Michigan Alcohol Screening Test

1. Abstract

The Michigan Alcohol Screening Test (MAST), originally devised by Dr. Melvin L. Selzer in 1971 at the University of Michigan Medical School, represents one of the earliest, most extensively evaluated, and widely implemented screening instruments for identifying alcohol dependence, hazardous drinking behaviors, and severe alcohol-related psychosocial and physiological consequences. Designed as a structured self-report questionnaire or clinician-administered interview, the instrument systematically queries the pervasive behavioral, social, occupational, medical, and legal disruptions characteristic of alcohol use disorder (AUD). Over the subsequent decades, the MAST framework has been adapted into multiple clinical and demographic variants, including the 10-item Brief MAST (Pokorny et al., 1972), the 13-item Short MAST (SMAST; Selzer et al., 1975), the Geriatric MAST (MAST-G; Joseph et al., 1995), and the 24-item adolescent and dual-substance adaptation assessing both alcohol and drug consumption patterns. Across diverse clinical and community populations, the instrument typically utilizes a dichotomous (Yes/No) response format, employing either an empirically derived differential point-weighting algorithm or unit-weighted scoring rubrics to quantify the severity of alcohol-related problems.

From a psychometric perspective, the MAST exhibits robust internal consistency across general adult, psychiatric outpatient, criminal justice, and adolescent cohorts, with Cronbach’s alpha (α) coefficients consistently ranging between .83 and .95, and test-retest reliability coefficients spanning .84 to .97 across short intervals. Construct and criterion validity are substantiated by pronounced correlations with DSM-IV-TR and DSM-5 diagnostic criteria for substance use disorders, clinical biochemical markers (such as gamma-glutamyl transferase, mean corpuscular volume, and carbohydrate-deficient transferrin), and collateral family reports. Exploratory and confirmatory factor analyses generally substantiate an overarching general factor of severity alongside multidimensional second-order facets, including Loss of Control, Interpersonal/Family Complications, Severe Somatic/Medical Consequences, and Legal/Occupational Disruptions. Despite vulnerability to social desirability bias in forensic and pre-employment settings, the MAST remains a foundational cornerstone in clinical addiction psychiatry, public health epidemiology, and adolescent substance abuse triage.

2. Keywords

Michigan Alcohol Screening Test, MAST, Alcohol Use Disorder, substance abuse screening, psychometrics, addiction assessment, clinical psychology, alcoholism diagnosis, behavioral screening, Brief MAST, SMAST, adolescent substance use

3. Authors

The foundational Michigan Alcoholism Screening Test was developed by Melvin L. Selzer, M.D., an eminent American psychiatrist and clinical researcher affiliated with the Department of Psychiatry at the University of Michigan Medical School and the Highway Safety Research Institute in Ann Arbor, Michigan, USA. Dr. Selzer pioneered empirical diagnostic instrumentation to address the high prevalence of undiagnosed alcoholism among clinical cohorts, psychiatric admissions, and drivers involved in fatal motor vehicle collisions.

Subsequent psychometric adaptations and shortened iterations were spearheaded by leading addiction scholars:

• Alex D. Pokorny, M.D., along with Byron A. Miller and Thomas E. Kanas, developed the Brief MAST (1972) at the Veterans Administration Hospital and the Department of Psychiatry, Baylor College of Medicine, Houston, Texas.
• Melvin L. Selzer, M.D., in collaboration with Aron Vinokur, Ph.D., and Louis van Rooijen, Ph.D. (1975), engineered and validated the Self-Administered Short Version of the Michigan Alcoholism Screening Test (SMAST) at the University of Michigan.
• The Geriatric adaptations (MAST-G and SMAST-G) were formulated by Frederic C. Blow, Ph.D., K. J. Joseph, M.S.W., and colleagues (1992, 1995) at the University of Michigan Alcohol Research Center and the Serious Mental Illness Treatment Research and Evaluation Center (SMITREC), Department of Veterans Affairs.
• The adolescent and dual-substance modifications (MAST-Adolescent/Substance Version) were systematized by multidisciplinary behavioral health researchers to bridge the developmental and contextual nuances of adolescent substance misuse, school disruption, peer dynamics, and polysubstance involvement.

