The relationship between working alliance and rehabilitation outcomes

by Daniel C. Lustig; David R. Strauser; N. Dewaine Rice; Tom F. Rucker

The relationship between working alliance and rehabilitation outcomes. Daniel C. Lustig; David R. Strauser; N. Dewaine Rice; Tom F. Rucker. Rehabilitation Counseling Bulletin, Fall 2002 v46 i1 p25(9) Full Text: COPYRIGHT 2002 Pro-Ed

A key factor in the development of an effective relationship between the client and counselor is the development of a strong working alliance (Bordin, 1979). Although considerable research has investigated the impact of the working alliance on counseling outcomes, no research has considered the effect of this alliance between rehabilitation counselors and clients within the context of the state--federal rehabilitation system. This study used existing survey data on 2,732 vocational rehabilitation clients during fiscal year 1999-2000. Results indicated that (a) employed clients had a stronger working alliance than unemployed clients, and (b) the working alliance is related to the client's perception of future employment prospects and his or her satisfaction with the current job. Implications for rehabilitation counselors are discussed.

The goal of the vocational rehabilitation system is to "empower individuals with disabilities to maximize employment" (1992 Amendments to the Rehabilitation Act). As Bolton, Bellini, and Brookings (2000) suggested, an important focus of research has been to determine variables that influence successful employment outcomes. Although a variety of factors have been considered (e.g., demographic variables, services provided, functional limitations), one factor that has not been considered is the effect of the client--counselor relationship. The 1998 Amendments to the Rehabilitation Act stated that individuals with disabilities served in the state--federal rehabilitation system must be "active and full partners" in the vocational rehabilitation process. Client involvement in this process has been viewed as important for increasing the likelihood of successful employment outcomes (Chan, Shaw, McMahon, Koch, & Strauser, 1997). A key factor helping the client become an active participant in this process is the development of a working alliance between the client and his or her rehabilitation counselor.

The construct of the working alliance was defined by Bordin (1979) as a collaboration between the client and the counselor based on the development of an attachment bond as well as a shared commitment to the goals and tasks of counseling. The working alliance is viewed as a collaborative effort in which the counselor and the client make equal contributions to the counseling relationship. It is thought that the working alliance makes it possible for the client to accept and follow through in the counseling process based on a sense of ownership (Horvath & Symonds, 1991). The working alliance is conducive to active participation between clients and counselors in the rehabilitation process.

Bordin (1979) theorized that this working alliance is the key to change in the client and its development is dependent on the level of collaboration between the client and the counselor. In counseling, problems associated with developing such an alliance are characteristic of the manner in which the client functions outside of counseling. The growth of a strong working alliance thus assists the client in overcoming self-defeating thoughts and behaviors outside of counseling as well.

Bordin (1979) conceptualized the working alliance as consisting of three interdependent components: goals, tasks, and bonds. Goals can be defined as the targets for interventions, and many times they are viewed as outcomes in the counseling process (Chan et al., 1997). The key, in terms of the working alliance, is the level of agreement or mutuality between client and counselor regarding counseling change goals. The process of reaching a mutually agreed change goal also assists in the development of counselor--client bonds. During the process of defining the counseling goals, the client begins to get a sense of the counselor's commitment to helping him or her and whether the counselor views the client as an equal participant in a collaborative counseling relationship (Bordin, 1994). Bordin also believed that the client's understanding of the change goal is therapeutic, sometimes providing him or her with the motivation to begin to change.

Tasks are the behaviors and cognitions engaged in by both the counselor and client while in counseling. For most counselors, the specification of the change goal to some extent prescribes the counseling tasks (Bordin, 1979). The relevance between the change goal and counseling tasks must be evident. Bordin stated that "the effectiveness of [the] tasks ... depends upon the vividness with which the therapist can link the assigned task to the patient's sense of difficulties and his wish to change" (p. 254). In a well-functioning counseling relationship, both participants perceive these tasks as relevant and efficacious. Although the relative responsibility of the performance of counseling tasks varies from one counseling approach to another, both the counselor and the client must accept some level of responsibility to perform these tasks.

