Influence of the supervisory working alliance on supervisee work satisfaction and work-related stress.
Clinical supervision has become increasingly important to the development of supervisees (Getz, 1999) and the advancement of the counseling profession. Although the practice of supervision originated in the 1920s as part of psychoanalytic training (Feltham, 2000; Fleming & Benedek, 1966), supervision as a specialty area has only emerged in the past two decades (Dye & Borders, 1990). While there is an extensive body of work on many facets of counselor supervision, there is very little that explores the supervisory working alliance (SWA) in professional counseling settings. Further, most of the supervision research has been conducted with trainees, not professional counselors. This study expands our understanding of the SWA by investigating how professional mental health counselors perceive the supervision relationship and how this relationship influences work satisfaction and work-related stress.
BRIEF OVERVIEW OF SWA
A central component of the clinical supervision process is the SWA, which Ladany, Ellis, and Friedlander (1999) identified as "potentially one of the most important common factors in the change process of supervision" (p. 447). The SWA model consists of three characteristics drawn from the therapeutic working alliance model: (a) mutual agreement on supervision goals, (b) specific tasks related to supervision goals, and (c) the development of bonds between supervisor and supervisee (Bordin, 1983). These three components of the therapeutic working alliance form the foundation for the SWA, which based on Bordin's model is defined as the relational bond that develops between supervisor and supervisee when they work together to achieve mutual goals through clearly identified tasks.
It is important to note several distinctions between the SWA and the supervisory relationship. The SWA focuses on establishing the relational bond within the supervision dyad in order to serve as a change agent. It differs from the supervisory relationship in that the focus is on the supervisee's goals (Bordin, 1983) rather than the supervisor's. Aspects of supervisory relationships that are not specifically captured in the SWA include the evaluative component that is central to the supervisory relationship, feedback to ensure that supervisees maintain client welfare and monitor client care, and the supervisor's role as gatekeeper (Bernard & Goodyear, 2004).
Quality of the SWA
Goals, tasks, and relational bonds define the SWA, but the degree to which change occurs depends on the strength of the collaboration between supervisor and supervisee (Bordin, 1983). The quality of the SWA is a distinctive characteristic reflective of the strength of the relationship and refers to the supervisee's perceived experience within the supervision dyad (cf. Worthington, 1984). The quality of the SWA was typically viewed as ranging from positive (high quality) to negative (low quality). A high-quality SWA in the current study was equivalent to a strong supervisory relationship and a low-quality SWA as a weak supervisory relationship.
Supervisor characteristics typically associated with a positive SWA were being nonjudgmental, providing validation, supporting exploration, imparting an empathic attitude, normalizing anxiety and tension, and strengthening the SWA (Worthen & McNeill, 1996). Numerous researchers (cf. Gray, Ladany, Walker, & Ancis, 2001; Nelson & Friedlander, 2001; Ramos-Sanchez et al., 2002) identified supervisor characteristics that led to negative experiences in supervision, including being rigid, overly critical, or untrustworthy; displaying a lack of respect; lacking openness; being unsupportive; limiting direct feedback; conveying little empathy; not respecting differences; lacking encouragement and praise; and being inattentive. (For further discussion on creating a strong SWA, see Worthen and McNeill, 1996, and Carifio and Hess, 1987).
SWA, WORK SATISFACTION, AND WORK-RELATED STRESS
There is an extensive business and general service literature that explores the relationship between supervision and work satisfaction (e.g., Eklund & Hallberg, 2000; Hyrkas, 2005; Trant, Larsen, & Feimer, 2000; Uys, Minnaar, Simpson, & Reid, 2005) and supervision and work-related stress (e.g., O'Driscoll & Beehr, 1994; Seltzer & Numerof, 1988), but not SWA, work satisfaction, and work-related stress simultaneously. Fincham and Rhodes (2005) defined work satisfaction as "the feelings or 'affective response' someone experiences in a job role" (p. 9). Gabriel and Liimatainen (2000) defined work-related stress as "the harmful physical and emotional response that occurs when the requirements of the job do not match the capabilities, resources, or needs of the worker" (p. 11). Even though these variables have been well defined, as have relations among various combinations of variables, it is unclear whether findings from other disciplines are comparable when exploring the role, dimensions, and philosophy of supervision in professional counseling settings. Understanding the influence that the SWA has on these work constructs is significant not only because of their impact on supervisees' ability to function effectively and productively in counseling but also because of the potential implications for the therapeutic working alliance and client treatment outcomes.
