THE CLINICAL EXPERIENCE OF THERAPISTS IN A TRAINING PROGRAM AS A PREDICTOR OF THE THERAPY ALLIANCE, CLIENT DROPOUTS, AND DURATION OF THERAPY Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. ______________________________ Austin W. Houghtaling ______________________________ Alexander T. Vazsonyi Professor Scott A. Ketring, Chair Associate Professor Human Development and Family Human Development and Family Studies ______________________________ Thomas A. Smith Associate Professor Human Development and Family Studies ______________________________ Studies ______________________________ Stephen L. McFarland Acting Dean Graduate School THE CLINICAL EXPERIENCE OF THERAPISTS IN A TRAINING PROGRAM AS A PREDICTOR OF THE THERAPY ALLIANCE, CLIENT DROPOUTS, AND DURATION OF THERAPY Austin W. Houghtaling A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 15, 2006 THE CLINICAL EXPERIENCE OF THERAPISTS IN A TRAINING PROGRAM AS A PREDICTOR OF THE THERAPY ALLIANCE, CLIENT DROPOUTS, AND DURATION OF THERAPY Austin W. Houghtaling Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions at their expense. The author reserves all publication rights. ______________________________ Signature of Author ______________________________ Date of Graduation iii iv THESIS ABSTRACT THE CLINICAL EXPERIENCE OF THERAPISTS IN A TRAINING PROGRAM AS A PREDICTOR OF THE THERAPY ALLIANCE, CLIENT DROPOUTS, AND DURATION OF THERAPY Austin W. Houghtaling Master of Science, December 15, 2006 (B.S., Brigham Young University, 2004) 66 Typed Pages Directed by Scott A. Ketring Little research has addressed the effect clinical experience during graduate training has on students? ability to improve therapeutic effectiveness. This study examined the relationship between student-therapists? clinical experience and three therapeutic factors shown in the literature to be associated with client outcome?the therapy alliance, the number of dropouts, and the number of completed sessions therapists have with their clients. Data were gathered from 24 student-therapists and 420 cases seen at the Marriage and Family Therapy Center at Auburn University. Therapy alliance scores, whether or not the client dropped out of treatment, and the number of completed sessions was tracked for each therapist?s cases seen at the university clinic. v Clinical experience was measured using two formats: time in the training program and the number of client-contact hours accrued by student-therapists throughout the program. Results showed a negative, significant relationship between clinical experience and the number of dropouts?with more clinical experience, therapists had fewer clients drop out of treatment. There was also a positive, significant relationship between clinical experience and the number of sessions completed with clients?therapists with more clinical experience had, on average, more completed sessions with their clients than therapists with less clinical experience. vi Style manual used: Publication Manual of the American Psychological Association, Fifth Edition. Computer software used: Microsoft Word, SPSS vii TABLE OF CONTENTS LIST OF TABLES.........................................................................................................viii I. INTRODUCTION ................................................................................................ 1 II. REVIEW OF LITERATURE ............................................................................... 4 III. METHODS ......................................................................................................... 17 IV. RESULTS ........................................................................................................... 26 V. DISCUSSION..................................................................................................... 42 REFERENCES ............................................................................................................... 49 APPENDIX..................................................................................................................... 53 viii LIST OF TABLES 1. Summary of Means and Standard Deviations for Fourth Session Therapy Alliance Scores and Sessions Completed...............................................27 2. Summary of Means and Standard Deviations for Male and Female Fourth Session Therapy Alliance Scores, Dropouts, and Completed Sessions, Separated Into Three Multi-Month Segments.......................................28 3. Summary of Means and Standard Deviations for Male and Female Fourth Session Therapy Alliance Scores, Dropouts, and Completed Sessions, Separated Into Five, 100-hour Segments ..............................................29 4. Correlation Coefficients for Therapy Alliance Scores and Variables of Interest ..............................................................................................................31 5. Correlation Coefficients for Therapy Alliance Scores, Dropout Rate, Number of Sessions Completed, and Variables of Interest ..................................32 6. Summary of Means and Standard Deviations for Males? Fourth Session Therapy Alliance Scores and Females? Ages ..........................................32 7. Correlation Coefficients for Therapy Alliance Scores and Variables of Interest ..............................................................................................................34 8. Correlation Coefficients for Dropout Rate, Number of Sessions Completed, and Variables of Interest....................................................................34 ix 9. Summary of Hierarchical Regression Analysis for Variables Predicting Number of Sessions Completed ..........................................................36 10. Summary of Hierarchical Regression Analysis for Variables Predicting Number of Sessions Completed ..........................................................37 1 I. INTRODUCTION Much time, labor, and money is invested in training mental health professionals in graduate programs across the country. It is reasonable to suggest that before spending these resources, training programs should be evaluated as to their effectiveness in producing competent therapists and assuring that during the course of the program student-therapists experience improvement in their ability to do therapy. Additionally, it seems graduate programs would have the desire to assess their program?s strengths and weaknesses and make adjustments in order to better assist students in their development as therapists. Specifically, there has been little or no attempt made at developing a way to assess the effect a student-therapist?s clinical experience has on therapeutic variables known to be related to more favorable client outcomes. As Davenport and Ratliff (2001) point out, ?No consensus exists about the appropriate scope and method of evaluating clinical competence? (p. 441-442). This kind of a study would help determine, in part, the value of graduate training. One literature review described the evidence supporting the value of graduate training as being ?indirect? (Stein & Lambert, 1995). The need for direct evidence exists. In contrast, considerable research has been done on factors correlated with positive client outcomes. Three of these factors which are relatively simple and efficient for a university clinic to assess are the therapy alliance, client dropout rates, and the 2 number of treatment sessions. For example, research supporting the relevance of these three factors in the therapy process includes the following: stronger therapy alliance scores and the number of treatment sessions a client attends are related to more favorable client outcomes (Hampson & Beavers, 1996; Howard, Kopta, Krause, & Orlinsky, 1986; Johanson, Silverberg, & Lilly, 1980; Johnson, Wright, & Ketring, 2002; Kopta, 2003; Messer & Boals, 1981; Symonds & Horvath, 2004). Similarly, greater therapist experience (usually defined by the amount of time a therapist has been practicing) is related to a lower client dropout rate (Atkins & Christensen, 2001; Burlingame, Fuhriman, Paul, & Ogles, 1989; Pekarik, 1994; Pekarik & Stephenson, 1988; Stein & Lambert, 1995; Sue, McKinney, & Allen, 1976) and higher therapy alliance scores (Davenport & Ratliff, 2001). Amid the literature in psychology?arguably the branch of mental health supported by the strongest research?there are very few instances in which graduate training programs do some type of self-evaluation, and when they do, they most often focus directly on client outcome (Atkins & Christensen, 2001; Messer & Boals, 1981; Stein & Lambert, 1995). Several years ago, Julian Ford (1977) wrote an article in which psychologists were encouraged to generate and use research which would make training programs more effective. This admonition is equally suited to the related field of marriage and family therapy. Yet, it appears that no research has been done addressing student-therapists going through these training programs. Moreover, there is a lack of research regarding ways a student-therapist?s changing abilities can be tracked throughout the course of a program. Research involving graduate programs has not 3 attempted to measure?let alone focus on measuring therapist experience as it relates to therapeutic variables. The present study proposes a method of evaluating student-therapist progress within a Commission on Accreditation for Marriage and Family Therapy Education (COAMFTE) approved master?s level training program by looking at the three specific therapeutic variables described above (which will be referred to throughout this paper as ?the three target factors?): the therapy alliance, client dropout rate, and the duration of therapy (as measured by the number of therapy sessions completed). Because little theory has been propounded concerning therapist development, the first step is to examine variables known to be related to good therapeutic outcomes. The next step is to compare them across different points throughout a therapy training program. Future research possibilities are also discussed. 4 II. LITERATURE REVIEW Much research in the fields of marriage and family therapy, counseling, and psychology has focused on client outcome and factors that contribute to greater therapeutic yield, and whether these factors stem from the client, the therapist, or a combination of the two. In the last several years advancements have been made which have helped identify some of these factors and put into practice therapeutic techniques, models, and skills that aid in the therapeutic process. With a growing list and understanding of what some of these helpful factors, skills and techniques are, one would assume that an emphasis be placed on assessing if clinicians are indeed learning them, putting them into practice, and thus becoming better therapists. This is particularly relevant in training programs, which breed the next generation of therapists. A review of the literature, however, shows a relative lack of research in the area of therapist training in general, and an even greater lack when looking specifically at factors known to be related to positive client outcomes. Moreover, the few studies in existence are cross-sectional; none have tracked therapists over the course of training. Because of the lack of research addressing students in graduate training, it was necessary to look at the literature outside of this specific topic in order to find associations between clinical experience or training and therapeutic variables related to client outcome. For example, several of the studies which will be examined in this review 5 of the literature address the present study?s three target factors as they relate to client outcomes, but in a therapy context other than graduate training. Several of the studies looking at therapist training assessed training of certain manualized or short-term therapy programs with already practicing therapists or counselors, and did not look at training in terms of initial schooling or clinical experience in general. This literature review focuses on the three target factors because they are easily measured in most therapy training programs while at the same time have substantial research supporting their therapeutic validity. While looking at these three areas is admittedly not a perfect measure of therapists? development during a training program, it is more than has been done in the field thus far and is a good place to start filling in a large gap in the research. Therapy Alliance The therapy alliance has been a topic of interest in counseling fields in recent years with many