Client Satisfaction Questionnaire (CSQ)

Abstract

The Client Satisfaction Questionnaire (most notably operationalized in its widely utilized 8-item format, the CSQ-8) is one of the premier standardized psychometric instruments designed to assess general, unidimensional client and patient satisfaction with health, mental health, and human services. Originally developed and empirically validated by Daniel L. Larsen, C. Clifford Attkisson, William A. Hargreaves, and Tuan D. Nguyen in 1979 at the University of California, San Francisco (UCSF), the instrument was crafted to overcome the severe methodological weaknesses of idiosyncratic, unstandardized satisfaction surveys that dominated clinical program evaluations. The CSQ-8 captures a client’s global evaluation of health care delivery, examining perceived service quality, alignment with subjective expectations, extent of unmet needs, willingness to recommend the service, adequacy of service dosage, functional efficacy in coping with life problems, overall affective satisfaction, and anticipated service re-utilization.

Comprising eight self-report items scored on a 4-point Likert-type scale with unique, item-specific anchor descriptions (yielding an aggregate score ranging from 8 to 32), the scale avoids generic agreement scaling to mitigate acquiescence bias and reduce ceiling effects. Psychometric evaluations across diverse outpatient, inpatient, medical, psychiatric, and substance use populations consistently substantiate a robust single-factor structure accounting for over 60% to 80% of common variance. The CSQ-8 exhibits exceptional internal consistency reliability, with Cronbach's alpha coefficients regularly falling between .83 and .94, alongside established concurrent, convergent, and predictive validity correlating with treatment completion, symptom alleviation, and therapeutic alliance. Translated into dozens of languages, the CSQ-8 remains an international benchmark in health services research and clinical outcome monitoring.

Keywords

Client Satisfaction Questionnaire, CSQ-8, patient satisfaction, psychometrics, program evaluation, health services research, clinical outcomes, treatment adherence, health care quality, unidimensional measurement, patient-reported outcome measures (PROMs)

Authors

The Client Satisfaction Questionnaire suite was conceptualized, developed, and validated through an extensive research initiative at the Department of Psychiatry, School of Medicine, University of California, San Francisco (UCSF).

• C. Clifford Attkisson, Ph.D.: Professor Emeritus of Medical Psychology, Department of Psychiatry, and former Dean of Graduate Studies at the University of California, San Francisco (UCSF). Dr. Attkisson is an internationally recognized pioneer in health services research, mental health program evaluation, and patient satisfaction methodologies. Email: [email protected].
• Daniel L. Larsen, Ph.D.: Clinical Psychologist and research investigator associated with the UCSF Department of Psychiatry during the formative development of the CSQ scales.
• William A. Hargreaves, Ph.D.: Professor Emeritus of Psychiatry, University of California, San Francisco; renowned methodologist in clinical evaluation, mental health services research, and psychiatric outcomes.
• Tuan D. Nguyen, Ph.D.: Research methodologist and behavioral scientist contributing to early psychometric validation and structural modeling of patient evaluation indices at UCSF.
• Rebecca Zwick, Ph.D.: Psychometrician and quantitative methodologist who collaborated extensively with Dr. Attkisson in publishing seminal empirical validations of the CSQ-8's psychometric properties and predictive relationships with treatment retention.
• Thomas K. Greenfield, Ph.D.: Senior scientist at the Public Health Institute and collaborator on extended variants of the CSQ family, including the Service Satisfaction Scale (SSS-30).

Purpose

The primary purpose of the Client Satisfaction Questionnaire (CSQ) is to provide a brief, standardized, psychometrically rigorous, and universally interpretable measure of overall consumer satisfaction with health, mental health, and human services. Before the development of the CSQ in the late 1970s, program evaluation suffered from widespread fragmentation. Most clinical settings relied on home-grown, non-validated questionnaires characterized by highly skewed distributions, prominent ceiling effects, unmeasured measurement error, and an absence of comparative normative data. The CSQ was deliberately constructed to establish a robust gold standard capable of benchmarking clinical service quality across diverse therapeutic environments.