Several researchers have posited that work satisfaction is a function of how supervisees perceive the SWA, assuming that supervisees' perceptions of it might influence not only how they function in the workplace but also their professional development (Ramos-Sanchez et al., 2002; Worthen & McNeill, 1996). Ramos-Sanchez and colleagues contended that as supervisees begin to question their career decisions, they might experience discouragement or despair and a sense of failure or loss of confidence in their ability to work in the profession. Holloway and Neufeldt (1995) noted that if supervision is efficacious (e.g., developing skills, establishing a therapeutic bond with the supervisee), there is a greater probability of improved job performance. In their study of the relations between supervision characteristics and job satisfaction, Kavanagh et al. (2003) reported that supervision might have a positive effect on client treatment and staff retention rates. Schroffel (1999) examined the association between work satisfaction and the frequency, quality, and style of current supervision and reported, among other things, higher work satisfaction when the style of supervision matched the supervisees' preferred supervisory style. Exploring the relations between clinical supervision, work satisfaction, and burnout with mental health and psychiatric nurses, Hyrkas (2005) found that supervisees who valued the clinical supervision experience viewed it as beneficial (e.g., lower burnout scores), as well as contributing to greater work satisfaction. Finally, one proposed benefit of supervision is that it serves as a mechanism to help supervisees cope with the stress associated with working in mental health occupations (Spence, Wilson, Kavanagh, Strong, & Worrall, 2001).
Research on counselor supervision has typically taken place in academic settings (cf. Gray et al., 2001; Ladany et al., 1999; Magnuson, Norem, & Wilcoxon, 2000; Nelson & Friedlander, 2001; Ramos-Sanchez et al., 2002). Ronnestad and Skovholt (1993) found that research on quality of clinical supervision often took place with supervisors-in-training, who usually had limited field experience. Nearly a decade later, Spence et al. (200l) similarly observed that "most of the research has focused on the training of specific clinical skills with interns.... [h]owever, minimal research has been conducted to determine the impact of supervision on the practice of clinicians in mental health settings" (p. 144).
Clinical supervision studies need to shift the preponderance away from academic settings to professional settings to better assess the association between supervisee work satisfaction and work-related stress. The purpose of the current study was to identify relations between several variables: (a) supervisees' perceptions of the quality of the SWA, (b) work satisfaction, (c) work-related stress for supervisees working in mental health agencies, (d) counseling setting, and (e) number of clients per week. Three alternative hypotheses were developed to test the relations between the variables.
[H.sub.1] : As supervisees' perceptions of the quality of the SWA increase, their perceptions of work satisfaction will also increase.
[H.sub.2]: As supervisees' perceptions of the quality of the SWA increase, their perceptions of work-related stress will decrease.
[H.sub.3]: Supervisees' work setting, caseload, perceptions of rapport with supervisor, and perceptions of supervisor focus on the client will correlate with supervisees' intrinsic and extrinsic work satisfaction, role ambiguity, role boundary, and role overload.
Participants were selected using a criterion-based random sample from the American Mental Health Counseling Association, a division of the American Counseling Association (ACA). This division was selected because members were likely to have engaged in clinical supervision as part of their professional training and advancement within the profession. Participation criteria were (a) work in professional mental health or drug and alcohol settings for at least one year, (b) a master's degree in counseling, and (c) either currently receiving clinical supervision or previously received supervision for a period of 12 months as part of their professional (post-master's degree) counseling work experience.
Participants who were not currently receiving supervision were advised to respond to all survey questions based on their experiences when they were last supervised.
There were 71 participants ranging in age from 29 to 73, with the average age of 51 (n = 71); 68 percent were female (n = 48) and 31% male (n = 22)--one person did not respond. As for race, 90% identified as Caucasian (n = 64), 4% as Latino American (n = 3), 3% as multi- or biracial (n = 2), 1% (1) as African American, and 1% (1) as Native American.
Eighty-three percent reported having a master's degree (n = 59) and 17% had completed a doctoral degree (n = 12). Participants reported working as supervisees for an average of 17 years (M = 17.01 years; SD = 9.53) after completing their master's degree. The number of years employed post-master's ranged from 3 to 38. Asked the number of years of work in their current organization, the responses ranged from 1 year to 30, with an average of just under 9 (M = 8.89 years; SD = 6.99 years).
Participants identified four main counseling work settings: 39% were in private practice (n = 28), 27% worked in a mental health agency (n = 19), 16% worked for a private, nonprofit agency (n = 11), 4% worked in a hospital (n = 3), and 14% identified other settings (n = 10), including colleges or universities, correctional facilities, a chemical dependency treatment center, a geriatric center, and a forprofit subcontractor.