studies looking at the importance of maintaining a strong therapeutic alliance with clients. The therapy alliance broadly refers to the relationship between the therapist and the clients. Pinsof and Catherall define the therapeutic alliance as ?that aspect of the relationship between the therapist system and the patient system that pertains to their capacity to mutually invest in, and collaborate on, the therapy? (1986, p. 139). While there is some difference of opinion as to what defines the therapy alliance there is also much agreement. Factors such as whether the clients feel comfortable with and trust the therapist, whether the clients feel the therapist can help them, and whether they feel that the therapist is warm, engaging, understanding, and empathic, are all commonly accepted as contributing to the alliance. Also referred to in the literature as the 6 ?working alliance,? a strong therapy alliance is now widely accepted as fundamental to successful therapy. A meta-analysis of the research done on the therapy alliance and its effect on therapy is beyond the scope of this paper. In one such analysis, Horvath and Symonds looked at 24 different studies (based on 20 distinct data sets) on the therapy alliance and its relation to client outcomes. They found a ?moderate but reliable? association between a good working alliance and positive therapy outcomes (Horvath & Symonds, 1991). There is much research showing how the therapy alliance positively affects client outcomes. Studies involving measures of the therapy alliance often use the Working Alliance Inventory (Horvath & Greenberg, 1989) or either the Couple or Family Therapy Alliance Scales (Pinsof & Catherall, 1986). In another study involving 43 families (81 participants), the therapy alliance was examined as it related to symptom distress. These families had been referred to therapy because they were at risk of having one or more of their children removed from the home due to abuse, neglect, or juvenile offenses. The therapists were all doctoral students doing cotherapy in in-home therapy settings. Family members completed outcome and therapy alliance questionnaires at the beginning and end of therapy. Analysis of the data revealed that the therapy alliance explained 19% of the variance in symptom distress changes for mothers, 55% for fathers, and 39% for adolescents (Johnson et al., 2002). Johnson et al. discussed a ?net? family therapeutic alliance with the therapist and the possibility that this net alliance might be more influential in outcome than the individual family members? alliance scores. 7 These findings coincide with the research of Symonds and Horvath in which they studied the relationship between alliance and outcome in therapy involving 47 couples. Alliance was measured using the Working Alliance Inventory after the first and third sessions. There were six therapists; all were either psychologists or clinical counselors who had already graduated with at least a Master?s Degree and worked in either a private practice or community mental health center. Results of the study indicated a significantly stronger correlation between alliance and outcome when the couple agreed about the strength of the alliance, when the males? alliance scores were higher than the females? and finally when the couples? alliance with the therapist strengthened during the course of therapy (Symonds & Horvath, 2004). In what is one of the only studies to note student-therapists? alliance scores cross- sectionally throughout a training program, Davenport and Ratliff (2001) found significant correlations between client-reported alliance ratings and number of clinical hours of experience in master?s level marriage and family therapy training. They noted that ?with increased experience, trainees became more consistently skilled at developing therapeutic alliance? (Davenport & Ratliff, 2001, p. 446). The sample in this study was a total of 51 clients and 14 student-therapists. The more clinical hours the trainees had acquired (meaning, the total client-contact hours at that point in their program and not hours spent with a particular client), the higher?on average?their alliance scores. In fact, a stepwise regression analysis revealed that the therapists? experience level was the only significant predictor of alliance scores. Of further import, research supports the idea that client- reported assessment of the therapy alliance is more predictive of treatment outcomes than 8 either the therapists? assessments or observers? reports; this finding strengthens Davenport?s study (Horvath & Symonds, 1991). Dropout Rate One would assume that a strong therapy alliance would be related to fewer dropouts, and that the inverse would also be true. In this sense, dropout rate could be considered a type of therapy alliance measure. Beckham (1992) did a study involving 93 patients at an outpatient mental health clinic and noted that patients were more likely to continue treatment if rapport was established early with the therapist. While rapport is not, by definition, the same thing as a therapeutic alliance, it is no doubt influential in establishing such an alliance. Probably one of the most researched explanations as to why clients dropout or continue treatment is that of the therapist?s experience level, often determined by the time he or she has been practicing therapy. However, some of the research on ?experience? refers specifically to a given therapy approach or intervention. Much research supports the idea that the more experienced the therapist is, the lower the dropout rate (Burlingame, Fuhriman, Paul, & Ogles, 1989; Pekarik & Stephenson, 1988; Stein & Lambert, 1995; Sue et al., 1976). Burlingame et. al. (1989) looked at training in time- limited treatment in their study involving 12 therapists and 57 clients. They found that clients working with more experienced therapists (as measured by time practicing therapy) had better outcome ratings and fewer dropouts than those working with less experienced therapists. Pekarik and Stephenson observed differences in adult and child dropouts in a study consisting of 330 participants and 35 therapists. While none of the variables they investigated significantly predicted children?s dropout rates, or 9 ?continuance,? through a multiple-regression analysis they found therapist experience to be one of the variables able to predict adult dropouts (Pekarik & Stephenson, 1988). Stein and Lambert?s (1995) review of the literature also indicated that therapists with more training and experience tend to have fewer dropouts and maintain clients in therapy for a longer period of time. Among the findings of the study done by Sue et al. (1976), which involved 13,450 clients, was the fact that paraprofessionals had higher dropouts than professionals. One of their conclusions was that additional training for paraprofessionals may be necessary. Atkins and Christensen reviewed the research literature on the effectiveness of paraprofessional counselors and the effect professional training has on client outcomes. They noted several studies suggesting that paraprofessionals were at least as effective as professionals if not more so. However, their thorough review pointed out that these studies had several methodological flaws that prohibited making definitive conclusions from the research. Overall, they found that while paraprofessional counselors were indeed helping people, professionals (more experienced or trained as a result of their degree) were better at client retention and obtained better overall client outcomes (Atkins & Christensen, 2001). Other studies yielded similar findings showing that client outcomes were better when treatment termination is mutually determined with the therapist rather than the client dropping out prematurely (Johanson et al., 1980) and that both therapist experience and the intensity of therapists? training were related to superior client outcomes (Burlingame, et al., 1989). Pekarik (1994) did a study that yielded similar findings. His is one of the only studies looking at the effect of a brief therapy training program. The study involved 22 10 practicing psychotherapists at three public clinics. Twenty-two psychotherapists attended a ten hour, all day brief therapy training program with a format similar to postgraduate training workshops. The therapists were randomly divided into either a training or control group. Pretraining assessments indicated no significant differences between the two groups or their clients (n=176). Three months after the day-long training, therapists were asked to select, over the next six months, between eight and twelve newly admitted clients to use in their study. Client measures of outcome (i.e., client satisfaction, client termination status, client termination reason, therapist termination status, and brief vs. standard therapy assignment) were given at the intake appointment and again at 10 weeks and five months after the intake. Pekarik found, among other things, that ?trained? therapists had fewer client-rated dropouts and greater treatment satisfaction among clients at both follow-ups, as well as better therapist ratings of outcomes. Bischoff and Sprenkle (1993) reviewed the literature on dropping out of marriage and family therapy. Like Pekarik, they found that therapists having fewer dropouts. They indicated in their review that therapist experience was one of only three significant results generated in the marriage and family therapy literature concerning therapist demographics and premature termination (the other two were sex and race of the therapist). The definition of therapist experience they pulled from the literature is unique in that they specifically say it is the ?number of family therapy training experiences that the therapist has,? and not ?the years of experience in family therapy? (Bischoff & Sprenkle, 1993, pp. 360). While they did not specifically mention what ?number of family therapy training experiences? meant, it seems likely that graduate training would 11 yield a high number of training experiences due to the intensity and frequency of supervision. A unique study done by White and Pollard (1982) looked at the relationship between the therapeutic effectiveness of therapist-trainees (rated by fellow trainees and supervisors) and no-show rates. The trainees consisted of 23 master?s and doctoral students. Results indicated a significant relationship between no-show rates and the peer- rated therapeutic effectiveness. This study suggests that not only do no-shows, likely to include dropouts, not only affect the potential clients? outcome, but could also have an influence on the therapist?s ability to do therapy effectively. White and Pollard suggest this is probably best explained by thinking of a no-show client as a missed opportunity to gain experience and practice. In this light, experience and practice are directly related to the number of sessions a student-therapist has with his or her clients. Number of Sessions Gaining experience and practice are goals of every student in a training program. The ability to keep one?s clients attending therapy sessions is a coveted skill. A common related question among students in training as well as experienced clinicians and researchers is, ?What is the optimal number of therapy sessions?? In a day when managed care is encouraging brief therapy mental health professionals are feeling pressure to find ways to do effective therapy in the least amount of time possible. There are also professionals who feel that keeping clients for an extended period of time is unnecessary, unproductive, and even unethical. Reviewing the literature, one quickly ascertains that there is no generally accepted measurement of extended period of time. Certainly, however, it could be agreed that there is some point at 12 which the benefits of therapy decrease. Regardless of whether this change means termination of the case, alteration of the mode of therapy, a referral to a different therapist, or some other change, recognizing this point in therapy can be difficult when the decision is influenced by varying opinions, laws, insurance companies, and research findings. There is, however, a body of research supporting the claim that, in general, the more treatment sessions a person has or the longer someone stays in therapy, the better? meaning, better client outcomes are reported. Hampson and Beavers (1996) examined various family and therapist characteristics as they relate to treatment outcomes. They used a fairly large sample of families (n=434) seen at a sliding-fee clinic. The therapists