The idea of bonds is concerned with the level of "partner compatibility" (Bordin, 1994, p. 16) between the counselor and the client. Bonding develops from the interaction between counselor and client in a shared activity. This bond can be expressed as liking, trusting, or a feeling of common purpose and understanding between counselor and client (Bordin, 1994; Horvath & Greenberg, 1989).

According to Bordin (1994), the working alliance is not a specific intervention but rather facilitates the use of specific counseling interventions. The working alliance thus is useful across theories. The relative importance of the components of the working alliance differ, depending on the counseling approach used by the counselor and the phase of the counseling process.

Factors that affect the development of a working alliance include the following: * the amount of psychological threat a client experiences in counseling,

* the extent and nature of the treatment goals,

* negative expectations for success,

* difficulty in maintaining social relationships,

* the difficulty of the problem that is being addressed in the counseling relationship, and

* the compatibility of the treatment demands with the client's emotional capabilities (Chan et al., 1997; Gelso & Carter, 1985; Horvath, 1994).

Interestingly, the severity of the client's symptoms do not affect the development of a positive therapeutic relationship (Horvath, 1994). Researchers have provided evidence that client ratings of the alliance are stronger predictors of treatment outcome than are counselor ratings (Connors, Carroll, DiClemente, Longabaugh, & Donovan, 1997; Horvath & Symonds, 1991; Luborsky, 1994).

Counselors view the working alliance as important. In a qualitative study, master therapists expressed a belief that the foundation of therapeutic change was a strong working alliance between the counselor and client (Jennings & Skovholt, 1999). Mallinckrodt and Nelson (1991) demonstrated that the level of counselor experience was related to working alliance ratings, with more-experienced counselors receiving higher ratings. Raue, Goldfield, and Barkham (1997) found that counseling sessions judged by counselors as particularly helpful were associated with higher counselor ratings of the working alliance. Research has supported the idea that the development of a strong working alliance can be facilitated by the counselor (Luborsky, 1994).

There is growing evidence to support the strong contribution of the working alliance to successful counseling outcomes (e.g., Al-Darmaki & Kivlighan, 1993; Connors et al., 1997; Goering, Wasylenki, Lindsay, Lemire, & Rhodes, 1997; Kivlighan & Shaughnessy, 2000; Kokotovic & Tracy, 1990; Mallinckrodt & Nelson, 1991), and this effect is found across counseling approaches (Horvath & Symonds, 1991). In a review of extant meta-analyses, Horvath (1994) found that the working alliance was related to positive counseling outcomes, with an average effect size of .26.

Although there is a growing body of evidence to suggest that the working alliance is important for successful counseling outcomes, no research has addressed the impact on vocational rehabilitation outcomes of the working alliance between client and counselor. This study considered the effect of this alliance on three important rehabilitation outcomes. The first two outcomes are directly related to goals delineated in the 1992 Amendments to the Rehabilitation Act. For the third question, individuals who were currently employed were analyzed separately from unemployed clients. The following research questions were addressed: 1. Is there a difference between clients who are employed and unemployed with respect to measured levels of the working alliance?

2. For employed clients, is there a relationship between working alliance and satisfaction with their current job?

3. For employed clients, is there a relationship between the working alliance and the clients' view of their future employment prospects?

4. For unemployed clients, is there a relationship between the working alliance and the clients' view of their future employment prospects?

METHOD

The participants for this study were clients of the Tennessee Division of Rehabilitation Services (TDRS) who were contacted by telephone during fiscal year 1999-2000. Clients were classified either Status 26 (employed) or Status 28 (unemployed). The researchers contacted 2,732 clients.

Demographic characteristics of participants are presented in Table 1. Participants ranged in age from 15 years to 75 years (M = 31.7, SD = 12.0), with 46% (n = 1,257) between the ages of 15 and 25, 28% (n = 758) between the ages of 26 and 40, and 26% (n = 717) older than 41. Most participants had never been married (61%; n = 1,489), with 19% (n = 468) married, 14% (n = 341) divorced, 5% separated (n = 113), and 1% (n = 37) widowed (see Note 1). Most respondents were Caucasian (78%; n = 2,133), with 21% (n = 584) African American, less than 1% (n = 8) American Indian, and less than 1% (n = 7) Asian and Pacific Islander. Participants could identify themselves as an individual of Hispanic origin (Cubans, Puerto Ricans, Mexicans, etc.) and also choose one of the racial categories. Forty-eight percent had completed less than a high school diploma (n = 1,321), while 39% (n = 1,052) had completed high school, 11% (n = 189) had completed post--high school education, and 2% (n = 58) were in special education (see Note 2). More than half (56%; n = 1,518) of the respondents were men. Most respondents were employed (67%; n = 1,822; see Table 1 for demographic characteristics of the employed and unemployed participants).