Respondents on average counseled 20 clients a week (M = 19.94; SD = 8.89). Seventy percent reported spending one hour a week in individual clinical supervision (n = 50) and 17% reported spending two hours in this activity (n = 12). Participant responses to this question ranged from 1 to 10 hours a week (M = 1.54; SD = 1.18).
This study was approved by the Institutional Review Board (IRB# 23386). An initial letter was mailed to 350 prospective participants directing interested participants to an Internet website where they completed the survey. A prospective participant who did not have access to the Internet or preferred not to complete the survey online could be sent a paper copy of the survey. All prospective participants received two follow-up postcards mailed two and four weeks after the initial mailing, consistent with the recommendations of Vaux (1996) to improve response rates. Due to cost, no incentives were provided to potential participants.
Of the 350 prospective participants, 348 received an invitation letter and follow-up postcards (two had invalid addresses). Of these, 81 responded directly to the online survey and 1 completed a paper-and-pencil version. A review of participant responses showed that 11 individuals had completed the demographic questionnaire but did not fill in responses for the three instruments; they were not included in the study. Thus 71 participated in the final sample, a response rate of 20%.
Supervisory Working Alliance Inventory-Trainee. The SWAI-T (Efstation, Patton, & Kardash, 1990) was developed to measure supervisee perceptions of the clinical supervision relationship. It contains 19 items on two scales: Rapport, 13 items (e.g., I feel free to mention to my supervisor any troublesome feelings I might have about him/her); and Client Focus, 6 items (e.g., My supervisor encourages me to take time to understand what the client is saying and doing). Responses are reported on a 7-point Likert scale ranging from "almost never" (1) to "almost always" (7). Scores on the SWAI-T can range from 19 to 133 (Efstation et al., 1990). Several studies have demonstrated good reliability and validity data for the SWAI-T (e.g., Efstation et al. 1990; Patton, Brossart, Gehlert, Gold, & Jackson, 1992). For the current study, reliability data were derived using Cronbach's alpha, which yielded internal consistency reliabilities of. 97 for the overall SWAI-T, .88 for the Client Focus subscale, and .97 for the Rapport subscale. Inter-item correlations ranged from .35 to .71 for Client Focus and from .32 to .91 for Rapport. Reliability data for this study were consistent with results from other studies.
Minnesota Satisfaction Questionnaire-Short Form. The MSQ (Weiss, Dawis, England, & Lofquist, 1967) was designed to measure employee satisfaction with their jobs. It was used in this study to measure the work satisfaction supervisees experienced when they last received clinical supervision and to determine how work satisfaction was influenced by the quality of the SWA. The MSQ measures satisfaction with several aspects of work and the work environment. The short form is composed of 20 items, one from each scale of the MSQ long form, on three scales: Intrinsic Satisfaction, 12 items (e.g., The chance to do something that makes use of my abilities); Extrinsic Satisfaction, 6 items (e.g., The praise I get for doing a good job); and General Satisfaction, all 20 items (e.g., The working conditions).
The MSQ short form uses a Likert-type scale with five (weighted) responses ranging from "very dissatisfied" (1) to "very satisfied" (5). Scores on the short form can range from 20 to 100. Several studies have demonstrated good reliability and validity data for the MSQ (e.g., Albright, 1972; Anderson, Hohenshil, & Brown, 1984; Bolton, 1986; Brown, Hohenshil, & Brown, 1998; Decker & Borgen, 1993; Guion, 1978; Levinson, Fetchkan, & Hohenshil, 1988). For the current study, reliability data for the short form revealed internal consistency reliabilities of .92 for the General Satisfaction scale, .89 for the Intrinsic scale, and .88 for the Extrinsic scale. Reliability data for this study were consistent with results from other studies.
Occupational Stress Inventory-Revised. The OSI-R (Osipow, 1998) is a measure of occupational adjustment on three dimensions: occupational stress, psychological strain, and coping resources. It is composed of three questionnaires (Occupational Roles Questionnaire, Personal Strain Questionnaire, and Personal Resources Questionnaire) with a total of 140 items. The instrument yields scores on 14 different scales. The OSI-R uses a 5-point Likert scale with responses ranging from "rarely or never" (1) to "most of the time" (5). Scores are determined by summing columns, which constitute raw scores for each scale. The Occupational Roles Questionnaire (ORQ), consisting of 60 items (e.g., At work I am expected to do too many different tasks in too little time), measures occupational stress through six scale descriptors (Role Overload, Role Insufficiency, Role Ambiguity, Role Boundary, Responsibility, and Physical Environment).
