A Systematic Review of Validated Measures of Supervisory Competence in Psychology and Counselling Practice

ABSTRACT

In recent years, supervisor competency frameworks have been developed to guide the provision of supervision for trainee psychologists and counsellors, and ongoing clinical supervision in the workplace. A range of instruments have been developed that focus on one or more of these competencies. The aim of this systematic review was to identify psychometrically validated measures of supervisory competence and to review which domains are assessed by each instrument. Three databases were systematically searched, PsycINFO, Scopus and Medline, from 2010 to 2025, to identify sources that met the inclusion criteria. Sixteen sources were included in the review, all of which had a cross‐sectional design, and four included a repeated‐measures component. Two reviewers independently assessed the risk of bias of these studies, and two reviewers mapped the instruments to the domains they assess. All instruments were evaluated across five criteria: type of rating measure, number of domains assessed, psychometric properties, who completes the instrument and number of instrument formats. All instruments were questionnaire‐based; most assessed at least five of 10 relevant domains. The majority were designed for supervisees, with a few having supervisor and supervisee versions, and only two instruments had multiple formats. Internal reliability ranged from moderate to high, and four instruments demonstrated high test–retest reliability. Adequate to good convergent validity was established on most instruments, and adequate to good divergent validity on a few. This review synthesises key aspects of supervisor competency instruments, enabling supervisors, training institutions and clinical practices to select tools most relevant to their settings.

Summary

• Supervisor competencies have recently been developed and apply to accredited psychology and counselling supervisors.

• Various measures of supervisor competency are available to support the assessment of supervisors by supervisees and supervisors.

• This review reports on psychometrically validated supervisor competency measures and the supervisory domains they assess.

1. Introduction

Competency‐based models have emerged as a preferred approach for the training and accreditation of psychologists, with frameworks and guidelines being developed over the previous two decades. Rodolfa et al. (2005, 2013), Fouad et al. (2009) and the American Psychological Association (APA 2015; revised in 2025) produced early attempts at defining the core competencies that psychologists should be proficient in, with more recent guidelines being produced by the Psychology Board of Australia (2024) and the British Psychological Society (2024). A hallmark of competency‐based pedagogies is that the ‘what competencies should be attained’ is both foundational and pivotal to subsequent ‘how–when–who’ aspects of training and supervision (APA 2025; Gonsalvez and Crowe 2014). To this end systematic, careful, thoughtful and thorough efforts are invested into designing competency frameworks that then serve as blueprints that inform all other aspects of education and training, including how these competencies are taught and assessed effectively, when these learning and assessment activities occur, and who should deliver the learning activities to ensure effective acquisition, consolidation and integration of competencies (APA 2025; Gonsalvez and Deane 2026).

Competency‐based frameworks for supervisors have also been developed, with early approaches defining and classifying core competencies supervisors should attain into frameworks (e.g., Pilling and Roth 2014) or guidelines (APA 2014; Psychology Board of Australia 2018). These approaches are fairly consistent in the competencies they identify and have recently been synthesised into 10 domains (Gonsalvez et al., forthcoming), four encapsulating competencies related to the supervisor's oversight of their supervisee's competence and six on the supervisor's demonstration of their own competence (see Table 1). The competency frameworks proposed by the American Psychological Association (2025), British Association for Counselling and Psychotherapy (BACP 2021) and the Psychology Board of Australia (2018) were entered into a spreadsheet before mapping the competencies onto a domain structure that integrated the approaches. Whilst the level of detail differed widely across the approaches, as expected, substantive overlap among the competencies was evident. A 10‐domain structure that adequately captured the competencies of the three frameworks was arrived at by consensus among the three authors. Details of the similarities and differences among the three frameworks are provided in (Gonsalvez et al., forthcoming).

TABLE 1.

Core domains of supervisor competence (Gonsalvez et al., forthcoming).

Focus of domain	Domain
Supervisor's oversight of supervisee	Oversee and ensure client welfare and safety, and ethical, legal and regulatory standards are adhered to Knowledge of the area of clinical/professional practice being supervised including knowledge of evidence‐based practices Effective teaching and/or skills training competencies, such as best practice formative, use of appropriate observational methods in supervision Ability to assess trainee's competence in knowledge, skills and attitude‐values, assessment of competence problems and remediation processes
Supervisor's demonstration and/or focus on their own competence	5Knowledge of supervision, models, methods, processes and evidence‐based supervisory interventions 6Supervisory relationship, including form and maintain healthy supervisor‐supervisee relationship and deal effectively with problems in that relationship 7Professionalism in managing, conducting and coordinating supervisory activities, including in concert with multidisciplinary teams 8Ethical, legal and regulatory considerations, including supervisor misconduct or breaches of ethics/legal codes 9Individual, social and cultural diversity, including demonstration of supervisor's cultural awareness and responsiveness about diversity and multicultural competence 10Self‐awareness, evaluation and reflective practice, including openness to or seeking feedback from supervisees, peers or experts

Wheeler and Barkham (2014) conducted a comprehensive review of instruments before recommending a core battery of instruments for supervisor competence research. They identified 49 unique instruments and assessed these according to six criteria to evaluate their usefulness and accessibility for future research: pan‐theoretical, relatively short, face validity, ideally measure both supervisor and supervisee perspective, have psychometric validity and reliability, and be suitable for repeated use. Following a rigorous evaluation and rating process, the resulting battery of instruments included one routine measure of supervision, the Brief Supervisory Alliance Scale (Rønnestad and Lundquist 2009), and four instruments that measure specific aspects of supervision, the Development of Psychotherapist Common Core Questionnaire (Orlinsky and Rønnestad 2005), the Helpful Aspects of Supervision Questionnaire (Llewelyn 1988), the Role Conflict and Role Ambiguity Inventory (Olk and Friedlander 1992) and the Supervisory Working Alliance (Efstation et al. 1990). Two measures were also recommended that were developed following their initial evaluation, the Supervisory Relationship Questionnaire (Palomo et al. 2010) and the Leeds Alliance in Supervision Scale (Wainwright 2010).

