The Supervisory Alliance: A Half Century of Theory, Practice, and Research in Critical Perspective
Abstract
Introduction
Two Generative Visions of the Alliance in Supervision
Freud’S Contributions and Beyond
The Fleming and Benedek Perspective on the Learning Alliance
The structure of both the analytic and supervisory situations is determined primarily by the goal which each participant expects to accomplish in their work together. These ultimate expectations, whether therapeutic or educational, orient the behavior of each member and guide their interactions through many vicissitudes. Expectations of giving and receiving help initiate a bond of trust and confidence between analyst and patient without which analytic work cannot proceed.… In supervision there exists the same necessity for acceptance of a mutually shared educational goal and the same need for confidence that the expectations of teacher and learner can be satisfied. The term learning alliance describes the essential characteristic of this relationship (pp. 52-53).
The Bordin Perspective on the Supervision Working Alliance
The Two Alliance Conceptualizations in Perspective
What Have We Learned from Supervision Alliance Research?
Context and Method
Study | Sample Characteristics | Measures/Assessment Used | Procedure | Analyses Used | Findings | Limitations |
---|---|---|---|---|---|---|
Bennett, BrintzenhofeSzoc, Mohr, & Saks (2008) | 72 MSW students in their field placements (26female, 5male; 80% White; mean age=32.5 years; demographics provided for only 30 individuals, not entire sample) | Working Alliance Inventory; Supervisory Styles Inventory; Relationship Scales Questionnaire; Relationship Structures Questionnaire (RSQ) | Survey packet containing questionnaires provided to trainees for completion | Multiple regression analyses | Supervision-specific attachments found to strongly predict supervision alliance perceptions; alliance rated least favorably by highly avoidant supervisees | Ex post facto design; focus on supervisees’ perspective only; self-report survey data;48% return rate |
Bhat & Davis (2007) | 119 supervisors (80 female, 39 male; 108 White, 10 African-American, 1 Latino; x age = 50.5 years; 90 master’s degree, 21 doctoral degree, 8 other) | White Racial Identity Attitude Scale or People of Color Racial Identity Attitude Scale; Perceptions of Supervisee Racial Identity for White or Perceptions of Supervisee Racial Identity for POC; Working Alliance Inventory-Supervisor Version; demographic questionnaire | Survey packet containing questionnaires provided to all supervisors for completion | ANOVA | Racial identity status of supervisor and supervisee related to strength of supervision alliance (where identity status high for both, alliance rated more favorably; converse where identity status low for both) | Ex post facto design; self-report survey data; supervisors provided all ratings; no supervisee ratings included |
Bilodeau & Lecomte (2010) | 31 supervisees (26 female, 5 male, x age=31.9 years), 13 supervisors (9 female, 4 male, x age=37.5); each student received five supervision sessions | Supervisory Working Alliance Inventory-T (SWAI-T; supervisee version); Supervisory Working Alliance Inventory-S (SWAI-S; supervisor version); Internalized Shame Scale (ISS) | At study’s outset, supervisees completed ISS and demographic questionnaire; supervisees again completed ISS after final supervision session; after each supervision session, supervisees and supervisors, respectively, completed SWAI-T and SWAI-S | Repeated measures ANOVAs | Supervisees rated alliance more favorably across all sessions compared to supervisors; supervisee shame-proneness had no effect on alliance ratings of either supervisee or supervisor | Only four individuals in high shame group; some supervisees shared same supervisor; one university sample; self-report survey data |
Bilodeau & Lecomte (2012) | 43 supervisees (36 female, 7 male, x age=30.1 years); each student received five supervision sessions | Supervisory Working Alliance Inventory-T (SWAI-T; supervisee version); Internalized Shame Scale (ISS); Session Evaluation Questionnaire (SEQ) | At study’s outset, supervisees completed ISS and demographic questionnaire; supervisees again completed ISS after final supervision session; after each supervision session, supervisees completed SWAI-T and SEQ | Independent samples t tests; repeated measures analysis of covariance | Shame proneness affected supervisees’ alliance ratings across five sessions; highly shame prone supervisees rated alliance more favorably after first session but not after fifth | One university sample; self-report survey data |