Only the ORQ was used in the current study because it measures components of work-related stress. All six descriptors are used to measure overall work-related stress as tested in hypothesis two. Role Overload (views workload as increasing, unreasonable, and unsupported by needed resources); Role Ambiguity (experiences conflicting demands from supervisors; unclear how they will be evaluated); and Role Boundary (feeling caught between conflicting supervisory demands and factions) were used to test hypothesis three due to their focus on work-related stress associated with supervision. Scores on the ORQ can range from 60 to 300. Several studies document good reliability and validity data for the OSI-R (see Osipow, 1998). For the current study, reliability data yielded an internal reliability coefficient of .93 for the ORQ subscale, .85 for Role Overload, .86 for Role Ambiguity, .84 for Role Boundary, .86 for Role Insufficiency, .78 for Responsibility, .62 for Physical Environment. Reliability data for this study were consistent with the results of other studies.
Demographic questionnaire. The purpose of the author-created demographic questionnaire was to describe the participants and to screen out those who did not meet the study criteria. It contained 10 items, such as "What is your highest level of education?" "What type of counseling setting do you work in (e.g., mental health agency, private nonprofit, hospital, private practice)?" "Are you currently receiving individual clinical supervision (supervision focusing on therapeutic issues, processes, and outcomes)? Do not include staff meetings, administrative supervision (supervision focusing on non-therapeutic issues, organizational issues, and evaluation), or peer supervision (mutual collaboration with peers on therapeutic issues)." "On average, how many clients do you work with per week (based on a 40-hour week)?"
Tabachnick and Fidell (2007) indicated that the number of participants needed to test the hypothesis using canonical correlation was approximately 10 cases for each variable, depending on the reliability of the variables. They added that "if the reliability is very high..., then a much lower ratio of cases to variables is acceptable" (p. 570). Given that the reliabilities for the variables used to test the hypothesis are well above .80, fewer cases are acceptable and the sample size of 71 is appropriate.
Missing data analysis. Before testing the hypotheses, the data were inspected to identify miscoded data, data points that were outside the theoretical range of scores established by each instrument, compliance with statistical assumptions, and any patterns of missing data. Several cases had missing data points. After careful review of these cases, it was determined that the pattern of missing data seemed to be the result of values missing at random. After the analysis was completed, a regression imputation method was considered the best option for replacing missing values. According to Hair, Anderson, Tatham, and Black (1998), this imputation method offers the best option when "moderate levels of widely scattered missing data are present and for which the relationships between variables are sufficiently established so that the researcher is confident that using this method will not impact the generalizability of the results" (p. 54). Following the imputation procedure, data were reevaluated to ensure that assumptions were met. Several variables required transformations to meet the assumption of normality. These were client focus, rapport, intrinsic satisfaction, role ambiguity, and role boundary. No within-cells multivariate outliers were identified at p < .001. The nonresponse rate for the SWAI-T was 5.25%, the average nonresponse rate for the MSQ was 15.63%, and the average nonresponse rate for the ORQ was 6.27%. The overall individual item nonresponse rate for the survey (n = 71) was 1.3%.
Results of the analysis for hypothesis one, which stated that as supervisees' perceptions of the quality of the SWA increase, their perceptions of work satisfaction will also increase, revealed that with an alpha level of .05, the Pearson correlation was statistically significant, r = .60, t = 6.23, p < .001. The shared variance between supervisees' perceptions of work satisfaction and perceptions of the quality of the SWA was 36%. As a result, when supervisees perceived a positive SWA, they also perceived greater work satisfaction.
In testing the second hypothesis, which stated that as supervisees' perceptions of the quality of the SWA increase, their perceptions of work-related stress will decrease, with an alpha level of .05, the Pearson correlation was statistically significant, r = -.56, t = -5.59, p < .001. The amount of shared variance between supervisees' perceptions of work-related stress and their perceptions of the quality of the SWA was 31%. As a result, when supervisees perceived the SWA as positive, they also perceived less work-related stress.
For hypothesis three (supervisees' work setting, caseload, perceptions of rapport with supervisor, and perceptions of supervisor focus on the client will correlate with supervisee intrinsic and extrinsic work satisfaction, role ambiguity, role boundary, and role overload), canonical correlation was performed between a set of supervision/setting variables and a set of work satisfaction/work stress variables. Higher numbers for rapport and client focus indicated that counselors' perceptions of their supervisory working relationship was strong. Higher scores on intrinsic and extrinsic satisfaction reflected that counselors perceived a sense of internal satisfaction and a sense of satisfaction derived from their work environment. Lower scores on role boundary, role ambiguity, and role overload reflect less work-related stress.