In clinical practice, the CSQ-8 fulfills critical continuous quality improvement (CQI) and outcome monitoring functions. Because consumer satisfaction directly reflects the client's subjective experience of care, routine administration of the scale enables clinics to detect institutional bottlenecks, interpersonal disconnects between clinicians and patients, perceived inadequacies in service volume, and structural mismatches between patient expectations and actual intervention delivery. Research has demonstrated that lower satisfaction scores serve as a reliable early warning indicator for premature treatment termination, treatment non-adherence, and missed appointments. Clinicians and clinical directors utilize these findings to tailor intervention plans, repair ruptures in the therapeutic alliance, and allocate clinical resources more equitably.

In research contexts, the CSQ-8 is extensively used as a primary or secondary Patient-Reported Outcome Measure (PROM) in randomized controlled trials (RCTs), comparative effectiveness research (CER), and community-based epidemiological studies. It provides an indispensable consumer-centered metric to complement objective clinical endpoints such as biological markers, diagnostic symptom inventories, or clinician-rated functioning scales. Because modern healthcare models prioritize patient-centered care, regulatory bodies, third-party payers, and accreditation organizations (such as The Joint Commission and CARF) require validated documentation of client perceptions—a role for which the CSQ-8 has remained an international cornerstone for more than four decades.

Psychological Construct

The construct measured by the CSQ-8 is general, unidimensional client satisfaction with professional services. Within modern behavioral and psychometric science, client satisfaction is defined as an individual's subjective cognitive and affective evaluation of the health care services received relative to their internalized standards, expectations, desires, and baseline needs. Rather than treating satisfaction as a fragmented, multifaceted composite of disparate environmental features (such as parking accessibility, facility decor, or administrative paperwork), the CSQ focuses specifically on the core interpersonal and outcome-driven encounter between the client and the clinical program.

Although the CSQ-8 functions mathematically and conceptually as a strictly unidimensional scale, it comprehensively samples eight distinct operational facets of the overarching construct of satisfaction:

• Perceived Service Quality (Item 1): Assesses the respondent's appraisal of the technical excellence, professional competence, and interpersonal warmth demonstrated by service providers. Example: rating overall service quality along a continuum from "Poor" to "Excellent."
• Service Congruence / Type Fulfillment (Item 2): Evaluates the degree of congruence between the actual modalities of care delivered and what the client initially sought or anticipated. Example: assessing whether the client obtained the precise type of service they desired.
• Need Fulfillment (Item 3): Captures the extent to which the intervention resolved the client's presenting complaints and subjective physical or psychological requirements. Example: evaluating whether "almost all" versus "none" of the client's needs were met.
• Advocacy and Social Endorsement (Item 4): Serves as a behavioral intention indicator measuring the client's willingness to endorse the program to a friend or family member experiencing similar difficulties.
• Adequacy of Service Volume / Dosage (Item 5): Evaluates whether the frequency, duration, and aggregate quantity of the intervention were perceived as sufficient or inadequate.
• Perceived Efficacy / Coping Enhancement (Item 6): Measures perceived functional improvement and symptom reduction, assessing whether the clinical contact empowered the client to handle everyday problems more effectively or conversely exacerbated their difficulties.
• Global Evaluative Satisfaction (Item 7): Measures holistic, affective satisfaction reflecting overall emotional and cognitive evaluation of the treatment episode.
• Future Service Utilization Intent (Item 8): Assesses client loyalty and behavioral commitment, specifically whether the individual would choose to return to the specific facility should future clinical needs arise.

Theoretical Framework

The design and interpretation of the CSQ are grounded in seminal sociological, psychological, and health services theories, most notably the Expectancy-Disconfirmation Model, Avedis Donabedian's Healthcare Quality Framework, and cognitive appraisal models of subjective well-being.