Bucky, Marques, Daly, Alley, & Karp (2010) | 87 clinical psychology doctoral students(74 female, 13 male; 72 White, 1 African-American, 5 Latino, 5 Asian-American; 64 students 20 to 30 years in age, remainder over 30; 26, 36, 16, and 9 students, respectively, identified as being in their second, third, fourth, and fifth year (or beyond) of training | Supervisee Evaluation of Supervisor Questionnaire | Students who had completed practicum or internship the previous year filled out online questionnaire | Frequencies determined | “The results of the study support the… importance of the supervisory relationship or working alliance…”(Bucky et al., 2010, p. 159). | Ex post facto design; frequencies only; self-report survey data from supervisees only |
Burke, Goodyear, & Guzzardo (1998) | 10 supervisor-supervisee dyads (5 female, 5 male supervisors; 7 female, 3 male supervisees) from mental health center or counseling center settings; supervisees all pre-doctoral psychology interns | Working Alliance Inventory-Modified (WAI-M; parallel versions for supervisor and supervisee); Working Alliance Inventory-Modified Short Form (WAI-MSF; parallel versions for supervisor and supervisee); Session Evaluation Questionnaire (SEQ); outcome rating | Supervisor and supervisee completed their respective versions of WAI-M after first and last of ten audiotaped sessions; for sessions 2-9, supervisor and supervisee completed their respective versions of WAI-MSF and SEQ-M | ANOVA, t tests, audiotape coding | Supervisor ratings of alliance considered more stable and consistent than supervisee ratings; supervisee experience level appeared to affect alliance weakening and repair events | Still the only study to examine alliance rupture/repair events in process; multiple case study design |
Chen & Bernstein (2000)1 | 1 high-alliance and 1 low-alliance dyad selected from 10 supervision dyads overall (9 White female and 1 White male supervisees in first counseling practicum; x age=36; 6 White female and 1 white male doctoral student supervisors [in supervision course], x age = 33) | Supervisory Styles Inventory; Critical Incidents Questionnaire; Supervisory Working Alliance Inventory (Trainee and Supervisor Versions); Complementarity Indices | Questionnaires completed by supervisor/supervisee dyads prior to first supervision meetingand during the three supervision meetings held | Chi-square tests and correlational data | Higher degree of complementary interaction found in “high-rated alliance” dyad as opposed to “low-rated alliance” dyad | Relatively small sample pool Limited evidence to support validity of Complementarity Indices in supervision research Age and experience differences in high-alliance versus low-alliance dyads |
Cooper & Ng (2009) | 64 supervisees (61 female, 3 male; 64% White, 28% African-Americans, 3% Latino, 2% Asian-Americans, 2% American Indian, 1% other; x age=33.8 years;) completing internship in community/agency-based sites; 64 supervisors (55 female, 9 male; 86% White, 9% African-Americans, 3% Latino, 2% Asian-Americans,; x age=46.7 years;) providing supervision to supervisee sample | Trait Emotional Intelligence Questionnaire-Short Form; Working Alliance Inventory-Modified (supervisee and supervisor versions) | Questionnaires packets provided to supervisor/supervisee dyads for completion | Hierarchical multiple regression analysis | Higher levels of emotional intelligence related to more favorable perceptions of supervisory alliance for both supervisee and supervisor, but no interaction effects found | Ex post facto design; self-report survey data; 36% and 26% response rate, respectively, for supervisees and supervisors; predominantly female convenience sample |
Culbreth & Borders (1999) | 370 supervised substance abuse counselors (202 female, 122 male, 36 no response; 282 White, 65 African-American, 13others; x age = 41 years; recovery status=235 non-recovering, 123 recovering, 2 no response; 40% held graduate degree) | Supervisory Styles Inventory; Supervisor Rating Form-SV; Working Alliance Inventory (where “supervision” substituted for “counseling” in all items); Barrett-Lennard Relationship Inventory; supervision satisfaction items | Survey packet containing questionnaires provided to all substance abuse counselors for completion | Series of MANOVAs | Counselor or supervisor recovery status had no effect on supervision relationship variables; supervision alliance bond rated highly by counselors | Ex post facto design; self-report survey data; supervisee perspective only; one state survey |
Davidson (2011) | 184 social work students enrolled in field placement (164 female, 20 male; 142 White/Caucasian; x age=29 years; respndents evenly split between foundation and advanced year) | Working Alliance Inventory-Trainee Version; Supervisor Self-Disclosure Index | Online questionnaires completed by all participating students | Correlations; multiple regression | More frequent supervisor selfdisclosures related to stronger supervisory alliance; certain types of supervisor disclosures (e.g., normalizing) most strongly associated with positive alliance ratings | Ex post facto design; self-report survey data; supervisee perspective only; 23% response rate |