used in this study were graduate students from various universities and disciplines (including psychology, marriage and family, social work, and psychiatry) doing their internship and who had training in the Beavers Systems Model of family therapy, or were post-graduates working towards certification in family therapy. One of the interesting findings from this study was that ?there was a striking relationship between lower goal attainment and lower number of visits? (Hampson & Beavers, 1996, pp. 354). Families that came to only one session (63 out of 79 families) said they had not attained any of their goals. When these families that attended only one session were removed from the analysis, the overall improvement rate was 86.6%. The improvement rate was 93.8% when families that attended fewer than four sessions were removed. In one of the few studies conducted at a university-based graduate training clinic, Messer and Boals (1981) assessed the student-therapists? psychotherapy effectiveness as measured by both therapist and client reports of the overall level of helpfulness of therapy 13 and the fulfillment of the therapy goals. The study included individual, marital, child and family therapy. The therapists in training were in either the school clinical or counseling psychology program. The sample was fairly small (n=90). The key finding was that ?the longer the therapy, the better the outcome? (Messer & Boals, 1981, pp. 790). Length of therapy and type of termination (whether mutual or one-sided?mutual yielding more positive results) were both significantly related to client outcome. This study lends good support for the present study in that, in a university-based clinic using student-therapists, more sessions did, in fact, yield better results. Again from the psychology literature, in what has been called the ?flagship study? estimating how much psychotherapy is enough (Kopta, 2003, pp.727), Howard et al. introduced the dose-effect model (dose being the number of sessions and effect being the percentage of patients improved) in 1986. When their article was published, managed health care was beginning to apply pressure towards time-limited psychotherapy. The authors were making one of the first efforts at ?empirically estimating how much psychotherapy is enough? (Kopta, 2003). They first noted that the effect of therapy is greater at the beginning of therapy and increases more slowly as therapy continues. However, their study did not show any decline in effectiveness, rather a more shallow growth curve. They then did a meta-analysis of 15 data samples (n=2,431) that covered a period of over 30 years and used ?traditional measures? of client improvement. From this meta-analysis they noted that about 15% of patients improved before the first therapy session, about 50% of patients improved by the eighth therapy session, about 75% by session 26, and about 85% at 52 sessions (Howard, et al., 1986). They were then able to use these findings to predict the course of improvement of patients. This study has great 14 implications for graduate-level training programs and supports the idea of seeing clients for much longer than most brief-therapy training would suggest prudent, when client outcome is considered. The large percentage of patients exhibiting improvement at sessions 26 to 52 in the study of Howard et al. suggests it would be unlikely that student- therapists could see clients beyond an undefined threshold of therapeutic effectiveness while in a graduate training program. Kopta (1994) was the principal investigator in another study using the dosage model to examine how many ?doses? are needed for relief of various distressing symptoms. They gave symptom checklists to 854 psychotherapy outpatients at intake and during treatment. Patients were typically seen weekly. The therapists included 141 psychologists, psychiatrists, social workers, or students training in these various fields of theoretical orientations. Sixty-two symptoms (i.e., ?feeling blue,? ?trouble concentrating,? ?thoughts of death,? etc.) were put into three categories: acute distress symptoms, chronic distress symptoms, and characterological symptoms. Fifty percent of patients recovered around 2.5 months, or 11 sessions. Researchers found that 58 sessions, or a little more than a year of weekly therapy, produced recovery in 75% of their patients. Another point to consider is that perhaps by putting a time-limit on treatment (especially on less experienced therapists) we may defeat the ?cost effect? intent of the time limit by inadvertently contributing to higher rates of attrition and recidivism (Burlingame, et al., 1989). Burlingame, et al. (1989) studied, among other variables, the relationship between treatment success and the therapist?s experience. The study involved 57 clients at a university counseling center. Exclusion criteria included the following: severely depressed, acutely psychotic, borderline personality, anger as main affect, no 15 consistent sense of self-identity, or unrealistic expectations of therapy. The therapists were either therapists-in-training or senior staff with degrees in counseling psychology or social work. Burlingame, et al. found that clients of more experienced therapists reported greater improvement than those seen by less experienced therapists. While the study did not specifically define ?experience,? it appears that experience was determined by the amount of time the therapist had been seeing clients (the range was between 1 and 15 years). They also noted a trend showing that the more experienced therapists saw clients for more sessions. In conclusion, research on graduate training in general is sparse, and even more so when narrowed down to a specific field, such as marriage and family therapy. Because of the lack of research in this area, it is necessary to review the literature outside of graduate training in order to address the area of interest for this study?how clinical experience in graduate training is associated with therapeutic variables such as the therapy alliance, dropouts, and the duration of therapy. These three variables are of particular interest, as they are known in the research to be related to better client outcomes and can be easily measured in training programs. Ironically, there has been no study to date addressing whether or not the clinical experience afforded during graduate training leads to better client outcomes. A practical first step is to determine if student-therapists achieve higher alliance scores, experience fewer dropouts, are able to retain clients for more therapy sessions as they progress through their training program. Future research could look at the possible moderating relationship supervision plays as it relates to these factors and client outcomes. 16 Hypotheses 1. An increase in clinical experience (measured by time in the program and client- contact hours) will be associated with an increase in therapy alliance scores. 2. An increase in clinical experience will be associated with fewer dropouts. 3. An increase in clinical experience will be associated with more therapy sessions completed with clients. 17 III. METHODS Procedure The study utilized data from the Auburn University Marriage and Family Therapy Center (MFT Center) which serves residents of eastern Alabama and is located on campus. It is the training clinic for the COAMFTE accredited master?s program and is staffed by student-therapists. The clinic records were used to identify 24 student- therapists from the graduating classes of the years 2003 to 2006, as well as 420 client cases. Each of the cases included in this study were seen at the clinic between the years 2002 and 2006. These 420 cases included those for which the following necessary data were available from the client records: therapy alliance scores, the date of the first session, the assigned therapist, and the number of completed sessions. Adolescent data was not included in this study due to the limited number of adolescent cases seen at the clinic. Therefore, the included cases consisted of adults, both individuals and those in committed relationships, and parents in family therapy. Clients regularly filled out self-report questionnaires throughout treatment. Clients completed paperwork at their intake appointment and at the fourth session which included the therapy alliance measure used in the present study. It is important to note that there were several instances?75 cases (17.9%)?where fourth session paperwork was not collected from clients. This was due to therapist noncompliance with paperwork administration and clients? refusal to complete paperwork. 18 At intake clients were assigned to a specific therapist who would provide services. The number of therapists? accrued clinical hours and the number of months in the program at the time of the clients? intake session was the measure of therapist experience. However, on 18 therapy cases (4.3% of all the cases) a therapist graduated or left the program and a new therapist was assigned to the case. To provide continuity and in order to be able to analyze the data, it was determined that the three target factor variables would be paired with the initial therapist. In the 18 transfer cases the mean number of sessions before the transfer was 12.8, while the mean number of sessions after transfer was 4.7. There were 13 cases in which the total number of sessions before the transfer was greater than the number after the transfer. There were five cases in which a transfer occurred but the client never returned for clinical services. A chi-square test showed that the difference between the number of sessions before and after transfers was not significant. Participants. The vast majority of clients seen in therapy were Caucasians, followed by African Americans. Of the males, 174 (79.5%) were Caucasian and 35 (16%) were African American, while 253 (76.7%) of females were Caucasian and 56 (17.0%) were African American. Twenty-eight percent (66) of males ranged from 18 to 25 years old, 47.9% (112) ranged from 26 to 40 years old, and 23.9% (56) ranged from 41 to 60 years old. Thirty percent (103), 45.7% (154), and 21.9% (74) of females fell into the 18 to 25, 26 to 40, and 41 to 60 year old age ranges, respectively. A large majority of both men and women reported having at least a high school education. Almost 34% (79) of males and over 36% (120) of females reported high school as being the highest level of education attained. Over 20% (53 males and 67 19 females) of both males and females reported having attained a Bachelor?s Degree, over 10% (24 males and 46 females) of both men and women earned an Associate?s or other 2-year degree, and 6% (14) of men and 10% (34) of women reported having a Master?s Degree. Income level varied greatly, with a fairly even distribution for both males and females indicated in each income bracket on the intake assessment. The brackets start at ?under $5,000? and go up to ?over $40,000? annually. Over 20% (54 men and 69 women) of both men and women reported their annual income as more than $40,000. The next most common income bracket was ?under $5,000? with 11.7% (26) of men and 18.4% (58) of women so reporting. Because the Marriage and Family Therapy Center is on a university campus and offers a discounted therapy cost for students, many of these people reporting in the ?under $5,000? bracket may be students, and would generally be more accurately described as middle to upper class. Another possible explanation for the number of people reporting annual earnings of under $5,000 is that individual adults in a committed relationship may not have reported the income of their financially supportive partner. Of those reporting marital status, 38.5% (87) of men and 33.2% (112) of women reported being married, 34.1% (77) of men and 33.5% (113) of women reported being separated, 20% (45) of men and 25.5% (86) of women reported being in a committed, heterosexual relationship, and 4.9% (11) of men and 6.2% (21) of women reported being in a committed, homosexual relationship. The participants included in this study consisted of 234 males (41%) and 337 females (59%), for a total of 571 people. These make up the 420 included cases and represent various therapy types?individual, couple, or family. The different therapy 20 types determined the paperwork given to each client. It did not refer directly to the number of people involved in that case. For example, if a case was marked as a ?couple,? there were not necessarily two different people being seen in therapy, since many individuals come to therapy wanting to