Multidimensional Health Locus of Control Scale

1. Abstract

The Multidimensional Health Locus of Control (MHLC) Scale is one of the most widely implemented psychometric instruments in health psychology, behavioral medicine, and clinical epidemiology. Developed by Kenneth A. Wallston, Brenda S. Wallston, and Robert DeVellis in 1978, the instrument was engineered to measure an individual’s generalized expectancies regarding where the control over their health outcomes resides. Grounded in Julian Rotter‘s social learning theory and locus of control paradigm, as well as subsequent multidimensional reformulations by Hanna Levenson, the MHLC departed from unidimensional conceptualizations of health-related mastery by partitioning attributional beliefs into three distinct orthogonal dimensions: Internal Health Locus of Control (IHLC), which captures beliefs that personal behaviors, lifestyle, and agency determine wellness and disease recovery; Powerful Others Health Locus of Control (PHLC), reflecting the conviction that health outcomes are directed by medical professionals, physicians, nurses, family members, or external authorities; and Chance Health Locus of Control (CHLC), operationalizing the fatalistic expectation that health, disease onset, and recovery are predominantly determined by luck, fate, chance, or uncontrollable external circumstances.

The standard instrument (comprising parallel Forms A and B, alongside condition-specific Form C developed in 1994) consists of 18 self-administered declarative items evenly distributed across the three dimensions, with each subscale measured via six items evaluated along an authentic 6-point Likert response scale ranging from 1 (“Strongly Disagree”) to 6 (“Strongly Agree”). Extensive empirical psychometric evaluations over more than four decades substantiate the instrument’s structural integrity, factorial stability, and cross-cultural transportability across diverse clinical cohorts, including patients suffering from chronic pain, diabetes mellitus, cardiovascular conditions, cancer, and rheumatologic disorders. Internal consistency estimates (Cronbach’s alpha) typically fall between .67 and .86 across normative and clinical samples, while confirmatory factor analytic studies consistently demonstrate the structural superiority of a three-factor oblique or orthogonal model over unidimensional and bidimensional alternatives. Criterion, convergent, and discriminant validity have been confirmed against operational measures of health literacy, psychological coping mechanisms, medication adherence regimens, self-efficacy, and biological prognostic indicators. The MHLC remains a cornerstone tool for elucidating patient agency, optimizing tailored behavioral interventions, and predicting therapeutic engagement across global healthcare settings.

2. Keywords

Multidimensional Health Locus of Control, MHLC, Kenneth A. Wallston, health psychology, internal health locus of control, powerful others locus of control, chance locus of control, psychometrics, health behavior, medical adherence, social learning theory, patient autonomy, chronic disease management

3. Authors

The Multidimensional Health Locus of Control Scale was created by a pioneering team of behavioral scientists and health psychologists based at Vanderbilt University:

• Kenneth A. Wallston, Ph.D.: Professor of Psychology in Nursing at Vanderbilt University School of Nursing, Senior Research Associate at the Institute for Public Policy Studies, and a foundational figure in health-related social learning theory and behavioral psychometrics.
• Brenda S. Wallston, Ph.D.: Associate Professor of Psychology at George Peabody College for Teachers and Vanderbilt University, whose scholarship shaped early psychological research into perceived control, patient decision-making, and gender-related health attitudes.
• Robert F. DeVellis, Ph.D.: Distinguished psychometrician and Emeritus Professor of Health Behavior at the University of North Carolina at Chapel Hill, known broadly for his seminal publications on scale development and measurement theory.

Subsequent modifications, including the condition-specific Form C designed for populations diagnosed with established chronic illnesses, were published in 1994 by Kenneth A. Wallston, Cheryl A. Stein, and Craig A. Smith at Vanderbilt University.

4. Purpose

The primary clinical and empirical purpose of the Multidimensional Health Locus of Control Scale is to measure the generalized attributional beliefs that individuals maintain concerning the factors that govern, maintain, or compromise their physical health status. Prior to the late 1970s, psychological research into patient perceptions relied predominantly on unidimensional locus of control assessments—such as Julian Rotter’s Internal-External (I-E) scale or Wallston and colleagues’ original 11-item Health Locus of Control (HLC) scale published in 1976. These early instruments measured agency along a continuous bipolar continuum, positioning an individual as either fundamentally internal (believing one has mastery over health) or external (believing that external agents or fortune govern well-being). However, empirical observations revealed that human belief architectures are far more intricate: an individual can concurrently endorse strong confidence in personal lifestyle modifications (internal control) while simultaneously expressing high deference to specialized medical professionals (powerful others control), without regarding these two beliefs as mutually exclusive.