Since the publication of Wheeler and Barkham's (2014) comprehensive work, there have been significant changes in the supervision landscape, including the development of new and revisions of established instruments. A more recent review provides a description and validation of 59 measures (Li et al. 2024). The current review extends the important work of past researchers by providing an update on instruments of supervisory competence. Importantly, the current review maps the instruments onto competency domains, thereby providing educators and supervisors a ‘ready reckoner’ resource that summarises (a) which instruments measure what competencies and (b) a quick description and evaluation of these instruments.

2. Method

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) guidelines (Page et al. 2021).

2.1. Search Strategy

The search strategy was designed to replicate the database searches conducted by Wheeler and Barkham (2014), with some necessary adjustments to accommodate changes in available databases in 2025 and refine terminology to more directly target suitable sources. Three databases, PsycINFO, Scopus and Medline, were systematically searched in November 2025, using terms similar to those used by Wheeler and Barkham, identifying published studies since the completion of the previous work (2010–2025). Searches were limited to English language and peer‐reviewed or edited publications, with key words searched across the databases. Search terms included supervis*, evaluat*, assess*, instrument*, measure*, questionnaire*, scale*, inventory*, reliability, valid* and develop*. The specific search process for each database is included in the Supplementary Data 1.

2.2. Selection

The inclusion criteria were studies that reported on an instrument that (i) assessed at least one of the 10 designated supervisory competence domains; (ii) evaluated clinical supervisors in the fields of psychology, counselling or mental health; (iii) was available in English and easy to access; (iv) reported psychometric assessment properties more rigorously than providing internal reliability measures; and (v) the instrument was developed to assess general supervisory competence in these fields. Exclusion criteria were scales measuring competence in a specific supervisory orientation (e.g., family therapy, CBT supervision, constructivist supervision), measures of line management, and measures developed specifically for disciplines (e.g., nursing) other than those specified in the inclusion criteria.

The results from the three databases searches were extracted to Endnote (Version 21, Clarivate, 2024) and combined. Following the removal of duplicates, all sources were reviewed by the first and second authors, screening the title and reviewing abstracts of sources that were potentially relevant to the study. The full text of all sources potentially relevant after the abstract review were retrieved and reviewed by the first and fourth authors. Studies were selected for review if they were primary research that reported on the development of an instrument that met the inclusion criteria. At each stage of study selection, any discrepancies were discussed by the authors and resolved according to whether the study met the inclusion criteria. Additionally, recently published studies, including those in recently published reviews (e.g., Li et al. 2024) were screened by the first and second authors, to identify studies missed in previous searches.

2.3. Data Extraction and Analyses

Studies that met the inclusion criteria were reviewed, and data were extracted and tabulated by the first author (see Table 2). Mapping of instruments to the 10 competency domains was independently conducted by the first and third authors. The initial level of agreement between raters was 90%; full agreement was reached after discussion between the two reviewers of the discrepant ratings, referring to specific scale items to resolve the differences. The mapping process involved both authors looking at each instrument and mapping individual items to the relevant domains. Neither author was involved in the development of any of the included instruments. One author has extensive experience in supervision of trainee clinicians, bringing expert knowledge of the competency domains, and the other author is a non‐clinical researcher, who could objectively evaluate each of the measures in the mapping process, without prior exposure to the instruments. Based on key themes that emerged from the descriptions of the studies, the authors developed a set of criteria that instruments could be categorised into, reflecting the various strengths and weaknesses of each and highlighting the most appropriate instruments for different types of assessments. The dichotomous criteria are (1) rating measure/observation assessment, (2) limited/comprehensive number of domains assessed, (3) strong/adequate psychometric assessment, (4) measure completed by supervisee or supervisor and (5) single/multiple forms of the instrument. These assessments are included in Table 2 and reported in more detail in the results section.

TABLE 2.

Included studies.

