgments, perceptions, and length of treatment clients receive, between therapists who have different degrees of adherence to community mental health ideology. A secondary purpose was to examine whether therapists with different degrees of adherence to CMH ideol- ogy were influenced differently by client socioeconomic factors. It was theorized that therapists with low levels of adherence to community mental health ideology have a greater tendency toward a personalistic bias in making attributions about the cause of psychopathology than therapists with high levels of adherence to CMH ideology. The personalistic tendency is a bias to attribute cause of behavior to the personality of the client and minimize the influence of situational factors. The personalistic bias was expected to 85 86 result in perceptions of more psychopathology by therapists with low levels of adherence to CMH ideology than therapists with high levels of adherence to CMH ideology. Consequent- ly, therapists with low levels of adherence to CMH ideology were expected to make more severe clinical judgments about clients than therapists with high levels of adherence to CMH ideology. In addition, the personalistic bias hypothesized for low CMH ideology therapists would be threatening to clients who focus on situational explanations for their problems. Consequently, low CMH ideology therapists were expected to have fewer therapy sessions with clients than high CMH ideology therapists. The dependent variables were therapists' ratings of (1) client motivation, (2) initial severity of disturbance, (3) severity of disturbance at a subsequent date, (4) ad- justed change scores for severity of disturbance ratings, (5) clients' reported degree of satisfaction and (6) dura- tion of therapy. Client Motivation was assessed with the Motivation for Treatment Scale, which was an agency derived, five-point Likert scale. The Motivation for Treatment scale was part of the agency initial interview summary form. Client sever- ity of disturbance was assessed with the Global Assessment Scale (GAS), which is a one hundred point rating scale. Therapists were required by agency policy to make GAS rat- ings of all clients after the initial interview, every three months thereafter and after termination of treatment. 87 Initial GAS ratings and subsequent GAS ratings were obtained from agency forms. Residual Gain Scores were calculated from the initial and subsequent GAS ratings. Initial GAS ratings were used in a regression equation to predict subse- quent GAS ratings. The Residual Gain Score for each client was the residual, the difference between the predicted score and the actual subsequent GAS rating. Residual Gain Scores have been suggested as a practical outcome measure because they result from control of statistical regression and ceil- ing effects of ratings. The clients reported satisfaction was assessed with the Client's Reported Degree of Satisfac- tion Scale, which is an agency derived scale of unknown origin. Duration of treatment was measured as the number of therapy sessions. Therapists'ladherence to community mental health ideol- ogy was measured with the Community Mental Health Ideology Scale (CMHI). The CMHI measures overall endorsement of beliefs and assumptions underlying the community mental health movement. Client years of education and family in- come were used as separate measures of socioeconomic back- ground. Twenty-four community mental health outpatient thera- pists volunteered for the study and took the CMHI. They were not informed about the hypotheses of the study. Ten clients were randomly selected from each therapistfis case- load. Dependent variable data was collected from the 88 client's case file, and when the file was incomplete, through interview with the therapist. The hypotheses tested in the study were stated in general form as follows: Hypothesis_l: Therapists with different degrees of adherence to CMH ideology perceive and evaluate clients differently. Hypothesis_2: The perceptions and evaluations of therapists with different degrees of adherence to CMH ideology are influenced to different degrees by client socioeconomic factors. Hypothesis 1 was postulated to examine general differ- ences in perceptions of clients between levels of therapist endorsement of community mental health ideology. A one-way analysis of variance was conducted for each dependent vari- able to test the first hypothesis. Hypothesis 2 was postu- lated to examine whether clients' socioeconomic background is associated with evaluations made by therapists with vary- ing degrees of therapist endorsement of community mental health ideology. Multiple regression was used for each dependent variable to test the second hypothesis. Two quasi variables were constructed that represented potential inter- action between levels of adherence to CMH ideology and the two client socioeconomic variables. The quasi variable EDCM represented the potential interaction of CMHI scores and client years of education on a dependent variable. The 89 quasi variable INCM represented the potential interaction of CMHI scores and client family income on a dependent vari- able. The statistical significance of the interactions were derived by examining the partial correlation of the quasi variable with the dependent variable. A series of two-way analysis of variance were performed to provide a comparison between the significance of the interactions tested with multiple regression and signifi- cance tested with traditional analysis of variance. The independent variables of therapist CMHI scores and client years of education were used in a two-way analysis of vari- ance for each of the six dependent variables. .A second two- way analysis of variance was performed for each of the six dependent variables using therapist CMHI scores and client family income for the independent variables. The results were summarized with a restatement of the hypothesis followed by the conclusions. The comparison between the combination of one-way analyses of variance and multiple regression with the two-way analyses of variance resulted in a different conclusion for therapists' initial GAS ratings. MotioationiorJroafment H01: There is no difference in therapists' ratings on the Motivation for Treatment Scale between high and low CMHI therapists. 90 Hal: Motivation ratings made by high CMHI therapists are higher than motivation ratings made by low CMHI therapists. High CMHI therapists rated clients' motivation as higher than low CMHI therapists. The mean for high CMHI therapists was 2.12 and the man for low CMHI therapists was 2.72. ‘The Motivation.for Treatment Scale was scaled such that the lower the number value, the higher the rating of motivation. H02: There is no linear relationship between the quasi variables and motivation ratings once the separate effects of the variables used in constructing the quasi variables are removed. 2: There is a linear relationship between the quasi vari- ables and motivation ratings once the separate effects of the variables used in constructing the quasi vari- ables are removed. The partial correlation coefficients for EDCM and INCM did not reach the p==.05 significance level. Therefore, null hypothesis two was not rejected. InitialfilobolJaaoasment W H03: There is no difference in initial GAS ratings between high CMHI therapists and low CMHI therapists. 