By disaggregating external attributions into Powerful Others (social external) and Chance (fatalistic external) vectors, the MHLC serves multiple critical purposes across healthcare systems and clinical trials:

• Assessment of Patient Engagement and Self-Management: In chronic diseases requiring complex daily self-care regimens—such as type 1 and type 2 diabetes mellitus, chronic kidney disease, hypertension, and congestive heart failure—the MHLC allows clinicians to identify whether patients perceive that their daily adherence directly impacts long-term health outcomes. Individuals exhibiting low internal health locus of control alongside elevated chance attributions are at heightened risk for non-adherence, passivity, and therapeutic nihilism.
• Predictive Modeling of Health-Promoting Behaviors: In primary prevention contexts, the MHLC serves as a key predictive variable in structural equation models examining dietary practices, physical activity, tobacco cessation, alcohol moderation, and cancer screening compliance (e.g., mammography, colonoscopy). Research indicates that high IHLC predicts health information seeking and preventive uptake, provided that the individual places a high value on health.
• Personalizing Behavioral and Educational Interventions: Clinicians and health educators can calibrate psychological counseling and psychoeducational programs based on a patient’s MHLC profile. For instance, a patient with high Powerful Others beliefs responds favorably to authoritative, directive medical recommendations, whereas a patient with high Internal control benefits more from autonomy-supportive, collaborative self-management goal setting.
• Investigating Coping Strategies in Acute and Chronic Illness: In life-threatening diagnoses such as oncology or organ failure, the MHLC clarifies how patients cope with functional decline. Extreme internal attribution in circumstances involving poor medical prognoses can occasionally precipitate unwarranted self-blame, psychological distress, and depressive symptomatology, whereas an adaptive balance between internal control and reliance on medical professionals facilitates cognitive adaptation.

5. Psychological Construct

The psychological construct operationalized by the MHLC is health locus of control, defined as a generalized expectancy regarding whether one’s health status, disease vulnerability, and therapeutic recovery are contingent upon internal behavioral factors or external environmental forces. Rather than conceptualizing health locus of control as a static personality trait, the construct is defined as a relatively enduring cognitive orientation that influences health-related decision-making, cognitive appraisal, and behavioral responses. The MHLC measures three distinct, non-mutually exclusive dimensions:

Internal Health Locus of Control (IHLC)

The Internal Health Locus of Control dimension measures the degree to which an individual believes that their physical wellness, biological vulnerability, and recuperative trajectory are primarily governed by their own actions, lifestyle decisions, and behavioral self-regulation. Individuals scoring high on the IHLC scale endorse statements reflecting personal responsibility, such as acknowledging that if they become ill, their own behavioral modifications dictate the speed of recovery, and that consistent self-care prevents somatic disease. Psychologically, high IHLC reflects personal agency, self-directed efficacy, and behavioral attribution. For example, a person with high IHLC facing an elevated blood pressure reading is likely to attribute the metric to dietary sodium consumption, psychological stress management, and physical inactivity, immediately seeking lifestyle interventions to normalize cardiovascular parameters.

Powerful Others Health Locus of Control (PHLC)

The Powerful Others Health Locus of Control subscale assesses the belief that an individual’s health status is dependent upon the actions, authority, and interventions of external human agents. In the general MHLC Forms A and B, these powerful others include licensed medical professionals (physicians, nurses, allied healthcare personnel) as well as close familial and social support systems. Individuals scoring high on the PHLC dimension demonstrate high reliance on clinical expertise and social support networks. For example, an individual scoring high on PHLC adheres strictly to pharmaceutical prescriptions, follows post-surgical protocols meticulously, and views regular medical consultations as the primary defensive barrier against chronic disease. Importantly, high PHLC does not inherently signify cognitive helplessness; rather, it often reflects trust in institutional healthcare delivery systems, although extreme scores can manifest as passive over-dependence on clinical practitioners.