Participants reported a primary and secondary (if any) disability, as well as the severity of their disability. Of participants reporting the severity of their disability, 90% (n = 2,448) reported a severe disability. Twenty-three percent (n = 557) of participants reported a secondary disability. Respondents reported the following primary disabilities: * 33% (n = 913) chronic medical conditions,

* 27% (n = 727) psychiatric disorders,

* 19% (n = 519) mobility and orthopedic impairments,

* 11% (n = 312) mental retardation,

* 5% (n = 130) visual impairments,

* 3% (n = 71) hearing impairment, and

* 2% (n = 60) traumatic brain injury.

A comparison of the group of clients who were employed with those who were unemployed revealed that more individuals with a psychiatric disability were in the unemployed group and more individuals with chronic medical conditions were in the employed group. In addition, the unemployed group had a higher percentage of African Americans and was older than the employed group.

Instrument

The Bureau of Business and Economic Research/Center for Manpower Studies (BBER/CMS) at The University of Memphis developed a 47-item questionnaire regarding clients' satisfaction with TDRS programs and services, current employment status, and wages and benefits. We used two versions of the survey in this study: One version was used with clients who were employed, and a modified version that did not include the questions about benefits or satisfaction with current employment was used with individuals who were unemployed.

For purposes of this study, specific questions within the BBER/CMS questionnaire were used to measure the construct of working alliance. Following guidelines proposed by DeVellis (1991), we developed a nine-item instrument, named the Working Alliance Survey (WAS), specifically for this investigation. We defined working alliance as a collaboration between the client and the counselor based on the development of an attachment bond as well as a shared commitment to the goals and tasks of counseling (Bordin, 1979). Specifically, the development of the instrument was guided by two factors: (a) the concept of the working alliance as delineated by H. Bordin and (b) expert ratings by counselor educators familiar with the concept.

A review of the relevant literature indicated that the working alliance consists of three interdependent components--goals, tasks, and bonds (H. Bordin, 1979). Items contained in the BBER/CMS questionnaire were analyzed with respect to their relevance to Bordin's conception. The principal researcher chose items that addressed the core ideas contained in Bordin's components, which resulted in a set of 11 items to be reviewed by six expert raters. Criteria for selection as an expert rater included (a) a doctoral degree in counseling or counseling psychology, (b) experience as a counselor educator, and (c) familiarity with the concept of the working alliance. All of the experts had taught graduate courses in counseling techniques.

The expert raters were asked to judge the relevance of the items to the concept of the working alliance and to choose the component (goals, bonds, or tasks) of the working alliance most closely associated with the item. The expert raters rated each item on a 5-point Likert scale (5 = very relevant, 1 = not related). Items with a mean rating of 4.0 or higher were retained. This procedure reduced the item pool from 11 items to 9 (see the Appendix for a description of the items). For each item, the percentage of raters specifying a particular component of the working alliance was calculated. For example, if four raters judged Item 2 to be associated with the bonds component and two raters assigned the item to the tasks component, then Item 2 would be proportionally assigned as .66 to the bonds component and .33 to the tasks component. This proportional partition was computed for each item and assigned into the appropriate component of the working alliance. For each component, a linear equation was calculated using the proportion assigned by each expert rater. For example, the bonds component was computed as .16 (Item 1) + .66 (Item 2) + 1.0 (Item 5) + .33 (Item 8) + .16 (Item 9) + .16 (Item 11). E. Bordin (1994) conceptualized the working alliance as consisting of three equally important, interdependent components. In order to weight each component of the working alliance equally, the scores for bonds, tasks, and goals were standardized. Finally, the three standardized scores were added together to produce a working alliance total score. For the WAS, and measures of a client's view of his or her future employment prospects and satisfaction with the current job, a low score indicated a stronger working alliance, a more positive view of future job prospects, and satisfaction with the current job. With respect to the standardized working alliance score, a score of zero was at the mean and a negative score indicated a stronger working alliance. The internal consistency reliability coefficient (Cronbach's [alpha]) for the working alliance scale in this study was .80.