The foundational conceptual model informing client satisfaction research is the Expectancy-Disconfirmation Theory (Oliver, 1980), adapted to clinical care by Larsen, Attkisson, and colleagues. Under this framework, clients approach therapeutic encounters with pre-existing, multifaceted expectations regarding how clinical care will proceed, how rapidly their distress will diminish, and how respectfully they will be treated. Satisfaction emerges not merely as a direct function of technical medical outcomes, but as a cognitive comparison between these initial expectations and the perceived performance of the service. When care matches expectations, confirmation occurs; when care exceeds initial expectations, positive disconfirmation leads to elevated satisfaction. Conversely, when clinical delivery falls below perceived standards, negative disconfirmation yields dissatisfaction. The CSQ-8 explicitly captures this disconfirmation dynamic by asking whether clients received the exact kind of service they wanted and whether their needs were fulfilled.

Furthermore, the CSQ operationalizes Avedis Donabedian's classic tripartite paradigm of healthcare quality evaluation: Structure (the physical setting, human resources, administrative organization), Process (the technical competence, interpersonal interaction, communication quality, and procedural empathy of the providers), and Outcome (the resultant change in the patient's health status, coping capacities, and subjective quality of life). The developers of the CSQ discovered during item development that while structural amenities (e.g., waiting room comfort) produce isolated noise, process (items 1, 2, and 5) and perceived outcome (items 3, 6, 7, and 8) coalesce into a single, dominant cognitive evaluation. Consequently, the scale acts as an integrated cognitive-affective synthesis of process quality and perceived outcome efficacy.

Validity

The validity of the Client Satisfaction Questionnaire has been documented through dozens of empirical investigations across international clinical and research settings.

Construct and Structural Validity

Construct validity was initially established by Larsen et al. (1979) and later reinforced by Attkisson and Zwick (1982). Item-total correlations for the CSQ-8 consistently range between .60 and .85, demonstrating that each item shares extensive variance with the overarching construct. When analyzed alongside multi-dimensional evaluation instruments, the CSQ-8 reliably loads onto a central patient satisfaction dimension, distinct from generalized psychological distress or trait neuroticism.

Convergent and Concurrent Validity

Convergent validity is illustrated by strong correlations between the CSQ-8 and alternative, independently developed satisfaction inventories, such as the Service Satisfaction Scale (SSS-30) and the Patient Satisfaction Questionnaire (PSQ-III), with correlation coefficients frequently exceeding $r = .70$ to $.80$. Moreover, CSQ-8 scores demonstrate moderate-to-strong positive associations with validated measures of the Working Alliance Inventory (WAI) ($r = .55 – .72$), indicating that patients who develop deep trust and collaborative alignment with their clinicians experience substantially higher satisfaction.

Predictive and Criterion-Related Validity

Extensive health services research demonstrates the clinical predictive utility of the CSQ-8. In seminal work by Attkisson and Zwick (1982), client satisfaction scores collected at early and midpoint stages of psychotherapy significantly predicted treatment retention and therapeutic completion. Patients reporting lower satisfaction demonstrated substantially higher hazard ratios for premature termination against medical advice and unilateral treatment drop-out. Furthermore, CSQ-8 scores consistently correlate with symptom reduction on the Symptom Checklist-90-Revised (SCL-90-R) and the Patient Health Questionnaire-9 (PHQ-9), as well as increased compliance with psychiatric medication regimens and follow-up medical appointments.

Discriminant Validity

The scale effectively discriminates between distinct service modalities and treatment settings. In controlled comparative program evaluations, the CSQ-8 reliably detects statistically significant differences in satisfaction between integrated collaborative care models versus traditional siloed outpatient clinics, as well as between specialized assertive community treatment (ACT) teams and standard case management, confirming its sensitivity to true variations in clinical quality.