91 Ha3: Initial GAS ratings made by high CMHI therapists are higher than initial GAS ratings made by low CMHI therapists. The null hypothesis was rejected when significance was tested with.alone-way analysis of variance. The mean for high CMHI therapists was 55.60 and 53.02 for low CMHI thera- pists. The GAS was scaled such that the lower the number value the greater the severity of disturbance. When the significance of the difference between the means of high and low CMHI therapists was tested with a two-way analysis of variance the difference between means had an associated p a .08 significance level. The difference between high CMHI therapists' initial GAS ratings and low CMHI therapists' initial GAS ratings did not reach the p:=.05 significance level required for rejection of the null hypothesis. Conse- quently, null hypothesis three was not rejected. H54: There is no linear relationship between the quasi variables and initial GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. E14: There is a linear relationship between the quasi variables and initial GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. The partial correlation coefficients for EDCM with initial GAS ratings and INCM with initial GAS ratings did 92 not reach the p = .05 level of significance. Consequently, null hypothesis four was not rejected. fioalolfiASIJotinoa H05: Therelis no difference in subsequent GAS ratings be- tween high CMHI therapists and low CMHI therapists. H 5: Subsequent GAS ratings made by high CMHI therapists are higher than subsequent GAS ratings made by low CMHI therapists. The null hypothesis was rejected. The mean for high CMHI therapists was 62.97 and 57.82 for low CMHI therapists. Ekfi: There is no linear relationship between the quasi variables and subsequent GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. H 6: There is a linear relationship between the quasi vari- ables and subsequent GAS ratings once the separate effects of the variables used in constructing the quasi variables are removed. The partial correlation coefficients for quasi vari- ables EDCM and INCM did not reach the p = .05 level of significance and null hypothesis six was not rejected. We H07: There is no difference in residual gain scores between clients of high CMHI therapists and low CMHI thera- pists. 93 Ha7: Residual gain scores for clients of high CMHI thera- pists are higher than residual gain scores for clients of low CMHI therapists. The null hypothesis was rejected. Clients of high CMHI therapists had a mean of 1.55 and clients of low CMHI thera- pists had a mean of -1.51. 8: There is no linear relationship between the quasi variables and residual gain scores once the separate effects of the variables used in constructing the quasi variables are removed. H 8: There is a linear relationship between the quasi variables and residual gain scores once the separate effects of the variables used in constructing the quasi variables are removed. Null hypothesis eight was not rejected because the partial correlation coefficients for EDCM with residual gain scores and INCM with residual gain scores did not reach the p = .05 significance level. Wain: H09: There is no difference in the number of therapy sessions received by clients of high CMHI therapists and low CMHI therapists. H 9: Clients of high CMHI therapists received more therapy sessions than clients of low CMHI therapists. The null hypothesis was not rejected. H 10: H 10: 94 There is no linear relationship between the quasi variables and number of therapy sessions once the separate effects of the variables used in construct- ing the quasi variables are removed. There is a linear relationship between the quasi variables and number of therapy sessions once the separate effects of the variables used in construct- ing the quasi variables are removed. Null hypothesis ten was not rejected. 21W W Ihgl: There is no difference in ratings on the Client Reported Degree of Satisfaction Scale between high CMHI therapists and low CMHI therapists. Hall: Client satisfaction ratings made by high CMHI thera- pists are higher than client satisfaction ratings made by low CMHI therapists. The null hypothesis was not rejected 3612: There is no linear relationship between the quasi variables and satisfaction ratings once the separate effects of the variables used in constructing the quasi variables are removed. Ha12: There is a linear relationship between the quasi variables and satisfaction ratings once the separate effects of the variables used in constructing the quasi variables are removed. 95 Null hypothesis twelve was not rejected. Conclusions The results cannot be interpreted as conclusively demonstrating that therapists with low adherence to com- munity mental health ideology are more biased toward making personalistic attributions about psychological disturbance than therapists with high adherence to community mental health ideology. However, the results do give some support for the theory that differences in clinical perceptions are a function of differences in attribution biases between therapists. l. The differences in residual gain scores and subsequent GAS ratings supports the conclusion that high CMHI ther- apists perceive clients as achieving more improvement than low CMHI therapists. 2. Support was also found for the hypothesis that low CMHI therapists perceive clients as less motivated for treat- ment than high CMHI therapists. The different percep- tions of client motivation between high and low CMHI therapists is consistent with the inference that the personal istic bias of low CMHI therapists would clash with clients' situational attributions about their prob- lems and results in low CMHI therapists perceiving clients as less motivated than high CMHI therapists. 96 Disoussion Twenty-four out of twenty-eight eligible therapists volunteered for the study. The high rate of participation supported the conclusion that the sample studied was repre- sentative of the population of adult outpatient therapists employed by the Eaton-Ingham-Clinton County Community Mental Health Board to treat nonpsychotic emotional disorders. The sample of therapists in the present study was not a random sample. Consequently, a question remains about whether the results may be generalized beyond the sample employed for the study. The composition of the high and low CMHI therapist groups were similar on variables of therapist sex, profes- sional discipline and years of experience (see Table 3.1). Clients of high and low CMHI therapists were also similar on variables of education, income and sex (see Appendix C). The similarity of the therapist groups and their respective clients on the variables mentioned about supported the as- sumption that the groups were comparable. Therefore, the differences in motivation and GAS ratings between high and low CMHI therapists were apparently not a result of dissimi- lar groups of therapists or their clients. There are two reasons for caution in the interpretation of the results found to be statistically significant. First, each of the six dependent variables were analyzed with a separate statistical technique. Second, clients were the units of statistical analysis rather than therapists, 97 even though therapists were the objects of interest in the present study. The consequence of both of these factors was that the probability'of erroneously