VISTAS articles are made available for historical reference only and are presented "as is." ACA does not guarantee or represent that the information is current, accurate or indicative of the original or intended quality. These materials are not maintained or updated and may contain outdated or incomplete information. Readers should exercise discretion and verify information independently before relying on it. We assume no responsibility for the use or interpretation of this content.

Article 38

The Supervisory Relationship

Its Impact on Trainee Personal and Skills Development

Stephanie Horrocks and Marlowe H. Smaby

Download PDF

Supervision is considered a primary method for imparting knowledge to counseling trainees (Holloway, 1992). However, the formal study and development of supervision as a specialty with a distinct set of requisite skills and knowledge gained momentum as recently as the 1980s (Dye & Borders, 1990). According to Bernard and Goodyear (2004), counselor supervision is an intervention in which a more senior member of the profession assists a more junior member of the profession to enhance professional functioning and ensure client welfare. The nature of supervision is evaluative and extends over time.

Models and theories of counseling supervision have emerged mostly as extensions of counseling practice. In practice, supervisors, like counselors, probably rely on a variety of supervision theories and techniques with trainees (Loganbill, Hardy, & Delworth, 1982). Because supervision and counseling share certain characteristics, the application of counseling theories to supervisory relationships seems logical. Bernard and Goodyear (2004) have described ways in which supervision sometimes resembles counseling, but they contrasted the counseling-related aspects of supervision with other roles that are unlike that of a counselor. For example, supervisors frequently act as teachers and consultants. They also evaluate trainees. In graduate programs, where the standards for initial teaching and training are a primary function in supervision, the gatekeeping role in terms of building counseling competencies may be especially pronounced (Lumadue & Duffey, 1999).

Researchers have begun to consider the more global, or pantheoretical, attributes of the supervisory relationship that affect supervision outcomes regardless of supervisors’ theoretical approaches to counseling and supervision. The supervisory working alliance (Bordin, 1983) is one such area of research. This study is an attempt to answer the question, “Does a strong supervisory working alliance contribute to counseling trainees’ personal growth and attainment of counseling skills?”

In 1983, Bordin generalized the therapeutic working alliance to the supervisory relationship. The supervisory working alliance is comprised of the same three components as the therapeutic working alliance (i.e., goals, tasks, and bonds), but the alliance is formed between supervisor and trainee instead of counselor and client. Bordin asserted that stronger supervisory working alliances are conducive to supervision outcomes, just as strong therapeutic alliances have been related to improved counseling outcomes (Horvath & Symonds, 1991).

The outcomes measured in this study relate directly to two supervisory goals suggested by Bordin. Specifically, counseling trainees’ levels of counseling skills mastery, understanding of self, and understanding of the client as measured by the Counseling Skills Personal Development – Rater Form (CSPD-RF; Torres-Rivera et al., 2002; Wilbur, 1991) will be tested according to high and low working alliance as measured by the Supervisory Working Alliance-Trainee version (SWAI-T; Efstation, Patton, & Kardash, 1990). The hypotheses of the current study are (a) trainee self- ratings of personal growth and skills attainment as measured by the CSPD-RF will be significantly different according to either high or low levels of supervisory working alliance, and (b) supervisor ratings of trainee personal growth and skills attainment as measured by the CSPD-RF will be significantly different according to either high or low levels of supervisory working alliance.

Method

Participants

Participants were 24 graduate-level counseling trainees (19 women and 5 men, median age = 30 years) who were in a semester-long comprehensive examination course at a midsized land grant, research- based university in the western United States. Two participants were minority students (one Hispanic/ Mexican-American; one Asian/Pacific Islander); three declined to state an ethnicity; and two reported an unknown ethnic status. All students were near the end of their counseling programs of study with various areas of concentration (community counseling, college student development, marriage and family therapy, combined community counseling and marriage and family therapy, school counseling, addictions counseling).

The course was a 3-credit-hour, semester-long course offered in a counselor education program accredited by the Council for Accreditation for Counseling and Related Educational Programs (CACREP). Data were gleaned from the instruments and assignments normally required for completion of the course.

Instrumentation

Supervision Working Alliance Inventory—Trainee Version(SWAI-T; Efstation et al., 1990). The purpose of the SWAI is to measure the relationship in counselor supervision, as perceived by the supervisor and the trainee. This measure of supervisory working alliance was selected because it is constructed such that each person in the supervisory dyad completes items according to their perception of the other attempting certain behaviors thought to be building blocks of the working alliance. Two factors, Rapport and Client Focus, were extracted through a factor analysis of the trainee version (Efstation et al., 1990). Rapport (Items 1–12) is a measure of supervisors’ efforts to build a bond or relationship with the trainee, as perceived by trainees. Client Focus (Items 13–19) is a measure of the degree to which trainees believe their supervisors encouraged focused efforts toward specific goals and tasks expected to benefit clients. Subscale scores on the SWAI-T are reported as means of the total scores on each factor. Therefore, averaged scores on the Rapport and Client Focus subscales range from 1 to 7, and overall scores range from 12 to 84 on the Rapport subscale and from 7 to 49 on the Client Focus subscale. Efstation et al. (1990) reported Chronbach alpha coefficients of .90 and .77 for the Rapport and Client Focus scales respectively.