Author, year, country	Title of instrument	Study design	Aim	Sample size and characteristics	Description of instrument Summary of instrument qualities	Domains assessed	Key findings
Choy‐Brown et al. 2023, United States	Evidence‐based Clinical Supervision Strategies (EBCSS) scale	Cross‐sectional Sample 1 confirmatory factor analysis (CFA) Sample 2 CFA to replicate 2‐factor structure, discriminant validity	To develop and evaluate a reliable, valid, and pragmatic measure of clinical super vision, assessing audit and feedback, and active learning	N = 154 clinicians working in outpatient mental health clinics working with young people N = 181 community mental health workers employed in an Assertive Community Treatment team	5 items completed by se on 5‐point Likert scale Items assess extent se experiences evidence‐based strategies in supervision session Rating measure Limited domains Strong psychometric assessment Se rated Single form	3 4	2‐factor structure: Audit and feedback Active learning Moderate internal reliability (α > 0.7) Good construct validity, examined with scores on agency climate of evidence‐based practice implementation Discriminant validity with a measure of supervisory alliance (BSWAI‐T)
Cliffe et al. 2016 England	Short Version of the Supervisory Relationship Questionnaire (S‐SRQ)	Repeated measures with small group of participants 2–4 weeks after initial measure, to examine test‐retest reliability	To produce a shorter version of the SRQ that retained its reliability and psychometric validity	N = 203 Trainee clinical psychologists enrolled in doctoral 18 programs N = 86 completed measure a second time	Short version of 67‐item SRQ (see below) 18 items completed by se on 7‐point Likert scale Rating measure Limited domains Strong psychometric assessment Se rated Multiple forms	2 3 4 6 7	3 subscales: Safe Base Reflective Education Structure High internal reliability (α = 0.96) High test–retest reliability (r = 0.94) Good convergent validity with measures of supervisory relationship (SWAI‐T and SRQ) Good divergent validity with not being correlated with personality constructs (short‐scale EPQ‐R) Good predictive validity for supervision effectiveness and satisfaction (SSQ)
Davis‐Wright et al. 2025, United States	Psychology Trainee Evaluation of Supervision Competencies (PTESC) Scale developed by Vas et al. (2015, 2021)	Cross‐sectional	To (a) identify dimensional structure of PTESC, (b) assess fit of PTESC's dimensional structure with 7 domains of APA (2014), and (c) examine PTESC's value in predicting trainees' acquisition of profession‐wide competencies	N = 110 Trainee psychologists enrolled in an urban academic medical centre	97 items completed by se on 5‐point Likert scale 7 domains (number of items in brackets): competence (31); diversity (15); supervisory relationship (12); professionalism (8); evaluation and feedback (15); professional competence (7); and ethical, legal and regulatory (8) Rating measure All domains Adequate psychometric assessment Se rated Single form	1 2 3 4 5 6 7 8 9 10	Single dimension All domains contribute to overall Supervision Competence dimension Evaluation and Feedback, Supervisory Relationship, and Supervisor Competence domains most strongly related to the single dimension High internal reliability on all domain scales (α ranging from 0.89 to 0.98) Construct validity assessed using exploratory graph analysis, with the model evaluated using confirmatory factor analysis Supervision competence predicted se reported growth in 9 profession wide competencies
Gonsalvez 2021, Australia	Supervision Evaluation and Supervisory Competence Scale—Short (SE‐SC Short)	Cross‐sectional	To determine if a short version of SE‐SC could reliably represent full scale, for ongoing use and routine estimates of a range of sr competencies	N = 122 Trainee clinical psychologists at a psychology Master's training program	8‐item short version of SE‐SC, completed by se on 7‐point Likert scale Rating measure Limited domains Strong psychometric assessment Se rated Multiple forms	2 3 5 6 7	Good internal reliability (α = 0.89) Good reliability with high correlations with original scale (r = 0.97) Adequate convergent validity with related constructs on SWAI and SSI Adequate divergent validity with dissimilar constructs on SWAI and SSI
Gonsalvez, Hamid, et al. 2017, Australia	Supervision Evaluation and Supervisory Competence Scale (SE‐SC)	Repeated measures with small group of participants 4–6 weeks after initial measure, to examine test‐retest reliability	To design a new scale to assess the supervisee's evaluation of supervision and ratings of supervisory competence, and to establish the psychometric credentials of the scale	N = 142 Trainee clinical psychologists in several tertiary programs N = 99 professional Master's program N = 17 doctoral degree N = 26 2‐year internship program N = 20 completed measure a second time	31 items completed on a 7‐point Likert scale by se and sr Rating measure Limited domains Strong psychometric assessment Se and sr rated Multiple forms	2 3 5 6 7	6‐cluster structure: Openness, caring and support Sr's knowledge/expertise as therapist Supervision planning and management Goal‐directed supervision Restorative competencies Insight and management of therapist‐client dynamics and reflective practitioner competencies High internal reliability (α = 0.75–0.92) High test–retest reliability (r = 0.81–0.93) Good convergent validity with measures of supervisory working alliance (SWAI and SSI) High concurrent validity, with 6 factors of instrument correlating with related construct subscales on SWAI and SSI
Hamilton et al. 2022, Australia	Generic Supervision Assessment Tool Supervisor (GSAT‐SR) and Supervisee (GSAT‐SE) versions	Cross‐sectional	To complete the development and psychometric validation of the GSAT	Phase I N = 447 trainees N = 479 supervisors Phase II N = 186 trainees N = 182 supervisors Participants from Australian and New Zealand tertiary education, in clinical education, psychotherapy, nursing, social work, occupational therapy, psychology, counselling	26 items sr version 21 items se version Tool designed for assessment of sr competencies across a range of professions Rating measure Comprehensive coverage of domains Adequate psychometric assessment Se and sr rated Single form	1 3 4 5 6 7 8 9 10	4‐factor structure in sr tool: Fundamental feedback processes Goals and tasks Respect Enabling supervisory practices 2‐factor structure in se tool: Foundational competencies enabling effective feedback Fundamental accountability, safety and awareness High internal reliability (α = 0.98 for GSAT‐SE; α = 0.97 for GSAT‐SR) High convergent validity with SWAI and SE‐SC
Lee et al. 2016, South Korea	Counseling Supervisor's Behavior Questionnaire (CSBQ)	Cross‐sectional	To develop the Counseling Supervisor's Behavior Questionnaire to identify specific supervisory behaviours with different functions that could describe what goes on during supervision	N = 164 Qualified counsellors in schools, couples/family therapy, career, spirituality, military/veteran, corporate, and addiction settings who have experience receiving supervision	20 items completed on 5‐point Likert scale by se Items assess extent sr exhibits supervisory behaviours during supervision sessions Rating measure Limited domains Adequate psychometric assessment Se rated Single form	2 3 4 7	5 dimensions of supervisory behaviours: Structuring Problem‐indicating Encouraging Exploring Decision‐making Internal reliability moderate to good on four of the five subscales (α = 0.76–0.86), and poor on the problem indicating scale (α = 0.67) Good concurrent validity with SSI
Maaß et al. 2024, Germany	Quality of Supervision Questionnaire (QSQ)	Cross‐sectional	To develop a brief questionnaire on supervisor effectiveness, suitable for evaluating the quality of a single supervision session, by both se and sr, to closely monitor the supervision process	N = 374 trainee psychotherapists N = 136 supervisors Participants from a range of tertiary education institutions in Germany	12‐item questionnaire with sr and se versions to measure the quality of a single supervisory session Tool separates the constructs of supervision effectiveness, satisfaction, and relationship Rating scale Limited domains Adequate psychometric assessment Se and sr rated Single form	2 3 6	3‐factor structure Effectiveness Procedural knowledge Relationship Se version: High internal reliability (α = 0.91) Moderate to strong convergent validity with SWAI and successful supervision (STEP‐SV) Sr version: Good internal reliability (α = 0.81) Weak to good convergent validity with SWAI and STEP‐SV
Palomo et al. 2010, England	Supervisory Relationship Questionnaire (SRQ)	Repeated measures, second measure on average 29.6 days after initial measure to examine test‐retest reliability	To develop a measure of the supervisory relationship from the supervisee perspective and test its initial psychometric properties	N = 284 Trainee clinical psychologists from 19 doctoral courses	67 item questionnaire completed on a 7‐point Likert scale by se Companion instrument to sr completed Supervisory Relationship Measure (Pearce et al. 2013) Rating scale Limited domains Strong psychometric assessment Se rated Multiple forms	2 3 5 6 7	Five factors identified: Safe base Structure Commitment Reflective education Role model High internal reliability (α = 0.98) Good test–retest reliability (r = 0.97) Good convergent validity with evaluation process (EPWSI), Supervisory role (RCRA), WAI‐Therapist, and Relationship Inventory Good divergent validity: differences in total and subscale scores and participant gender
Pearce et al. 2013, England	Supervisory Relationship Measure (SRM)	Repeated measures with small group of participants approximately 2 weeks after initial measure, to examine test‐retest reliability	To develop a valid and reliable measure of the supervisory relationship from the perspective of sr	N = 267 Supervisors in clinical psychology training programs N = 134 completed measure a second time	51 items, completed by sr, companion instrument to SRQ (Palomo et al. 2010) Rating measure Limited domains Adequate psychometric assessment Sr rated Single form	2 3 4 6 7 10	Five factors identified: Safe base Sr commitment Trainee contribution External influences Sr investment Good internal reliability (α = 0.90) Good test–retest reliability (r = 0.94) Good convergent validity with WAI, PRS‐R, and SSI
Presseau et al. 2021, United States	Alliance and Negotiation Scale‐Supervision Version (ANS‐S)	Cross‐sectional	To develop a form of the ANS focused on measuring supervisory relationship dynamics from the perspective of the supervisee	N = 113 Mental health clinical trainees currently enrolled in psychology graduate university programs, practicing under a licensed sr	12‐items responded to by se on a 7‐point Likert scale Rating measure Limited domains Adequate psychometric assessment Se rated Single form	6 9 10	Single factor scale High internal reliability (α = 0.93) Moderate convergent validity with Supervisory Multicultural Competence Inventory (SMCI) Moderate discriminant validity with SMCI. Positive correlation with WAI/Supervision‐short scale demonstrates the scale fails to adequately differentiate negotiation from general working alliance
Rousmaniere and Ellis 2013, United States	Collaborative Supervision Behaviors Scale (CSBS)	Cross‐sectional	To create and assess a measure that assesses the extent of collaborative behaviours in supervision from the supervisee's perspective, and secondly, to test empirically the relation between CS and the supervisory alliance	N = 252 Trainee doctoral or master programs: Clinical psychology (42.9%) Counselling psychology (37.3%) School counselling (6.7%) Clinical social work (4.0%) Marriage and Family Therapy (4.0%) Other psychology or counselling programs (5.2%)	5 items responded to by se on 4‐ or 5‐point Likert scale Scale designed to measure collaboration in clinical supervision, related but distinct from supervisory working alliance Rating scale Limited domains Strong psychometric assessment Se rated Single form	3 6	High internal reliability (α = 0.93) Convergent validity, being related to but distinct from SWAI‐T Discriminant construct validity, as CSBS and SWAI‐T both correlated with Collaborative Supervisory Alliance, an empirically derived model representing overall quality of collaborative supervisory relationship CSBS correlated positively with SWAI‐T, indicating collaborative supervision may be an important aspect of supervision for strengthening supervisory working alliance Divergent validity indicated by collaborative supervision not being related to sr or se related characteristics (e.g., age, gender, se's training level)
Sangganjanavanich and Black 2011, United States	Multicultural Supervision Scale (MSS)	Cross‐sectional	To explore the dimensions of competent multicultural supervision through the development and initial validation of the Multicultural Supervision Scale	N = 304 Counsellor educators and clinical supervisors	39 items responded to by sr on an unspecified number Likert‐type scale Rating scale Limited domains Limited psychometric assessment Sr rated Single form	9	Three factors identified: Supervisory skills Sr's attitudes and beliefs Stereotypes toward diverse populations Moderate internal reliability (α = 0.76) Construct validity based on principal components analyses (PCA)
Schultz 2011, United States	Supervisor Principle Ethics Scale (SPES)	Cross‐sectional	To create a measure of rehabilitation counsellor sr's use of ethical principles within the supervisory relationship, and assess the construct validity and reliability	N = 87 Participants were rehabilitation counsellors employed in a state‐designated vocational rehabilitation agency and receiving supervision	22 items responded to by se on 7‐point Likert scale Scale measures ethical principles: fidelity, beneficence, veracity, justice, autonomy Rating scale Limited domains Adequate psychometric assessment Se rated Single form	6 7 8 9	No factor analysis of scale conducted to examine the designed subscales High internal reliability (α = 0.97) Good construct validity with SWAI‐T
Swank et al. 2021, United States	Supervision Competencies Scale (SCS)	Cross‐Sectional	Psychometric validation of the SCS	N = 416 Counsellor supervisors, predominantly faculty members (46%) and practitioners (27%), remainder students, administrators, site supervisors, others Majority practitioners in mental health (63%) Others in marriage, couple, and family therapy, and other	65 items responded to by sr on a 5‐point Likert scale Rating scale Comprehensive coverage of domains Adequate psychometric assessment Sr rated Single form	1 2 3 4 5 6 7 8 9	Four factors identified: Supervision skills Supervision laws and ethics Supervision modalities Supervision knowledge High internal reliability (α = 0.97) Construct validity based on PCA
Tsong and Goodyear 2014, United States	Supervision Outcome Scale (SOS)	Cross‐sectional	To describe development and initial psychometric evaluation of the SOS, designed to assess outcome of supervision from se's perspective	N = 357 Counselling and clinical psychology doctoral students	7 items responded to by se on a 5‐point Likert scale Rating scale Limited domains Adequate psychometric assessment Se rated Single form	3 9	Two factors identified: Clinical Multicultural Correlation of the two factors was 0.68, indicating low discriminant validity between them High internal reliability (α = 0.90) Good concurrent validity with SWAI