concluding that there was a significant difference between the groups was greater than the probability statements reported. However, the probability statements concerning such an error reported for motivation ratings, subsequent GAS ratings and residual gain scores were much smaller than the p a .05 level of signifi- cance adopted for the present study. In each case, the significance levels actually obtained concerning an errone- ous conclusion were less than p==.fil so that the actual probability of an erroneous conclusion may still have been less than a p = .05 level. The results were consistent with the findings in the Del Gaudio study, in which high CMHI therapists gave higher ratings of prognosis than low CMHI therapists. The results of the Del Gaudio study lends support for the conclusion that therapists*with low'community mental health ideology judge clients more harshly than therapists with high adher- ence to community mental health ideology. Del Gaudio con- cluded that his results were due to more Optimism in high CMHI therapists than low CMHI therapists. However, the question of what might cause the difference in Optimism is unanswered. The differences in motivation and GAS ratings between high and low CMHI therapists were consistent with the postulated theory that low CMHI therapists perceive more intrapersonal psychopathology in clients than high CMHI 98 therapists because of a personalistic attribution bias. However, it is possible that belief in CMH ideology is related to therapists' theoretical orientation and that the high and low CMHI therapists in the present study differed on the basis of theoretical orientation. Differences be- tween the high and low CMHI therapists on the basis of theoretical orientation does not necessarily negate the theory that attributional bias is the source of therapists' different clinical perceptions of clients. Langer and Abelson (1974) found that psychodynamically oriented clini- cians rated clients as significantly more maladjusted than behaviorally'oriented'clinicians and concluded that clinical judgment bias was a function of theoretical orientation. Snyder (1977) reanalyzed Langer and AbelsonFs data and found that attributional bias was related to theoretical orienta- tion. In addition, theoretical orientation and attribution bias were related to perceived severity of clients malad- justment. Psychodynamically oriented therapists were more inclined to make a personalistic attribution about the locus of clients' problems and judged clients as more maladjusted than behaviorally oriented therapists. If differences in theoretical orientation were present between high and low CMHI therapists in the present study, the orientation dif- ferences were possibly associated with differences in attri- bution biases between the high and low CMHI therapists. Snyder tentatively concluded that theoretical orientation has an effect on whether the therapist perceives the locus 99 of the client's problem as person centered or situation centered. However, his data does not support such a causal inference. It is not possible to determine from Snyder's data whether attribution bias is a function of theoretical orientation or whether theoretical orientation is a function of attribution bias. Consequently, the differences found on the dependent variables between high and low CMHI therapists in the present study'may have been a function of therapist theoretical orientation or attribution bias. The size of the differences found between high and low CMHI therapists was not large and may have resulted from splitting the sample at the median of CMHI scores to form high and low CMHI groups. The effect of adherence to com- munity mental health ideology on perceptions of clients would be expected to be more pronounced at the extremes of high and low CMHI adherence than in the middle range of adherence. Both high and low CMHI groups contained thera- pists whose CMHI score were close to the median (see Appen- dix). Consequently, this middle group of therapists, who were present in both high and low CMHI groups, may have been more alike than different in their perceptions of clients. Consequently, therapists from the middle range may have tended to offset the influence of therapists at the ex- tremes, which would have decreased the differences between high and low CMHI means for motivation and GAS ratings. The failure to find a significant difference for ini- tial GAS ratings between high and low CMHI therapists when 100 tested with a two-way analysis of variance when a signifi- cant difference was found in the one-way analysis of vari- ance results from the change in the mean square values. The mean square for the therapists was reduced by a large value in the two—way analysis of variance while the mean square for within groups was reduced by only a small value and resulted in a smaller F value than in the one-way analysis of variance. The most likely interpretation for this change is that the use of more components in the two-way analysis partitioned the overall sum of squares more completely and accounted for more variance in the initial GAS ratings than the one-way analysis of variance. Consequently, the results of the two-way analysis of variance was a more precise analysis of the data. High and low CMHI therapists ap- parently perceived the same degree of client disturbance at intake. The difference, between high and low CMHI thera- pists, on subsequent GAS ratings was a result of differences in the amount of client improvement perceived by therapists. The failure to find a difference in the number of therapy sessions between high and low CMHI therapists may have been due to agency centered forces. The agency empha- sized brief therapy and the emphasis on brief therapy was reinforced by peer review. An additional reinforcement of brief therapy was that therapists were required to schedule a fixed number of new clients each month. Full time thera- pists scheduled twelve intakes and half time therapists scheduled six intakes per month. The agency emphasis on 101 brief therapy may have overridden therapist inclinations for long term therapy and is consistent with Carpenter's inter- pretation that therapist variables alone have little impact on the duration of therapy. The absence of a significant difference in the clientis reported degree of satisfaction may have been due to two sources. First, many therapists chose to not fill out the client satisfaction scale on most of their clients or checked the unknown option of the scale. Many therapists commented that they responded this way because they believed that they had no direct comment from the client about their satisfaction. The overall mean for the client satisfaction scale was 1.96 with a standard deviation of .77 which would indicate that most clients who received a rating were rated as satisfied to some degree. Consequently, there was very little room for variance, which would have limited reliabil- ity and reduced the power of the statistical tests. A possible explanation for reluctance to fill out the client satisfaction scale is that therapists were generally reluc— tant to report clients as dissatisfied, which would reflect negatively on therapists. Those clients who were perhaps perceived as dissatisfied may have been the clients who received no rating. The failure to find an interaction between CMHI and