Chance Health Locus of Control (CHLC)

The Chance Health Locus of Control subscale operationalizes fatalistic beliefs that health, disease onset, and physical recovery are governed by luck, fate, destiny, accident, or unpredictable random events. Individuals scoring high on the CHLC subscale endorse items affirming that illness occurs irrespective of personal preventative behavior, that good health is largely an accident of fortune, and that recovery depends on circumstances beyond human influence. Psychologically, elevated CHLC is associated with learned helplessness, externalized passivity, and reduced preventive health engagement. For example, a smoker scoring high on CHLC may rationalize tobacco use with the belief that developing lung cancer is predetermined by genetic destiny or sheer bad luck, concluding that behavioral cessation would exert negligible influence over long-term mortality.

6. Theoretical Framework

The theoretical architecture of the Multidimensional Health Locus of Control Scale is derived from Julian B. Rotter’s Social Learning Theory (1954, 1966) and the later multidimensional attributional models pioneered by Hanna Levenson (1973, 1974). In Rotter’s classical formulation, the potential for a particular behavior to occur in a specific situation ($BP_{x,s_1,r_a}$) is a joint function of the expectancy ($E_{x,r_a,s_1}$) that the behavior will lead to a particular reinforcement and the subjective reinforcement value ($RV_{a,s_1}$) of that outcome:

BP = f(E & RV)

Within this framework, generalized expectancies become dominant determinants of behavior when individuals encounter novel, ambiguous, or complex scenarios where specific behavioral-outcome associations have not yet been stabilized. Rotter originally operationalized locus of control as an omnibus, generalized personality variable across diverse domains of life. However, generalized locus of control scales frequently demonstrated weak predictive validity when applied to highly specialized behavioral environments, such as preventative health choices, medical adherence, and symptom reporting.

Recognizing this empirical limitation, Kenneth Wallston and Brenda Wallston postulated that domain-specific expectancy measures would demonstrate significantly higher predictive utility. They developed the original unidimensional Health Locus of Control (HLC) scale in 1976. However, contemporaneous work by Hanna Levenson demonstrated that Rotter’s unidimensional I-E construct conflated two fundamentally disparate external orientations: the belief that the world is inherently chaotic and unpredictable (Chance), versus the belief that the world is ordered and predictable, but governed by powerful individuals who wield institutional authority (Powerful Others).

Levenson’s tripartite framework (Internal, Powerful Others, Chance; IPC) provided the direct conceptual framework for the construction of the MHLC in 1978. Wallston, Wallston, and DeVellis integrated Levenson’s formulation into health-specific behavioral environments. The MHLC assumes that:

• Health locus of control orientations are multidimensional rather than positioned on a single continuum; individuals possess independent, varying levels of internal mastery, institutional medical trust, and fatalistic attribution.
• These attributional beliefs operate as generalized expectancies across health domains, but interact with specific outcome expectancies, health values, and perceived self-efficacy (as subsequently articulated in Albert Bandura’s Social Cognitive Theory).
• An attributional belief alone is insufficient to predict health-promoting behaviors; it must be evaluated in conjunction with the subjective value an individual places on maintaining physical health. An individual with high IHLC who assigns low value to physical longevity is unlikely to adopt rigorous dietary or athletic routines.

7. Validity

The construct, convergent, discriminant, and predictive validity of the MHLC has been examined across empirical investigations spanning general population cohorts, outpatient clinics, and specialized inpatient medical settings.

Construct and Factorial Validity

The tripartite construct validity of the MHLC was initially demonstrated in the seminal validation study by Wallston, Wallston, and DeVellis (1978). In a developmental sample comprising 115 adult community members and an analytical cross-validation sample of 85 adults, principal factor analyses with orthogonal (varimax) and oblique rotations demonstrated that the 18 items loaded cleanly onto three distinct factors, corresponding directly to the theoretical constructs of IHLC, PHLC, and CHLC. The three-factor solution explained substantial total variance, with factor intercorrelations confirming theoretical expectations: IHLC demonstrated modest to negligible correlations with PHLC (typically ranging between $r = -.10$ and $r = .15$) and weak negative correlations with CHLC ($r = -.15$ to $-.30$). In contrast, PHLC and CHLC consistently exhibited mild to moderate positive intercorrelations ($r = .20$ to $.35$), corroborating the proposition that both represent externalized attributional paradigms while maintaining psychometric distinctiveness.