4. Purpose

The primary clinical and psychometric purpose of the Michigan Alcohol Screening Test is to serve as an objective, standardized, and rapid screening instrument engineered to identify individuals suffering from alcohol abuse, alcohol dependence, or pathological drinking patterns that disrupt major life domains. Historically, clinical detection of alcoholism in medical, psychiatric, and institutional settings relied almost exclusively on unstructured clinical interviews, which were notoriously plagued by high rates of false negatives stemming from patient denial, minimized disclosure, lack of diagnostic consensus, and physician reluctance to confront suspected substance use pathology. The MAST was systematically designed to bypass these diagnostic bottlenecks by presenting structured, behavioral inquiries centered on the tangible manifestations and biographical sequelae of disordered drinking.

From a functional perspective, the MAST does not simply query current volumetric ethanol consumption or daily unit intake (such as grams of absolute alcohol per day); rather, it measures the persistent life disruption, loss of behavioral control, subjective distress, and socio-environmental friction generated by an individual's drinking patterns over their lifetime or over recent developmental intervals. Consequently, the instrument addresses:

• Diagnostic Stratification and Secondary Prevention: Differentiating social, normative drinkers from problem drinkers and severely dependent individuals requiring structured detoxification, inpatient rehabilitation, or specialized outpatient pharmacotherapy and psychotherapy.
• Screening Across Diverse Settings: Providing primary care clinics, general medical wards, psychiatric emergency services, and mental health centers with a validated questionnaire to flag hidden alcohol-related morbidity that masquerades as primary depression, anxiety, insomnia, or unexplained gastrointestinal illness.
• Forensic and Occupational Evaluations: Quantifying substance-related impairment for forensic assessments, court-mandated evaluations following Driving Under the Influence (DUI) and Driving While Intoxicated (DWI) convictions, employee assistance programs (EAPs), and occupational fitness-for-duty determinations.
• Adolescent and Dual-Substance Screening: In its modified adolescent and combined alcohol/drug variations, the instrument evaluates youth-specific systemic impairment—including academic deterioration, peer relationship fallout, parental conflict, and polysubstance use—prior to the emergence of end-stage chronic medical morbidities.

By shifting the assessment paradigm from subjective clinical impression to standardized, observable behavioral markers, the MAST established an empirical foundation that continues to inform modern screening, brief intervention, and referral to treatment (SBIRT) protocols globally.

5. Psychological Construct

The psychological construct captured by the MAST is Alcohol Use Disorder Severity, conceptualized as a chronic, multidimensional behavioral syndrome characterized by impaired self-regulation over substance intake, persistent cognitive rationalization, escalating neurobiological tolerance and withdrawal, and the continued consumption of alcohol despite mounting interpersonal, occupational, legal, physical, and psychiatric harm. The operationalization of this construct spans several key theoretical and behavioral dimensions:

1. Impaired Control and Inability to Moderate

Central to the conceptualization of addiction is the executive failure of self-regulation. This dimension evaluates the individual’s subjective perception of normative control versus the reality of behavioral dysregulation. Items assess whether the respondent perceives themselves as a "normal drinker/user," whether they can terminate drinking after one or two drinks without an internal psychological struggle, and whether they can discontinue consumption at will. Psychometrically, an endorsement of inability to stop reflects diminished prefrontal inhibitory control over reward-seeking pathways.

2. Cognitive and Emotional Distress (Guilt and Subjective Awareness)

This facet captures the internal psychological fallout of alcohol misuse. Disordered drinking is frequently accompanied by profound affective dissonance, manifest as subjective guilt, remorse, self-reproach, and perceived loss of self-efficacy following drinking episodes. Queries targeting whether the individual "feels bad" or guilty about drinking tap into internal conflict and demoralization, which frequently coexist with escalating dependence.