client socioeconomic background may have been related to the use of separate demographic variables of education and in- come. These variables have not received much support as 102 valid indicators of client socioeconomic background. Re- search that has found a connection between client socio- economic background and treatment has generally used the Hollingshead criteria for socioeconomic status. ImolioatioanonsommuniLLusntalmm The results and conclusions of the present study cannot confidently be generalized to other therapists and other agencies because the sample was not random and was obtained from only one community mental health board. Generalization of the results requires that the study be replicated with a new sample from other community mental health agencies. The conclusion that therapists in the present study who were low in adherence to community mental health ideology perceived clients as less motivated for treatment, more disturbed and less improved than therapists with high adherence to CMH ideology has potentially important implications that.warrant further study. An implication of these conclusions is that therapists with low adherence to CMH ideology may perceive the problems of clients as less amenable to remediation than therapists with high adherence to CMH ideology. The accu- racy of the perceptions of either therapist group was not studied in the present research and*would not necessarily matter. The perception that problems are not readily amen- able to treatment may conflict with the emphasis on brief therapy present in many community mental health centers. A consequence of such a conflict between therapist perception 103 and agency policy may be therapist apathy and pessimism. The presence of therapist apathy, pessimismtor some other similar affective response may negatively influence the employee-employer relationship and the quality of services provided to clients. Consequently, continued research on the influence of therapist adherence to community mental health ideology and therapist performance is recommended. NW The theory that low CMHI therapists are more subject to personalistic bias than high CMHI therapists was not directly tested. Future research along this line would be strengthened with a dependent measure of therapists' belief that client personnel disposition is the source of psychological disturbance rather than situational factors. Such methodology could be patterned after methodology used in research in attribution theory. Therapist theoretical orientation may be associated with level of community mental health ideology. Theoretical orientation of therapists needs to be considered in future research on therapists'ldegree of adherence to community mental health ideology. Future research concerning bias toward clients from different socioeconomic backgrounds would be improved if clients' socioeconomic background were evaluated using the Hollingshead criteria for socioeconomic status, rather than separate demographic variables. In 104 addition, the social class of therapists' family of origin needstx>be considered because social class of origin has been linked to therapists' receptiveness to working with clients from lower socioeconomic classes (Kandel, 1966: Mitchell & Namenek, 1970). Any interac- tion found between levels of CMHI and client social class may also be explained by therapists' social class of origin. The question of whether the difference in perception really makes any difference in the quality of treatment could be addressed by using the Barrett—Lennard Rela- tionship Inventory as a dependent variable. Another area worth investigating is the relationship between adherence to community mental health ideology and job satisfaction and burnout. APPENDICES APPENDIX A COMMUNITY MENTAL HEALTH IDEOLOGY SCALE PLEASE NOTE: Copyrighted materials in this document have not been filmed at the request of the author. They are available for consultation, however, in the author's university library. These consist of pages: Appendix A, pages 105-109 Appendix 8, pages 110-111 University? Micro Ilms lntemational 330 N. ZEEB RD. ANN ARBOR, Ml 48106 (313) 761.4700 Instruction: APPENDIX A BAKER-SCHULBERG CHMI SCALE Please read each of the statements carefully, in the order in which it appears, and for each one, indicate to what extent you personally agree or disagree with it. You should do this by circling, next to each statement, the ONE of six symbols which best represents your own feeling about the statement. Circle AAA, if you STRONTLY agree Circle AA, if you MODERATELY agree Circle A, if you SLIGHTLY agree Circle DDD, if you STRONGLY disagree Circle DD, if you MODERATELY disagree Circle D, if you SLIGHTLY disagree 3. Every mental health center'should.have formally'associated with it a local citi- zen's board assigned significant respon- sibilities. Our time-tested pattern of diagnosing and treating individual patients is still the optimal way for us to function profes- sionally. with out limited professional resources, it makes more sense to use established knowledge to treat the mentally ill rather than trying to deal with the so- cial conditions which may cause mental illness. 105 Strongly Agree AAA AAA AAA Moderately Agree AA AA AA Slightly Agree Slightly Disagree Moderately Disagree DD DD DD Strongly Disagree DDD DDD DDD 5. 9. 10. 11. 12. 106 Our responsibility for patients extends beyond the contact we have with them in the mental health center. A significant part of the psychiatrist's Job consists of finding out who the mentally disordered are and where they are located in the community. Such public health programs as primary preventive services are still of little value to the mental health field. A mental health program should direct particular attention to groups of people who are potentially vulnerable to upset- ting pressures. The planning and operation of mental health programs are professional func- tions which should not be influenced by citizen pressures. Mental health programs should give a high priority to lowering the rate of new cases in a community by reducing harmful environmental conditions. The mental health specialist should seek to extend his effectiveness by working through other people. A mental health professional can.only'be responsible for the mentally 111 who come to him; he cannot be responsible for those who do not seek him out. Our program emphasis should be shifted from the clinical model, directed at specific patients, to the public health model, focusing upon populations. Strongly Agree AAA AAA AAA AAA AAA AAA AAA AAA AAA Moderately Agree AA AA AA AA AA AA AA AA AA Slightly Agree Slightly Disagree U Moderately Disagree DD DD DD DD DD DD DD DD Strongly Disagree DDD DDD DDD DDD DDD DDD DDD DDD DDD 13. 1A. 15. 16. 17. 18. 19. 20. 21. 