Deal, Bennett, Mohr, & Hwang (2011) | 100 field instructors (FIs)/supervisors: (93 female, 7 male; 70% White, 21% African-Americans, 6% Latino/Hispanic, 1% Pacific Islander, 2% biracial or other; x age =41.2 years;) 48 received training in Developmental-Relational Approach to Field Supervision, 52 did not (control group); 64 supervisees (59 female, 5 male; 80% White, 9.4% African-Americans, 3% Latino/Hispanic, 3% Asian-Americans, 5% biracial or other; x age =31.1 years) enrolled in year-long practicum with with supervisors | Time 1 supervisor assessments: demographic questionnaire, Working Alliance Inventory-Supervisor Version (WAI-SVOR), Competency-Based Evaluation (CBE); time 2 and 3 supervisor assessments WAI-SVOR and CBE; Time 1 supervisee assessments: demographic questionnaire, Relationship Scales Questionnaire (RSQ), Positive and Negative Affect Scale (PANAS), Working Alliance Inventory-Supervisee Version WAI-SVEE), Supervisee Levels Questionnaire-Revised (SLQ-R); time 2 and 3 supervisee assessments WAI-SVEE, PANAS, and SLQ-R | Supervisors receiving training completed assessments time 1 at beginning of first training session, time 2 at end of last training session, and time 3 at end of academic year; control group supervisors received mailed questionnaires at same times; supervisees also received their questionnaires at same times | Hierarchical linear modeling | Training led to more favorable perceptions of supervisory alliance on the part of supervisors but not supervisees; “approach appeared to help FIs improve the supervisory alliance from their perspective” (Deal et al., 2011, p. 722); supervisee attachment style had no moderating effect | Intervention study; randomized controlled trial; increased likelihood of Type I errors; lack of statistical power possible with some student sample results; self-report data |
Dickson, Moberly, Marshall, & Reilly (2011) | 259 clinical psychology trainees (229 female, 25 male, 5 unspecified; mean age=28.6 years; 71% White British; 104 1st year trainees, 81 2nd year, 70 3rd year, 14th year, and 3 unspecified | Working Alliance Inventory; Measure of Parental Style; Reciprocal Attachment Questionnaire; Relationship Questionnaire | Online data collected from trainees across 28 clinical doctoral programs in Britain | MANOVAs; structural equation modeling | Trainees’ ratings of supervisory working alliance related to their perceptions of supervisor attachment style; replication of Riggs & Bretz (2006) | Ex post facto design; focus on supervisee perceptions only; self-report survey data; return rate not reported |
Efstation, Patton, & Kardash (1990)1 | 185 supervisors (69 female, 114 male, 2 unspecified; x age= 42 years; 122 clinical and 45 counseling psychologists, 12 other; outpatient clinics 33%, university/college counseling centers 31%; VAs 13%, psychiatric hospitals 21%); 178 supervisees (103 female, 73 male; 2 not specified; x age = 30 years) | Supervisory Working Alliance Inventory; Supervisory Style Inventory; Self-Efficacy Inventory | Questionnaire packets mailed out | Principal components factor analysis; hierarchical regression | Positive relationship found between supervisee self-efficacy expectations and perceptions of supervisory working alliance | Advanced practicum and intern-level students only |
Gatmon, Jackson, Koshkarian, Martos-Perry, Molina, Patel, & Rodolfa (2001)1 | 289 predoctoral psychology interns (203 female, 86 male; 73.4% European-American, 6.6% African-American, 5.9% Asian-American, 5.2% Chicano/Latino, 8% other) | Working Alliance Inventory; Supervision Questionnaire—Revised; Cultural variables questions | Questionnaire packets distributed | Frequency and correlational analyses, ANOVAs and MANOVAs | Positive relationship found between discussions of cultural variables in supervision and supervisees’ reported satisfaction with supervision and supervisory working alliance | Ex post facto design; self-report data; focused exclusively on perceptions of supervisees |