3. Interpersonal, Familial, and Social Disruption

Substance use disorders exert catastrophic effects on systemic attachment and social networks. The MAST operationalizes this construct by examining external conflict: marital disputes, parental worry, severed friendships, domestic complaints, and family members seeking outside professional or ministerial counseling to cope with the respondent's substance consumption. Importantly, this domain incorporates collateral societal feedback, querying whether friends or relatives view the individual's drinking as abnormal, thereby counterbalancing internal minimization.

4. Functional, Occupational, and Academic Role Failure

This operational domain maps directly onto the functional role-impairment criteria articulated across contemporary psychiatric taxonomies. It assesses the behavioral neglect of fundamental responsibilities—such as missing school or work for two or more consecutive days, experiencing disciplinary action at school, getting reprimanded at work, or being terminated from employment. In adolescent cohorts, this dimension heavily weights academic decline, truancy, and behavioral infractions in educational environments.

5. Neurobiological Adaptation, Withdrawal, and Somatic Pathology

At advanced levels of the severity continuum, the construct taps severe physiological adaptation. Items query neurovegetative and cognitive disturbances such as alcohol-induced anterograde amnesia (blackouts), morning drinking or consumption prior to noon to stave off withdrawal, severe tremors, hallucinations, delirium tremens (DTs), and permanent organ damage such as hepatic cirrhosis. These manifestations signify moderate-to-severe physiological neuroadaptation within gamma-aminobutyric acid (GABA) and N-methyl-D-aspartate (NMDA) receptor systems.

6. Behavioral Extinction Failures: Help-Seeking and Legal/Forensic Repercussions

The final facet operationalizes an individual’s encounter with institutional societal boundaries. Endorsement of attendance at Alcoholics Anonymous (AA) or Narcotics Anonymous (NA) meetings, seeking formal professional or pastoral help, prior psychiatric or medical hospitalizations for drinking, and repeated interactions with law enforcement (such as non-DUI behavioral arrests, alcohol-related physical fights, and DUI/DWI vehicular citations) serve as definitive indicators of clinically severe addiction.

6. Theoretical Framework

The design and scoring architecture of the Michigan Alcohol Screening Test were directly influenced by the mid-twentieth-century evolution of addiction conceptualizations, primarily integrating the Disease Concept of Alcoholism, early behavioral conditioning paradigms, and systemic social deviance theories.

The Disease Model and Jellinek's Typology

The foundational bedrock of Selzer’s work is rooted in the medicalization of alcoholism advanced by E. M. Jellinek in his seminal 1960 treatise, The Disease Concept of Alcoholism. Jellinek posited that alcoholism is not a moral failure or simple lack of willpower, but a progressive, distinct clinical disease entity characterized by predictable symptomatological phases: the pre-alcoholic symptomatic phase, the prodromal phase (marked by blackouts and secretive drinking), the crucial phase (characterized by complete loss of control, social friction, and rationalizations), and the chronic phase (marked by prolonged intoxications, physical tremors, delirium tremens, and end-organ systemic degeneration).

Selzer explicitly mapped the MAST questions onto Jellinek’s progressive phases. In the original 1971 weighting schema, items reflecting prodromal and crucial stages (e.g., blackouts, spousal complaints) were assigned 1 or 2 points, whereas pathognomonic markers of late crucial and chronic alcoholism—such as experiencing delirium tremens, seeking formal psychiatric hospitalization for alcohol problems, or attending Alcoholics Anonymous—were assigned arbitrary heavy weights (5 points each). Under Selzer's classical model, a single endorsement of a 5-point item was considered definitively diagnostic of alcoholism.