107 understanding of the community in which we work should be made a central focus in the training of'mental health profes- sionals. The control of mental illness is a goal that can only be attained through psy- chiatric treatment. A mental health professional assumes responsibility not only fer his current caseload, but also for unidentified potentially maladjusted people in the community. Our current emphasis upon the problems of individual patients is a relatively ineffective approach for easing a commu- nity's total psychiatric problem. Our professional mandate is to treat individual patients and not the harmful influences in society. Our efforts to involve citizens in men- tal health programs have not produced sufficient payoff to make it worth our while. The locus of mental illness must be viewed as extending beyond the indivi- dual, and into the family, the communi- ty, and the society. Mental health professionals can be con- cerned for their patient's welfare only when having them in active treatment. Mental health consultation is a neces- sary service which we must provide to community caregivers who can help in the care of the mentally’ill. m m a) H a) (D Di 0) H :3 H U1 U? a: >w «c p: >1 0) >1 r4 .u F4 or m +) r: H .1: O (1) ON H 'U "-1 +’ 0 r4 (D 2: to AAA AA A AAA AA A AAA AA A AAA AA A AAA AA A AAA AA A AAA AA A AAA AA A AAA AA A Slightly Disagree Moderately Disagree DD DD DD DD DD DD DD DD DD Strongly Disagree DDD DDD DDD DDD DDD DDD DDD DDD DDD 22. 23. 2h. 25. 26. 27. 28. 29. 108 Caregiving agents who worked with the patient before and during his contact at the mental health center should be in- cluded in the formulation of treatment plans. A psychiatrist can only provide useful services to those people with whom he has direct personal contact. Skill in collaborating with nonmental health professionals is relatively unim- portant to the success of our work with the mentally ill. The mental health center in only part of a comprehensive community mental health program. Mental health professionals should only provide their services to individuals whom society defines as mentally ill or who voluntarily seek these services. We should deal with people who are not yet sick by helping them to develop ways for coping with expected life difficul- ties. we should not legitimately be concerned with modifying aspects of our patient's environment but rather in bolstering his ability to cope with it. It is a poor treatment policy to allow non-psychiatrists to perform traditional psychiatric functions. Strongly Agree AAA AAA AAA AAA AAA AAA AAA AAA Moderately Agree AA AA AA AA AA AA AA AA Slightly Agree Slightly Disagree Moderately Disagree DD DD DD DD DD DD DD DD Strongly Disagree DDD DDD DDD DDD DDD DDD DDD DDD 30. 31. 32. 33. 3". 35. 36. 37. 38. 109 Since we do not know enough about pre- vention, mental health programs should direct their prime efforts toward treat- ing the mentally ill rather than.devel- oping prevention programs. The hospital and community should strive for the goal of each participating in the affairs and activities of the other. Social action is required to insure the success of mental health programs. In view of the professional manpower shortage, existing resources should be used for treatment programs rather than prevention programs. Each mental health center should Join the health and welfare counsel of each community it serves. The respnosible mental health profes- sional should become an agent for social change. We can make more effective use of our skills by intensively treating a limited number of patients instead of working indirectly with many patients. By and large, the practice of good psy- chiatry does not require very much knowb ledge about sociology and anthropology. Community agencies working with the patient should not be involved with the different phases of a patient's hospi- talization. Strongly Agree AAA AAA AAA AAA AAA AAA AAA AAA AAA Moderately Agree AA AA AA AA AA AA AA AA AA Slightly Agree Slightly Disagree Moderately Disagree DD DD DD DD DD DD DD DD DD Strongly Disagree DDD DDD DDD DDD DDD DDD DDD DDD DDD APPENDIX B GLOBAL ASSESSMENT SCALE APPENDIX B GLOBAL ASSESSMENT SCALE (GAS) Robert L. Spitzer, M.D., Miriam Gibbon, M.S.W., Jean Endicott, Ph.D. Rate the client's lowest level of functioning on a hypothet- ical continuum of mental health-illness. The rating is based on observed and reported functioning for the geek .ast_gonta§t, Rate actual functioning inde- pendent of diagnosis or treatment. 100 Superior functioning in a wide range of activities, | life‘s problems never seem to get out of hand, is sought out by others because of his warmth and 91 integrity. No symptoms. 90 Good functioning in all areas, many interests, socially ‘ effective, generally satisfied with life. There may or may not be transient symptoms and "everyday‘ worries 81 that only occasionally get out of hand. 80 No more than slight impairment in functioning, varying ‘ degrees of "everyday" worries and problems that sometimes get out of hand. Minimal symptoms may or may 71 not be present. 70 Some mild symptoms (e4b, depressive mood and mild insomnia) OR some difficulty in several areas of functioning, but generally functioning pretty well, has some meaningful interpersonal relationships and most 61 untrained people would not consider him "sick.“ 60 Moderate symptoms OR generally functioning with some difficulty (e.g., few friends and flat affect, depressed mood and pathological self-doubt, euphoric mood and pressure of speech, moderately severe anti- 51 social behavior). 50 Any serious symptomatology or impairment in functioning that most clinicians would think obviously requires treatment or attention (eéh, suicidal preoccupation or gesture, severe obsessional rituals, frequent anxiety attacks, serious antisocial behavior, compulsive 41 drinking, mild but definite manic syndrome). 110 40 31 30 21 20 11 10 111 Major impairment in several areas, such as work, family relations, judgment, thinking or mood (3&L' depressed woman avoids friends, neglects family unable to do housework), OR some impairment in reality testing or communication (e.g., speech is at times obscure, illogical or irrelevant), or single suicide attempt. Unable to function in almost all areas (64%, stays in bed all day) OR behavior is considerably influenced by either delusions or hallucinations OR serious impair- ment in community (e.g., sometimes incoherent or unre- sponsive) or judgment (equ acts grossly inappropri- ately). Needs some supervision to prevent hurting self or others, or to maintain personal hygiene (64L, repeated suicide attempts, frequently violent, manic excitement, smears feces), OR gross impairment in communication (e.g., largely incoherent or mute). Needs constant supervison for several days to prevent hurting self or others (e4L, requires an intensive care unit with special observation by staff), makes no attempt to maintain minimal personal hygiene, or serious suicide act with clear intent and expectation of death. APPENDIX C DISTRIBUTION OF CLIENTS FOR HIGH AND LOW CMHI THERAPISTS APPENDIX C DISTRIBUTION OF CLIENTS FOR HIGH AND LOW CMHI THERAPISTS Table A1. Client Distribution by Income for High and Low CMHI Therapists Family Income Frequency LQfl QMHI Therapists 1 Less than $1,000 22 2 $ 1,000 - 1,999 1 3 $ 2,000 - 2,999 28 4 $ 3,000 - 3,999 7 5 $ 4,000 - 4,999 7 8 $ 7,000 - 7,999 3 9 $ 8,000 - 8,999 2 10 $ 9,000 - 9,999 3 11 $10,000 - 11,999 5 12 $12,000 - 14,999 9 13 $15,000 - 24,999 l4 14 $25,000 - 49,999 10 WW 1 Less than $1,000 20 2 $ 1,000 - 1,999 0 3 $ 2,000 - 2,999 18 4 $ 3,000 - 3,999 9 5 $ 4,000 - 4,999 7 6 $ 5,000 - 5,999 3 7 $ 6,000 - 6,999 7 8 $ 7,000 - 7,999 2 9 $ 8,000 — 8,999 4 10 $ 9,000 — 9,999 4 11 $10,000 - 11,999 9 12 $12,000 - 14,999 10 13 $15,000 - 24,999 19 14 $25,000 - 49,999 7 N = 239 Chi Square = 8.32 Significance = .83 112 113 Table A2. Distribution of Clients by Education for High and Low CMHI Therapists Client Years of Education