Procedures

Each month the TDRS provided the BBER/CMS at The University of Memphis with a list of clients. Staff at the BBER/CMS contacted clients by telephone 60 days after closure and administered the questionnaire by phone. If the initial attempt to contact the client was unsuccessful, six additional attempts were made. The BBER/CMS attempted to contact 10,387 clients. Of this number, 46% (n = 4,754) were contacted and completed the questionnaire. Ninety-three percent of the individuals who completed the questionnaire were clients, whereas parents completed the questionnaire for 5% of the clients, and family members and guardians completed the questionnaire for 2% of the clients. The BBER/CMS was unable to contact 4,913 individuals, and 722 were contacted but refused to respond. Approximately 43% of the questionnaires were unusable due to missing data and frequency of items marked "not sure," "does not apply," and "no response" answers. A final sample of 2,732 participants were used for analysis.

Data Analysis

A t test was conducted on the continuous variable of the WAS in order to compare clients who were employed with clients who were unemployed regarding the first research question, "Is there a difference between clients who are employed and unemployed with respect to measured levels of working alliance?" A Pearson product-moment correlation between the continuous variables of the WAS and satisfaction with their current job was calculated for clients who were employed regarding the second research question, "For employed clients, is there a relationship between working alliance and satisfaction with their current job?" A Pearson product-moment correlation between the continuous variables of the WAS and future employment prospects was calculated for clients who were employed regarding the third research question "For employed clients, is there a relationship between working alliance and the client's view of their future employment prospects?" Finally, a Pearson product-moment correlation between the continuous variables of the WAS and future employment prospects was calculated to deal with the fourth research question, "For unemployed clients, is there a relationship between working alliance and the client's view of their future employment prospects?" An alpha level of .05 was used for hypothesis testing.

Because this study used an ex post facto research design, specific demographic and disability-related factors were considered as potential sources of error in the results. Based on previous research (Bolton et al., 2000; Wilson, 2000) four variables were considered: 1. the participant's disability category,

2. whether the disability was severe or not severe,

3. the existence of a secondary disability, and

4. the participant's ethnicity.

Three of the variables (disability category, disability severity, and secondary disability) were considered inappropriate for statistical control for two reasons. First, they are characteristics of the population, and to analyze the data "as if" they were controlled would not represent real life (Stevens, 1992). Second, for the variables of disability category, disability severity, and secondary disability, when an ANOVA or t test was used, no significant statistical or meaningful difference was found on the continuous variable of working alliance. Although a significant difference was found on disability category, the effect size was minimal, F(6,725) = 3.77, p < .001; [[eta].sup.2] = .008. In addition, only the comparison between the individuals who were visually impaired and individuals with traumatic brain injury was found to be significantly different. Although a significant difference was found between the group of individuals with severe disabilities and individuals with nonsevere disabilities on the variable of working alliance, the effect size was minimal, t(2446) = 3.11, p < .01; [[eta].sup.2] = .004. The difference between individuals with a secondary disability and without a secondary disability was not significant, t(2446) = .77, p > .05; [[eta].sup.2] < .000. The final variable considered for statistical control was ethnicity. This variable was collapsed into two groups, Caucasian and non-Caucasian. The difference between these two groups was not significant on the continuous variable of working alliance, t(2730) = -.49, p > .05; [[eta].sup.2] < .000. The four variables thus were not considered to be appropriate for statistical control.

RESULTS

The use of a t test indicated a significant difference on the variable of working alliance for the employed group (M = -.65; SD = 2.26) versus the unemployed group (M = 1.29; SD = 3.26), t(2730) = -18.08, p < .001, with an effect size of d = .73. For clients who were employed, the measure of working alliance correlated significantly with satisfaction with current job (r = .15, p < .001). The measure of working alliance also correlated significantly with these clients' views of their future employment prospects (r = .51, p < .001). For clients who were unemployed, the measure of working alliance correlated significantly with their views of their future employment prospects (r = .52, p < .001).