Behavioral and Conditioning Paradigms

Concurrently, the MAST incorporates principles of operant and classical conditioning. Substance use behaviors are initiated and maintained via positive reinforcement (euphoria, social facilitation) and negative reinforcement (alleviation of anxiety, suppression of withdrawal dysphoria, coping with life stressors). Morning drinking ("drinking before noon") represents a classic negative reinforcement paradigm designed to terminate autonomic hyperarousal following nocturnal alcohol clearance. Furthermore, items evaluating escalating conflict and repeated arrests capture the behavioral hallmark of addiction: the persistence of drug-seeking behavior despite severe, contingent aversive stimuli (punishment insensitivity).

Social Learning and Family Systems Theory

The instrument also integrates constructs from social learning theory and family systems theory. Addictive behaviors do not occur in an environmental vacuum; they profoundly reshape family homeostasis. The inclusion of questions addressing whether family members worry, complain, or seek outside guidance reflects the systemic accommodation and codependent crisis interventions that emerge as the family struggles to maintain equilibrium in the face of progressive behavioral volatility.

7. Validity

The psychometric validity of the MAST and its derivatives has been rigorously scrutinized across five decades of clinical, community, and forensic research.

Criterion and Predictive Validity

In his initial validation study, Selzer (1971) administered the 25-item weighted MAST to five distinct cohorts: a hospitalized alcoholic validation cohort ($N = 116$), individuals convicted of driving under the influence ($N = 99$), drivers involved in fatal vehicular crashes ($N = 98$), individuals undergoing psychiatric outpatient treatment ($N = 101$), and a non-clinical control group ($N = 51$). The instrument demonstrated extraordinary discriminative power: 98% of the hospitalized alcoholic sample scored $ge 5$ points (the designated diagnostic threshold), whereas 95% of non-clinical controls scored within the non-alcoholic range (0–3 points). Subsequent investigations confirmed that the MAST accurately classified clinical alcoholics with sensitivity rates ranging from 84% to 98% and specificity rates from 78% to 95% when evaluated against structured clinical interviews based on DSM-III, DSM-IV, and DSM-5 criteria.

Predictive validity has been repeatedly demonstrated in longitudinal studies evaluating treatment outcomes and recidivism. High baseline MAST scores correlate significantly with higher rates of post-discharge relapse, recurrent DUI offenses, and elevated long-term all-cause mortality. In a meta-analytic review of psychiatric inpatients by Teitelbaum and Mullen (2000), the MAST yielded an aggregate effect size ($d > 1.4$) for differentiating substance-abusing psychiatric patients from non-substance-abusing counterparts, confirming its predictive utility even amidst confounding severe mental illnesses like schizophrenia and bipolar disorder.

Convergent and Discriminant Validity

The MAST exhibits substantial convergent validity when correlated with alternative screening instruments and biological markers. Studies evaluating concurrent administration have reported high product-moment correlations between the MAST and the Alcohol Use Disorders Identification Test (AUDIT; $r = .72$ to $.88$), the CAGE Questionnaire ($r = .65$ to $.79$), and the MacAndrew Alcoholism Scale of the MMPI ($r = .58$ to $.71$). Furthermore, elevated MAST scores correlate moderately but significantly with biological markers of chronic excessive alcohol consumption, including serum gamma-glutamyl transferase (GGT; $r = .38$ to $.52$), erythrocyte mean corpuscular volume (MCV; $r = .32$ to $.45$), and carbohydrate-deficient transferrin (CDT; $r = .41$ to $.59$).

Discriminant validity has been demonstrated by showing lower correlations between the MAST and measures of generalized state anxiety ($r < .25$), trait neuroticism ($r < .30$), or unipolar major depression when controlled for co-occurring addictive pathology, confirming that the tool specifically measures alcohol-related behavioral disruption rather than generalized diffuse psychological distress.