Frequency WW: 4 1 5 1 6 0 7 l 8 0 9 4 10 5 ll 8 12 59 13 16 14 10 15 1 16 8 17 1 18 4 19 1 20 0 MH T ' : 4 0 5 1 6 0 7 0 8 6 9 3 '10 9 ll 6 12 43 13 18 14 12 15 7 16 6 17 2 l8 4 19 1 20 1 N = 239 Chi Square = 18.50 Significance = .24 114 Table A3. Client Distribution by Sex for High and Low CMHI Therapists Client Sex Frequency MH ‘1' ' : Male 36 Female 84 H MH T : Male 39 Female 80 N = 239 Chi Square = .21 Significance = .64 APPENDIX D CONSENT FORM 115 APPENDIX D CONSENT FORM TO PARTICIPATE IN RESEARCH I understand that this study is being conducted by Robert Smith, M.A., for the purpose of exploring therapists' belief in the community mental health movement. I understand that an area being studied is therapists' perceptions and clinical decisions. I understand that I will be completing a measure regarding degree of belief in assumptions underlying the com- munity mental health movement. I understand that my participation will remain strictly confidential. I understand that upon completion of the data collection procedure the scored test containing my name will be returned to me and feedback provided. I understand that all information will be recorded in such a way that my identify cannot be determined from any of these materials. I know that I am under no obligation to participate in this study and will not be penalized if I decline to participate. I also know that I may withdraw my consent to participate at any time by con- tacting Mr. Smith. Signature Date APPENDIX E COVER LETTER AND INSTRUCTIONS FOR THERAPISTS APPENDIX E COVER LETTER AND INSTRUCTIONS FOR THERAPISTS June 13, 1983 Dear Colleague: I have obtained permission from the Director of the Center to conduct a research study under the supervision of Dr. William Farquhar. The area I will be looking at is the relationship between therapists' belief in assumptions underlying the community mental health movement and therapists' perceptions of and clinical decisions about clients. There are a limited number of therapists within the Tri- County system appropriate for this study. Consequently, it is important that as many of you participate as possible for the study to be feasible. However, this research is for my dissertation and you are not obligated by CMH to participate. As one of you, I know that your schedule is already overloaded. Consequently, I have kept the demand on your time and energy to an absolute minimum. Your part in the study will only require about 15 minutes of your time to complete the 38-item Community Mental Health Ideology Scale and answer a few brief questions about yourself. Your name is needed on the test form because this informa- tion must be matched with clients you have seen. THIS IS NOT A STUDY OF INDIVIDUAL THERAPISTS. However, matching of therapist scores with information on clients is a fundamental part of the analysis procedure. I will have the only access to a temporary record connecting names and data. Once I have transferred the information to separate forms, your scored test form will be returned to you with a brief explanation. This means that prior to any analysis and interpretation of the data, there will be no record left, connecting names and data. (So, I had better do it right the first time!) 116 117 June 13, 1983 Cover Letter Please feel free to contact me at Ingham (374-8000) or home (351-2744) if you want clarification. At the completion of the study, I will send staff a summary of the results. Enclosed in this packet is a: . Demographic Cover Sheet (attached to CMHI Scale) The Community Mental Health Ideology Scale Consent Form Pre-addressed, stamped envelope DNNH o e 0 Please complete the three forms, seal them in the pre- addressed envelope and drop it in the regular mail. If you decide to participate, I need to receive this information by June 24, 1983. However, if you are late in sending this material, it may still be possible for me to use your materials, so please don't hesitate to send the forms. I thank you for your time and your help. Sincerely yours, Robert Smith APPENDIX F THERAPIST DEMOGRAPHIC QUESTIONNAIRE 118 APPENDIX F DEMOGRAPHIC COVER SHEET Name: Sex: Male Female Highest Professional Degree (check one): MSW BSW Ph.D. Clinical Psychology Ph.D. Counseling Psychology M.A. Clinical Psychology_____ M.A. Counseling PsychOIOgy_____ Other (£111 in) Please list other related professional degrees: Number of years of paid experience as a therapist: APPENDIX G CASE HISTORIES FOR RELIABILITY STUDY OF MOTIVATION FOR TREATMENT SCALE AND GLOBAL ASSESSMENT SCALE 119 APPENDIX G CASE 1 Mary, 31, stated that she was not quite sure what to do or where to start. She was somewhat passive and quiet. Mary has been having frequent attacks of anxiety with symptoms in- volving palpitations, smothering feelings, pains in her chest and dizziness, with fear of dying or going crazy for the past 6 months, but with a history going back 5 years. She is separated from her husband, who still supports her and her 7 year old son from a previous marriage. She does not work and spends little time outside the home. In the past week, she has only been out once, for lunch with a girlfriend and then an afternoon of shopping. Otherwise, she stays at home, looking after the child, preparing meals, etc., and shopping locally. She admits that she couldn't care less about how the apartment looks and doesn't bother to keep it clean and tidy. She is generally nervous, but this does not interfere with her daily activities. The attacks have been occurring about once a week. This past week was uneventful. Her husband visited on the weekend to give her some money but left soon afterwards. She spent the rest of the day watching television. On Monday afternoon, she had a mild attack. She thought she was going to have another attack on Wednesday, but all she felt was a few palpitations. On Thursday morning, when she was alone, another attack developed. She spoke to a neighbor about her attacks. The neighbor had previously been seen at Community Mental Health and suggested she phone for an appointment. GAS Motivation for Treatment: High Motivation Low Motivation 1 2 3 4 5 120 APPENDIX G CASE 2 Robin, 47, came on the advice of her sister. Her speech was crisp and expression taut. She has come for treatment because she has had trouble controlling her temper toward her 14 year old hyperactive son. Her son has always been difficult, but she has been less able to cope with him re- cently. This difficulty has apparently developed since her husband had a coronary 6 months ago. In the past few weeks, she awakens early in the morning feeling tired and tense, and is extremely irritable with her son until she goes to work.. She puts in a full day at the family owned store, in addition to doing the shopping and housework, but during the evening she is continuously irritated by her son, snapping at him about his not doing homework, his messing up the house, his playing music loud. She yells at him, gets angry, and goes to bed exhausted and depressed but cannot get sleep. She feels hopeless about changing the situation. GAS Motivation for Treatment: High Motivation Low Motivation 1 2 3 4 5 121 APPENDIX G CASE 3 Dave, a 43 year old school teacher and ex-patient, came for his yearly check-up. In the last week, he has returned home directly from school every evening feeling rather tired. His wife greets him with a ready made martini; he