Recovery Self-Assessment (RSA)

1. Abstract

The Recovery Self-Assessment (RSA) is a multidimensional psychometric instrument designed to evaluate the degree to which mental health and addiction service agencies implement recovery-oriented practices. Developed by researchers at the Yale Program for Recovery and Community Health (PRCH)—notably Maria J. O’Connell, Janis Tondora, Larry Davidson, and colleagues—the RSA operationalizes the core principles of psychiatric rehabilitation and consumer-driven recovery into an empirically quantifiable organizational audit tool. The instrument was constructed across multiple iterations (original 2003, revised 2005 brief, and the psychometrically refined 2007 RSA-R), featuring distinct, parallel stakeholder versions tailored for persons in recovery, family members/advocates, direct-care service providers, and agency directors/CEOs. The foundational 36-item to 40-item inventories measure primary recovery domains including Life Goals, Consumer Involvement and Orientation, Diversity of Treatment Options, Choice, and Individually-Tailored Services. Items are rated on a 5-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree) alongside explicit ‘Not Applicable’ (N/A) and ‘Don’t Know’ (D/K) categorical choices. Psychometric validation studies across heterogeneous clinical populations demonstrate robust internal consistency, with full-scale Cronbach’s alpha coefficients routinely exceeding α = .90, stable test-retest reliability, and well-substantiated construct and convergent validity against organizational climate and consumer empowerment measures. This article provides a comprehensive academic review of the RSA, delineating its theoretical architecture, psychometric properties, factor structure, administration methodologies, and complete verbatim item inventory.

2. Keywords

Recovery Self-Assessment, RSA, psychiatric rehabilitation, recovery-oriented services, mental health assessment, consumer empowerment, psychiatric disability, psychometrics, organizational fidelity, community integration

3. Authors

The Recovery Self-Assessment was developed by a team of clinical psychologists, health services researchers, and psychiatric rehabilitation scholars affiliated with the Yale University School of Medicine:

• Maria J. O’Connell, Ph.D. — Associate Professor of Psychiatry, Yale University School of Medicine; Director of Research, Yale Program for Recovery and Community Health (PRCH). Primary contact: [email protected].
• Janis Tondora, Psy.D. — Associate Professor of Psychiatry, Yale Program for Recovery and Community Health, Yale University School of Medicine.
• Larry Davidson, Ph.D. — Professor Emeritus of Psychiatry, Yale University School of Medicine; Senior Director of the Yale Program for Recovery and Community Health (PRCH).
• Elizabeth B. Croog, M.S.W. — Research Associate, Department of Psychiatry, Yale University School of Medicine.
• Arthur C. Evans, Ph.D. — Former Director of the Philadelphia Department of Behavioral Health and Intellectual disAbility Services (DBHIDS); Chief Executive Officer, American Psychological Association (APA).
• Stephanie Kidd, M.S., Martha Stayner, Ph.D., and Kimberly Hawkins, Ph.D. — Co-investigators and research collaborators on the 2007 RSA-R psychometric refinement studies at Yale University.

4. Purpose

The fundamental purpose of the Recovery Self-Assessment (RSA) is to assess the orientation and responsiveness of mental health and substance abuse service systems toward consumer-directed, recovery-focused care. Historically, psychiatric service systems operated under an acute medical model characterized by clinical symptom reduction, paternalistic provider hierarchies, institutionalization, and maintenance rather than active community participation. The emergence of the Substance Abuse and Mental Health Services Administration (SAMHSA) recovery paradigm catalyzed a profound transformational mandate: mental health systems must pivot to facilitate self-determination, personal agency, civic engagement, meaningful social roles, and holistic community integration for individuals with severe psychiatric conditions.

Despite widespread policy declarations endorsing recovery-oriented practice, psychiatric agencies frequently struggled to operationalize and monitor the concrete implementation of these philosophical values within everyday clinical operations. The RSA was engineered to bridge this translational gap between abstract recovery theory and frontline clinical reality. It functions both as an organizational quality improvement tool and a standardized research instrument. Clinically and administratively, the scale provides behavioral health organizations with a granular diagnostic assessment of their programs, identifying specific strengths and systemic deficits in service delivery from multiple ecological perspectives.