DISCUSSION

Four research questions guided this study. First, the results indicated that clients who were employed measured stronger on the working alliance with their counselor than did clients who were unemployed, as measured by the WAS. According to Cohen (1988), the effect size (d = .73) can be considered a medium effect. Second, for clients who were employed, the results indicated that the stronger the measured level of working alliance, the more satisfied the clients were with their current job. The correlation coefficient as measured by a Pearson product-moment correlation (r = .15) can be considered a small effect, per Cohen. Third, for clients who were employed, the results indicated that the stronger the measured level of working alliance, the more positive the view these clients held of their employment future. According to Cohen, the effect size as measured by the Pearson product-moment correlation (r = .51) can be considered a large effect. Finally for clients who were unemployed, the results indicated that the stronger the measured level of working alliance, the more positive the view these clients held of their employment future. This result can also be considered a large effect (r = .52).

Overall, the results produced two large effects, one medium effect, and one small effect. The importance of these effect sizes is illustrated by comments by Wampold (2001) and Meyer et al. (2001). In a review of meta-analytic studies looking at the relationship between working alliance and therapy outcomes, Wampold characterized the .26 aggregated correlation as a "robust relationship" (p. 151). Meyer et al., in comprehensive review of the relationship between the Pearson product-moment correlation and psychological interventions, stated that many psychological interventions and constructs produce correlations in the range of .15 to .30. The measured effect sizes in the current study thus are similar to effect sizes deemed robust by Wampold and typical by Meyer et al. In sum, the results suggest that the working alliance may be an important aspect of vocational rehabilitation services that can lead to positive outcomes, specifically, employment, satisfaction with the current job, and a positive perspective concerning the client's employment future. These results are congruent with previous research that showed a relationship between a strong working alliance and positive counseling outcomes (Al-Darmaki & Kivlighan, 1993; Connors et al., 1997; Goering et al., 1997; Horvath, 1994; Kivlighan & Shaughnessy, 2000; Kokotovic & Tracy, 1990; Mallinckrodt & Nelson, 1991).

Implications for Rehabilitation Counselors

The results of this study provide evidence that rehabilitation counselors may be able to improve outcomes by facilitating a strong working alliance with their clients. Bordin (1979) suggested that the importance of this alliance is pantheoretical. Although rehabilitation counselors in the state--federal rehabilitation system may employ different counseling approaches, most work within a brief counseling framework. In brief counseling, goals are limited and the counselor tends to take a more active approach than in most long-term counseling relationships (Safran & Muran, 1998a). In addition to its brief nature, counseling within the state--federal rehabilitation system tends to be more intermittent than that in other settings. Within the context of a brief, intermittent counseling approach, rehabilitation counselors can facilitate the development of a working alliance with clients by adhering to the following principles.

First, the bonds component of the working alliance can be facilitated by the counselor expressing warmth toward, respect for, and interest in the client (Safran & Muran, 1998b). This expression is important in both long-term and short-term counseling, but it is particularly important in the state--federal rehabilitation setting. The counselor and his or her client are more likely to have a strained relationship because caseloads constrict the amount of time a counselor can devote to a client. As a result, the counselor must be more active in facilitating movement toward counseling goals.

Second, the brief, intermittent nature of counseling within the state--federal system suggests that (a) the counseling tasks and goals should be determined early in counseling and (b) a more didactic approach to discussing the tasks and goals should be used (Safran & Muran, 1998b). Rehabilitation counselors in this system thus may spend time during the first or second meetings establishing goals within the context of the amendments to the Rehabilitation Act and discussing the range of in-counseling and extra-counseling activities that may be appropriate. The counselor and the client must also agree on realistic and focused goals. In order to increase the likelihood that counseling will produce positive outcomes, the scope of issues that are considered pertinent within the context of the rehabilitation counselor--client alliance must be delineated and issues that are more appropriately referred to outside agencies should be identified. The realistic framing of relevant counseling issues facilitates the development of the bonds, tasks, and goals of the working alliance.

Third, rehabilitation counselors within the state-federal rehabilitation system must balance directive responses with reflective responses (Safran & Muran, 1998b; Watson & Greenberg, 1998). Although the short-term, intermittent nature of the counseling process may suggest an increased reliance on directive counseling responses, a combination of directive and reflective responses may better serve to develop a strong bond between the counselor and his or her client.