drinks quietly while she tells him of the events of the day. He blew up last week when his wife told him how she dealt with a minor transgression by his 12 year old son. Feeling that she had been too lenient, he called the boy down from his homework and chastised him until the boy broke into tears and ran from the room. He turned on the wife and accused her of leniency and started to berate her. She refused to be treated that way, left the room and told him she would speak to him again when he was rational. He sat by himself for half an hour, reviewing the situation, and agreed with his wife that he had overreacted. He sought her out and apologized to his son, who was still very upset. The rest of the evening passed with some hostility in the family directed toward him, but the following morning the incident had been forgotten. Although he has been irritable with the children in his class, he hasn't reacted violently either verbally or otherwise. He gets along well with his colleagues at school, and there have been no complaints about his behavior in the classroom. Most evenings are spent reading or watching television; during the week-nights he does not like to involve himself in family activities unless he has to. On the weekend he relaxes. While his wife is busy in the house, he and his two sons clean up their boat and go for a sail. On Sunday, his wife and son joined him watching the 16 year old son play soccer. His wife cooked goodies for the post-game party and he acted as linesman. GAS Motivation for Treatment: High Motivation Low Motivation 1 2 3 4 5 122 APPENDIX G CASE 4 Pam is a 42 year old housewife. She is pleasant, if not a rather bland woman, who does not complain of problems. She is occupied with raising her two school-age children and taking care of the house, and gets considerable enjoyment from both, although she does express some concern that she is too quick to criticize the children and gets angry with them over minor matters, e.g., their failure to keep their rooms neat. This lead her to become involved in a group of women who meet weekly. She describes her marriage as satisfactory, although it seems that she is closer to her two women friends than to her husband. She has no strong interests outside of her family, and other than some minimal and obligatory PTA participation, her activities outside of the home consist of frequent coffee-klatches with her women friends, and occasional movies, parties, etc. GAS Motivation for Treatment: High Motivation Low Motivation 1 2 3 4 5 123 APPENDIX G CASE 5 Anne, 65 and retired, feels guilty that she stole some paper clips and stationery supplies from her place of employment many years ago (as did others). She believes that this was a crime-- a matter for the police, and she wishes they would come and arrest her. In the last week, she has been getting up earlier and earlier in the morning after only 5 or 6 hours of restless sleep. She reluctantly gets dressed and sometimes stays in her night clothes. She tries to prepare breakfast, but becomes agitated and starts pacing around, forgetting what is on the stove. When she sits down to eat, she is up in a moment or two, hardly touching her food. When she is forced to wash, she does so with many interruptions, complaining about her problems. When she is confined, as in a car, she asks to be let out, but if that isn't possible, she continues to sit twisting her hands in her lap and complaining on and on about her guilt. The family has tried to prevent her from talking to friends and neighbors as she expresses her problems too openly, but if she can buttonhole someone, she does so. She will occasionally nap for an hour or so in the afternoon, but upon awakening, resumes her pacing and agitated complaining. This continues throughout the evening. She tries to find things to do to take her mind off her problems, but is not successful in com- pleting simple household tasks. GAS Motivation for Treatment: High Motivation Low Motivation 1 2 3 4 5 REFERENCES 124 References Abramowitz, C.V. & Dokecki, P.R. "The politics of clinical judgment: Early empirical returns!‘ Psychology Bulletin, 1977, 84(3), 460-476. Abramowitz, S.I., Abramowitz, C.V., Jackson, C. a Gomes, G. 'The politics of clinical judgment: What nonliberal examiners infer about women who do not stifle themselves.‘ Journal of Consulting and Clinical Psychology: 1973, 41, 385-391. Adorno, T.W., Frenkel-Brunswik, E., Levinson, D.J. & Sanford, R.N. The authoritarian personality. New York: Harper and Row, 1950. Amira, S., Abramowitz, S.I. & Comes-Schwartz, B. "Socially- charged pupil and psychologist effects on psychoedu- cational decisions." The Journal of Special Education, 1977, 11(4), 435-440. Baekeland, F. & Lundwall, L. "Droppint out of treatment: A critical review." Psychological Bulletin, 1975, 82(5), 738-783. ' I Baekeland, F., Lundwall, L.:& Shanahan, T.J. "Correlates of patient attrition in the outpatient treatment of alco- holism." The Journal of Nervous and Mental Disease, 1973, 157(2), 99-107. Baker, F. & Schulberg, H.C. "The development of a community mental health ideology scale." Community Mental Health Journal, 1967, 3(3), 216-225. Baker, F. & Schulberg, H.C. "Community mental health ideoloqy, doqmatism, and political-economic conservatism." Com- munity Mental Health Journal, 1969, 5(6), 433-436. Berzins, J.I "Therapist-patient matching." In A.S. Gurman and A.M. Razin (Eds.)v Effective Psychotherapy. New York: Pergamon Press, 1977. Brown, R.D. "Expereinced and inexperienced counselors' first impressions of clients and case outcomes: Are first impressions lasting?" Journal of Counselinqusycholoqy, 1970, 17(6), 550-558. Brenner, D. & Howard, K. "Clinical Judgment as a function of experience and information." Journal of Clinical Psy- chology, 1976, 32(3), 721-728. 125 Canter, J.H. "Individual needs and salient constructs in interpersonal perception." Journal of Personality and Social Psychology, 1976, 34(3), 519-525. Carpenter, P.J., Del Gaudio, A.C., Morrow, G.R. & Ritzler, ' B.A. "Who keeps the first outpatient appointment?" American Journal of PsychiatEX: 1981, 138, 102-105. Carpenter, P.J. & Range, L.M. "Predicting psychotherapy duration from therapists' sex, professional affiliation, democratic values, and community mental health ideology." Journal of Clinical PsycholOng 1982, 38(1), 90-91. Cohen, J. "Partialed products are interactions: Partialed powers are curve components." Psychological Bulletin, 1978, 85, 858-856. De Gaudio, A.C., Stein, L.S., Ansley, M.Y. & Carpenter, P.J. "Attitudes of therapists varying in community mental health ideology and democratic values." Journal of Consulting and Clinical Psychology: 1976, 44(4), 646-655. Duckro, P., Beal, D. & George, C. "Research on the effects of disconfirmed client role expectations in psychotherapy: A critical review." Psychological Bulletin, 1979, 86(2), 260-275. Edelwich, J & Brodsky, A. Burn-out. Stages of Disillusionment in the Helping Professions. New York: Human ScIences Press, 1980. Eisenthal, S., Ferrell, R. & Lazare, A. "Attitude of psychiatric residents toward patient requests." American Journal of Psychiatry, 1976, 133(9), 1079-1081. Endicott, J., Spitzer, R., Fleiss, J. & Cohen, J. "The Global Assessment Scale: A procedure for measuring overall severity of psychiatric disturbance." Archives of General Psychiatry, 1976, 33, 766-771. Fiske, D.W. "Methodoloqical issues in research on the psycho- therapist." In A.S. Gurman & A.M. Razin (Eds.), Effective Psychotherapy. New York: Pergamon Press, 1977. Fiske, D., Hunt, H. Luborsky, L. Orne, M., Parloff, M., Reiser, M. & Tuma, A.H. "Planning research on effectiveness of psychotherapy." Archives of General Psychiatry, 1970, 22, 22-32. Fisher, J.D., Nadler, A. & Whitcher-Alagna, 8. "Recipient reactions to aid." Psychological Bulletin, 1982, 91(1), 27-54. 