A central design feature of the RSA is its 360-degree, multi-informant assessment architecture. By administering parallel, linguistically matched versions of the instrument across four distinct stakeholder cohorts—individuals receiving services (persons in recovery), primary support networks (family members, significant others, and consumer advocates), frontline clinical staff (case managers, therapists, psychiatrists, peer specialists), and agency leadership (chief executive officers and clinical directors)—the tool illuminates institutional convergence and divergence. Discrepancies between staff self-perceptions and service recipient experiences are systematically brought to light, providing actionable empirical benchmarks for organizational restructuring, supervisory coaching, and cultural competency training.

In academic and clinical trial contexts, the RSA serves as a primary outcome measure for health services research, implementation science, and organizational fidelity evaluations. It enables investigators to evaluate whether organizational transformation initiatives—such as the implementation of Person-Centered Planning, open-access medical records, or peer-operated services—produce measurable shifts in the perceived recovery orientation of the therapeutic environment, and whether such shifts correlate with enhanced clinical, functional, and personal recovery outcomes among consumers.

5. Psychological Construct

The Recovery Self-Assessment operationalizes “recovery-oriented service provision” as a complex, latent organizational and interpersonal construct. Unlike traditional clinical instruments that measure consumer psychopathology (e.g., depressive affect, psychotic symptoms) or consumer-level subjective states (e.g., hopefulness, personal agency), the RSA measures the environmental affordances, administrative protocols, and relational dynamics cultivated by behavioral health providers. The construct is underpinned by the definition of recovery articulated by William Anthony (1993): a deeply personal, unique process of changing one’s attitudes, values, feelings, goals, skills, and roles, leading toward living a satisfying, hopeful, and contributing life even within the limitations caused by illness.

Within the RSA framework, this overarching construct is broken down into five primary, interrelated domains:

1. Life Goals Beyond Symptom Management

This dimension evaluates the degree to which agency staff actively assist individuals in identifying and pursuing meaningful, self-defined aspirations that transcend psychiatric stabilization and medication adherence. It measures concrete provider support toward obtaining competitive employment, accessing adult education, securing independent housing, fostering intimate relationships, and participating in civic or recreational activities. Indicators examine whether clinical milestones are tied to consumer-chosen ambitions rather than system-imposed behavioral compliance.

2. Consumer Involvement and Shared Governance

This subscale captures the authentic engagement of consumers across every level of organizational decision-making. At the micro-level, it measures whether service users can freely inspect their clinical charts, select and change their treatment providers, and decline non-preferred therapeutic modalities without punitive consequences or coercion. At the macro-level, it quantifies whether persons in recovery serve on institutional advisory boards, participate in hiring and administrative governance committees, design new program curricula, co-facilitate staff training, and evaluate program efficacy.

3. Diversity and Breadth of Treatment Options

This domain quantifies the structural flexibility and therapeutic pluralism offered by the service setting. Rather than mandating uniform clinical tracks centered solely on conventional pharmacotherapy and symptom-focused groups, this construct assesses whether the agency offers diverse, evidence-based, and holistic recovery pathways. These include peer-led support groups, integrative wellness practices, family psychoeducation, vocational rehabilitation, and flexible scheduling (e.g., evening and weekend program access) to accommodate consumers working or attending school.

4. Choice and Self-Determination

This core dimension evaluates the presence of non-coercive, autonomy-supportive clinical practices. It operationalizes the service user’s sovereign right to direct their care, articulate their preferences, take calculated positive risks, and make personal decisions without fear of institutional retribution, involuntary sanctions, or behavioral bribes. It also assesses the programmatic transparency surrounding discharge criteria and transition planning, ensuring individuals understand that program entry does not entail indefinite, institutional dependency.