Reliability

The CSQ-8 demonstrates outstanding psychometric reliability across diverse populations, service delivery models, and administrative formats.

Internal Consistency Reliability

In the pioneering validation studies conducted across community mental health centers by Larsen et al. (1979) and Attkisson and Zwick (1982), the CSQ-8 yielded a Cronbach's alpha ($lpha$) of .93. Subsequent investigations across inpatient psychiatric wards, outpatient medical clinics, substance use treatment facilities, and telehealth platforms have corroborated these metrics, with alpha coefficients consistently observed within the narrow band of .83 to .94. The McDonald's omega ($\omega_t$) composite reliability coefficients mirror these findings, demonstrating minimal measurement error and exceptional internal coherence.

Test-Retest Stability

Because client satisfaction is a state-sensitive evaluative construct subject to changing health conditions and service experiences, test-retest reliability must be measured over short, stable intervals where clinical services have ceased. In post-discharge medical and psychiatric cohorts evaluated across a two- to three-week retest window without intervening treatment changes, Pearson correlation coefficients have consistently ranged from $r = .78$ to $.88$, confirming satisfactory longitudinal stability.

Cross-Cultural and Linguistic Reliability

The CSQ-8 has undergone rigorous cross-cultural validation and translation into Spanish, French, German, Dutch, Portuguese, Japanese, Chinese, Turkish, and Scandinavian languages. Cross-national psychometric evaluations confirm that high internal consistency is maintained across linguistic adaptations, with Cronbach's alpha values universally exceeding the accepted clinical research threshold of .80.

Factor Analysis

Both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) have extensively evaluated the latent dimensional architecture of the CSQ-8.

Exploratory Factor Analysis

In initial exploratory factor analytic investigations utilizing principal components and principal axis factoring with both orthogonal (Varimax) and oblique (Promax) rotations, eigenvalues from unconstrained models consistently yield a stark, single-factor solution. The primary factor accounts for 60% to 80% of the total variance across all 8 items, with an eigenvalue typically exceeding 4.5. Subsequent factors consistently display eigenvalues well below 0.70, showing a steep scree plot drop-off that provides unequivocal evidence of essential unidimensionality.

Factor loadings for individual items onto this primary satisfaction dimension are exceptionally uniform and robust:

• Item 1 (Quality of services): $lambda = .75 – .84$
• Item 2 (Kind of service wanted): $lambda = .68 – .78$
• Item 3 (Met needs): $lambda = .74 – .86$
• Item 4 (Recommend to friend): $lambda = .78 – .88$
• Item 5 (Amount of help): $lambda = .70 – .82$
• Item 6 (Dealt effectively with problems): $lambda = .64 – .77$
• Item 7 (Overall satisfaction): $lambda = .82 – .91$
• Item 8 (Come back if seeking help again): $lambda = .75 – .87$

Confirmatory Factor Analysis (CFA)

Modern structural equation modeling (SEM) and CFA studies have tested the single-factor measurement model against alternative multidimensional configurations (e.g., separating process items from perceived outcome items). The unidimensional model demonstrates superior parsimony and excellent global goodness-of-fit indices across large samples:

• Comparative Fit Index (CFI): $.96 – .99$
• Tucker-Lewis Index (TLI): $.95 – .98$
• Root Mean Square Error of Approximation (RMSEA): $.042 – .065$ (with 90% confidence intervals spanning .030 to .075)
• Standardized Root Mean Square Residual (SRMR): $.025 – .040$

While minor residual covariances occasionally emerge between Items 4 and 8 (both capturing future behavioral intentions), specifying a single general latent factor remains the empirically recommended and theoretically sound representation of the scale.

Instrument / Measurement Tool

The CSQ-8 is a self-administered or interviewer-administered questionnaire designed for rapid clinical and research administration.