work on relational problems, and were therefore given the paperwork for an ?adult in a committed relationship.? Of the 420 cases, 120 (28.6%) were for individual therapy, 222 (52.4%) for couple therapy, and 78 (18.6%) for family therapy. It is important to note that a few clients did not answer every question on the given paperwork or assessments and could not be included in the sample. Also, there were instances in which a spouse or partner was not present in the first or fourth session and did not fill out the corresponding paperwork. Therefore, the total number of cases used for analyzing the three target factors sometimes differed. Therapists consisted of 24 students from four cohorts from the classes of 2003 to 2006. There were five males and nineteen females. The average age of the male student therapists at the beginning of the training program was 24.8 years old, and the average age of the female student therapists at the same time period was 24.7 years old. Twenty (83.3%) of the student therapists were White, two were Asian (8.3%), one was African American (4.2%), and one was Native American (4.2%). Research design. Therapist experience was grouped by the number of accrued clinical hours and by the number of months in the program. The number of accrued clinical hours was evaluated by both the total number of clinical hours as well as by dividing the number of hours into segments of 100 hours. The therapists? total clinical hours recorded at the Auburn University Marriage and Family Therapy Center and the 21 total clinical hours acquired at all other clinical sites was evaluated in relation to the three target factors. During their second year in the training program, student-therapists work at an internship site in addition to seeing clients at the university clinic. One therapist in every cohort is assigned to work primarily at the university clinic for their internship, while the others are assigned to work primarily in a community mental health agency. However, all therapists acquire client-contact hours at both sites. All the student-therapists kept track of their client-contact hours accrued at both the university clinic as well as their off- campus placements. The study?s three target factors were collected exclusively at the campus marriage and family therapy center. However, all student-therapists gained experience at their community mental health placements. Therefore, client-contact hours were collected separately for both placement types. The reason client-contact hours from the students? off-campus internships were included in this study was to see if these hours, and the added experience accompanying them, played a uniquely significant role in predicting therapy alliance, dropouts, and number of sessions. In addition to keeping a running total of monthly client-contact hours, clinical experience was also analyzed by the number of months in the training program. This time variable (the number of months in the program) was evaluated in two formats. First, the program was divided into three segments. Second, it was divided into five segments. Each academic year included three semesters, each of which were approximately four months in duration. The fall semester starts in August and ends in December, the spring semester starts in January and ends in April, and the summer semester starts in May and ends in July. The three segment approach for dividing clinical experience 22 included, out of a 24-month program, months 1 to 11 in one segment?more months due to the fact that clinical months did not start until month seven, months 12 to 17 in another, and months 18 to 24 in a third segment. The rationale behind this specific segmentation was as follows: the earliest month that any therapist in this study saw clients was January of his or her second semester, month six of the program. Month six to month 11 makes up six months of clinical experience. Also, in the summer the students work under a new supervisor, so this segment was a natural break. Segment two, month 12 to month 17, makes another six months of clinical experience and terminates at the end of the fall semester, in December. Segment three, months 18 through 24, make seven more months of clinical experience and ends when the students graduate. The last month in the program?July?is when the graduating student-therapists transfer their cases to the following cohort. Therefore, this last month?month 24?was limited in the number of hours accrued by the therapists. The five-segment approach was made up of the five semesters in which therapists saw clients, starting with the spring semester of their first year in the program. This segmentation was an attempt to assess the three target factors by semester. This was the most natural way to measure time in the program. By using time in the program or client-contact hours as the independent variable, a regression was implemented to predict therapy alliance scores, client dropouts, and the number of therapy sessions. We would thus be able to compare our predicted values with our actual values and have a better understanding of the role clinical experience plays in a graduate program and its impact on therapeutic variables known to be related to positive client outcomes. While this study focused on clinical training affecting therapy outcomes, 23 it was determined that client and therapist demographic variables would be correlated to the outcome variables to better understand the factors impacting client outcomes. These significant relationships were incorporated into a regression equation. Cross tabulations and chi-square tests were used in the analysis of the dropouts and number of sessions completed. This allowed for the comparison between observed and expected counts in each cell. These differences between observed and expected values are called residuals. Adjusted residuals are normalized estimates of the residuals? standard error. Adjusted residuals that were less than or equal to -2 or greater than or equal to +2 were reported and the level of significance was set at 0.05. Adjusted residuals that are well above +2 or below -2 indicate findings that are substantially greater than is probable by chance alone. Measures Therapy alliance. The therapy alliance was measured by client self-report questionnaires filled out at the fourth session using the Therapy Alliance Scale (Pinsof & Catherall, 1986). This scale has been shown to be a reliable way to measure the therapy alliance (Heatherington & Friedlander, 1990). Moreover, alliance quality as reported by clients (as opposed to therapist or observer report) has been shown to be the most predictive of treatment outcomes (Horvath & Symonds, 1991). This scale uses a Likert-type format with a 7-point balanced scale ranging from ?completely agree? (7) to ?completely disagree? (1) with a ?neutral? midpoint (4). Around half of the items on each scale are phrased positively (e.g. ?The therapist cares about me as a person.?) and half are phrased negatively (e.g. ?The therapist does not understand me.?). Both the balanced scale and the positive and negative items are used in 24 order to minimize the chances that test responses will be subject to experimental bias (Pinsof & Catherall, 1986). The Therapy Alliance Scale developed by Pinsof and Catherall is comprised of three subscales: bonds (n=10 items), tasks (n=13 items), and goals (n=6 items). Studies using the Therapy Alliance Scale have shown overall test-retest reliability with a range of .72 to .84, with a p-value of less than .005 (Pinsof & Catherall, 1986). Heatherington and Friedlander (1990) examined Pinsof and Catherall?s measures of the Therapy Alliance Scale with an even larger sample (N=66) than Pinsof?s original samples (where N was between 12 and 24) and found the internal consistency reliability alpha scores ranging from .93 to .94. The subscales of bonds, tasks, and goals were first suggested by Bordin (1979), and have since been used in the development of therapy alliance and working alliance scales. The bonds subscale has to do with the quality of the relationship between the therapist and the client and contains statements like, ?The therapist cares about me as a person.? The tasks subscale includes statements like, ?The therapist and I are in agreement about the way in which the therapy is being conducted,? and is concerned with methods and techniques of therapy and whether or not the client believes they are relevant to his or her difficulties. Pinsof and Catherall (1986) did a study in which they broadened the tasks subscale to include the client?s belief that the therapist has the power and skills to help him or her in therapy. The goals subscale examines whether or not the client and therapist agree on the goals of therapy and contains statements such as, ?The therapist and I are not in agreement about the goals of this therapy.? 25 In the current study, the therapy alliance was measured at the fourth session, or when fourth session paperwork was not available, the next available assessment, not including paperwork beyond the eighth session. This is consistent with previous research. Heatherington and Friedlander examined the therapy alliance scores from the third to sixth session and noted that this time frame for data collection was chosen because it has been found that, at least in individual treatment, ?alliance measures administered at this time were predictive of outcome? (Heatherington & Friedlander, 1990, p. 301). They went on to explain that, ?Sampling conducted earlier than the third session may be misleading. Conversely, delaying sampling much beyond this point is likely to result in skewed ratings? (1990, p. 301). Dropouts. Dropouts were defined as clients that attended only one therapy session. This is similar to the measure Allgood and Crane (1991) used in their study on predicting marital therapy dropouts. Duration of therapy. Duration of therapy refers to the number of sessions in which a client was seen at the Marriage and Family Therapy Center before termination, including instances when termination was mutually determined as well decided only by the client. 26 IV. RESULTS This study investigated the relationship between clinical experience and therapy alliance, client dropouts, and duration of therapy for student-therapists in a marriage and family therapy training program. It was hypothesized that with increasing clinical experience, student-therapists would experience an increase in therapy alliance scores and number of sessions completed, while demonstrating fewer client dropouts. The Therapy Alliance Scale (Pinsof & Catherall, 1986) was used to assess the therapy alliance at the fourth session. Dropouts were determined based on clients attending only one therapy session. Duration of therapy was calculated as the number of sessions clients attended before terminating therapy. Means and standard deviations of the three target factors are presented first. Second, the three target factors and demographic variables were examined and put into three categories: therapist variables, client variables, and clinical variables. Correlations between these three target factors and therapist, client, and clinical variables were examined. Third, a regression model is presented. Next, the effect of dividing clinical experience into different segments is discussed as it relates to this study?s three target factors. Finally, the three hypotheses presented in this study are analyzed. 27 Research Findings Means and standard deviations. The means and standard deviations of males? and females? fourth session therapy alliance scores as well as sessions completed are listed in Table 1. A standard t-test revealed no significant difference between males? and females? therapy alliance scores. Table 1 Summary of Means and Standard Deviations for Fourth Session Therapy Alliance Scores and Sessions Completed. N Mean Std. Dev. Male 4 th Alliance 104 213.15 37.84 Female 4 th Alliance 132 216.24 34.28 Of the 420 cases included in this study, 99 (23.6%) were counted as clinical dropouts, as the participants attended only one session. The average number of total client-contact hours associated with clients that dropped out of treatment and those that did not, was 167.20 (N=99) and 186.27 (N=321), respectively. Clients attending four sessions or less accounted for 54.8% (230) of the total number of completed sessions. Approximately 90% (377) of the total completed sessions attended 12 or fewer therapy sessions. The average number of sessions for all cases included in this study was 5.79. When broken into segments?grouping months together?the training program was better analyzed. The three-segment variation for measuring clinical experience was examined. A summary of means and standard deviations can be seen in Table 2 for males? and females? fourth session therapy alliance scores, dropouts, and sessions completed, using this segmented approach. 