Note: se, supervisee; sr, supervisor; BSWAI‐T, Brief Supervisory Working Alliance Inventory‐trainee form (Sabella et al. 2020); EPWSI, Evaluation Process with Supervision Inventory (Lehrman‐Waterman and Ladany 2001); EPQ‐R, Eysenck Personality Questionnaire Revised (Eysenck et al. 1985); PRS‐R, Personal Reaction Scale Revised (Holloway and Wampold 1984); Relationship Inventory (Schacht et al. 1988); RCRA, Role Conflict and Role Ambiguity inventory (Olk and Friedlander 1992); STEP‐SV, Questionnaire to Evaluate Supervision (Zarbock et al. 2009); SE‐SC, Supervision Evaluation and Supervisory Competence Scale (Gonsalvez, Hamid, et al. 2017); SMCI, Supervisory Multicultural Competence Inventory (Inman 2006); SSQ, Supervisory Satisfaction Questionnaire (Ladany et al. 1996); SRQ, Supervisory Relationship Questionnaire (Palomo et al. 2010); SSI, Supervisory Styles Inventory (Friedlander and Ward 1984); Supervisory Working Alliance Inventory (Efstation et al. 1990); SWAI‐T, Supervisory Working Alliance Inventory—Trainee Version (Bahrick 1989); WAI, Working Alliance Inventory (Bahrick 1989); WAI/Supervision‐short, Working Alliance Inventory/Supervision‐short (Ladany et al. 2013); WAI‐Therapist, Working Alliance Inventory‐Therapist (Bahrick 1989).