Gnilka, Chang, & Dew (2012) | 232 supervisees (200 female, 30 male, 2 transgender; 78% White, 11% African-American, 3% Latino, 3% Asian/Pacific Islander, 5% biracial; x age=32.8 years) participating in practicum/internship across varied settings | Demographic sheet; Working Alliance Inventory-Short Form (WAI-S); Supervisory Working Alliance Inventory-Trainee Version (SWAI-TV); Perceived Stress Scale (PSS); Coping Resources Inventory for Stress-Short Form (CRIS-SF) | Questionnaires answered via web link by supervisees | Stepwise multiple regression analyses | Supervisee stress (where rated more highly) found to negatively impact perceptions of the supervisory alliance; coping resources (where rated more highly) found to positively impact supervisory alliance perceptions | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Gunn & Pistole (2012) | 480 supervisees (393 female, 80 male, 83% Caucasian, 3,5% African-American, 2% Asian-American; x age=30 years; from master’s and doctoral counseling and clinical programs | Demographis questionnaire; Experiences in Supervision Scale; Supervisory Working Alliance Inventory-Trainee Version; Disclosure in Supervision Scale | Questionnaires distributed to graduate students via the Web | Structural equation modeling | Higher supervisory alliance ratings and higher disclosure in supervision related to higher supervisee attachment security to supervisor | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Ladany, Brittan-Powell, & Pannu (1997) | 105 supervisees (81 female, 23 male; 1 unspecified; 71% White, 11% African-American, 5% Asian-American, 11% Latino, 3% other; x age = 29.9 years; clinical psychology 17%; counseling psychology/counselor education 71%;doctoral 43%, master’s 50%; university/college counseling center 38%, mental health center 22%, schools 27%) | Cultural Identity Attitude Scale; White Racial Identity Attitude Scale; Perceptions of Supervisor Racial Identity; Working Alliance Inventory-Trainee Version; Cross-Cultural Counseling Inventory—Revised | Questionnaire packets distributed to graduate training programs | Factorial MANOVA | Expected relations found between supervisees’ perceptions of racial identity interaction and supervisory working alliance | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Ladany, Ellis, & Friedlander (1999)1 | 107 supervisees (72 female, 35 male; 86% White, 7% African-Americans, 3% Latino, 2% Asian-Americans, 3% unspecified; x age =29.9 years; clinical psychology 36%, counseling psychology/counselor education 59%; doctoral 71%, master’s 29%; university/college counseling center 40%, mental health center 25%, VA 22%) | Working Alliance Inventory-Trainee Version; Self-Efficacy Inventory; Trainee Personal Reaction Scale—Revised | Questionnaire packets distributed | Multivariate multiple regression analysis | Emotional bond component of alliance significantly related to supervisee satisfaction with supervision | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Ladany & Friedlander (1995)1 | 123 supervisees (81 female, 42 male; 85% White, 8% African-American, 2.4% Latino, 1.6 Asian-American, 2.4% unspecified; x age = 30 years; 54% counseling psychology, 37% clinical psychology; 68% doctoral 27% masters; university/college counseling center 41%, mental health center 23%, VA 20%) | Working Alliance Inventory-Trainee Version; Role Conflict and Role Ambiguity Inventory | Questionnaire packets distributed | Multivariate multiple regression analysis | Expected relations found between supervisory working alliance and supervisees’ perceptions of role conflict and role ambiguity | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees; advanced sample of supervisees |
Ladany & Lehrman-Waterman (1999)1 | 105 supervisees (82 female, 23 male; 84 White, 12 African-American, 5 Hispanic, 1 unspecified; x age = 30.4 years; clinical psychology 30%, counseling psychology/counselor education 67%) | Supervisor Self-Disclosure Questionnaire; Supervisor Self-Disclosure Index; Supervisory Styles Inventory; Working Alliance Inventory-Trainee Version | Questionnaire packets distributed | Univariate and multivariate regression analyses | Positive relationship found between supervisor selfdisclosure frequency and supervisory working alliance components (goals, tasks, and bond) | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisee; used supervisee recall |
Ladany, Lehrman-Waterman, Molinaro, & Wolgast (1999)1 | 151 supervisees (114 females, 36 males, 1 unspecified; 121 White, 12 African-American, 4 Latino, 1 Native American, 4 unspecified; x age = 31.5 years; clinical psychology 26%; counseling psychology 68%; 58% doctoral, 36% master’s; university/college counseling center 41%, mental health center 18%; schools 9%; prisons 4%; private practice 2%) | Supervisor Ethical Practice Questionnaire; Supervisor Ethical Behavior Scale; Working Alliance Inventory-Trainee Version; Supervisee Satisfaction Questionnaire | Questionnaire packets distributed to graduate training programs and training sites | Multivariate multiple regression analysis | Expected relations found between supervisee perceptions of supervisors’ ethical behaviors and working alliance components | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Ladany, Mori, & Mehr (2013) | 