Contextual Validity and Response Distortion

A notable limitation of the MAST involves its susceptibility to impression management and social desirability. Because the face validity of items is exceptionally transparent (e.g., asking directly about AA attendance, arrests, or DTs), respondents in high-stakes forensic or employment evaluations can deliberately manipulate their responses. Research in driver licensing restoration and child custody disputes has demonstrated an elevation in false-negative rates, leading psychometricians to recommend supplementing the self-report MAST with collateral interviews, urine toxicological screens, and indirect psychometric indices in forensic contexts.

8. Reliability

Extensive psychometric investigations have established that the MAST and its brief iterations possess high internal consistency and temporal stability across diverse clinical, non-clinical, and demographic cohorts.

Internal Consistency

In his foundational review, Gibbs (1985) aggregated reliability data from over thirty peer-reviewed studies, documenting that Cronbach’s alpha (α) coefficients for the full-length MAST consistently range between .83 and .95 across adult medical inpatients, psychiatric admissions, and outpatient substance abuse treatment populations. For the 10-item Brief MAST (Pokorny et al., 1972), internal consistency remains robust, with alpha coefficients typically falling between .78 and .88. The 13-item SMAST (Selzer et al., 1975) demonstrates comparable internal reliability ($lpha = .81$ to $.90$). Even in vulnerable sub-populations, such as geriatric outpatients evaluated with the MAST-G or SMAST-G (Joseph et al., 1995; Hirata et al., 2001), alpha coefficients exceeding .84 have been consistently observed. Split-half reliability coefficients calculated using the Spearman-Brown prophecy formula generally parallel these figures, yielding split-half reliability estimates between .84 and .92.

Test-Retest Stability

The temporal stability of the instrument is exceptionally high, a reflection of the fact that the MAST predominantly queries cumulative lifetime historical events (e.g., "Have you ever been arrested…") rather than transient, fluctuating emotional states. Test-retest reliability coefficients across intervals spanning from one day to several months have yielded Pearson product-moment correlations and intraclass correlation coefficients (ICCs) between .84 and .97. Selzer and colleagues documented that re-administration to alcoholic outpatients after several weeks yielded stable diagnostic categorizations in over 92% of participants. Nevertheless, when administering the instrument longitudinally to evaluate treatment efficacy, investigators must attend to temporal instructions, as cumulative "lifetime" item phrasing restricts the instrument's sensitivity to detect recent post-intervention behavioral changes unless explicitly modified to query a specific time frame (such as the past 6 or 12 months).

9. Factor Analysis

Although Melvin Selzer initially conceptualized the MAST as a clinically unidimensional index of alcoholism severity, several decades of exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) have uncovered a robust multidimensional architectural framework.

Exploratory Factor Analyses (EFA)

Early psychometric deconstruction by Skinner (1979) and Zung (1980) challenged pure unidimensionality. Zung conducted an EFA utilizing principal components analysis with varimax rotation in a large psychiatric outpatient cohort ($N = 1,185$), identifying four to five clinically interpretable latent factors explaining over 52% of the total variance:

• Factor 1: Loss of Control / Pathological Drinking Pattern: Characterized by high loadings (> .60) on items assessing perceived normal drinking, inability to terminate drinking after one or two drinks, morning drinking, and frequent subjective guilt.
• Factor 2: Help-Seeking and Treatment History: Dominated by items measuring formal interventions, including attendance at Alcoholics Anonymous, consultations with physicians or counselors, and prior hospitalizations for drinking.
• Factor 3: Severe Somatic and Neuropsychiatric Consequences: Capturing advanced physiological damage, heavily loading on alcohol withdrawal delirium tremens, hallucinations, and chronic hepatic cirrhosis.
• Factor 4: Interpersonal and Familial Conflict: Emphasizing spousal, marital, and parental complaints, loss of social relationships, and domestic strife.
• Factor 5: Legal and Occupational Sequelae: Defined by alcohol-related arrests (DUI and public drunkenness), physical altercations, work or school disciplinary actions, and loss of employment.