126 Garfield, S.L. "Research on client variables in psychotherapy." In S.L. Garfield & A.E. Bergin (Eds.), Handbood of Psygho- therapyiand Behavior Change: An Empirical Analysis. New York: John Wiley & Sons, 1978. Henrey, W.E. "Personal and Social Identities of Psychotherapists." In A.S. Gurman & A.M. Razin (Eds.), Effective Psychotherapy. New York: Pergamon Press, 1977. Herman, S.E. "Global assessment scales: Southeastern region reliability study." Unpublished manuscript, 1981. (Available from Research and Evaluation Division, Michigan Department of Mental Health, Lewis Cass Building, Lansing, Michigan 48926). Hoffnung, R. a Taylor, A., Tylim, I. & Koretsky, M. "Endorsement of a community mental health ideology: A guide for inservice staff training." In Community Mental Health and Behavioral- Ecology, New York: Plenum Press, 1982. Hornstra, R.K., Lubin, B., Lewis, R.V. & Willis, B.S. "Worlds apart: Patients and professionals. Archives of General Psychiatry. 1972, 27(10), 553-557. Howard, K.I., Orlinsky, D.E. & Trattner, J.H. "Therapist orien- tation and patient experience in psychotherpay." Journal of Counselinngsycholqur 1970, 17(3), 263-270. Jeger, A.M. & Slotnick, R.S. "Community mental health: Toward a behavioral-ecological perspective." In A.M. Jeger and R.S. Slotnick (Eds.), Community Mental Health and Behavioral- Ecology. New York: Plenum Press, 1982. Johnson, D.W. & Matross, R.P. "Interpersonal influence in psychotherapy: A social psychological View." In A.S. Gurman, and A.M. Razin (Eds.), Effective Psychotherapy. New York: Pergamon Press, 1977. Kandel, D.E. "Status homophily, social context, and participation in psychotherapy." American Journal of Sociology: 1966, 71, 640-650. Kaplan, M.F. "Dispositional effects and weight of information in impression formation." Journal of Personality and Social Psychology, 1971, 18, 279-284. Kaplan, M.F. "Measurement and generality of response disposi- tions in person perception." Journal of Personality, 1976, 44, 179-194. Kaplan, M.F. "Information integration in social judgment: Interaction of judge and informational components." In M.F. Kaplan and S. Schwartz (Eds.) Human Judgment and Decision Processes. New York: AcademIc Press, 1975. 127 Kurtz, N., Kurtz, R. & Hoffnung, R. "Attitudes toward the lower and middle-class psychiatric patient as a function of authoritarianism among mental health students." Journal of Consultiagyand Clinical Psychology, 1970, 35(3), 338-341. Langer, E.J. & Abelson, R.P. "A patient by any other name...: Clinician group difference in labeling bias." Journal of Consulting and Clinical PsychongYr 1974, 42, 4-9. Langston, R.D. "Community mental health centers and community mental health ideology." Community Mental Health Journal, 1970, 6(5), 387-392. Lerner, B. & Fiske, D.W. "Client attributes and the eye of the beholder." Journal of Consulting and Clinical Psychology, 1973, 40(2), 272-277. Lerner, B. Therapy in the ghetto: Political impotence and personal disintegration. Baltimore: Johns Hopkifis Press, 1972. Lerner, B. "Democratic values and therapeutic efficacy: A construct validity study." Journal of Abnormal PsycholOgy, 1973, 82, 491-498. Mintz, J. "The role of the therapists in assessing psychotherapy outcome." In A.S. Gurman & A.M. Razin (Eds.), Effective Psychotherapy: A HandbookcnfResearch. New YOrk: Pergamon Press, 1977. Mintz, J. "What is 'success' in psychotherapy?" Journal of Abnormal Psychology, 1972, 80, 11-19. Mitchell, K.M. & Namenek, T.M. "A comparison of therapist and client social calss." Professional Psychology, 1970, 1, 225-230. Orlinsky, D.E. & Howard, K.I. Varieties of Psychotherapeutic Experience. New York: Teachers College Press, 1975. Parloff, M.B., Waskow, I.E. & Wolfe, B.E. "Research on therapist variables in relation to process and outcome." In S.L. Garfield & A. E. Gergin (Eds.), Handbook of Psychotherapy and Behavior Change: An Empirical Analysis. New York: John Wiley, 1978. Penn, N., Baker, F. &.Schulberg, H. "Community mental health ideology and personality preferences of social work students." Journal of Community Psychology: 1976, 4, 335-336. 128 Pettit, I.E., Pettit, T.F., & Welkowitz, J. "Relationship between values, social class, and duration of psychotherapy." Journal of Consulting and Clinical Psyghology: 1974, 42(4), 482-490. - Rabkin, J.G. "Therapists' attitudes toward mental illness and health." In A.S. Gurman & A.M. Razin (Eds.), Effective Psychotherapy. New York: Pergamon Press, 1977. Robin, 8. & Wagenfeld, M. "The nature and correlates of community mental health ideology in community mental health centers." Journal of Community Psychologlr 1976, 4, 335-346. Ross, L. "The intuitive psycholoqist and his shortcomings: Distortionijlthe attribution process." In L. Berkowitz (Ed.), Advances in Experimental Social Psycholoqy (Vol. 10). New York: Academic Press, 1977. Schulberg, H.C & Baker, F. "Varied attitudes to community mental health." Archives of General Psychiatry, 1967, 17(12), 658-663. Schulberg, H.C. & Baker, F. The Mental Hospital and Human Services. New York: Behavioral Puincations, Inc., 1975. Schwartz, J.M. & Abramowitz, S.I. "Value-related effects of psychiatric judgment." Archives of General Psychiatry, 32, 1525-1529. Sharaf, M. & Levinson, D.J. "Patterns of ideology and role dif- ferentiation among psychiatric residents." In M. Greenblatt, D.J. Levinson & R.H. Williams (Eds.), The Patient and the Mental Hospital. Glencoe, Ill.: The Free Press, 1957. Sharf, R.S. & Bishop, J.B. "Counselors' feelings toward clients as related to intake judgments and outcome variables." Journal of Counseling Psychology, 1979, 26(3) 267-269. Sheinfeld, S.M. & Weirich, T.W. "Ideology and performance: Service delivery in a community mental health center." Public Administration Review, 1981, 41, 63-72. Snyder, C.R. "A patient by any other name" revisited:‘ Malad- justment or attributional locus of the problem? Journal of Consulting and Clinical Psychology, 1977, 45, 101-103. Strupp, H.H. Psychotherapy: Clinical, research, and theoretical issues. New York: Jason Aronson, Inc., 1973. StrupP: H.H. & Williams, J.V. "Some determinants of clinical evaluations of different psychiatrists." Archives of General Psychiatry, 1960, 2(4), 434-440. 129 Trachtman, J.P. "Socio-economic class bias in Rorschach diag- nosis: Contributing psychosocial attributes of the clinician." Journal of Projective Techniques and Personality Assessment, 1971, 35, 229-240, 1971. Wallach, M.S. & Strupp, H.H. "Psychotherapists' clinical judg- ments and attitudes towards patients." Journal of Consulting and Clinical Psychology, 1960, 24(4), 316-323. Wills, T.A. "Perceptions of clients by professional helpers." Psychological Bulletin, 1978, 85(5), 968-1000. Wright, J.A. & Hutton, B.O. "Influence of selected behaviors, attitudes, and decisions of counselors." Journal of Counseling Psychology: 1977, 24(6), 527-530. Ill“! 93 03175 0262 I" I“ "I I“ "I I E“ H n H 1 31 KIHHWIHH