128 supervisees (100 females, 27 males, 1 unspecified; 109 White, 5 African-American, 8 Latino/Hispanic, 3 Asian-American/Pacific Islander, 2 other, 1 unspecified; x age = 35.4 years; doctoral programs represented 58% clinical psychology, 29% counseling psychology, 5% school psychology, 9% other) | Supervisee evaluation of supervisor form; Working Alliance Inventory/Supervision-Short Form; Supervisory Styles Inventory; Supervisor Self-Disclosure Index; Trainee Disclosure Scale; Evaluation Process Within Supervision Inventory | Participants responded to questionnaires via web link; asked to reflect upon a “best supervisor” and a “worst supervisor” in responding | Series of multivariate analyses | Behaviors of best and worst supervisors identified; results supported supervision relationship as “foundational competency”, “important influence on supervisee learning”; concluded that effective supervisors foremost work toward developing a strong supervisory alliance | Primarily ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Ladany, Walker, & Melincoff (2001)1 | 137 supervisors (80 female, 55 male, 1 other; 119 White, 6 African-American, 3 Latino, 1 other; x age = 45 years; clinical psychology 18%; counseling psychology/counselor education 68%; 110 doctoral, 27 master’s; university/college counseling center 33%, mental health center 15%; academic 15%) | Supervisory Styles Inventory; Working Alliance Inventory-Supervisor Version; Supervisor Self-Disclosure Inventory | Questionnaire packets mailed | Multivariate multiple regression analysis | Positive relationship found between supervisory style and working alliance components | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisors; could not determine return rate |
Livni, Crowe, & Gonsalvez (2012) | 37 supervisees (22 female, 7 male, 8 no response; modal age=45; 16 nurses, 5 psychologists, 1 social worker, 3 case workers, 2 addictions counselors, 10 other/unidentified); 10 supervisors (5 female, 5 male; 5 psychologists, 2 managers, 1 nurse, 2 unspecified); all participants from Area Health Service in New South Wales, Australia | Demographic questionnaire; Supervisory Working Alliance Inventory (parallel forms for supervisor and supervisee); Supervision Evaluation Questionnaire; Maslach Burnout Inventory; Intrinsic Job satisfaction Scale; Scales of Psychological Well-Being; California Psychotherapy Alliance Scale-Group-Modified | Supervisors received either individual, group, or individual/group supervision training; supervisees randomly assigned to individual or group supervision conditions; supervisees completed measures at baseline, again right before the start of their supervision (with the trained supervisors), and six months later | Repeated measures within groups design | Stronger supervisory alliance correlated with greater perceived supervision effectiveness; stronger alliance negatively related to burnout and positively related to job satisfaction and well-being in individual supervision condition | Infrequency of supervision sessions; self-report survey data; focused exclusively on perceptions of supervisees; absence of independent control group |
Mena & Bailey (2007)1 | 51 supervisors (47 female, 4 male; 49 White, 1 African-American, 1 other); 80 workers (all female; 69 White, 7 African-American, 3 Hispanic, 1 other) | Supervisory Working Alliance Inventory (Supervisor and Worker Versions); Minnesota Satisfaction Questionnaire; Maslach Burnout Inventory | Questionnaire packets distributed | Hierarchical linear modeling | Positive relations between alliance rapport and workers’ job satisfaction; negative relations found between alliance rapport and workers’ emotional exhaustion and depersonalization | Ex post facto design; self-report survey data; one point in time sampled |
Mehr, Ladany, & Caskie (2010) | 204 supervisees (172 females, 28 males, 4 unspecified; 181 White, 2 African-American, 5 Latino/Hispanic, 2 Native American/Alaskan Native, 7 Asian-American/Pacific Islander, 7 other/unspecified; x age=29.4 years; graduate programs represented 67% clinical psychology, 23% counseling psychology; practicum settings represented 28% university/college counseling center, 21% mental health center, 21% hospitals) | Demographic questionnaire; Supervisee Nondisclosure Survey; Trainee Disclosure Scale; Working Alliance Inventory/Supervision-Short (Trainee Version); Trainee Anxiety Scale | Participants responded to questionnaires via web link | Chi-square analyses; multivariate multiple regression analysis | “…strong supervisory working alliance was related to a lower amount of trainee nondisclosure and a higher overall willingness to disclose in a single supervision session” (p. 110). | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Newgent, Davis, & Farley (2004) | 15 doctoral students enrolled in supervision course (8 female, 7 male; 11 White, 4 non-White); each student received a certain number of individual, group, and triadic supervision of supervision sessions | Working Alliance Inventory (WAI; for measuring supervision of supervision alliance); Supervisory Styles Inventory (SSI); Supervisory Working Alliance Inventory-Supervisee Form (SWAI); Supervision of Supervision Evaluation (SSE) | Measures completed on each type of supervision received (individual, group, triadic) | ANOVAs | No significant alliance differences found between triadic and individual models or triadic and group models | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees; small sample size; supervisee exposure to individual and triadic supervision highly variable |
Patton & Kivlighan (1997)1 | 75 graduate-level pre-practicum supervisees (53 female, 22 male; 64 European-American, 11 African-American; x age=27.7 years); 77 clients (59 female, 16 male; 69 European-American, 8 African-American; x age=20 years; compensated undergraduate volunteers) | Working Alliance Inventory; Supervisor Working Alliance Inventory (Supervisee Form); Vanderbilt Therapeutic Strategies Scale | “Clients” seen by beginning graduate-level students for 4 50-minute sessions; “supervisees” then seen for supervision session after each client meeting; time-limited dynamic psychotherapy the treatment focus | Hierarchical linear modeling | Positive relationship found between supervisees’ perceptions of supervisory working alliance and clients’ perceptions of therapeutic working alliance | Focused on self-report perceptions of supervisees and clients; “clients” were compensated undergraduate volunteers who had not actually sought treatment; videotape session data also rated |
Quarto (2002)1 | 72 supervisees (78% female, 22% male; 86% White, 6% African-American, 4% Hispanic, 4% Other; x age=33.5 years; university counseling centers 28%, agencies 21%, psychology clinics 3%, other 15%); 74 supervisors (61% female, 39% male; x age=44.4 years; 73% professionals in counselor education, counseling psychology, or clinical psychology, 23% other) | Supervision Interaction Questionnaire; Supervisory Working Alliance Inventory (Trainee and Supervisor versions) | Questionnaire packets distributed | Exploratory factor analysis One-factor ANOVA | Negative relationship found between supervision conflict and supervision working alliance | Ex post facto design; self-report survey data; perceptions sampled at only one point in time |
Ramos-Sanchez, Esnil, Goodwin, Riggs, Touster, Wright, Ratanasiripuns, & Radolfa (2002)1 | 126 practicum students and interns (73% female, 27% male; 79% European-American, 21% Other; x age =30.7 years) | Relationship Questionnaire; Working Alliance Inventory; Supervisee Levels Questionnaire-Revised | Questionnaire packets distributed | Correlational, MANOVA, and qualitative analyses | Expected relations found between supervisee developmental level, negative supervisory events, and perceptions of supervisory working alliance | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisee |
Renfro-Michel & Sheperis (2009) | 117 graduate students (102f); even distribution of entry, prac and internship levels; variety of programs (school, counseling, rehabilitation, community, mental health) | Supervisory Working Alliance Inventory; Relationship Questionnaire; measured at mid-semester and end of semsester | Survey questionnaires completed | ANOVAs | Supervisee attachment related to alliance, with secure reporting better alliance than insecure at both mid and end of semester; supervisee development and working alliance unrelated | Ex post facto design; self-report survey data; of email signups 67.3% return rate in first semester, 54% return rate in second semester of data collection |
Riggs & Bretz (2006) | 87 psychology interns (66female, 20male; 78% Caucasian; mean age=32.6 years; 78% clinical psychology interns, 17% counseling psychology interns, 2% school psychology interns) | Working Alliance Inventory; Measure of Parental Style; Reciprocal Attachment Questionnaire; Relationship Questionnaire | Online data collection | MANOVAs; latent variable path analysis | Perceived supervisor attachment style had most direct impact on supervision alliance; supervisees who viewed supervisor as being securely attached viewed supervisory bond and task more positively | Ex post facto design; self-report survey data; focus on supervisee perceptions only; 50% return rate |