The first canonical correlation was .68 (46% overlapping variance); the second was .36 (13% overlapping variance); the third was .27 (7% overlapping variance); and the fourth was .12 (2% overlapping variance). With all four canonical correlations included, F (20, 206.6) = 3.04, p < .001. With the first canonical correlation removed, the other three were not statistically significant. Only the first pair of canonical variates accounted for the significant relation between the two sets of variables.
Data on the first pair of canonical variates are highlighted in Table 1. Based on a cutoff correlation of .3, variables in the supervision/setting set that were correlated with the first canonical variate were counseling setting, client focus in supervision, and rapport with supervisor. Work satisfaction/stress variables that correlated with the first canonical variate included intrinsic and extrinsic satisfaction, role ambiguity, role boundary, and role overload.
The first pair of canonical variates indicate that counseling setting (.54), client focus in supervision (.87), and rapport with supervisor (.95) are associated with intrinsic (.74) and extrinsic (.92) satisfaction, role ambiguity (-.87), role boundary (-.61), and role overload (-.50). These variates suggest that the setting in which the counselor is receiving supervision and the counselor's perception of the supervisor's focus on client issues and perceived rapport with the supervisor are highly positively correlated with the counselor's perception of what is personally satisfying and fulfilling at work and with feeling satisfied with the work environment. The setting and counselor perceptions of the supervisor's client focus and rapport were negatively associated with counselor perceptions of role boundary, role ambiguity, and role overload. The association between number of clients per week and perceptions of supervision, satisfaction, and work-related stress variables was well below the cutoff correlation.
SWA and Work Satisfaction
Supervisees who were more satisfied with the clinical supervision relationship were more satisfied with their work. This finding is consistent with results from other studies that investigated the relationship between clinical supervision and work satisfaction and found that a positive SWA is associated with higher work satisfaction (Hyrkas, 2005; Schroffel, 1999). Participant responses indicate that clinical supervision may be one factor in how they perceive their satisfaction with work. The amount of variance explained by the SWA on work satisfaction showed that supervision is an important component in supervisee perceptions of work satisfaction.
SIVA and Work-related Stress
The results provided evidence that supervisee perceptions of their SWA influenced how they respond to work-related stress. Supervisees experienced decreased work-related stress when they perceived a strong working alliance with their clinical supervisor. This outcome is supported by several studies that found when supervisees perceived the supervisory relationship as supportive, they tended to experience less stress and burnout (Coady, Kent, & Davis, 1990; Himle, Jayaratne, & Thyness, 1989). In addition, results of this study support the position of Rabin, Feldman, and Kaplan (1999) that how supervisees respond to work-related stress is contingent on how they perceive the clinical supervisory alliance.
Because a significant majority of participants reported a strong SWA, there were not enough data to analyze the responses of those who perceived a weak SWA. Due to the small number of participants reporting low scores on the SWAI-T (n = 7), no statistical analysis was conducted to compare high versus low SWAs. Visual inspection of the data for these seven participants revealed that their scores measuring quality of the SWA (SWAI-T) were approximately two standard deviations below the overall sample mean of the SWA measure, while their scores measuring work-related stress (ORQ) were two standard deviations above. This preliminary finding indicates that dissatisfaction with clinical supervision might result in higher work-related stress.
SWA, Setting, Caseload, Work Satisfaction, and Work-related Stress
The first two hypotheses demonstrated how counselors' perceptions of the SWA as a whole influenced their perceptions of work satisfaction and work-related stress. Hypothesis three attempted to better understand how specific variables correlated to each other. The first pair of canonical variates indicated that counselor setting (.54), counselor perceptions of the supervisor's focus on the client (.87), and their perception of the rapport they had with their supervisor were positively associated with personal satisfaction in their work with clients (.74) and satisfaction with their work environment (.92). Counseling setting, supervisor client focus, and rapport, however, were inversely related to counselor perceptions of conflicting supervisory demands or lack of clarity around lines of authority (role boundary -.61), uncertainty about what is expected and how they will be evaluated (role ambiguity -.87), and being unsupported by the supervisor, unreasonable expectations by the supervisor, and an increasing workload (role overload -.50). Although it was hypothesized that counselor caseload would reflect a higher correlation with the counselor work satisfaction/stress set, it yielded a very low correlation (.07), indicating that counselor perceptions of the number of clients they work with each week was not significantly associated with their perceived personal or work satisfaction or work-related stress.