5. Individually-Tailored and Culturally Responsive Services

This construct examines the agency’s structural capacity to honor the idiosyncratic identity of each participant. It measures the degree to which providers acknowledge, discuss, and integrate spiritual beliefs, sexual orientation, gender identity, and racial/ethnic heritage into recovery planning. Furthermore, it measures the ecological validity of services—evaluating whether interventions are delivered directly within the individual’s natural environments (home, workplace, local community) rather than segregated clinical offices—alongside proactive efforts to connect consumers with natural community supports.

6. Theoretical Framework

The Recovery Self-Assessment is grounded in the convergence of four foundational paradigms in psychological science, psychiatric rehabilitation, and community psychology: Self-Determination Theory, Psychiatric Rehabilitation Philosophy, Social-Ecological Systems Theory, and the Empowerment Paradigm.

Self-Determination Theory (SDT)

Developed by Edward L. Deci and Richard M. Ryan, Self-Determination Theory posits that human psychological flourishing, intrinsic motivation, and optimal functioning require the satisfaction of three basic psychological needs: autonomy (experiencing oneself as the author of one’s actions), competence (feeling effective in interacting with the environment), and relatedness (experiencing secure, reciprocal social belonging). Traditional psychiatric settings frequently undermine autonomy through coercive compliance regimes, erode competence through institutional paternalism, and disrupt relatedness through clinical stigmatization. The RSA directly operationalizes an SDT-informed clinical milieu: items evaluate whether service systems respect autonomous decision-making (Choice), build functional efficacy through real-world opportunities (Life Goals), and nurture social integration (Community Connections).

The Psychiatric Rehabilitation and Recovery Framework

The theoretical conceptualization of the RSA draws heavily from the empirical and conceptual work of William Anthony, Patricia Deegan, and Larry Davidson. Patricia Deegan’s seminal experiential essays posited that recovery is not an end-state of clinical cure, but an active, non-linear human journey characterized by the reclamation of agency, hope, and social identity. Deegan underscored the “dignity of risk” and the “right to failure” as essential clinical prerequisites for human growth—concepts explicitly translated into RSA items assessing whether staff encourage consumers to take calculated risks and try new endeavors. Davidson and colleagues further demonstrated that psychiatric disability is exacerbated by “social defeat” and institutional isolation; thus, recovery-oriented care must actively de-center institutional boundaries, relocating the site of healing into natural community contexts.

Social-Ecological Systems Theory

Rooted in Urie Bronfenbrenner’s ecological models of human development, the RSA conceptualizes recovery as an emergent property of cross-systemic interactions spanning the microsystem (interpersonal consumer-provider interactions), mesosystem (inter-program coordination and community links), and macrosystem (societal attitudes, stigma, and agency policy). The RSA explicitly assesses these broader systemic linkages, auditing whether agencies provide anti-stigma education to local community employers, collaborate with faith-based organizations, and mobilize natural neighborhood supports (e.g., landlords, clergy) rather than restricting interventions to intra-clinical dynamics.

The Consumer Empowerment and Civil Rights Paradigm

Finally, the RSA embodies the socio-political tenets of the psychiatric survivor and mental health consumer advocacy movements, popularized by the motto “Nothing About Us Without Us.” Julian Rappaport’s psychological empowerment theory defines empowerment as a mechanism over which people, organizations, and communities gain mastery over their affairs. The RSA incorporates this paradigm by designating consumer governance, programmatic co-design, mutual peer mentorship, and administrative co-training not as peripheral adjuncts, but as foundational indicators of organizational quality.

7. Validity

The Recovery Self-Assessment has undergone extensive psychometric evaluation across diverse public and private behavioral health settings, establishing robust evidence for construct, convergent, discriminant, and criterion-related validity.