Convergent and Discriminant Validity

Convergent validity has been established through systematic correlations with related psychological constructs:

• Levenson’s IPC Scales: Wallston et al. (1978) demonstrated strong convergent associations between the MHLC subscales and Levenson’s generalized Internal ($r = .56$), Powerful Others ($r = .28$), and Chance ($r = .55$) scales, establishing that the MHLC successfully captured the broader cognitive dimensions within a health-specific sphere.
• General Self-Efficacy and Health Self-Efficacy: Subsequent studies (e.g., Schwarzer & Fuchs, 1996) reported robust positive correlations between IHLC and generalized self-efficacy ($r = .40$ to $.52$), alongside negative associations between CHLC and perceived self-efficacy ($r = -.35$ to $-.48$).
• Discriminant Validity from Social Desirability: Correlations between the three MHLC subscales and the Marlowe-Crowne Social Desirability Scale have repeatedly been established as non-significant or trivial ($r < .15$), verifying that responses on the MHLC are not confounded by self-presentation bias or the desire to project social conformity.

Predictive and Criterion Validity

Criterion-related validity is evidenced by the MHLC’s capacity to predict clinically relevant health outcomes, behavioral adherence, and physiological markers. In studies involving patients with type 2 diabetes mellitus, high IHLC scores combined with low CHLC scores systematically predict lower glycosylated hemoglobin ($HbA_{1c}$) levels, greater self-monitoring of blood glucose, and structured dietary adherence. In oncology settings, elevated CHLC scores correlate with higher scores on the Hospital Anxiety and Depression Scale (HADS) and lower perceived quality of life, reflecting the detrimental impact of pervasive fatalism during intensive medical interventions. Conversely, elevated PHLC scores reliably predict adherence to complex pharmacological regimens and attendance at clinical follow-up appointments among renal transplant recipients and cardiac rehabilitation participants.

8. Reliability

Psychometric evaluations across diverse demographic groups, healthy cohorts, and chronic illness populations have established that the MHLC demonstrates acceptable to high levels of internal consistency and temporal stability.

Internal Consistency Reliability

In the original instrument development studies (Wallston et al., 1978), Cronbach’s alpha coefficients for the two parallel forms (Form A and Form B) established the baseline internal consistency of the 6-item subscales:

• Internal Health Locus of Control (IHLC): Form A $\alpha = .77$; Form B $\alpha = .71$. In subsequent normative and international adaptations, Cronbach’s alpha coefficients for the IHLC subscale typically range from $.72$ to $.86$.
• Powerful Others Health Locus of Control (PHLC): Form A $\alpha = .67$; Form B $\alpha = .72$. In specialized clinical samples where healthcare provider interactions are prominent, the alpha coefficient frequently increases to values between $.75$ and $.83$.
• Chance Health Locus of Control (CHLC): Form A $\alpha = .75$; Form B $\alpha = .75$. Cross-cultural evaluations have consistently shown that the CHLC subscale maintains alpha coefficients ranging from $.70$ to $.82$.

When Form A and Form B are combined into 12-item composite subscales (e.g., IHLC A+B), the internal consistency coefficients rise to $.83$ for IHLC, $.83$ for PHLC, and $.84$ for CHLC, reflecting the psychometric benefits of increased scale length.

Test-Retest Reliability and Temporal Stability

Because locus of control orientations represent generalized cognitive expectancies rather than momentary affective states, the MHLC demonstrates substantial temporal stability over longitudinal testing intervals. Initial test-retest reliability assessments conducted across a 6-week interval in community adults yielded Pearson correlation coefficients of $r = .71$ for IHLC, $r = .74$ for PHLC, and $r = .72$ for CHLC. Longitudinal evaluations spanning three to six months in stable medical cohorts (such as non-progressive osteoarthritis or stable hypertension) have shown intraclass correlation coefficients (ICCs) between $.65$ and $.78$. In interventions explicitly designed to restructure health attributions—such as cognitive-behavioral psychoeducation or chronic disease self-management programs—IHLC scores exhibit statistically significant, systematic increases, confirming the scale’s sensitivity to cognitive and behavioral change over time.