28 Table 2 Summary of Means and Standard Deviations for Male and Female Fourth Session Therapy Alliance Scores, Dropouts, and Completed Sessions, Separated Into Three Multi-Month Segments. Time Segment Male 4 th TA Female 4 th TA Dropout Completed Sessions 1 237.00 (38.30) 221.91 (17.58) .26 (.44) 5.16 (6.37) 2 214.02 (38.95) 219.17 (35.69) .26 (.44) 5.42 (6.05) 3 209.00 (36.16) 212.34 (35.10) .19 (.39) 6.46 (5.99) Total 213.15 (37.84) 216.24 (34.28) .23 (.42) 5.79 (6.06) In addition to being broken into multi-month segments, client-contact hours were divided into segments of 100 hours. Of the clinical experience variations segmented into client-contact hours, the 100-hour segments best showed the relationship between clinical experience and therapy alliance, dropouts, and completed sessions (See Table 3). Table 3 Summary of Means and Standard Deviations for Male and Female Fourth Session Therapy Alliance Scores, Dropouts, and Completed Sessions, Separated Into Five, 100-hour Segments. 100-hr. segment Male 4 th TA Female 4 th TA Dropout Completed Sessions 1 214.65 (39.16) 219.61 (32.60) .25 (.44) 5.34 (5.78) 2 223.71 (40.31) 222.00 (38.14) .26 (.44) 5.51 (6.58) 3 205.33 (47.74) 207.59 (30.94) .27 (.45) 6.16 (5.89) 4 210.64 (30.22) 211.71 (31.88) .10 (.31) 7.64 (6.21) 5 208.80 (39.55) 221.13 (50.75) .28 (.46) 4.89 (5.64) Total 213.15 (37.84) 216.24 (34.28) .24 (.43) 5.79 (6.06) Significant correlations. Preliminary analysis of the data showed that the distribution of the number of completed sessions was not representative of a linear relationship. Therefore, the number of completed sessions was transformed using the transformation Lg10(1 + sessions completed). The positive skew value was thus reduced 29 from 2.342 to .429. This transformation allowed for more accurate correlations to be made and for linear regression to be possible. Correlations were first examined between the three target factors and therapist variables such as the sex, age, and race of the therapist, as well as whether or not the therapist was the resident therapist assigned to the campus clinic. Table 4 shows these correlation values. Only one correlation from the therapist demographic information was significant?the therapists? age and dropout rate. Results indicated a negative relationship between age and dropout rate. The average age of student-therapists working with clients that did not drop out of treatment was 26.35 years old, while the average age of therapists working with clients that dropped out of therapy was 24.76 years old. Table 4 Correlation Coefficients for Therapy Alliance Scores and Variables of Interest. Variable 12 34 56 7 1. Male TA ? .47 ** .03 .07 .02 -.01 2. Female TA ? .03 -.05 -.14 -.13 3. Dropout Rate ? -.70 ** -.04 .04 -.10 * 4. # of Sessions ? -.03 -.04 .03 5. Resident Therapist ? .32 ** .42 ** 6. Sex of Therapist ? .07 7. Therapist Age ? * p < .05 ** p < .01 Male TA=Male Therapy Alliance Score at Session 4, Female TA=Female Therapy Alliance Score at Session 4. Next, correlations were examined between the three target factors and client variables such as male and female education levels, their age, and their race. Table 5 summarizes the correlations found between these variables. Three significant correlations 30 of interest were found. Two of the significant correlations were related to the male?s education level. The other correlation was related to the female?s age. Basically, male clients reporting higher levels of education, on average, attended more therapy sessions and were less likely to drop out of therapy. A ? 2 analysis revealed that males with either an Associate?s or Bachelor?s Degree had less than half the expected amount of dropouts, with a standard residual of -2.3. Males? education was divided into the following segments: (1) grade school, junior high school, and those that obtained a GED, (2) high school graduates, (3) vocational or technical school, Associate Degree or other two-year degree, and a Bachelor?s Degree, and (4) Master?s Degree or other higher education. In addition, older female clients seemed to be paired with men reporting lower fourth session therapy alliance scores. Table 6 summarizes the means of the males? therapy alliance scores with the correlated female partners? ages. 31 Table 5 Correlation Coefficients for Therapy Alliance Scores, Dropout Rate, Number of Sessions Completed, and Variables of Interest. Variable 1 2 3 4 5 6 7 8 9 10 1. Male TA ? 1.00 .47 ** .10 .13 -.25 * -.19 .11 -.02 2. Female TA ? .03 -.05 .08 -.14 -.18 -.11 -.21 3. Dropouts ? -.70 ** -.02 -.16 * -.03 -.02 .06 .07 4. # of Sessions ? .02 .22 ** .03 .05 -.07 -.03 5. Female Edu. ? .43 ** .10 .05 .06 .09 6. Male Edu. ? .07 .12 .14 .06 7. Female Age ? .82 ** .04 .01 8. Male Age ? .04 -.04 9. Female Race ? .54 ** 10. Male Race ? * p < .05 ** p < .01 Male TA=Male Therapy Alliance Score at Session 4, Female TA=Female Therapy Alliance Score at Session 4 Table 6 Summary of Means and Standard Deviations for Males? Fourth Session Therapy Alliance Scores and Female Partners? Ages. Age Range (in years) N Mean 18-22 13 242.00 23-26 16 209.12 27-30 11 225.82 31-34 9 202.78 35-38 12 216.17 39-65 11 202.55 Total 72 216.99 The third category of correlations involved the three target factors and clinical variables. These variables included variations of clinical experience. For instance, time in the program was broken down into various multi-month segments?three segments and 32 five segments. This form of measuring clinical experience seemed to have the best relationship to the variables of interest. Clinical experience was also divided into various groupings of client-contact hours. The grouping that best showed a correlation was client- contact hours broken into segments of 100. Total client contact hours were also examined, as were the subtotals of hours acquired at both the campus clinic as well as the off-campus community placements. Summaries of these various divisions of clinical experience and their correlations with the variables of interest are shown in Tables 7 and 8. The only significant correlation of meaning was a positive correlation between clinical experience?as measured using the three-segment approach?and the number of completed sessions, suggesting that, on average, the more clinical experience a student- therapist had, the more sessions were obtained with their clients (r = .11, p < .05). Indeed, the average number of sessions during each of the three time periods was 5.16 (N=38), 5.42 (N=224), and 6.46 (N=158), respectively. 33 Table 7 Correlation Coefficients for Therapy Alliance Scores and Variables of Interest. Varible 1 23456 7 1. Male TA ? .47 ** -.10 -.03 -.08 -.16 -.07 2. Female TA ? .00 -.12 -.05 -.11 -.09 3. OCPH ? .47 ** .90 ** .77 ** .87 ** 4. CCH ? .82 ** .70 ** .81 ** 5. Tot. Hrs. ? .86 ** .98 ** 6. 3 Seg. ? .84 ** 7. 100 hr. Seg ? * p < .05 ** p < .01 Male TA=Male Therapy Alliance Score at Session 4, Female TA=Female Therapy Alliance Score at Session 4, OCPH=Off Campus Placement Total Hours, CCH=Campus Clinic Total Hours, Tot. Hrs.=Total Client Contact Hours. Table 8 Correlation Coefficients for Dropout Rate, Number of Sessions Completed, and Variables of Interest. Varible 1 23456 7 1. Dropouts ? -.70 ** -.05 -.05 -.06 -.08 -.05 2. # of Sessions ? .09 .02 .07 .11 * .07 3. OCPH ? .47 ** .90 ** .77 ** .87 ** 4. CCH ? .82 ** .70 ** .81 ** 5. Tot. Hrs. ? .86 ** .98 ** 6. 3 Seg. ? .84 ** 7. 100 hr. Seg ? * p < .05 ** p < .01 OCPH=Off Campus Placement Total Hours, CCH=Campus Clinic Total Hours, Tot. Hrs.=Total Client Contact Hours. After these three categories of correlations were examined, there was one other significant correlation which was not originally hypothesized?the number of completed sessions was significantly correlated with males? highest level of education, r = .22, p 34 <.01. Males? education was grouped into the following categories: grade school through high school (including those receiving their GED), vocational/technical school and an Associate?s Degree or other two-year degree, Bachelor?s Degree, and finally a Master?s Degree or other specified education. This result indicated that, on average, male clients with higher levels of education stayed in therapy for more sessions. Regression models. The variables yielding significant correlations were used to predict the regression model of best fit. The correlated variables, when examined independently, accounted for more of the variance than a combination of the variables, thus indicating a high degree of covariance. Two variables?males? highest level of education, and time in the program (as measured by the three, multi-month segments)? were significant when entered into a regression. Table 9 summarizes the regression analysis of males? education level and time in the program as independent variables and the number of completed sessions as the dependent variable. This model accounted for 4.9% of the variance. A positive yet weak relationship exists between time in the program, male education, and the number of sessions completed. This means that, on average, the longer a student-therapist is in the program, the more sessions he or she will have with clients. It also means that males? level of education is predictive of the number of sessions completed?the more educated the male is, on average, the more sessions he attends. 35 Table 9 Summary of Hierarchical Regression Analysis for Variables Predicting Number of Sessions Completed. Variable B Std. Error ? t p Time in Program (3 seg.) .06 .03 .13 1.94 .05 Males? Highest Level of Edu. .05 .02 .17 2.62 .01 A second regression was performed, also using the highest level of males? education, but using a different measure of clinical experience. As was mentioned, this study also looked at clinical experience as number of client-contact hours. These hours were kept separately for both the campus clinic as well as student-therapists? community mental health placements. Correlations were made between the total number of client- contact hours as well as sub-totals from both clinical sites and the three target factors. A significant correlation was noted between the highest level of males? education and the number of client-contact hours acquired at the student-therapists? off-campus placement. Also, there was a correlation between these client-contact hours and the number of sessions completed, r = .09, p = .06. In contrast with the clinical hours accrued at the students? off-campus placements, the hours accrued at the campus clinic were not significantly correlated with either male education or number of sessions completed. This was also true for the summed total of clinical hours obtained from both clinical sites. These variables?highest level of males? education, number of sessions completed, and the number of client-contact hours obtained at the community mental health site?were entered into a regression and accounted for 5.3% of the variance. The regression analysis was significant, F (2, 230) = 6.48, p ? .01 (See Table 10). This regression indicates that client-contact hours accrued at the students? off-campus clinical site and the males? 36 highest level of education can be used to predict the number of sessions attended by clients. Another finding was that the total number of client-contact hours accrued at both clinical sites was not significantly related to the number of sessions completed. However, client-contact hours accrued at the student-therapists? community mental health placements were significantly correlated with the number of sessions completed. Table 10 Summary of Hierarchical Regression Analysis for Variables Predicting Number of Sessions Completed. Variable B Std. Error ? t p OCPH .00 .00 .14 2.10 .04 Males? Highest Level of Edu. .04 .01 .17 2.61 .01 OCPH=Off Campus Placement Total Hours Hypothesis 1?An increase in clinical experience (measured by time in the program and client-contact hours) will be associated with an increase in therapy alliance scores. An analysis of the data showed no significant correlations between clinical experience and therapy alliance scores. This was true regardless of how clinical experience was measured. There were also no significant differences associated with the clinical experience site. Minor fluctuation in alliance scores was noted throughout the students? training program. Hypothesis 2 and 3?An increase in clinical experience will be associated with fewer dropouts and more therapy sessions completed with clients. Regardless of how clinical experience was measured, the lowest dropout rate was achieved when therapists had accrued between 300 to 400 client-contact hours. However, the relationship between clinical experience and dropout rate was not statistically significant, ? 