Schultz, Ososkie, Fried, Nelson, & Bardos (2002) | 111 employed rehabilitation counselors in supervision (68 female, 43 male; 101 White, 5 Hispanic, 3 Native American, 2 Asian/Pacific Islander) | Demographic questionnaire; Supervisory Working Alliance Inventory-Trainee Form; Rahim Leader Power Inventory | Questionnaire packets mailed to counselor/super-visees in two western states | ANCOVA and MANCOVAs | More time spent in supervision related to stronger supervisory alliance; higher supervisor scores on Expert and Referent power bases related to stronger alliance ratings | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Sterner (2009)1 | 71 mental health counselors either now being supervised or having been supervised post degree (48 female, 22 male, 1 no response; 90% White, 4% Latino-American, 1% Native American; x age=51; 83% held master’s degree, 17% doctoral; 39% private practice, 27% mental health agency, 16% private, nonprofit agency, 4% hospital, 14% other) | Supervisory Working Alliance Inventory-Trainee Version; Minnesota Satisfaction Questionnaire-Short Form; Occupational Stress Inventory-Revised | Survey questionnaire completed on Internet website by American Mental Health Counseling Association members (AMHCA) | Correlational analyses and canonical correlational analysis | Positively perceived supervisory working alliance related to greater work satisfaction and less work-related stress | Ex post facto design; self-report survey data; focused exclusively on perceptions of AMHCA supervisees; unclear how many “supervisees” were not actually in supervision and how long they had not been |
Sumerel & Borders (1996) | 40 graduate-level counseling students (26 female, 14 male; 20 beginning and 20 advanced trainees; 37 White) | Supervisory Working Alliance Inventory-Trainee Version (SWAI-T); Impact Message Inventory IMI); Session Evaluation Questionnaire | Trainees watched either a videotape of supervisor addressing interfering supervisee personal issues or of supervisor addressing supervisee skill deficits in supervision; participants then completed three measures | Correlations; ANOVA; MANOVA | Negative relationship between Rapport (SWAI-T) and Dominance (IMI) | Analogue study; ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
Walker, Ladany, & Pate-Carolan (2007)1 | 111 female graduate student supervisees (91 White, 9 African-American, 4 Asian, 3 biracial, 3 Latina, 1 other; x age =31 years; 18% clinical psychology, 70% counseling psychology; university/college counseling centers 58%, mental health centers 18%, schools 6%; VAs 5%; state hospitals 5%) | Gender-Related Events Survey; Working Alliance Inventory-Trainee Version; Trainee Disclosure Scale | Questionnaire packets distributed | Multivariate multiple regression analyses | Supportive gen derrelated events found to be positively related to supervisory working alliance; converse found for non-supportive gender-related events | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees; could not determine response rate |
Webb & Wheeler (1998) | 96 counselors in supervision (75 females, 20 males, 1 unspecified; 44 in diploma or master’s training; all registered in British Association for Counselling) | Sensitivity to disclosing in supervision assessment tool; Supervisory Working Alliance Inventory (supervisee version) | Questionnaire packets mailed to counselors for completion | Correlations and factor analysis | Positive correlation found between supervisees’ ratings of alliance rapport and willingness to disclose in supervision | Ex post facto design; self-report survey data; focused exclusively on perceptions of counselor/supervisees; 44% return rate |
Wester, Vogel, & Archer (2004)1 | 103 male psychology interns (93 White, 9 Hispanic, 1 African-American; x age =33.3; 64 doctoral students 35 master’s) | Gender Role Conflict Scale; Supervisory Working Alliance Inventory-Trainee Version; Counseling Self-Estimate Inventory | Questionnaire packets distributed | t tests ANOVAs | Male supervisees working with male as opposed to female supervisors perceived supervisory working alliance less favorably, lending initial support to possible male socialization explanation | Ex post facto design; self-report survey data; focused exclusively on perceptions of supervisees |
White & Queener (2003) | 67 supervision dyads: Supervisees–56female, 11male; 59 master’s students, 8 doctoral students; 80% first year of training through internship; Supervisors–47female, 20male; 55 licensed professionals, 12 doctoral students | Supervisory Working Alliance Inventory; Adult Attachments Scale; Social Provisions Scale | Supervisees and supervisors completed survey packet containing questionnaires | Simultaneous regression analyses | Supervisors’ ability to form close, healthy relationships predictive of supervisees’ and supervisors’ perceptions supervisory working alliance; weaker relational ability and unfavorable alliance perceptions related and vice versa | Ex post facto design; self-report survey data; 50% return rate |