2.4. Risk of Bias Assessment

The evaluation of ‘risk of bias’ of included studies was assessed using the Critical Appraisal Skills Programme (CASP) for descriptive/cross‐sectional studies (CASP 2024). The first and fourth authors independently completed the assessments, with initial agreement between raters high (89% agreement). Discrepancies between raters were discussed and consensus reached following a second joint review of the publications in question.

3. Results

The selection process is illustrated in a PRISMA diagram (Figure 1). Initially, 345 sources were located in the searches, and after duplicates were manually removed, 296 publication titles were reviewed. Each reviewer independently completed the title review then compared results to determine which studies would be reviewed at abstract level. Of the 296 titles, the abstracts of 38 sources were independently reviewed, then results were compared to determine the studies that were assessed to be likely to meet the inclusion criteria, and 18 studies progressed to a full‐text review. This process was repeated with the 59 instruments included in Li et al.'s (2024) review, with the exception of 12 sources that were also located in the initial 345 sources. Of the 47 remaining sources reviewed by Li et al., the authors determined that 13 studies potentially met the inclusion criteria and the full texts were reviewed. One further study was located in a recent review. In total, the full text of 32 studies was reviewed by the first and fourth authors and consensus was reached with discussion with the second author, with a final decision that 16 instruments met the inclusion criteria. Table 2 provides an overview of the included studies.

FIGURE 1.

PRISMA diagram detailing the study selection process Source: Page et al. (2021).

An additional 10 studies were excluded following full text review of the articles. Of the excluded studies, six met the inclusion criteria with the exception that they pertained to supervisory competence in specific theoretical orientations. Of the six excluded studies, two pertained to supervision of CBT (Milne et al. 2011; Moeller et al. 2020), two targeted family and marital therapy supervision (Avila et al. 2016, 2018), one targeted trauma‐informed supervision (Cook et al. 2023) and one measured constructivist supervision (Halligan Avery et al. 2017). In each instance, the exclusion was determined by the authors' acknowledgement that the scale targeted a specific theoretical orientation. In addition, four studies were excluded because of inadequate information on the psychometric properties of the instrument (Bearman et al. 2019), or they were short versions of excluded scales (Reiser et al. 2018) or because the measure was designed for a specific course (Butler et al. 2021; Pérez‐Rojas et al. 2017).

3.1. Risk of Bias Assessment

Most of the studies were considered to have a low risk of bias. All studies identified their focus, used an appropriate methodology, recruited participants appropriately, collected data that addressed the issue, presented the results clearly and offered research that was valuable in the field of assessment of supervisor competency. Most studies accurately reduced measurement bias, analysed the data appropriately and produced results applicable to the local population. The primary methodological concerns are a failure to provide a justification for the adequacy of the sample size for all analyses (eight studies) and, in the discussion section, not making a clear statement on the findings (four studies). See Supplementary Data 1 for the Risk of Bias Assessments for individual studies.

3.2. Study Characteristics

All studies used cross‐sectional designs to develop and psychometrically validate the instruments, and four studies also included repeated measures with a subset of their sample to examine test–retest reliability (Cliffe et al. 2016; Gonsalvez, Hamid, et al. 2017; Pearce et al. 2013; Palomo et al. 2010). Sample sizes ranged from 87 to 447 supervisees and 136 to 479 supervisors. Eight studies were conducted in the United States (Choy‐Brown et al. 2023; Davis‐Wright et al. 2025; Presseau et al. 2021; Rousmaniere and Ellis 2013; Sangganjanavanich and Black 2011; Schultz 2011; Swank et al. 2021; Tsong and Goodyear 2014), three in the United Kingdom (Cliffe et al. 2016; Pearce et al. 2013; Palomo et al. 2010), three in Australia (Gonsalvez 2021; Gonsalvez, Hamid, et al. 2017; Hamilton et al. 2022), one in Germany (Maaß et al. 2024) and one in South Korea (Lee et al. 2016). Study settings were predominantly tertiary education institutions (N = 11), with 10 studies involving participants from several training institutions, indicating the instruments were assessed for use in diverse training programs. Other study settings included outpatient and community mental health services, vocational rehabilitation agencies, and studies that drew their sample from a wide range of counselling professionals.

3.3. Summary of Competency Domains

For brevity, Table 2 lists only the domain numbers. The full description of each domain is included in Table 1 and is noted briefly here. Domains relating to the supervisor's oversight of the supervisee are as follows: (1) oversee and ensure client welfare and safety; (2) knowledge of area of clinical/professional practice being supervised; (3) effective teaching and/or skills training; and (4) ability to assess supervisee's competence in knowledge, skills and attitude‐values. Domains relating to the supervisor's demonstration and/or focus on their own competence are as follows: (5) knowledge of supervision, models, methods, processes and evidence‐based supervisory interventions; (6) supervisory relationship; (7) professionalism in managing, conducting and coordinating supervisory activities; (8) ethical, legal and regulatory considerations; (9) individual, social and cultural diversity; and (10) self‐awareness, evaluation and reflective practice.