9. Factor Analysis

The underlying dimensionality of the Multidimensional Health Locus of Control Scale has been extensively evaluated through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) across numerous cultural adaptations and clinical groups.

Exploratory Factor Analysis (EFA)

During the developmental psychometric procedures outlined by Wallston et al. (1978), a pool of 80 potential items was subjected to principal axis factoring and principal component analysis. Iterative item reduction strategies based on factor loadings, conceptual clarity, and item-to-total correlations resulted in the retention of 18 items. When rotated using both orthogonal (varimax) and oblique (promax/oblimin) criteria, the eigenvalues and scree plot inspection confirmed a clear three-factor structure. Items designed to assess personal responsibility loaded robustly onto Factor 1 (IHLC, with factor loadings ranging from $.45$ to $.78$). Items operationalizing the intervention of doctors, nurses, and family members loaded onto Factor 2 (PHLC, with factor loadings from $.41$ to $.75$). Statements attributing health outcomes to luck, fortune, and fate loaded onto Factor 3 (CHLC, with factor loadings from $.52$ to $.79$). Cross-loadings across alternative factors were minimal, generally remaining below $.25$.

Confirmatory Factor Analysis (CFA) and Model Fit Indices

Subsequent contemporary structural validation using Confirmatory Factor Analysis has rigorously tested competing models across various populations, including healthy adults, chronic illness cohorts, and non-Western clinical groups. The primary competing models evaluated in the psychometric literature include:

1. A One-Factor General Model: Specifying that all 18 items load onto a single omnibus health mastery factor. This model routinely demonstrates unacceptable fit indices (e.g., $\chi^2/df > 6.0$, RMSEA $> .12$, CFI $< .60$).
2. A Two-Factor Model: Partitioning items into an Internal dimension and a unitary External dimension (combining Powerful Others and Chance items). This model likewise fails to meet standard structural criteria (e.g., $\chi^2/df > 4.5$, RMSEA $> .09$, CFI $< .75$).
3. A Three-Factor Independent/Correlated Model: Preserving the distinct latent constructs of IHLC, PHLC, and CHLC. Structural equation modeling across large clinical cohorts (e.g., Ward, 1993; Malcarne et al., 2005) confirms that the three-factor correlated model exhibits superior fit indices across modern goodness-of-fit benchmarks:

Representative global fit parameters for the three-factor correlated model reported across peer-reviewed CFA literature demonstrate:

• Comparative Fit Index (CFI): $.92$ to $.96$, exceeding the established $.90$ threshold for acceptable model fit.
• Tucker-Lewis Index (TLI): $.91$ to $.95$, confirming adequacy relative to null baseline models.
• Root Mean Square Error of Approximation (RMSEA): $.045$ to $.062$ (with 90% confidence intervals spanning $.038$ to $.071$), satisfying the benchmark for close approximate fit ($< .08$).
• Standardized Root Mean Square Residual (SRMR): $.048$ to $.065$, demonstrating minimal residual covariance.

Multigroup invariance analyses across gender, age brackets, and diagnostic categories have substantiated metric and scalar invariance, verifying that the 18 items measure the three theoretical constructs equivalently across diverse demographic groups.