These results suggest that relations between SWA, work satisfaction, and work-related stress for professional counselors may help explain the specific function clinical supervision serves for supervisees. Supervision provides experiences for supervisees to address both work variables, which means that they perceive clinical supervision as not only contributing to their professional development and work satisfaction but also as moderating work-related stressors. Since this study intended to explore relationships between key variables, other variables such as supervisee self-efficacy, supervisee self-reliance, development of coping strategies, degree of autonomy, quality of graduate level training, life and career experiences, length of time in the workforce, and individual qualities and traits may account for some of the variance in work satisfaction and work-related stress.
Limitations of the Study
Since participant response rate was lower than desired, the findings may not fully represent the larger population of AMHCA members. Low response rates were anticipated because mental health counselors were surveyed during the summer, as well as because of concerns about work demands and other time commitments.
Survey data relied on participant self-reporting. A recent positive or negative event with a clinical supervisor may have distorted how participants evaluated their overall experiences with supervision. Individuals who did not respond may have disregarded the survey because of frustrations with their clinical supervisor. Individuals who did not have access to the Internet may have chosen not to participate, despite opportunities to receive a paper copy of the survey. Another concern with the online survey was that it made no provision for counterbalancing instruments. Finally, efforts were made to ensure a racially/ethnically diverse sample, but participant selection options were limited based on the need for a random sample of individuals meeting the study criteria.
IMPLICATIONS FOR PRACTICE, TRAINING, AND RESEARCH
Outcomes from this study can help clinical supervisors better understand their role and the importance of the relationship they establish with their supervisees. Supervisors need to be aware that they can have a significant influence on how supervisees perceive their workplace as well as their ability to adapt to changing demands and expectations. Another implication for practice is the need for supervisors to convey the importance of clinical supervision within mental health agencies and the benefits for supervisees associated with this relationship.
When clinical supervisors are able to create a strong SWA, supervisees perceive the supervisory relationship as beneficial, possibly resulting in a sense of personal satisfaction. McMahon and Patton (2000) reported that when supervisees view the supervisory relationship as helpful it provides secondary benefits, such as emotional balance and a sense of well being. Finally, supervisees can gain a better understanding of clinical supervision outcomes and what impact they will have on their work, professional growth and development, and development of skills to moderate work-related stress.
Those educating counselors need to convey which supervision attributes can lead to a positive SWA and which may negatively affect the quality of the SWA. Training should help students identify what constitutes intrinsic and extrinsic satisfaction while preparing them to deal with work-related stressors as a professional counselor. One concern is that many supervisors in professional settings may have limited or no training in clinical supervision with professional counselors, which could affect the quality of the SWA. In addition, the findings from this study may give counselor educators and supervisors a better understanding of the role of the SWA in professional settings, especially since most of the literature has considered this construct only in academic settings.
Implications for supervisees include helping them find the supervision that best meets their professional needs. They can also gain a better understanding of the SWA and the benefits associated with establishing this alliance. Finally, they can make better use of their administrative supervision experiences by incorporating components of the SWA into these relationships.
This study gives researchers an opportunity to understand the relations between clinical supervision, work satisfaction, work-related stress, and other setting variables. Findings from this study extend the research on this topic, creating much needed data on supervisees working in mental health settings. Since this was an exploratory study, future research should attempt to understand what factors (e.g., personal attributes) contribute to the remaining variance between study variables. Further, a criterion-based stratified random sample might have improved representation of diverse groups. Another research implication is understanding whether administrative supervision has the same influence on work satisfaction and on work-related stress, especially with counseling professionals who typically do not receive clinical supervision (e.g., those in schools).
The results of the current study show relationships between supervisee perceptions of the quality of the SWA, work satisfaction, and work-related stress. Consistent with studies in other disciplines exploring relations among these constructs, the current study demonstrated that supervision plays an important role in how supervisees perceive their work. The importance of the current study is not just with the explicit findings, especially given the fact that no study has explored the relations between these variables in the counseling field, but also with the realization that implicit factors are likely significant components of work satisfaction and work-related stress. These results highlight the importance of the relations between these constructs and provide a foundation on which to apply other constructs so as to better understand the role of supervision.
Albright, L. E. (1972). A review of the Minnesota Satisfaction Questionnaire. In O. K. Burns (Ed.), The seventh mental measurements yearbook, Vol. II (pp. 1492-1494). Highland Park, NJ: Gryphon Press.
Anderson, W. T., Hohenshil, T. H., & Brown, D. T. (1984). Job satisfaction among practicing school psychologists: A national study. School Psychology Review, 13, 225-230.
Bernard, J. M., & Goodyear, R. K. (2004). Fundamentals of clinical supervision (3rd ed.). Boston: Pearson.