2 (4, N = 420) = 7.83, p > .05. Also, the residuals for therapists? first 299 37 client-contact hours were unremarkable (greater than or equal to -2 or less than or equal to +2). Nevertheless, when dropouts were considered after breaking clinical experience into segments of 100 client-contact hours, large adjusted residuals were observed representing the 300-400 client-contact hours mark. The number of observed dropouts during this time period was 7, with an expected count of 15.8, and an adjusted residual of -2.8. The number of observed cases in which dropout did not occur was 60, with an expected count of 51.2, and an adjusted residual of 2.8. These large residuals offered support for Hypothesis 2?more clinical experience is related to fewer client dropouts. With a larger sample, this relationship may have been found statistically significant. Completed sessions were also at a high point around the 400-hour mark, with the highest average number of sessions being approximately 7.6. Several measures of clinical experience were utilized in this analysis. When clinical experience was separated into three, multi-month segments, the average number of sessions went from 5.16 at segment one to 6.46 at segment three, showing a slight increase. This increase reached statistical significance, t = 1.94, p = .05. This lent initial support for Hypothesis 3. Further support of Hypothesis 3 was obtained with the same measure of clinical experience. As part of the analysis, the number of sessions completed was divided into five groups: those completing only 1 session?constituting this study?s dropouts (23.6% of all cases completed at the Marriage and Family Therapy Center), 2 to 3 sessions (22.6%), 4 to 7 sessions (28.8%), 8 to 11 sessions (12.8%), and 12 or more sessions (10.1%). These groups were then compared to therapists? clinical experience as measured in 100-hour segments, ? 2 (16, N = 420) = 31.17, p = .01. Several significant values were observed. First, therapists with 0 to 99 clinical hours had 12 observed cases in which 38 clients attended 12 or more sessions, with an expected count of 20.3 and an adjusted residual of -2.5. Second, therapists with 200 to 299 clinical hours had 11 observed cases compared to an expected 6 cases?with a residual of 2.4?in which clients attended 12 or more sessions. Third, therapists with between 300 and 399 clinical hours had 7 cases in which clients attended only one session, with an expected count of 15.8 cases and an adjusted residual of -2.8. This finding offered support for Hypothesis 2?there were significantly fewer dropouts than statistically expected. Finally, therapists with between 300 and 399 clinical hours had 19 observed cases compared to an expected 11.6 cases in which clients attended between 8 and 11 sessions. Similar results were found when measuring clinical experience by semester in the program. There were five semesters in which student-therapists saw clients. The number of sessions completed was again examined using a chi-square test for the five groupings previously mentioned, ? 2 (16, N = 420) = 28.80, p = .03. Large residuals (less than or equal to -2 or greater than or equal to +2) were observed which signified important findings. Several significant findings were noted, with the strongest support for Hypothesis 3 being associated with students? fourth and fifth clinical semesters. This was not a surprise, as these semesters closely correspond with the therapists? 300 to 400 client-contact hours mark. First, student-therapists in their second clinical semester had an observed 4 cases compared to the expected 9.8 cases?with a residual of -2.2?in which clients attended 12 or more sessions. Second, student-therapists in their fourth clinical semester experienced 22 cases in which clients attended only one session?dropped out of treatment?compared to the 29.9 expected cases, with a residual of -2.0. In fact, the 39 fourth clinical semester is the first semester in which there were fewer dropouts than statistically expected. This finding further supported Hypothesis 2. Third, while still in the fourth clinical semester, student-therapists had an observed 21 cases in which clients attended either 2 or 3 sessions compared to an expected 28.7 cases, with a residual of - 2.0. Fourth, again in the fourth clinical semester, student-therapists had 24 observed cases in which clients attended therapy for 12 or more sessions. This is far more than the expected 15.4 expected cases, with a residual of 2.8. Lastly, students in their last clinical semester had almost two times as many observed cases (13) than expected cases (7.0) in which clients attended 2 or 3 sessions, with a residual of 2.7. At the same time, students had less than half (3) of the expected cases (7.5) in which clients attended between 4 and 7 sessions, with a residual of -2.0. Students in their last clinical semester had less than the expected count for all cases attending four or more sessions. This significant finding was of interest because it represents a different trend than the previous findings. Therapists had fewer sessions with clients in their last semester than in the previous semester. This finding was consistent with the findings using other measures of clinical experience? regardless of how clinical experience is measured, there is a slight decrease in the average number of sessions completed, as well as in the average therapy alliance scores, immediately prior to the students? graduation. 40 V. DISCUSSION Previous research supports the idea that, with increased clinical experience, therapists are better able to achieve higher therapy alliance scores, experience fewer dropouts, and have more sessions with their clients. These factors, which have been referred to in this study as the three target factors, have all been shown in previous research to be related to more positive client outcomes. Up to this point, there has been very little research on therapists-in-training, and no research specifically attempting to measure student-therapists? development during the course of their training program. This study was an attempted first step in this direction, measuring the relationship between student-therapists? clinical experience and these three target factors. This study succeeded in finding statistically significant results for one of the three hypotheses?Hypothesis 3?supporting the idea that with increased clinical experience, student-therapists experienced more sessions with their clients. Support was also noted for Hypothesis 2; however, in this sample the relationship was not statistically significant. Dropouts appeared to decrease and the number of sessions increased as student-therapists advanced in the training program. Student-therapists in their fourth clinical semester, or around 400 client-contact hours, achieved the most optimal results found in this study. Clients often express, particularly at the beginning of therapy, the awkwardness of attending therapy, or their lack of initial comfort in therapeutic settings. These findings suggest that more experienced student-therapists are better able to relieve these tensions, 41 and assure that clients continue attending therapy despite their initial feelings of discomfort. The fact that student-therapists were better able to assure clients? continuation in therapy as they progressed through their training program suggests that they also became more skilled at developing therapeutic relationships with their clients. The Therapy Alliance Scale is one tool that has been used to objectify this relationship. Surprisingly, in this study there were no significant results observed concerning the relationship between the therapy alliance scores as reported by clients, and clinical experience. However, the strength of the findings supporting the other two hypotheses concerning dropouts and number of completed sessions suggests that therapists were indeed forming stronger therapeutic relationships with their clients. Common sense suggests that without a good therapeutic relationship, clients would not continue therapy. Most marriage and family therapy cases are characterized by periods of strain, or ruptures in the therapeutic relationship. The ability of the therapist to overcome and address these therapeutic impasses would likely ensure client continuation in therapy and reduce the number of dropouts. While this ability was not directly examined in this study, its development can be inferred by the increasing ability of therapist-trainees to not experience client dropouts. The use of the Therapy Alliance Scale in this study or similar future studies needs to be considered. Perhaps it is not the best tool for measuring student-therapists? relationships with their clients. There is no other study in the human sciences literature where the Therapy Alliance Scale was used with therapists-in-training. However, to date there are no measures for the therapy alliance that have been developed specifically for training programs. 42 All three of the target factors were thought to be dependent on therapists? clinical experience. Clinical experience was examined in various ways including amount of time in the program (number of months), and number of accrued client-contact hours. Both of these measures for clinical experience were analyzed in different segments. Various multi-month divisions of the training program were grouped and tested for correlations with the dependent variables. This was also done with various groupings of client-contact hours. Client-contact hours and time in the program were kept for both clients seen at the Marriage and Family Therapy Center as well as the students? community mental health internship sites. It was thought that since most of the student-therapists accrued the bulk of their contact hours outside of the university clinic it would be important to examine the impact of each therapy site individually as well as summed together in relation to the three target factors. The clinical hours accrued at the off-campus placements were significantly related to the number of sessions completed, but not significantly related to either therapy alliance scores or number of dropouts. Clinical hours accrued at the Marriage and Family Therapy Center were not significantly correlated with any of the three target factors. Hypothesis 1?An increase in clinical experience (measured by time in the program and client-contact hours) will be associated with an increase in therapy alliance scores. The first hypothesis regarding increased clinical experience being associated with more favorable therapy alliance scores was not supported in this study. The Therapy Alliance Scale had been used at the Marriage and Family Therapy Center for several years to assess the therapeutic relationship because of its reliability and simplicity. The 43 validity of the TAS has been supported in previous studies, some of which were mentioned in the review of literature. However, because the TAS was the only measure used in this study assessing the client-therapist relationship, it is possible that this study failed to capture the relationship between clinical experience and the therapy alliance. It is also possible that, in fact, there is no relationship between clinical experience and therapeutic relationships. It may be that students? focus is on learning the various therapy models or approaches that are presented during their training program instead of on developing and maintaining a relationship with their clients. This seems unlikely though, since virtually every therapy model emphasizes the necessity of creating such a positive relationship if therapeutic gains are to be made. Another explanation for the lack of significant findings supporting Hypothesis 1 stems from a lack of variability in therapy alliance scores. As was mentioned, the Therapy Alliance Scale is made up of 40 questions using a Likert-type format, ranging from 1 to 7 (7 being most favorable). Data analysis revealed that clients? average score on each question was a 5, which is higher than average. Because the majority of clients reported higher than average therapy alliance scores, there was very little variance in