Confirmatory Factor Analyses (CFA) and Modern Structural Equation Modeling

Contemporary psychometric evaluations employing CFA (e.g., Thurber et al., 2001) have systematically compared alternative structural models: a strict single-factor model, an uncorrelated orthogonal multi-factor model, and a hierarchical (bifactor or higher-order) model. The empirical consensus indicates that a higher-order factor model or a bifactor model exhibits superior goodness-of-fit indices:

• Comparative Fit Index (CFI) $ge .94$
• Tucker-Lewis Index (TLI) $ge .93$
• Root Mean Square Error of Approximation (RMSEA) $le .048$ (90% CI [.041, .055])
• Standardized Root Mean Square Residual (SRMR) $le .051$

In this hierarchical structure, an overarching general factor—representing global Alcohol Problem Severity—subsumes distinct, inter-correlated second-order dimensions (Impaired Control, Socio-Occupational Distress, Help-Seeking, and Medical/Forensic Complications). Modern Item Response Theory (IRT) analyses, utilizing two-parameter logistic (2PL) models, further substantiate that items representing formal treatment and severe withdrawal have high item discrimination parameters ($lpha > 1.8$) and locate at the extreme positive end of the latent severity continuum ($ heta > +2.0$), while items assessing family complaints and subjective guilt operate effectively at lower levels of the latent severity continuum ($ heta pprox 0.0$ to $+1.0$).

10. Instrument / Measurement Tool

• Instrument Name: Michigan Alcohol Screening Test (MAST) [including Adolescent and Substance-Adapted 24-Item Variant]
• Original Developer: Melvin L. Selzer, M.D. (1971)
• Assessment Type: Standardized Clinical Screening Tool / Self-Report Questionnaire / Structured Interview
• Target Population: Adults, young adults, and adolescents suspected of alcohol or substance misuse across clinical, outpatient, forensic, educational, and research settings
• Administration Time: Approximately 8 to 12 minutes
• Item Count: 24 substantive items (in the standardized adolescent and substance-adapted formulation presented herein)
• Response Scale: Dichotomous format (Yes / No)
• Scoring and Scoring Rules:
  • Directionality: For the majority of items, an affirmative (Yes) response indicates problematic substance-related behavior, interpersonal conflict, or somatic impairment.
  • Reverse-Scored Items: Items 1, 4, 6, and 7 assess self-perceived normative control and self-regulation (e.g., "Do you feel you are a normal drinker/user?"; "Can you stop drinking/using without a struggle…?"; "Do friends or relatives think you are a normal drinker/user?"; "Are you always able to stop drinking/using when you want to?"). For these specific items, a negative (NO) response indicates substance pathology.
  • Scoring Algorithms:
    • Unit-Weighted Scoring: Each response indicative of substance misuse is assigned 1 point (Total Score range: 0–24 points). In adolescent and general primary care settings, unit-weighted scoring is frequently utilized for ease of clinical administration: scores of 0–2 indicate low risk / normative use; 3–4 suggest borderline / mild risk warranting further monitoring; and scores $ge 5$ indicate significant risk / probable substance use disorder.
    • Weighted Scoring (Classical Selzer Model): In the classical adult Selzer framework, differential point values (1, 2, or 5 points) are assigned to items based on diagnostic discriminative power. A cumulative score of 0–3 indicates no problem; 4 points indicates suggestive problem drinking; and 5 or more points is considered diagnostic of alcoholism/severe substance use disorder. A single positive response to a 5-point item (e.g., prior hospitalization, delirium tremens, or AA/NA attendance) historically triggers a presumptive positive screen.

11. Permissions & Fee and Test Year

The Michigan Alcoholism Screening Test was originally published by Dr. Melvin L. Selzer in the American Journal of Psychiatry in 1971. As a seminal assessment developed under academic and public research auspices, the original MAST, SMAST, Brief MAST, and their clinical adaptations have historically entered the public domain for clinical, educational, non-profit, and academic research purposes. No royalty fees or mandatory commercial purchasing licensing models are required to administer, score, or interpret the instrument in standard behavioral healthcare or scholarly research.