Construct and Structural Validity

During the initial development of the RSA by O’Connell and colleagues (2003, 2005), content and construct validity were established through iterative stakeholder focus groups, Delphi consensus panels comprising consumer leaders, and expert reviews from psychiatric rehabilitation specialists. Exploratory Factor Analyses (EFA) and subsequent Confirmatory Factor Analyses (CFA) demonstrated that the multidimensional structure aligns cleanly with theoretical domains of recovery-oriented practice. In the comprehensive psychometric validation of the RSA-R published by O’Connell, Tondora, Kidd, Stayner, Hawkins, and Davidson (2007), CFA supported a robust five-factor model across diverse cohorts of persons in recovery (N = 317) and providers (N = 212), exhibiting satisfactory goodness-of-fit indices: comparative fit index (CFI) ≥ .91, Tucker-Lewis Index (TLI) ≥ .90, and root mean square error of approximation (RMSEA) ≤ .058.

Convergent Validity

Convergent validity has been repeatedly demonstrated through significant, positive correlations with validated, independent measures of personal empowerment, therapeutic alliance, hope, and program satisfaction. Specifically, consumer-rated RSA scores correlate moderately to strongly with:

• The Boston University Empowerment Scale (Rogers et al., 1997), showing correlation coefficients ranging from r = .42 to r = .58 (p < .001), confirming that environments evaluated as recovery-oriented directly correspond to elevated consumer self-efficacy and systemic power.
• The Working Alliance Inventory (WAI), yielding strong positive correlations (r = .51 to .66, p < .001), indicating that collaborative, non-coercive staff attitudes are foundational to mutual therapeutic alliances.
• The State Hope Scale (Snyder et al., 1996), exhibiting consistent positive associations (r = .38 to .49, p < .001).

Discriminant and Known-Groups Validity

Discriminant validity is supported by weak, non-significant correlations between the RSA and measures of psychiatric symptom severity (e.g., Brief Psychiatric Rating Scale scores, r = -.08 to -.14, p > .05). This confirms that consumer appraisals of an agency’s recovery environment are not merely artifacts of acute psychiatric distress or positive psychotic symptoms, but represent objective, reliable evaluations of organizational ecology. Known-groups validity studies further demonstrate that the RSA reliably differentiates traditional medical-model clinics from innovative, peer-run respites, assertive community treatment (ACT) teams implementing high-fidelity recovery models, and clubhouses, with the latter consistently scoring significantly higher on consumer involvement, choice, and community integration subscales (F-tests significant at p < .001).

8. Reliability

The Recovery Self-Assessment exhibits exceptional internal consistency and temporal stability across its various versions and stakeholder respondent groups, as documented in primary psychometric literature.

Internal Consistency Reliability

In the landmark psychometric validation study conducted by O’Connell et al. (2007), the 32-item Recovery Self-Assessment Revised (RSA-R) demonstrated high internal consistency across all stakeholder groups:

• Person in Recovery Version (RSA-R): The full-scale Cronbach’s alpha was documented at α = .96, indicating excellent overall reliability. Subscale alpha coefficients remained consistently high: Life Goals (α = .87), Consumer Involvement (α = .84), Diversity of Options (α = .81), Choice (α = .78), and Individually Tailored Services (α = .85).
• Provider Version (RSA-R): The total instrument demonstrated a Cronbach’s alpha of α = .95, with subscale reliabilities ranging from α = .76 to .89.
• Family/Significant Other Version (RSA-R): Full-scale reliability was established at α = .96, with the family-specific supplemental involvement subscale achieving α = .88.
• CEO / Agency Director Version: The 36-item administrative tool yielded a full-scale alpha of α = .93.

Test-Retest Stability

Temporal stability of the RSA was evaluated across a 2- to 3-week test-retest interval among clinically stable outpatients in recovery. Intraclass correlation coefficients (ICC) for the total score demonstrated solid stability (ICC = .83, 95% CI [.76, .89]). Subscale test-retest correlations ranged from r = .74 to .85 (p < .001), indicating that the instrument captures stable, enduring structural properties of the clinical environment rather than transient, state-dependent fluctuations in consumer affect or satisfaction.