10. Instrument / Measurement Tool

• Test Type: Standardized self-report psychological rating scale; domain-specific cognitive expectancy inventory.
• Target Population: Adults (18 years and older); applicable to both community populations and clinical patients diagnosed with acute or chronic medical conditions.
• Administration Format: Self-administered paper-and-pencil questionnaire, digital/web-based computerized survey, or structured clinical interview.
• Total Number of Items: 18 declarative statements.
• Subscales:
  • Internal Health Locus of Control (IHLC): 6 items (Items 1, 6, 8, 12, 13, 17)
  • Powerful Others Health Locus of Control (PHLC): 6 items (Items 3, 5, 7, 10, 14, 18)
  • Chance Health Locus of Control (CHLC): 6 items (Items 2, 4, 9, 11, 15, 16)
• Authentic Response Scale: 6-point Likert scale:
  • 1 = Strongly Disagree
  • 2 = Moderately Disagree
  • 3 = Slightly Disagree
  • 4 = Slightly Agree
  • 5 = Moderately Agree
  • 6 = Strongly Agree
• Scoring Rules:
  • Each of the three subscales is scored independently by calculating the arithmetic sum of its respective 6 items.
  • No items are reverse-scored; all items are scored directly according to their numerical response weight (1 through 6).
  • Theoretical score range for each subscale: Minimum = 6, Maximum = 36.
  • The instrument produces three independent, continuous subscale scores. Scores across the three dimensions must never be summed into a total aggregate score, as they represent distinct, orthogonal constructs.
  • Alternatively, researchers may utilize median splits or cluster analyses across all three dimensions to categorize participants into typology profiles (e.g., “Pure Internal”, “Pure External”, “Double External”, “Believer in Control”, or “Fatalist”).
• Completion Time: Approximately 5 to 10 minutes.

11. Permissions & Fee and Test Year

The Multidimensional Health Locus of Control Scale was originally published in 1978 by Kenneth A. Wallston, Brenda S. Wallston, and Robert DeVellis in the Health Education Monographs (now Health Education & Behavior). Under the licensing and distribution philosophy established by Dr. Kenneth A. Wallston and Vanderbilt University, the MHLC (Forms A, B, and the subsequent condition-specific Form C published in 1994) is situated in the public domain for academic, clinical, and non-commercial research use.

No formal licensing fees, royalties, or written permissions are mandated for scholarly investigations, university-based dissertations, or non-profit healthcare evaluation protocols. The developer requested that researchers cite the foundational publications accurately in all dissemination materials and share empirical findings regarding scale properties. For commercial usage, clinical trial sponsorship by commercial pharmaceutical entities, or proprietary mobile application integration, interested organizations should contact the Vanderbilt University Office of Technology Transfer and Commercialization.

12. References

• Levenson, H. (1973). Multidimensional locus of control in psychiatric patients. Journal of Consulting and Clinical Psychology, 41(3), 397–404. https://doi.org/10.1037/h0035357
• Levenson, H. (1974). Activism and powerful others: Distinctions within the concept of internal-external control. Journal of Personality Assessment, 38(4), 377–383. https://doi.org/10.1080/00223891.1974.10119988
• Malcarne, V. L., Drahota, A., & Hamilton, N. A. (2005). Children’s health locus of control: Factor structure and validation of the Children’s Health Locus of Control Scale. Journal of Pediatric Psychology, 30(5), 437–448. https://doi.org/10.1093/jpepsy/jsi034
• Rotter, J. B. (1954). Social learning and clinical psychology. Prentice-Hall. https://doi.org/10.1037/10788-000
• Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, 80(1), 1–28. https://doi.org/10.1037/h0092976
• Wallston, B. S., Wallston, K. A., Kaplan, G. D., & Maides, S. A. (1976). Development and validation of the Health Locus of Control (HLC) Scale. Journal of Consulting and Clinical Psychology, 44(4), 580–585. https://doi.org/10.1037/0022-006X.44.4.580
• Wallston, K. A., Stein, C. A., & Smith, C. A. (1994). Form C of the MHLC Scales: A condition-specific measure of locus of control. Journal of Personality Assessment, 63(3), 534–553. https://doi.org/10.1207/s15327752jpa6303_11
• Wallston, K. A., & Wallston, B. S. (1981). Health locus of control scales. In H. M. Lefcourt (Ed.), Research with the locus of control construct: Vol. 1. Assessment methods (pp. 189–243). Academic Press.
• Wallston, K. A., Wallston, B. S., & DeVellis, R. (1978). Development of the Multidimensional Health Locus of Control (MHLC) Scales. Health Education Monographs, 6(2), 160–170. https://doi.org/10.1177/109019817800600107
• Ward, S. E. (1993). The ceiling on the Internal Health Locus of Control scale: A psychometric artifact. Research in Nursing & Health, 16(5), 375–381. https://doi.org/10.1002/nur.4770160509

13. Items of the Scale