Obviously, the development of a working alliance is more likely to be successful when the counseling is longer term (Safran & Muran, 1998b). Consequently, it is more likely that a rehabilitation counselor and his or her client will experience problems in the development of strong goals, tasks, and bonds when the counseling is brief and intermittent. Clients may feel overwhelmed by the process of finding a job and believe they are not ready to start vocational rehabilitation at the pace suggested by the counselor (Newman, 1998). Disagreement about counseling goals and tasks will negatively affect the establishment of a positive personal attachment between the counselor and the client. In order to minimize the breakdown in the development of the working alliance, counselors should elicit feedback from their clients so as to facilitate the forward movement of the client toward the counseling goals.

Limitations

Conclusions about the results are limited by several considerations. First, this study utilized an ex post facto design. A limitation of this type of design is the difficulty in determining a causal link between variables. Other factors may have affected the outcome. For example, clients may have refused service when the initial counselor contact was problematic in terms of the development of a working alliance. Some preselection of client and counselor thus may have occurred. Other factors that could affect the outcome include family and financial support, training of the counselor, and the duration of service. Second, the BBER/CMS was unable to contact slightly less than half (47%) of the potential respondents, and another 7% were contacted but refused to reply. In addition, of those who responded to the questionnaire, approximately 50% did not participate in the study due to missing data. It is unclear whether nonrespondents and respondents with missing data differed significantly from respondents. Third, interviews were completed during the 1999-2000 fiscal year with TDRS clients. The interpretation of the results therefore should be limited to the sample examined at the time of the study. Fourth, although care was taken to provide evidence of the reliability and validity of the WAS, more evidence is needed to substantiate its reliability and validity. Finally, only the client's view of the strength of the working alliance was ascertained. Although research has provided evidence that client ratings of the alliance are stronger predictors of treatment outcome than are counselor ratings (Connors et al., 1997; Horvath & Symonds, 1991; Luborsky, 1994), the validity of the measure of working alliance may be increased if both the counselor and the client are asked for their perception of the working alliance.

Future Research

The results of this study suggest a number of areas for future research. Although the current data did not indicate a significant effect related to specific demographic factors and type of disability on the level of working alliance and rehabilitation outcomes, it would be informative to investigate the effect of these factors with other samples. Do race or gender matter in the measured level of working alliance? Are the level of working alliance and associated rehabilitation outcomes different for individuals with psychiatric disorders than for persons with chronic medical conditions?

More broadly, the concept of working alliance has not been studied within the context of the state-federal vocational rehabilitation system. For this study, we developed a measure of working alliance based on existing survey questions, but a number of other measures of working alliance do exist (see Horvath, 1994; Horvath & Greenberg, 1989). A study utilizing an existing instrument with a vocational rehabilitation population thus may be useful. The following areas also need to be investigated. 1. What is the process of development of the working alliance during the initial, middle, and later phases of counseling?

2. What is the impact of client factors (e.g., disability, race, gender, age, education, past employment) and counselor factors (e.g., degree or certification, length of time with vocational rehabilitation, race, gender) on the development of the working alliance? In addition, what is the effect of the interaction of these factors on the development of the working alliance?

3. What are effective methods for improving a counselor's development of an effective working alliance?

4. Bordin theorized that the components of the working alliance (goals, bonds, and tasks) are of equal importance and interdependent, but it would also be constructive to ask the following question: What are the links between specific components and rehabilitation outcomes?

Investigation of these areas would increase our understanding of the impact of the working alliance on rehabilitation outcomes.

The development of an effective working alliance between a rehabilitation counselor and client requires the counselor to be able to demonstrate a relatively high level of counseling skill (Egan, 1998; Gelso & Carter, 1994; Jennings & Skovholt, 1999). Building the foundation for the development of these skills typically starts in a graduate-level counseling program. The importance of these graduate level counseling skills in the field of rehabilitation counseling is underscored by (a) research that has indicated the effectiveness of counselors with counseling and rehabilitation counseling degrees (Szymanski, 1991, 1992), (b) the importance placed on certification of rehabilitation counselors (Leahy & Holt, 1993; Szymanski, Leahy, & Linkowski, 1993), and (c) the Rehabilitation Services Administration (2000) implementation of the Comprehensive System of Personnel Development requiring state-federal rehabilitation counselors to possess the highest licensing, certification, or registration standard in the state or to be a certified rehabilitation counselor. If possessing a graduate degree in counseling or rehabilitation counseling is important, then it should be empirically demonstrated that high-level counseling skills positively affect important rehabilitation counseling outcomes. This study provides preliminary evidence that a specific counseling skill--the development of a working alliance between the counselor and the client--does positively affect the employment of the client, as well as his or her view of future employment prospects and his or her satisfaction with the current job.