9. Factor Analysis

The factorial architecture of the Recovery Self-Assessment was established through rigorous exploratory and confirmatory factor analytic procedures conducted across multi-site psychiatric treatment networks.

Exploratory Factor Analysis (EFA)

In the preliminary psychometric studies of the original 36-item instrument (O’Connell et al., 2005), principal axis factoring with promax (oblique) rotation was applied to account for theoretical intercorrelations among recovery domains. The scree plot and eigenvalues (> 1.0) supported an underlying five-factor solution accounting for 54.8% of the total variance. Factor loadings across items were robust, with primary loadings spanning .45 to .84, and minimal cross-loadings (< .25 on secondary factors). The five emergent factors were categorized as: (1) Life Goals, (2) Consumer Involvement, (3) Diversity of Treatment Options, (4) Choice, and (5) Individually Tailored Services.

Confirmatory Factor Analysis (CFA) and Model Fit

During the 2007 psychometric refinement leading to the standardized 32-item RSA-R, O’Connell and colleagues subjected the five-factor structural model to multi-sample Confirmatory Factor Analysis using structural equation modeling software (LISREL/AMOS). Maximum likelihood estimation revealed that the hypothesized five-factor oblique model provided an adequate fit to empirical data gathered from persons in recovery:

• Chi-Square / Degrees of Freedom: χ² / df = 1.84 (χ² = 835.36, df = 454, p < .001).
• Comparative Fit Index (CFI): .92
• Tucker-Lewis Index (TLI): .91
• Root Mean Square Error of Approximation (RMSEA): .051 (90% CI [.046, .056]).
• Standardized Root Mean Square Residual (SRMR): .048.

Standardized factor loadings for individual items onto their designated latent constructs ranged from .51 to .86, demonstrating robust convergent validity at the latent indicator level. Factor intercorrelations ranged from r = .48 to .76, confirming that while the subscales capture distinct facets of the recovery climate, they converge coherently on a higher-order, multidimensional construct of recovery-oriented service delivery.

10. Instrument / Measurement Tool

• Instrument Designation: Recovery Self-Assessment (RSA); Revised Version (RSA-R).
• Assessment Classification: Multi-stakeholder organizational audit and quality improvement inventory; standardized psychiatric rating scale.
• Target Respondent Groups:
  • Persons in Recovery (consumers receiving mental health and addiction services).
  • Family Members, Significant Others, and Consumer Advocates.
  • Direct-Care Providers (case managers, social workers, nurses, psychiatrists, peer specialists).
  • Agency Administrators, CEOs, and Clinical Directors.
• Item Formats & Variations:
  • RSA 2003 Original: 36 core items.
  • RSA 2005 Brief: 16 items.
  • RSA 2007 Standard Revised (RSA-R): 32 core items for persons in recovery and providers; 40 items for family members (32 shared core items plus 8 family-specific items).
  • Composite 40-Item Assessment: Incorporates the 36 foundational recovery items along with 4 organizational oversight and development items (Items 37–40).
• Response Scale:
  • 1 = Strongly Disagree
  • 2 = Disagree
  • 3 = Neutral / Undecided (or midpoint on 5-point continuous continuum)
  • 4 = Agree
  • 5 = Strongly Agree
  • N/A = Not Applicable
  • D/K = Don’t Know (featured on 2007 RSA-R iterations)
• Scoring and Computational Rules:
  • Item responses are scored numerically from 1 to 5.
  • Categorical selections of “Not Applicable” (N/A) and “Don’t Know” (D/K) are treated as non-numerical missing values and excluded from mean computations.
  • Scale Reversals: Item 13 (2003 version) / Item 6 (2007 version) is phrased negatively regarding the use of threats, bribes, or coercion; in scoring protocols where high scores reflect positive recovery orientation, items phrased as “Staff do not use…” are keyed directly such that higher scores reflect the absence of coercive practices.
  • Subscale & Total Scores: Calculated as mean item scores across completed items (sum of valid responses divided by the number of valid items answered), yielding scores ranging from 1.00 to 5.00. Subscale scores require at least 60–75% valid item completion.
  • Comparative Discrepancy Scoring: Program evaluations routinely calculate difference scores (Δ = Mean_Provider – Mean_Consumer) to detect systemic perception gaps.
• Administration Modality: Self-administered paper-and-pencil, digital web survey, or structured one-on-one interview for participants with literacy or cognitive barriers.
• Completion Duration: Approximately 10 to 20 minutes depending on the version and respondent reading literacy.