3.4. Rating vs. Observer‐Based Instruments

All instruments used a rating‐based scale, completed as a self‐assessment by the supervisor or a supervisee assessment of their supervisor.

3.5. Number of Supervisor Competency Domains Assessed

A list of 10 domains informed by competency frameworks proposed by reputed professional bodies (Gonsalvez et al., forthcoming) was used to indicate the assessment scope of each instrument. The number of domains assessed by the instruments ranges from 1 to 10, with the majority of instruments (N = 11) assessing competencies that correspond to both the supervisor's competence to supervise the supervisee and the supervisor's own professional competence. Four instruments focus solely on domains associated with the supervisor's own professional competence (Presseau et al. 2021; Rousmaniere and Ellis 2013; Sangganjanavanich and Black 2011; Schultz 2011), and one instrument only assesses domains relating to the supervisor's competence to supervise the supervisee (Choy‐Brown et al. 2023).

One instrument assesses all 10 domains and two instruments assess nine. The 97‐item PTESC (Davis‐Wright et al. 2025) was developed to assess the seven clinical supervisor competencies specified by the APA (2014) and also comprehensively assesses the 10 synthesised domains. The SCS (Swank et al. 2021) is also a relatively long instrument with 65 Likert‐scale questions, evaluating all but self‐awareness, evaluation and reflective practice. A much shorter instrument, the GSAT (Hamilton et al. 2022) comprises 26 items for the supervisor version and 21 items for the supervisee version, with Knowledge of Clinical Practice the only domain not covered.

Five of the instruments assess five or six of the domains: the full and short versions of the SE‐SC (Gonsalvez 2021; Gonsalvez, Hamid, et al. 2017), the SRM (Pearce et al. 2013) and the full and short versions of the SRQ (Cliffe et al. 2016; Palomo et al. 2010). Most of these scales assess knowledge of clinical practice, teaching skills, the supervisory relationship and professionalism. This indicates that when scales are developed that aim to broadly assess supervisory competence, there is a uniform view of the key domains that should be assessed. Three of these scales also assess knowledge of supervision, models, methods and processes (Gonsalvez 2021; Gonsalvez, Hamid, et al. 2017; Palomo et al. 2010). The S‐SRQ (Cliffe et al. 2016) and SRM (Pearce et al. 2013) also evaluate ability to assess supervisee's competence in knowledge, skills and attitude‐values. Few of the instruments assess the supervisor's self‐awareness, evaluation and reflective practice, with the SRM being one of the few scales that does (Pearce et al. 2013).

Eight scales assess between one and four domains. The CSBQ (Lee et al. 2016) and SPES (Schultz 2011) each assess four domains, with the former assessing various supervisor behaviours and the latter focusing on domains assessing the supervisor's professionalism and ethics. The QSM (Maaß et al. 2024) and ANS‐S (Presseau et al. 2021) assess three domains, with the QSM designed to evaluate a single session of supervision. The remaining four scales, the CSBS (Rousmaniere and Ellis 2013), SOS (Tsong and Goodyear 2014), EBCSS (Choy‐Brown et al. 2023) and the MSS, (Sangganjanavanich and Black 2011) cover one or two domains, as outlined in Table 2.

Evaluation of the supervisory relationship is the domain that is included in the most instruments (N = 12), and the focus on the supervisory relationship is consistent with several of the measures selected in the toolkit or recommended for use by Wheeler and Barkham (2014).

There is a trade‐off between comprehensively assessing all or most of the domains and the length of the scale. If a brief scale is a priority, four measures have 31 items or less and capture at least five competence domains (S‐SRQ, Cliffe et al. 2016; SE‐SC, Gonsalvez, Hamid, et al. 2017; SE‐SC‐Short, Gonsalvez 2021, and GSAT, Hamilton et al. 2022). The GSAT stands out as an exception, with a relatively low number of items in the scales (supervisor version, N = 26, supervisee version, N = 21) that cover nine domains, making this scale both brief and comprehensive. However, this scale was not specifically developed for the assessment of the competence of counselling or psychology supervisors; rather, it is a more general instrument that can be used in a range of allied health tertiary education training institutions, potentially lacking some specificity to these fields.

3.6. Psychometric Assessment of Instruments

3.6.1. Reliability of Instruments

Internal reliability (Cronbach's α) was high for most measures (S‐SRQ, Cliffe et al. 2016; GSAT‐SR, GSAT‐SE, Hamilton et al. 2022; SRQ, Palomo et al. 2010; ANS‐S, Presseau et al. 2021; SPES, Schultz 2011; SCS, Swank et al. 2021; SOS, Tsong and Goodyear 2014), good for the SE‐SC Short scale (Gonsalvez 2021), QSQ trainee version (Maaß et al. 2024), SRM (Pearce et al. 2013), and CSBS (Rousmaniere and Ellis 2013), and moderate for the EBCSS (Choy‐Brown et al. 2023), QSQ supervisor version (Maaß et al. 2024), and MSS (Sangganjanavanich and Black 2011). Three instruments did not report α‐values for the full scale, but reported them for the subscales, with α‐values high for the PTESC (Davis‐Wright et al. 2025), ranging between moderate to high for the subscales of the SE‐SC (Gonsalvez, Hamid, et al. 2017), and between poor and good for the CSBQ (Lee et al. 2016). It should be noted that very high α‐values are indicative of item redundancy.

Test–retest reliability was assessed in four scales using a subset of the initial sample. Gonsalvez, Hamid, et al. (2017) reported high reliability of the SE‐SC at 4–6 weeks after the initial measure. Cliffe et al. (2016) reported the S‐SRQ has high test–retest reliability, while also demonstrating sensitivity to change, at 2 to 4 weeks after initial completion. Palomo et al. (2010) reported the SRQ has high test–retest reliability at between 12 and 78 days (mean 29.6 days). Pearce et al. (2013) reported good reliability for the SRM between 9 and 53 days, mean interval 17.2 days.