AUTHORS' NOTE

The authors would like to thank the Tennessee Division of Rehabilitation Services and the Bureau of Business and Economic Research/Center for Manpower Studies at The University of Memphis for their assistance in survey development and implementation. APPENDIX: WORKING ALLIANCE SURVEY ITEMS

1. Did the vocational rehabilitation counselor and staff seem committed to helping you find a job?

2. Did your counselor try to match your skills with the jobs available at the time?

3. Did your counselor try to understand your problems and needs?

4. Did your counselor help you try to solve your problems?

5. Did your counselors and staff treat you with dignity and respect?

6. Did you feel that you received all the services specified in your rehabilitation plan?

7. How involved were you in developing your vocational goals?

8. How involved were you in selecting your program services?

9. How involved were you in developing your service providers?

NOTES

(1.) The marital status numbers do not add up to 2,732 due to missing data.

(2.) The education numbers do not add up to 2,732 due to missing data.

REFERENCES

Al-Darmaki, F., & Kivlighan, D. M. (1993). Congruence in client-counselor expectations for relationship and the working alliance. Journal of Counseling Psychology, 40, 379-384.

Bolton, B., Bellini, J., & Brookings, J. (2000). Predicting client employment outcomes from personal history, functional limitations, and rehabilitation services. Rehabilitation Counseling Bulletin, 44, 10-21.

Bordin, E. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research and Practice, 16, 252-260.

Bordin, E. (1994). Theory and research on the therapeutic working alliance: New directions. In A. Horvath & L. Greenberg (Eds.), The working alliance: Theory, research and practice (pp. 13-37). New York: Wiley.

Chan, F., Shaw, L., McMahon, B., Koch, L., & Strauser, D. (1997). A model for enhancing rehabilitation counselor-consumer working relationships. Rehabilitation Counseling Bulletin, 41, 122-134.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Connors, G., Carroll, K., DiClemente, C., Longabaugh, R., & Donovan, D. (1997). The therapeutic alliance and its relationship to alcoholism treatment participation and outcome. Journal of Consulting and Clinical Psychology, 65, 588-598.

DeVellis, R. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage.

Egan, G. (1998). The skilled helper (6th ed.). Pacific Grove, CA: Brooks/Cole.

Gelso, C., & Carter, J. (1985). The relationship in counseling and therapy: Components, consequences and theoretical antecedents. The Counseling Psychologist, 13, 155-243.

Gelso, C., & Carter, J. (1994). Components of the psychotherapy relationship: Their interaction and unfolding during treatment. Journal of Counseling Psychology, 41, 296-306.

Goering, P., Wasylenki, D., Lindsay, S., Lemire, D., & Rhodes, A. (1997). Process and outcome in a hostel outreach program for homeless clients with severe mental illness. American Journal of Orthopsychiatry, 67, 607-617.

Horvath, A. (1994). Research on the alliance. In A. Horvath & L. Greenberg (Eds.), The working alliance: Theory, research and practice (pp. 259-286). New York: Wiley.

Horvath, A., & Greenberg, L., (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology, 36, 223-233.

Horvath, A., & Symonds, D. (1991). Relation between the working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149.

Jennings, L., & Skovholt, T. (1999). The cognitive, emotional, and relational characteristics of master therapists. Journal of Counseling Psychology, 46, 3-11.

Kivlighan, D., & Shaughnessy, P. (2000). Patterns of working alliance development: A typology of client's working alliance ratings. Journal of Counseling Psychology, 47, 362-371.

Kokotovic, A., & Tracy, T. (1990). Working alliance in the early phase of counseling. Journal of Counseling Psychology, 37, 16-21.