• Instrument Name: Client Satisfaction Questionnaire (CSQ-8)
• Primary Author: C. Clifford Attkisson, Ph.D., and colleagues (1979, 1982)
• Measurement Format: Paper-and-pencil questionnaire, computerized/tablet survey, or telephone interview
• Administration Time: Approximately 3 to 5 minutes
• Target Population: Adolescents and adults (ages 12+) receiving healthcare, behavioral health, psychiatric, psychological, or social services
• Number of Items: 8 items
• Response Format: 4-point Likert-type scale with customized, item-specific anchor labels. Crucially, the response continuum alternates between negative-to-positive and positive-to-negative orientations across items to attenuate habituation and acquiescence response sets.
• Scoring Methodology:
  • Each individual item is scored from 1 to 4, where 1 indicates the lowest level of satisfaction and 4 indicates the highest level of satisfaction.
  • Reverse Scoring: Items whose text anchors start from highest satisfaction to lowest (such as Items 1, 3, and 7 in certain standard layouts) must be properly recoded so that 4 always reflects maximum satisfaction and 1 reflects minimal satisfaction.
  • Total Score Calculation: The individual item scores are summed to produce a composite score ranging from 8 to 32. Higher aggregate scores indicate greater overall client satisfaction.
  • Missing Data Handling: If a respondent omits 1 or 2 items, researchers commonly impute the mean of the completed items; questionnaires missing 3 or more items are typically deemed invalid.

Permissions & Fee and Test Year

The Client Satisfaction Questionnaire (CSQ-8) was formally introduced in peer-reviewed literature in 1979 (Larsen et al., 1979) with extensive psychometric properties detailed by Attkisson and Zwick in 1982. The instrument, along with its sister scales (including the CSQ-18, CSQ-4, and the Service Satisfaction Scale), is protected by international copyright owned by Dr. C. Clifford Attkisson and the CSQ-Scales operational group.

Although the items are widely published in academic dissertations and public health reports, formal clinical, commercial, or funded research utilization of the CSQ scales requires obtaining appropriate licensing permissions and paying nominal user fees based on institutional scale and sample size. Inquiries regarding official authorized language translations, licensing agreements, electronic administration integration, and manual documentation should be directed to Dr. C. Clifford Attkisson via the University of California, San Francisco (UCSF) or the official CSQ-Scales administrative portal ([email protected]).

References

• Attkisson, C. C., & Greenfield, T. K. (1996). The Client Satisfaction Questionnaire (CSQ) Scales and the Service Satisfaction Scale-30 (SSS-30). In L. I. Sederer & B. Dickey (Eds.), Outcomes assessment in clinical practice (pp. 120–127). Williams & Wilkins.
• Attkisson, C. C., Roberts, R. E., & Pascoe, G. C. (1983). The Evaluation Ranking Scale: Clarification of methodological and procedural issues. Evaluation and Program Planning, 6(3–4), 349–358. https://doi.org/10.1016/0149-7189(83)90013-X
• Attkisson, C. C., & Zwick, R. (1982). The client satisfaction questionnaire: Psychometric properties and correlations with service utilization and psychotherapy outcome. Evaluation and Program Planning, 5(3), 233–237. https://doi.org/10.1016/0149-7189(82)90074-X
• Donabedian, A. (1988). The quality of care: How can it be assessed? JAMA, 260(12), 1743–1748. https://doi.org/10.1001/jama.1988.03410120089033
• Larsen, D. L., Attkisson, C. C., Hargreaves, W. A., & Nguyen, T. D. (1979). Assessment of client/patient satisfaction: Development of a general scale. Evaluation and Program Planning, 2(3), 197–207. https://doi.org/10.1016/0149-7189(79)90094-6
• Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460–469. https://doi.org/10.1177/002224378001700405
• Pascoe, G. C. (1983). Patient satisfaction in primary health care: A literature review and analysis. Evaluation and Program Planning, 6(3–4), 185–210. https://doi.org/10.1016/0149-7189(83)90002-2

13. Items of the Scale (Questionnaire)