these scores. This lack of variance did not allow for generalizations to be made regarding therapist development concerning therapy alliance. Reasons why clients seen at the Marriage and Family Therapy Training Center rated the therapy alliance as higher than average throughout the program need to be considered. One reason may be that clients are aware that the student-therapists are being supervised and may feel that indicating lower scores may have negative academic repercussions on the student-therapist. Another 44 reason may be that clients that would be likely to report lower therapy alliance scores are those that drop out of treatment and never actually report these lower scores. The findings related to Hypotheses 2 and 3 provide circumstantial evidence that indeed therapists improve in their ability to establish positive relationships with their clients. Perhaps by using other methods to measure the therapeutic relationship significant findings would have been obtained. The development and use of such methods could be the subject of a future research study. Hypothesis 2?An increase in clinical experience will be associated with fewer dropouts. Hypothesis 2 was supported in that more clinical experience?measured by either the total number of months in the program or total client-contact hours?was associated with fewer client dropouts. Large residuals indicated a vast difference between the observed and expected number of dropouts at around 300 to 400 client-contact hours. The p-value was greater than 0.05, and therefore not statistically significant. However, it is believed that with a larger sample, this finding may have yielded statistical significance. This finding might suggest the importance of graduate programs allowing their students to begin seeing clients as early as possible, and thus gain additional experience by the end of their program. A study comparing programs that have students see clients within the first couple semesters with those that start seeing clients later on in their program would further test this assumption. Clients? motives for dropping out were not assessed in this study. There is currently no method in place at the university clinic for measuring dropout motive or rationale. If such a method were utilized, the reasons clients were dropping out could be made a focus of training, and perhaps reduce the number of dropouts. 45 Hypothesis 3?An increase in clinical experience will be associated with more therapy sessions completed with clients. Hypothesis 3 was the only statistically supported hypothesis in this study. Results offered strong support that student-therapists experienced a significant increase in the number of sessions had with their clients as they gained experience throughout the program. Chi-square tests and cross tabulations were used to analyze the relationship. These analyses allowed for the comparison of the expected number of sessions completed with the observed number of sessions completed. Adjusted residuals?defined as the standardized difference between the expected and observed count for the variables of interest?were considered large if they were less than or equal to -2 or greater than or equal to +2. Clinical experience measured both by time in the program as well as by the number of client-contact hours yielded significant findings. Time in the program was broken into five clinical semesters. These semesters began with the spring of the students? first year in the program and ended with the students last semester, which was the summer semester of their second year. Analysis of the data revealed that as the students advanced through the program instances where the expected number of sessions completed was significantly different than the observed number of sessions completed were more common. For instance, during the first clinical semester, small adjusted residuals were observed, with none being less than or equal to -1 or greater than or equal to +1. However, during the students? second clinical semester, all but one (representing cases where the clients attended only one therapy session) of the adjusted residuals were less than or equal to -1 or greater than or equal to +1. For cases in 46 which clients attended 12 or more therapy sessions, the adjusted residual was -2.2, indicating less than half of the expected number. During the first three clinical semesters the number of cases seen for one to three therapy sessions was greater than the expected count. While the residuals associated with these semesters were not significantly large, they nevertheless help to illustrate the trend toward improvement when contrasted with the fourth clinical semester. The fourth clinical semester yielded very promising results. There were significantly fewer cases than expected that were seen for only one to three therapy sessions, with residuals of -2.0. There were more cases in which clients attended at least four sessions than were expected, with cases completing 12 or more sessions yielding an adjusted residual of 2.8. This means that, on average, therapists were seeing their clients for almost 9 more sessions than expected at that time period. The results found in this fourth clinical semester most strongly support Hypothesis 3. The last clinical semester showed an interesting trend. For cases seen in therapy for two to three sessions, the observed count was almost twice the expected count, with an adjusted residual of 2.7. This indicated that student-therapists were experiencing significantly more two- to three-session cases than expected. In contrast, cases in which clients attended four to seven therapy sessions were significantly fewer than expected? less than half?with an adjusted residual of -2.0. Also, while not statistically significant, this trend of having fewer observed sessions completed than expected continued for cases in which clients attended eight therapy sessions or more. This increase in cases seen between one and three sessions, and the decrease in cases seen for at least four sessions is opposite that observed just the semester before. Two branches of possible explanations 47 seem most probable?either the therapists are able to help clients resolve their issues in fewer sessions or therapists are in a hurry to terminate with their clients. The latter possibility seems likely, as there is much going on at the end of the students? program, such as graduation and finishing theses. Perhaps clients can sense the therapists? preoccupation and resist coming back to therapy. Another possible reason that therapists may be terminating with clients in fewer sessions is reluctance on the part of either the therapist or the clients about transferring to a new therapist. This finding would be beneficial for clinical supervisors working with students during the last clinical semester?an emphasis might be placed on therapy continuation and care taken so as not to terminate with clients prematurely. A measure of client satisfaction with therapy would help ascertain if clients are satisfied with the therapy process during this last semester. In addition to the results found by measuring clinical experience as time in the program, measuring clinical experience as client-contact hours yielded similarly significant results. These results offered strong support that as student-therapists accrued client-contact hours, they experienced significantly more therapy sessions with their clients. Client-contact hours were separated into five segments of 100 hours each. Chi- square analyses and cross tabulations showed similar trends as when measuring time in the program. Adjusted residuals were again examined in the same manner as previously described. Particularly within the first 100 client-contact hours significantly fewer sessions than statistically expected (with an adjusted residual of -2.5) were noted for cases in which clients attended at least 12 therapy sessions. During the second 100-hour segment, only cases in which clients attended 8 to 11 sessions had an adjusted residual (-1.8) that 48 approached significance. These first 200 client-contact hours indicate that student- therapists have some difficulty keeping clients in therapy for longer than eight sessions. Also, during the first 300 client-contact hours, there were slightly more one- to three- session-only cases, although this number was not significant. When therapists had acquired between 200 and 299 client-contact hours, they experienced almost twice the expected number of cases in which clients attended at least 12 sessions, with an adjusted residual of 2.4. Between 300 and 399 client-contact hours, therapists experienced similar results as those found during the therapists? fourth clinical semester. Less than half of the expected instances of cases in which clients attended therapy for only one session were observed, with an adjusted residual of -2.8. During the same period there were significantly more observed than expected cases in which clients attended between 8 and 11 therapy sessions. It appears that therapists during this period are better at engaging and developing a therapeutic relationship with their clients such that these clients continue on for treatment for more sessions than when therapists had less clinical experience. One significant unexpected finding resulted from this study. While examining the clients? demographic information, a strong significant relationship between the number of sessions completed and the males? highest level of education was observed. On average, the higher the males? level of education, the more therapy sessions were completed. This was true regardless of the therapy type, meaning, for individual male clients as well as couples. Speculations about this finding can be made. Perhaps males with more education have more confidence in careers requiring further education, such as professional counseling or family therapy. They may trust in professional help more than someone 49 that does not share the same confidence in education. This may be especially true when coming to a training clinic on a university campus. The educational environment may be more comfortable to someone that experienced more education themselves. When the relationship was tested for female clients? highest level of education, the result was insignificant. Why this relationship was not shared by female clients remains unclear. Minor fluctuations were apparent in each of the three target factors? relationship with clinical experience. These may be associated with the academic calendar. The nature of the training program is that each semester brings a new emphasis in training, a different therapy model is taught and encouraged to be practiced. Student-therapists that were becoming comfortable and more skilled at using one therapy model or approach are then encouraged to try new models and approaches. Perhaps this explains the roller- coaster effect of therapy alliance scores, dropout rates, and number of sessions completed. Each of the target factors experienced a drop in the last section or last 100 client- contact hours in the program. During this period, regardless of how clinical experience was measured, a decrease in therapy alliance scores and number of sessions completed, and an increase in dropouts was observed. There are various possible explanations. One probable explanation is that at the end of the program students are anticipating graduation, experiencing high stress and anxiety about meeting deadlines and working on theses, and therefore may pay less attention and put less effort into their clinical work. Another explanation is that student-therapists may feel an urge to reach termination with their clients rather than having to transfer them to another therapist. 