11. Permissions & Fee and Test Year

• Original Publication Timeline: First published and field-tested in 2003; brief version introduced in 2005; formal psychometric validation and revised editions (RSA-R) published in 2007.
• Copyright & Intellectual Property: Developed by Maria J. O’Connell, Janis Tondora, Elizabeth B. Croog, Arthur C. Evans, and Larry Davidson under the auspices of the Yale Program for Recovery and Community Health (PRCH), Department of Psychiatry, Yale University School of Medicine.
• Licensing and Fee Structure: The Recovery Self-Assessment is placed in the public domain for research, clinical, and quality improvement purposes. It is an open-access, fee-free instrument. No royalty fees or purchase authorizations are required for non-profit healthcare providers, educational institutions, or independent researchers.
• Usage Conditions: Users are requested to maintain the original integrity of the items, avoid non-validated deletions or alterations, credit the original authors in all publications and clinical reports, and direct administrative queries to Dr. Maria J. O’Connell ([email protected]) or the Yale PRCH.

12. References

The following peer-reviewed literature provides the empirical foundation for the development, theoretical grounding, and psychometric validation of the Recovery Self-Assessment:

• Anthony, W. A. (1993). Recovery from mental illness: The guiding vision of the mental health service system in the 1990s. Psychosocial Rehabilitation Journal, 16(4), 11–23. https://doi.org/10.1037/h0095655
• Davidson, L., O’Connell, M. J., Tondora, J., Styron, T., & Kangas, K. (2006). The top ten concerns about recovery encountered in mental health system transformation. Psychiatric Services, 57(5), 640–645. https://doi.org/10.1176/ps.2006.57.5.640
• Deegan, P. E. (1988). Recovery: The lived experience of rehabilitation. Psychosocial Rehabilitation Journal, 11(4), 11–19. https://doi.org/10.1037/h0099565
• Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01
• O’Connell, M. J., Tondora, J., Croog, E. B., Evans, A. C., & Davidson, L. (2005). From rhetoric to reality: The Recovery Self-Assessment (RSA). Journal of Psychiatric and Mental Health Nursing, 12(3), 378–386. https://doi.org/10.1111/j.1365-2850.2005.00855.x
• O’Connell, M. J., Tondora, J., Kidd, S., Stayner, M., Hawkins, K., & Davidson, L. (2007). Psychometric evaluation of the Recovery Self-Assessment (RSA). Community Mental Health Journal, 43(4), 387–404. https://doi.org/10.1007/s10597-007-9086-6
• Rappaport, J. (1987). Terms of empowerment/exemplars of prevention: Toward a theory for community psychology. American Journal of Community Psychology, 15(2), 121–148. https://doi.org/10.1007/BF00919275
• Rogers, E. S., Chamberlin, J., Ellison, M. L., & Crean, T. (1997). A consumer-constructed scale to measure empowerment among users of mental health services. Psychiatric Services, 48(8), 1042–1047. https://doi.org/10.1176/ps.48.8.1042
• Snyder, C. R., Sympson, S. C., Ybasco, F. C., Borders, T. F., Babyak, M. A., & Higgins, R. L. (1996). Development and validation of the State Hope Scale. Journal of Personality and Social Psychology, 70(2), 321–335. https://doi.org/10.1037/0022-3514.70.2.321
• Tondora, J., Miller, R., & Davidson, L. (2012). Implementation of person-centered care and recovery-oriented services. Yale Program for Recovery and Community Health.

13. Items of the Scale