Williams, Helm, & Clemens (2012) | 131 mental health therapists employed full time in community mental health agencies (83 female, 48 male; 106 White, 19 Hispanic, 1 Native American, 2 multiethnic, 3 other; 50 social workers, 11 marriage and family therapists, 40 professional counselors; 7 psychologists, 17 unlicensed professionals); received at least one hour of supervision per month | Childhood Trauma Questionnaire; Five Factor Wellness Inventory-Form A; Supervisory Working Alliance Inventory-Supervisee Form; Job Satisfaction Survey; Quantitative Workload Inventory; Trauma and Attachment Belief Scale | Questionnaires administered (in random order) to participants’ at their work sites | Path analysis | Anticipated partial mediating effect of supervisory alliance on vicarious traumatization not significant | Ex post facto design; self-report survey data; limited range of scores on alliance measure may have negatively affected results; very limited amount of supervision upon which to base alliance ratings |
Author(s) | Ex Post Facto, Cross-Sectional | Self-Report Survey Data | Supervisee Perspective | Supervisor Perspective | Other Study Features/Comments |
---|---|---|---|---|---|
Bennett et al. (2008) | X | X | X | – | |
Bhat & Davis (2007) | X | X | – | X | |
Bilodeau & Lecomte (2010) | X | X | X | Five supervision sessions involved; self-report data gathered prior to first session and across all five sessions | |
Bilodeau & Lecomte (2012) | X | X | – | Five supervision sessions involved; self-report data gathered prior to first session and across all five sessions | |
Bucky et al. (2010) | X | X | X | – | |
Burke et al. (1998) | X | X | X | Ten supervision sessions involved; self-report data gathered across all ten sessions; interaction data also coded | |
Chen & Bernstein (2000) | X | X | X | Three supervision sessions involved; self-report data gathered at four separate points; interaction data also coded | |
Cooper & Ng (2009) | X | X | X | X | |
Culbreth & Borders (1999) | X | X | X | – | Workplace study |
Davidson (2011) | X | X | X | – | |
Deal et al. (2011) | X | X | X | Randomized controlled trial; data gathered at three separate points; strong study | |
Dickson et al. (2011) | X | X | X | – | |
Efstation et al. (1990) | X | X | X | X | Validation study for Supervisory Working Alliance Inventory |
Gatmon et al. (2001) | X | X | X | – | |
Gnilka et al. (2012) | X | X | X | – | |
Gunn & Pistole (2012) | X | X | X | – | |
Ladany et al. (1997) | X | X | X | – | |
Ladany et al. (1999) | X | X | X | – | |
Ladany & Friedlander (1995) | X | X | X | – | |
Ladany & Lehrman-Waterman (1999) | X | X | X | – | |
Ladany, Lehrman-Waterman et al. (1999) | X | X | X | – | |
Ladany et al. (2013) | X | X | X | – | Two-item qualitative questionnaire also included |
Ladany et al. (2001) | X | X | – | X | |
Livni et al. (2012) | X | X | X | Up to eight supervision sessions involved; data gathered at three separate points; well done workplace study | |
Mena & Bailey (2007) | X | X | X | X | Workplace study |
Mehr et al. (2010) | X | X | X | – | |
Newgent et al. (2004) | X | X | X | – | |
Patton & Kivlighan (1997) | X | X | – | Four therapy and four supervision sessions involved; only study in which client perspective assessed | |
Quarto (2002) | X | X | X | X | |
Ramos-Sanchez et al. (2002) | X | X | X | – | |
Renfro-Michel & Sheperis (2009) | X | X | – | Assessments done at two separate points | |
Riggs & Bretz (2006) | X | X | X | – | |
Schultz et al. (2002) | X | X | X | – | Workplace study |
Sterner (2009) | X | X | X | – | Workplace study |
Sumerel & Borders (1996) | X | X | X | – | |
Walker et al. (2007) | X | X | X | – | |
Webb & Wheeler (1998) | X | X | X | – | |
Wester et al. (2004) | X | X | X | – | |
White & Queener (2003) | X | X | X | – | |
Williams et al. (2012) | X | X | X | – | Workplace study |
Review Results and Discussion
Setting/Sample Characteristics
Measures Used
Procedures
Analyses
Findings/Conclusions
Study Limitations
Discussion
Investigation of the Supervisory Alliance in Process
Tapping Multiple Perspectives When Measuring Alliance
Taking a Methodologically Diverse and Diversified Approach to Alliance Research
Other Factors
… comparing the impact of supervision and the alliance on outcomes among different groups of professionals (e.g., counselors, nurses, social workers), as well as between professionals and training students could highlight differences in important components of supervision across the field (p. 184).
Conclusion
References
Information & Authors
Information
Published In
History
Keywords:
Authors
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
There are no citations for this item
View Options
View options
PDF/ePub
View PDF/ePubGet Access
Login options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).