3.6.2. Validity of Instruments

The studies used a variety of validity measures (construct, convergent, divergent and predictive validity) with several providing strong evidence of the validity of the instrument against known and previously validated scales. Instruments with good evidence of convergent or construct validity with measures of supervisory relationship or supervisory working alliance include the full and short versions of the SRQ (Cliffe et al. 2016; Palomo et al. 2010), the SRM (Pearce et al. 2013), the SE‐SC (Gonsalvez, Hamid, et al. 2017), the GSAT (Hamilton et al. 2022), the CSBQ (Lee et al. 2016), the SPES (Schultz 2011) and the SOS (Tsong and Goodyear 2014). In addition, Cliffe et al. (2016) report the S‐SRQ has good predictive validity for supervision effectiveness and satisfaction. In terms of cross validation in the sample of instruments in this review, Hamilton et al. (2022) reported high convergent validity of their supervision assessment measure with Gonsalvez, Hamid, et al. (2017) measure of supervisory competence (SE‐SC).

Instruments with adequate or moderate convergent validity with measures of supervisory working alliance, supervisory styles or supervisory competence include the SE‐SC–Short (Gonsalvez 2021), the ANS‐S (Presseau et al. 2021) and the CSBS (Rousmaniere and Ellis 2013). The QSQ (Maaß et al. 2024) demonstrated moderate to strong convergent validity for the supervisee version and weak to good for the supervisor version.

Divergent or discriminant validity tended to be examined through a negative association with dissimilar constructs of the scales used to assess convergent validity, with only one instrument finding good evidence of divergent validity, the short version of the SRQ (Cliffe et al. 2016). Scales with adequate or moderate divergent validity include the EBCSS (Choy‐Brown et al. 2023), the SE‐SC–Short (Gonsalvez 2021), the SRQ (Palomo et al. 2010), the ANS‐S (Presseau et al. 2021) and the CSBS (Rousmaniere and Ellis 2013).

Construct validity of the PTESC was assessed by confirmatory factor analysis (Davis‐Wright et al. 2025), and by principal components analysis for the MSS (Sangganjanavanich and Black 2011) and SCS (Swank et al. 2021), with all scales providing strong evidence that they reflect the underlying construct they are designed to measure. The internal structure of the subscales and their relationships with one another were also examined through hierarchical cluster analyses for the SE‐SC (Gonsalvez, Hamid, et al. 2017).

3.7. Instruments Completed by Supervisee or Supervisor

All instruments were designed for use by either the supervisor, trainee/supervisee or both. No general supervisor competence instruments were located that facilitate assessment by a rater other than the supervisor or a supervisee.

Three supervisory competency instruments have both a supervisee‐report and a supervisor self‐report version. The GSAT (Hamilton et al. 2022) has short supervisor and supervisee versions, designed for use across a range of professions. The SRM (Pearce et al. 2013) is the supervisor self‐rated companion version of the supervisee‐rated SRQ (Palomo et al. 2010). The QSQ (Maaß et al. 2024) also has two versions of the instrument, developed for rating a single supervision session.

Of the other 12 supervision assessment instruments, the majority (N = 10) are completed by supervisees, either by trainees or qualified practitioners receiving supervision in the workplace. Two instruments were developed solely for supervisor self‐evaluation: the MSS (Sangganjanavanich and Black 2011) and SCS (Swank et al. 2021).

3.8. Complementary Forms of the Instrument

Two full‐length instruments offer short versions of the scale, the SRQ and the SE‐SC. The SE‐SC‐Short consists of eight items, covering five domains with a 6‐factor structure and has a high correlation with the original scale. The Short Version of the SRQ comprises 18 items across four domains and has three subscales that correlate highly with items of the original scale.

4. Discussion

The development of validated instruments to assess supervisor competence is an important step that should follow the extensive and commendable work on competency frameworks. The current review aimed to provide a critical overview of instruments of supervisory competence and, importantly, to map the instruments onto competency domains, thereby providing educators and supervisors a ‘ready reckoner’ resource that identifies the competency domain(s) targeted by each instrument together with a quick description and evaluation of these instruments.

Of the 16 instruments identified in this review, three covered nine or 10 of the competency domains. The supervisor and supervisee versions of the GSAT (Hamilton et al. 2022) are short measures allowing for repeated use, but due to their brevity they do not appear to provide in‐depth assessment of supervisor competencies in each domain. The other comprehensive scales are the SCS (Swank et al. 2021) and PTESC (Davis‐Wright et al. 2025); however, both are long instruments (65 items and 96 items, respectively) and, therefore, are time‐consuming to complete, potentially restricting how often they can be used by supervisors as a self‐rating tool.

This review is not a comprehensive overview of all available instruments. Other potentially valuable instruments (e.g., Supervisor Competency Self‐Assessment scale, Falender et al. 2016) have content validity, but we found no published data to establish their psychometric credentials.

There are several instruments that have validated supervisor and supervisee measures. These allow for assessments of aspects of supervision in a common language for comparison between a supervisor's self‐reported strengths and weaknesses and the experiences of their supervisee. The importance of incorporating supervisees' perceptions of supervisor competence cannot be understated. Firstly, supervisees rely on professional guidance, training, support, and a role model from their supervisor in the early stages of their professional practice (Todd and Storm 2014). While supervisee feedback can be invaluable to the supervisor, it is limited to only what the supervisee observes or their opinion of certain aspects of the supervisor's competence. Secondly, the incidence of inadequate or harmful supervision reported by supervisees is high (Ellis et al. 2014). Supervisees need an avenue to express if they are not receiving adequate supervision to support their professional development, yet providing honest feedback can be difficult for supervisees. For example, there is a tendency to avoid providing negative ratings if the supervisee is concerned their supervisor will be aware of their assessment; therefore, the supervisee's anonymity is important (Gonsalvez, Wahnon, and Deane 2017). Thirdly, self‐assessments can be inaccurate or biased (Walfish et al. 2012), thus, seeking the views of supervisees can help supervisors to develop a more balanced assessment of their competence and provide direction for further training.