Bolton, B. (1986). A review of the Minnesota Satisfaction Questionnaire. In D. J. Keyser & R. C. Sweetland (Eds.), Test critiques, Vol. V (pp. 255-265). Kansas City, MO: Test Corporation of America.
Bordin, E. S. (1983). A working alliance based model of supervision. The Counseling Psychologist, 11, 35-42.
Brown, M. B., Hohenshil, T. H., & Brown, D. T. (1998). Job satisfaction of school psychologists in the United States. School Psychology International, 19, 79-89.
Carifio, M. S., & Hess, A. K. (1987). Who is the ideal supervisor? Professional Psychology." Research and Practice, 18, 244-250.
Coady, C. A., Kent, V. D., & Davis, P. W. (1990). Burnout among social workers working with patients with cystic fibrosis. Health and Social Work, 15, 116-124.
Decker, P. J., & Borgen, F. H. (1993). Dimensions of work appraisal: Stress, strain, coping, job satisfaction, and negative affectivity. Journal of Counseling Psychology, 40, 470-478.
Dye, H. A., & Borders, L. D. (1990). Counseling supervisors: Standards for preparation and practice. Journal of Counseling & Development, 69, 27-29.
Efstation, J. F., Patton, M. J., & Kardash, C. M. (1990). Measuring the working alliance in supervisee supervision. Journal of Counseling Psychology, 37, 322-329.
Eklund, M., & Hallberg, I. R. (2000). Factors influencing job satisfaction among Swedish occupational therapists in psychiatric care. Scandinavian Journal of Caring Sciences, 14, 162-171.
Feltham, C. (2000). Counselling supervision: Baselines, problems and possibilities. In B. Lawton & C. Feltham (Eds.), Taking supervision forward: Enquiries and trends in counselling and psychotherapy (pp. 5-24). London: Sage.
Fincham, R., & Rhodes, P. (2005). Organizational behaviour (4th ed.). London: Oxford University Press. Retrieved from http://www.oup.com/uk/booksites/content/0199253978/ student/glossary/glossary.htm#J Fleming, J., & Benedek, T. F. (1966). Psychoanalytic supervision: A method of clinical teaching. New York: Grune & Stratton.
Gabriel, P., & Liimatainen, M.-R. (2000). Mental health in the workplace. Retrieved from http://www-ilo-mirror.cornell.edu/public/english/employment/skills/ disability /download/execsums.pdf
Getz, H. G. (1999). Assessment of clinical supervisor competencies. Journal of Counseling & Development, 77, 491-497.
Gray, L. A., Ladany, N., Walker, J. A., & Antis, J. R. (2001). Psychotherapy trainees' experience of counterproductive events in supervision. Journal of Counseling Psychology, 48, 371-383.
Green, S. B. (1991). How many subjects does it take to do a regression analysis? Multivariate Behavioral Research, 26, 499-510.
Guion, R. M. (1978). A review of the Minnesota Satisfaction Questionnaire. In O. K. Buros (Ed.), The eighth mental measurements yearbook, Vol. H (pp. 1679-1680). Highland Park, NJ: Gryphon Press.
Hair, J. F., Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Upper Saddle River, NJ: Prentice Hall.
Himle, D. P., Jayaratne, S., & Thyness, P. (1989). The effects of emotional support on burnout, work stress and mental health among Norwegian and American social workers. Journal of Social Service Research, 13, 27-45.
Holloway, E. L., & Neufeldt, S. A. (1995). Supervision: Its contribution to treatment efficacy. Journal of Consulting and Clinical Psychology, 63, 207-213.
Hyrkas, K. (2005). Clinical supervision, burnout, and job satisfaction among mental health and psychiatric nurses in Finland. Issues in Mental Health Nursing. 26, 531-556.
Kavanagh, D. J., Spence, S. H., Strong, J., Wilson, J., Sturk, H., & Crow, N. (2003). Supervision practices in allied mental health: Relationships of supervision characteristics to perceived impact and job satisfaction. Mental Health Services Research, 5, 187-195.
Ladany, N., Ellis, M. V., & Friedlander, M. L. (1999). The supervisory working alliance, trainee self-efficacy, and satisfaction. Journal of Counseling & Development, 77, 447-455.
Levinson, E. M., Fetchkan, R., & Hohenshil, T. H. (1988). Job satisfaction among practicing school psychologists revisited. School Psychology Review, 17, 101-112.
Magnuson, S., Norem, K., & Wilcoxon, S. A. (2000). Clinical supervision of prelicensed supervisees: Recommendations for consideration and practice. Journal of Mental Health Counseling, 22, 176-188.