Leahy, M., & Holt, E. (1993). Certification in rehabilitation counseling: History and process. Rehabilitation Counseling Bulletin, 37, 71-80.

Luborsky, L. (1994). Therapeutic alliances predictors of psychotherapy outcomes: Factors explaining the predictive success--New directions. In A. Horvath & L. Greenberg (Eds.), The working alliance: Theory, research and practice (pp. 38-50). New York: Wiley.

Mallinckrodt, B., & Nelson, M. (1991). Counselor training level and the formation of the psychotherapeutic working alliance. Journal of Counseling Psychology, 38, 133-138.

Meyer, G., Finn, S., Eyde, L., Kay, G., Moreland, K., Dies, R., et al. (2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56, 128-165.

Newman, C. (1998). The therapeutic relationship and alliance in short-term cognitive therapy. In J. Safran & J. Muran (Eds.), The therapeutic alliance in brief psychotherapy (pp. 95-122). Washington DC: American Psychological Association.

Raue, P., Goldfield, M., & Barkham, M. (1997). The therapeutic alliance in psychodynamic-interpersonal and cognitive-behavioral therapy. Journal of Consulting and Clinical Psychology, 65, 582-587.

Rehabilitation Act Amendments, Pub. L. No. 102-569 (codified as amended at 29 U.S.C. 701 et seq (1992)).

Rehabilitation Services Administration. (2000). Comprehensive system of personnel development: Common questions and answers. Washington DC: Author.

Safran, J., & Muran, J. (Eds.). (1998a). Negotiating the therapeutic alliance in brief psychotherapy. In J. Safran & J. Muran (Eds.), The therapeutic alliance in brief psychotherapy (pp. 3-14). Washington DC: American Psychological Association.

Safran, J., & Muran, J. (1998b). The therapeutic alliance in brief psychotherapy: General principles. In J. Safran & J. Muran (Eds.), The therapeutic alliance in brief psychotherapy (pp. 217-229). Washington DC: American Psychological Association.

Stevens, J. (1992). Applied multivariate statistics for the social sciences. Hillsdale, NJ: Erlbaum.

Szymanski, E. (1991). Relationship of level of rehabilitation counselor education to rehabilitation client outcome in the Wisconsin Division of Vocational Rehabilitation. Rehabilitation Counseling Bulletin, 35, 23-37.

Szymanski, E. (1992). The relationship of rehabilitation counselor education to rehabilitation client outcome: A replication and extension. Journal of Rehabilitation, 58(1), 49-56.

Szymanski, E., Leahy, M., & Linkowski, D. (1993). Reported preparedness of certified rehabilitation counselors in rehabilitation counseling knowledge areas. Rehabilitation Counseling Bulletin, 37, 146-162.

Wampold, B. (2001). The great psychotherapy debate: Models, methods, and findings. Mahwah, NJ: Erlbaum.

Watson, J., & Greenberg, L. (1998). The therapeutic alliance in short-term humanistic and experiential therapies. In J. Safran & J. Muran (Eds.), The therapeutic alliance in brief psychotherapy (pp. 123-145). Washington DC: American Psychological Association.

Wilson, K. (2000). Predicting vocational rehabilitation acceptance based on race, education, work status, and source of support at application. Rehabilitation Counseling Bulletin, 43, 97-105.

Daniel C. Lustig, PhD, is an assistant professor in the Department of Counseling, Educational Psychology, and Research at The University of Memphis and coordinator of the Rehabilitation Counseling Program. His current research interests include families, career thoughts, and working alliance. David R. Strauser, PhD, is the director of the Center for Rehabilitation and Employment Research and an associate professor in the Department of Counseling, Educational Psychology, and Research at The University of Memphis. His research interests are career development for individuals with disabilities, work personality, and working alliance. N. Dewaine Rice, EdD, is an associate professor in the Department of Counseling, Educational Psychology, and Research at The University of Memphis. He is interested in college student development and pragmatic applications of hypnosis. Tom E Rucker, MS, is a research assistant with the Bureau of Business and Economic Research/Center for Manpower Studies at The University of Memphis. His primary research interest is college student development. Address: Daniel C. Lustig, Department of Counseling, Educational Psychology, and Research, University of Memphis, 100 Bath Hall, Memphis, TN 38111-9890; e-mail: dlustig@memphis.edu