4.1. Measures of Supervisory Competency: Some Concerns

This review reveals several concerns pertaining to measurement of supervisor competency. First, the lack of instruments designed for multisource feedback is glaring. In addition to supervisees, supervisory competence is often of direct and critical importance to other stakeholders, including peers, employers, line managers, and educational and training experts. The domains of Client Welfare, Professionalism, and Ethical, Legal, and Regulatory Considerations all directly involve multiple stakeholders. However, the review has not produced instruments designed and used for multisource appraisals and feedback of supervisor competency. This is not the case in other allied fields. For instance, a systematic review found that multisource feedback instruments have adequate reliability, validity, and utility in leadership development of health practitioners (Andrews et al. 2013).

There is a need to develop observational rating tools designed for multi‐respondent assessment of a recorded supervision session by the supervisor, supervisee, peer, and manager, enabling comparisons across perspectives on a supervisor's performance and constructive assessment of areas for further development. We are not aware of such a rating tool for the assessment of general supervisor competency. However, the obvious need for observer‐based rating scales of overall supervisor competence does not apply uniformly across all domains and theoretical orientations. Specifically, observation rating scales have been developed for the assessment of CBT supervision (e.g., Milne et al. 2011; Reiser et al. 2018), and the Supervisor Observational Coding System (Chapman et al. 2022) for the assessment of evidence‐based treatment supervision.

While supervisor self‐report and supervisee‐completed assessments are paramount for assessing supervisor competence, supervisors would also benefit from professional input on their strengths and areas that they can improve, from managers, supervisors, and other professionals or peers. In this review, no purpose‐made validated questionnaire or interview‐based instruments were identified for supervisors to seek this type of feedback, although we note that the supervisor self‐report measures could be used by third‐party raters to provide feedback to supervisors.

While acknowledging the importance of the voice of the supervisee in the evaluation of supervisory competence, the limitations of relying primarily on supervisee feedback are well documented. Supervisee feedback may be compromised by a range of factors, including their limited knowledge and experience associated with their early developmental level, personalised prior supervisory experience, or lenient rating tendencies stemming from cultural backgrounds or supervisee‐supervisor power differentials (Bernard and Goodyear 2019). Li et al. (2024) found widespread limitations concerning the psychometric testing of supervision measures used in the counselling field, with approximately half of the instruments included in the review having undergone validation at the time of their development, and only half of these instruments had further cross‐validation with other samples. In this current review, only six studies reported more than one type of validity assessment of the instrument. Further, while it is a strength of many of these studies that the development and initial validation of the scales used multiple tertiary institution samples, there is a lack of analysis on the validity of the instruments with non‐majority cultures or diverse samples. This current limitation in the research literature is a concern also shared with Li and colleagues, particularly validation studies using different samples than those involved in the development of the measure.

Finally, the extent to which the psychometric characteristics of these instruments, including their underlying dimensional structure, remain stable across varying levels of supervisory experience is uncertain. Most studies reviewed (11 of 16) relied on data from master's or doctoral trainees, with only five studies focusing on post‐licensure supervision. Examining developmental variations in supervisory competence and distinguishing foundational from advanced competencies represents important directions for future research.

4.2. Strengths and Limitations of the Current Review

This study builds on the growing literature identifying current limitations on research into measures of supervisory competence, such as a lack of instruments that have undergone extensive validation and have empirical support in multiple contexts and diverse samples. A strength of this study is assessing the instruments against a comprehensive list of competence domains that, together, are relevant to all stakeholders. Instruments can be selected according to the domains they assess, offering nuanced measures that are suitable for the purpose of the assessment. Secondly, the study reviews the psychometric assessments of the instruments to provide confidence for the user that the instrument is a valid measure and, in some cases, has test–retest reliability.

The current study used the search strategy adopted by Wheeler and Barkham (2014), to extend the findings of their comprehensive work. While this approach has merits, limitations and biases that apply to the original search strategy would also apply to the current review. Furthermore, the purpose of this review was not to identify every published instrument, but to identify a set of psychometrically validated instruments likely to be of practical value to most educators and supervisors in the field. It is also noted that the taxonomy used in this review was derived from the authors' synthesis of frameworks by major professional bodies (e.g., APA 2025; BACP 2021; Psychology Board of Australia 2018) but has not been independently validated. Other alternative categorisations or conceptual organisations of supervisor competency are possible.

4.3. Directions for Future Research

Whilst all instruments identified for consideration in this review have at least preliminary validation, most instruments require further research with diverse samples and in different settings to facilitate a better understanding of their structure, the similarity and unique elements of their subscales, and their application. There are very limited instruments specifically designed for objective ratings of supervisory competence (e.g., expert evaluations of videotapes of supervision sessions). The bulk of the data on supervisory competence is generated from supervisee's evaluations that are vulnerable to leniency (when supervisee anonymity is not assured) and other biases. The development and validation of multisource feedback tools is essential if supervisors are to receive comprehensive assessments of their strengths and weaknesses. Finally, the development and concurrent use of self‐rating and other‐rating versions of the same instrument are needed. Self‐other comparisons yield information about the direction (over or under‐estimation of one's competence), magnitude and nature (variability across competency domains) of self‐assessment that is pivotal for reflective practice and continued enhancement of supervisory competence.

5. Conclusion

This systematic review has identified a range of instruments useful for assessing supervisor competency in counselling and psychology. Each instrument has been assessed against the common supervisor competency domains developed by regulatory and professional bodies, providing a practical reference for selecting the most appropriate instrument for the assessment purpose. The profile of strengths and limitations of the instruments is summarised in a user‐friendly manner that will hopefully advance informed utilisation of the instruments in both practice and research. Several gaps in the research literature are evident, highlighting the need for studies to validate the instruments more broadly with different population groups and settings, and to develop multisource feedback instruments to provide a more comprehensive assessment of a supervisor's strengths and weaknesses.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting information.

Data Availability Statement

All data used in the present systematic review is presented in Table 2, and additional documentation of the processes of this review can be found in the Supporting Information.