McMahon, M., & Patton, W. (2000). Conversations on clinical supervision: Benefits perceived by school counsellors. British Journal of Guidance & Counselling, 28, 339-352.
Nelson, M. L., & Friedlander, M. L. (2001). A close look at conflictual supervisory relationships: The trainee's perspective. Journal of Counseling Psychology, 48, 384-395.
O'Driscoll, M. P., & Beehr, T. A. (1994). Supervisor behaviors, role stressors and uncertainty as predictors of personal outcomes for subordinates. Journal of Organizational Behavior, 15, 141-155.
Osipow, S. H. (1998). Occupational Stress Inventory Revised Edition (OSI-R) professional manual. Lutz, FL: Psychological Assessment Resources, Inc.
Patton, M. J., Brossart, D. F., Gehlert, K. M., Gold, P. B., & Jackson, A. P. (1992, August). The Supervisory Working Alliance Inventory: A validity study. Paper presented at the annual meeting of the American Psychological Association, Washington, DC. (ERIC Document Reproduction Service No. ED360358).
Rabin, S., Feldman, D., & Kaplan, Z. (1999). Stress and intervention strategies in mental health professionals. British Journal of Medical Psychology, 72, 159-169.
Ramos-Sanchez, L., Esnil, E., Goodwin, A., Riggs, S., Touster, L. O., Wright, L. K., et al. (2002). Negative supervisory events: Effects on supervision satisfaction and supervisory alliance. Professional Psychology: Research and Practice, 33, 197-202.
Ronnestad, M. H., & Skovholt, T. M. (1993). Supervision of beginning and advanced graduate students of counseling and psychotherapy. Journal of Counseling & Development, 71, 396-405.
Schroffel, A. (1999). How does clinical supervision affect job satisfaction? The Clinical Supervisr, 18, 91-105.
Seltzer, J., & Numerof, R. E. (1988). Supervisory leadership and subordinate burnout. Academy of Management Journal, 31, 439-446.
Spence, S. H., Wilson, J., Kavanagh, D., Strong, J., & Worrall, L. (2001). Clinical supervision in four mental health professions: A review of evidence. Behaviour Change, 18, 135-155.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston: Allyn and Bacon.
Traut, C. A., Larsen, R., & Feimer, S. H. (2000). Hanging on or fading out? Job satisfaction and the long-term worker. Public Personnel Management, 29, 343-351.
Uys, L. R., Minnaar, A., Simpson, B., & Reid, S. (2005). The effect of two models of supervision on selected outcomes. Journal of Nursing Scholarship, 37, 282-288.
Vaux, A. (1996). Conducting mail surveys. In F.T.L. Leong and J.T. Austin (Eds.), The psychology research handbook, (pp. 127-138). Thousand Oaks, CA: Sage.
Weiss, D. J., Dawis, R. V., England, G. W., & Lofquist, L. H. (1967). Manual for the Minnesota Satisfaction Questionnaire. Minneapolis: University of Minnesota.
Worthen, V., & McNeill, B. W. (1996). A phenomenological investigation of "good" supervision events. Journal of Counseling Psychology, 43, 25-34.
Worthington, E. L., Jr. (1984). Empirical investigation of supervision of supervisees as they gain experience. Journal of Counseling Psychology, 31, 63-75.
William R. Sterner is affiliated with Youngstown State University. Correspondence concerning this article should be addressed to One University Plaza, Youngstown, OH 44555. E-mail: firstname.lastname@example.org. The author would like to thank JoLynn Carney and Richard Hazier for their feedback and support.
Table 1. Summary Canonical Correlation Data Between Supervision/Setting and Work Satisfaction/Stress Variables and their Corresponding Variates First Canonical Variate Correlation Coefficient Supervision/setting set Counseling setting .54 .27 Clients/week .07 -.04 Client focus in supervision (squared) .87 .24 Rapport with supervisor (squared) .95 .69 Percent of variance .49 Redundancy .23 Work satisfaction/stress set Intrinsic satisfaction (squared) .74 .17 Extrinsic satisfaction .92 .58 Role ambiguity (logarithm) -.87 -.46 Role boundary (logarithm) -.61 .21 Role overload -.50 -.13 Percent of variance .26 Redundancy .55 Canonical correlation .68
|Printer friendly Cite/link Email Feedback|
|Author:||Sterner, William R.|
|Publication:||Journal of Mental Health Counseling|
|Date:||Jul 1, 2009|
|Previous Article:||Supervising trainees who counsel clients with borderline personality characteristics: implications for training and practice.|
|Next Article:||The experiences of transgendered persons in psychotherapy: voices and recommendations.|