50 Limitations The single greatest limitation of this study was the sample size. Because there were several different factors that should have been controlled for (many more than were able to be taken into account in this study?i.e., the supervisor, the therapist, the therapy model being practiced, etc.), a much larger sample would be needed in order to produce the statistical power necessary to yield significant findings should those relationships exist. With a larger sample of both therapists and client cases, it would be possible to track individual therapists? progress throughout the program with each client rather than use averages from all therapists and clients together. This would be a more highly accurate way of following therapists? progress. It would also provide the field with much needed longitudinal data, as almost all data thus far is cross-sectional. Another limitation was that in the case of many of the clients seen at the Marriage and Family Therapy Center in the last four years, therapists did not give clients the required fourth session paperwork. This resulted in several files not having the needed data required for inclusion in this study. Specifically, the therapy alliance scale was missing in several files. In others, therapy alliance scales were only partially completed, and were therefore not able to be used in this study. In order for a study like this to produce robust findings, precision in assessment and paperwork administration is requisite. Lastly, the students? off-campus placements did not use the same assessments, and records were not kept concerning dropouts and sessions completed. It was therefore impossible to adequately compare the off-campus placements and the experience gained there with that gained at the campus clinic. Most therapists acquired the majority of their 51 client-contact hours at their off-campus placements. Had the Therapy Alliance Scale been used at the community mental health centers, perhaps a significant relationship would have been noted. However, this is unlikely since all of the therapists were also concurrently seeing clients at the campus clinic throughout the program, and these alliance scores were being assessed. Future Research There remains a great need to assess the effectiveness of graduate training in counseling fields. The possibility exists that at some time in the future, licensing agencies or accrediting bodies will require training programs to demonstrate effectiveness in producing competent therapists. Longitudinal research is much needed in order to track therapist development or progressive competency during graduate training. With a larger sample of therapists it would be possible to track individual therapists against themselves throughout the course of their training. Because of the small number of therapists included in this study, it was necessary to pool the data from all therapists and look at them as a group. Further defining clinical experience and experimenting with ways to analyze it could be an important future study. For instance, student-therapists in accredited marriage and family therapy programs require 500 hours of direct client-contact hours for students to graduate. It is possible to include up to 100 hours of ?team? hours, which is where one student-therapist observes another student-therapist from behind a one-way mirror. Future research could investigate if there is a significant different between experience gained by personal, face-to-face client contact versus working as a team member. 52 Duplicating the kind of research attempted in this study, but with a larger sample and an added client outcome measure would help ascertain program effectiveness. Also, it would be beneficial for multiple graduate programs to either develop their own research design for a similar study, or for them to work together and pool their data in an effort to create a larger sample. Lastly, other ways of measuring student-therapists? development or improved competence as therapists during graduate training are needed. While this study failed to provide many statistically significant results, it did serve as a pioneering attempt to answer the question regarding the necessity and efficiency of graduate training in improving student-therapists? competency and effectiveness. This was done by quantitatively measuring student-therapists? alliance scores, client dropout rate, and average number of completed sessions as they related to the therapists? clinical experience. The need exists to measure student-therapists longitudinally throughout their training program as opposed to measuring the averages of a group of trainees. Results from this study were inconclusive and suggest the need for further investigation with a larger sample of both clients and therapists. 53 REFERENCES Allgood, S.M., & Crane, R.D. (1991). Predicting marital therapy dropouts. Journal of Marital and Family Therapy, 17, 73-79. Atkins, D.C., & Christensen, A. (2001). Is professional training worth the bother? A review of the impact of psychotherapy training on client outcome. Australian Psychologist, 36, 122-131. Beckham, E.E. (1992). Predicting patient dropout in psychotherapy. Psychotherapy: Theory, Research, Practice, Training, 29, 177-182. Bischoff, E.S., & Sprenkle, D.H. (1993). Dropping out of marriage and family therapy: A critical review of research. Family Process, 32, 35-42. Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, and Practice, 16, 252-260. Burlingame, G.M., Fuhriman, A., Paul, S., & Ogles, B.M. (1989). Implementing a time- limited therapy program: Differential effects of training and experience. Psychotherapy: Theory, Research, Practice, Training, 26, 303-313. Davenport, B.R., & Ratliff, D. (2001). Alliance ratings as a part of trainee evaluations within family therapy training. Contemporary Family Therapy, 23, 441-454. Davis, H., & Dhillion, A.M. (1989). Prediction of early attrition from couple therapy. Psychological Reports, 65, 899-902. 54 Dodd, J. (1970). A retrospective analysis of variables related to duration of treatment in a university psychiatric clinic. The Journal of Nervous and Mental Disease, 151, 75-84. Ford, J. (1977). Training in clinical psychology: A reappraisal based on recent empirical evidence. Clinical Psychologist, 30, 14-16. Greenspan, M., & Kulish, N.M. (1985). Factors in premature termination in long term psychotherapy. Psychotherapy: Theory, Research, Practice, Training, 22, 75-82. Hampson, R.B., & Beavers, W.R. (1996). Measuring family therapy outcome in a clinical setting: Families that do better or do worse in therapy. Family Process, 35, 347- 361. Heatherington, L., & Friedlander, M.L. (1990) Couple and Family Therapy Alliance Scales: Empirical considerations. Journal of Marital and Family Therapy, 16, 299-306. Horvath, A.O., & Greenberg, L.S. (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology, 36, 223-233. Horvath, A.O., & Symonds, B.D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149. Howard, K.I., Kopta, S.M., Krause, M.S., & Orlinsky, D.E. (1986). The dose-effect relationship in psychotherapy. American Psychologist, 41, 159-164. Johanson, S.L., Silverberg, R., & Lilly, R.S. (1980). Therapy improvement in two out- patient mental health clinics. Journal of Community Psychology, 8, 34-44. 55 Johnson, L.N., Wright, D.W., & Ketring, S.A. (2002). The therapeutic alliance in home- based family therapy: Is it predictive of outcome? Journal of Marital and Family Therapy, 28, 93-102. Kopta, S.M., Howard, K.I., Lowry, J.L., & Beutler, L.E. (1994). Patterns of symptomatic recovery in psychotherapy. Journal of Consulting and Clinical Psychology, 62, 1009-1016. Kopta, S.M. (2003). The dose-effect relationship in psychotherapy: A defining achievement for Dr. Kenneth Howard. Journal of Clinical Psychology, 59, 727- 733. Messer, S.B., & Boals, G.F. (1981). Psychotherapy outcome in a university-based psychology training clinic. Professional Psychology: Research and Practice, 12, 785-793. Pekarik, G. (1994). Effects of brief therapy training on practicing psychotherapists and their clients. Community Mental Health Journal, 30, 135-144. Pekarik, G., & Stephenson, L.A. (1988). Adult and child client differences in therapy dropout research. Journal of Clinical Child Psychology, 17, 316-321. Pinsof, W.M., & Catherall, D.R. (1986). The integrative psychotherapy alliance: Family, couple, and individual therapy scales. Journal of Marital and Family Therapy, 12, 137-151. Propst, A., Paris, J., & Rosberger, Z. (1994). Do therapist experience, diagnosis, and functional level predict outcome in short term psychotherapy? Canadian Journal of Psychiatry, 39, 168-176. 56 Stein, D.M., & Lambert, M.J. (1995). Graduate training in psychotherapy: Are therapy outcomes enhanced? Journal of Consulting and Clinical Psychology, 63, 182-196. Sue, S., McKinney, H.L., & Allen, D.B. (1976). Predictors of the duration of therapy for clients in the community mental health system. Community Mental Health Journal, 12, 365-375. Symonds, D., & Horvath, A.O. (2004). Optimizing the alliance in couple therapy. Family Process, 43, 443-455. White, G.D., & Pollard, J. (1982). Assessing therapeutic competence from therapy session attendance. Professional Psychology: Research and Practice, 13, 628- 633. 57 APPENDIX Therapy Alliance Scale Instructions: The following statements refer to your feelings and thoughts about your therapist and your therapy right NOW. Please work quickly. We are interested in your FIRST impressions. Your ratings are CONFIDENTIAL. They will not be shown to your therapist or other family members and will only be used for research purposes. Although some of the statements appear to be similar or identical, each statement is unique. PLEASE BE SURE TO RATE EACH STATEMENT. Each statement is followed by a seven-point scale. Please rate the extent to which you agree or disagree with each statement AT THIS TIME. If you completely agree with the statement, circle number 7. If you completely disagree with the statement, circle number 1. Use the numbers in-between to describe variations between the extremes. Completely Agree Strongly Agree Agree Neutral Disagree Strongly Disagree Completely Disagree 7 6 5 4 3 2 1 1. The therapist cares about me as a person 7 6 5 4 3 2 1 2. The therapist and I are not in agreement about the goals for this therapy. 7 6 5 4 3 2 1 3. My partner and I help each other in this therapy. 7 6 5 4 3 2 1 4. My partner and I do not feel the same ways about what we want to get out of this therapy. 7 6 5 4 3 2 1 5. I trust the therapist. 7 6 5 4 3 2 1 6. The therapist lacks the skills and ability to help my partner and myself with our relationship. 7 6 5 4 3 2 1 7. My partner feels accepted by the therapist. 7 6 5 4 3 2 1 8. The therapist does not understand the relationship between my partner and myself. 7 6 5 4 3 2 1 9. The therapist understands my goals in therapy. 7 6 5 4 3 2 1 10. The therapist and my partner are not in agreement about the about the goals for this therapy. 7 6 5 4 3 2 1 11. My partner cares about the therapist as a person. 7 6 5 4 3 2 1 12. My partner and I do not feel safe with each other in this therapy. 7 6 5 4 3 2 1 13. My partner and I understand each other?s goals for this therapy. 7 6 5 4 3 2 1 14. The therapist does not understand the goals that my partner and I have for ourselves in this therapy. 7 6 5 4 3 2 1 58 15. My partner and the therapists are in agreement about the way the therapy is being conducted. 7 6 5 4 3 2 1 16. The therapist does not understand me. 7 6 5 4 3 2 1 17. The therapist is helping my partner and me with our relationship. 7 6 5 4 3 2 1 18. I am not satisfied with the therapy. 7 6 5 4 3 2 1 19. My partner and I understand what each of us is doing in this therapy. 7 6 5 4 3 2 1 20. My partner and I do not accept each other in this therapy. 7 6 5 4 3 2 1 21. The therapist understands my partner?s goals for this therapy. 7 6 5 4 3 2 1 22. I do not feel accepted by the therapist. 7 6 5 4 3 2 1 23. The therapist and I are in agreement about the way the therapy is being conducted. 7 6 5 4 3 2 1 24. The therapist is not helping me. 7 6 5 4 3 2 1 25. The therapist is in agreement with the goals that my partner and I have for ourselves as a couple in this therapy. 7 6 5 4 3 2 1 26. The therapist does not care about my partner as a person. 7 6 5 4 3 2 1 27. My partner and I are in agreement with each other about the goals of this therapy. 7 6 5 4 3 2 1 28. My partner and I are not in agreement about the things that each of us needs to do in this therapy. 7 6 5 4 3 2 1 29. The therapist has the skills and ability to help me. 7 6 5 4 3 2 1 30. The therapist is not helping my partner. 7 6 5 4 3 2 1 31. My partner is satisfied with the therapy. 7 6 5 4 3 2 1 32. I do not care about the therapist as a person. 7 6 5 4 3 2 1 33. The therapist has the skills and ability to help my partner. 7 6 5 4 3 2 1 34. My partner and I are not pleased with the things that each of us does in this therapy. 7 6 5 4 3 2 1 35. My partner and I trust each other in this therapy. 7 6 5 4 3 2 1 36. My partner and I distrust the therapist. 7 6 5 4 3 2 1 37. The therapist cares about the relationship between my partner and myself. 7 6 5 4 3 2 1 38. The therapist does not understand my partner. 7 6 5 4 3 2 1 39. My partner and I care about each other in this therapy. 7 6 5 4 3 2 1 40. The therapist does not appreciate how important my relationship between my partner and myself is to me. 7 6 5 4 3 2 1