Counseling Skills and Personal Development— Rating Form(CSPD-RF; Wilbur, 1991). The CSPD- RF consists of 20 items that are statements related to the performance and personal development of the counselor-in-training. Ten items were designed to measure skills development, and 10 were designed to measure personal development (Wilbur, 1991). The intent of the developers of the CSPD-RF was to develop an instrument that allowed for the identification of components of counseling that enhance the integration of counseling skills and trainees’ personal development. Originally, the CSPD items were rated on a 7-point Likert-type scale from 1 (unacceptable) to 7 (outstanding). However Torres-Rivera et al. (2002) modified the CSPD-RF scale such that the midpoint response was removed, thus converting the instrument to a forced-choice format for validity exploration. Items are rated on a 6-point Likert-type scale (1 = unacceptable and 6 = outstanding). The revised forced- choice version of the scale was used in this study. Therefore, total scores may range from 20 to 120 and are indicative of the quality with which the skill or level of personal awareness in the counseling situation was demonstrated.

The Cronbach’s alpha coefficient for all 20 items on the CSPD-RF was .91, and the split-half reliability coefficients were .83 and .84 (Torres-Rivera et al., 2002). Four factors were identified on the CSPD-RF with eigenvalues greater than 1.0, accounting for 58.4% of the total variance (Torres-Rivera et al., 2002). Emotional sensitivity (items 1, 2, 3, 4, 5, and 20) is a measure of the trainee’s communication and clarification skills, awareness of emotional states, congruency, empathy, sensitivity, self-disclosure, and overall honesty in the counseling session. Factor loadings for emotional sensitivity ranged from .46 to .76. Total emotional sensitivity scores may range from 6 to 36 points.

Basic listening skills (items 6, 7, and 15) are a measure of the trainee’s ability to listen accurately to clients. Items in this factor are related to basic communications skills such as attending behaviors, paraphrasing, summarizing, providing feedback, and interpreting client statements. Total factor loadings for basic listening skills ranged from .71 to .77, and total scores for basic listening skills may range from 3 to 18 points.

Multicultural skills (items 8, 12, 14, 16, 18, and 19) are a measure of the demonstrated ability of the trainee to recognize and respond to diversity, racism, and prejudice issues that affect the counseling relationship and session. Factor loadings ranged from .48 to .79, and total scores for multicultural skills may range from 6 to 36 points.

Influencing skills (items 10, 11, 13, and 17) are a measure of the trainee’s ability to affect the choices and decision making of the client in the session. It is related to the trainee’s ability to inspire decisions and a plan for change from the client. As such, it seems to be a measure of interpersonal influence. Factor loadings ranged from .56 to .72, and total scores for influencing skills may range from 4 to 24 points.

Emotional sensitivity and basic listening skills comprise the counseling skills measure of the instrument. Multicultural skills and influencing skills comprise the personal development dimension of the instrument.

Procedure

Participants attended a 1-hour orientation in which they reviewed course expectations and were given a syllabus. As part of the comprehensive examination course, participants were required to (a) videotape a minimum of two actual counseling sessions with clients from their respective internship sites, (b) meet with a PhD-level supervisor for at least two 1- to 1.5-hour supervision sessions, (c) complete self-ratings using the CSPD-RF prior to supervision sessions, and (d) select one videotaped counseling session for final submission. After the preestablished deadline for submission of the comprehensive exams tapes, each tape was viewed and rated by the supervisor using the CSPD-RF.

Supervisors were eight PhD-level counseling students who had taken a course in counseling supervision (6 women and 2 men, median age = 35 years). Trainees were assigned to supervisors with whom they had not previously received counseling supervision and did not associate with outside of the university setting. As such, true random assignment of participants was not possible.

Trainees completed the SWAI–T after the second supervision session and were asked to submit it in an envelope directly to a graduate assistant for the course. Trainees were encouraged to complete the SWAI–T openly and honestly, as it would not be part of their final grade in their course or part of their final consideration for graduation.

Results

High and low SWAI-T scores were established using a median point split. Scores higher than the median points of each scale were considered high; scores lower than the median points of each scale were considered low SWAI-T (client focus median = 6.57; rapport median = 6.58; overall SWAI-T median = 6.68). These results are based on a preliminary analysis of data.

Due to the small, nonrandomized sample, Mann- Whitney U statistics for two independent means were calculated to test for differences in the outcomes measures according to high or low levels of supervisory working alliance (see Table 1).

Table 1

Mean Ranks for Measured Supervision Outcomes by High or Low Working Alliance

*p< .05

The most consistent association of supervision outcomes according to the supervisory working alliance was in supervisors’ ratings of trainees’ skills development. Trainee personal development also differed according to high and low supervisory working alliance, but rapport in the supervisory relationship seemed to be the most related aspect of the alliance.

Discussion

The hypotheses of this study were partially supported. These findings generally support Bordin’s (1983) proposition that strong supervisor working alliances are conducive to supervisory outcomes. High and low levels of supervisory working alliance were predictive of trainee skill attainment and personal growth. However, the supervisor ratings varied more than the trainee ratings. Trainee perceptions of their own personal growth and skills attainment according to high or low working alliance varied less than supervisor ratings. This was an unexpected finding of the study. It is possible that the supervisor and trainee experience in supervision differs more than Bordin proposed.

There are several limitations of the current study. The sample size was small, and random assignment of participants was not possible. The generalizability of findings may be limited. Additionally, the supervisory working alliance ratings were gathered from only the trainee perspective. Given the differential reports from supervisors and trainees on personal growth and skills attainment, future studies might gather perceptions from both the supervisor and trainee. Studies that include the theoretical orientation of the supervisor might further test the pantheoretical assumption of Bordin’s (1983) theory as well.