Clinicians and investigators must, however, maintain intellectual property integrity by providing full academic attribution to Dr. Melvin L. Selzer and the University of Michigan Medical School in all professional publications, clinical manuals, and derivative clinical systems. Commercial integration into proprietary electronic health record (EHR) systems or software platforms intended for commercial resale should respect relevant trademark and original publication copyright protocols established by the American Psychiatric Association and respective publishing bodies.

12. References

• Blow, F. C., Brower, K. J., Schulenberg, J. E., Demo-Dananberg, L. M., Young, J. P., & Beresford, T. P. (1992). The Michigan Alcoholism Screening Test – Geriatric Version (MAST-G): A new elderly-specific screening instrument. Alcoholism: Clinical and Experimental Research, 16(2), 372.
• Bradley, K. A., Boyd-Wickizer, J., Powell, S. H., & Burman, M. L. (1998). Alcohol screening questionnaires in women: A critical review. JAMA: The Journal of the American Medical Association, 280(2), 166–171. https://doi.org/10.1001/jama.280.2.166
• Gibbs, L. E. (1985). Validity and reliability of the Michigan Alcoholism Screening Test: A review. Drug and Alcohol Dependence, 12(3), 279–285. https://doi.org/10.1016/0376-8716(83)90019-3
• Hirata, E. S., Almeida, O. P., Funari, R. R., & Klein, E. L. (2001). Validity of the Michigan Alcoholism Screening Test (MAST) for the detection of alcohol-related problems among male geriatric outpatients. The American Journal of Geriatric Psychiatry, 9(1), 30–34. https://doi.org/10.1097/00019442-200102000-00006
• Jellinek, E. M. (1960). The disease concept of alcoholism. Hillhouse Press.
• Joseph, K. J., Blow, F. C., & Wallace, S. P. (1995). The Short Michigan Alcoholism Screening Test – Geriatric Version (SMAST-G). Alcoholism: Clinical and Experimental Research, 19(Suppl 2), 69A.
• Maisto, S. A., Connors, G. J., & Allen, J. P. (1995). Contrasting self-report screens for alcohol problems: A review. Alcoholism: Clinical and Experimental Research, 19(6), 1510–1516. https://doi.org/10.1111/j.1530-0277.1995.tb01016.x
• Pokorny, A. D., Miller, B. A., & Kanas, T. E. (1972). The Brief MAST: A shortened version of the Michigan Alcoholism Screening Test. American Journal of Psychiatry, 129(3), 342–345. https://doi.org/10.1176/ajp.129.3.342
• Selzer, M. L. (1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127(12), 1653–1658. https://doi.org/10.1176/ajp.127.12.1653
• Selzer, M. L., Vinokur, A., & van Rooijen, L. (1975). A self-administered Short Michigan Alcoholism Screening Test (SMAST). Journal of Studies on Alcohol, 36(1), 117–126. https://doi.org/10.15288/jsa.1975.36.117
• Skinner, H. A. (1979). A multivariate evaluation of the Michigan Alcoholism Screening Test. Journal of Studies on Alcohol, 40(9), 831–844. https://doi.org/10.15288/jsa.1979.40.831
• Teitelbaum, L., & Mullen, B. (2000). Validity of the MAST in psychiatric settings: A meta-analytic integration. Journal of Studies on Alcohol, 61(2), 254–261. https://doi.org/10.15288/jsa.2000.61.254
• Thurber, S., Snow, M., Lewis, D., & Hodgson, J. M. (2001). Item characteristics of the Michigan Alcoholism Screening Test. Journal of Clinical Psychology, 57(1), 139–144. https://doi.org/10.1002/1097-4679(200101)57:1<139::aid-jclp14>3.0.co;2-a
• Zung, B. J. (1980). Factor structure of the Michigan Alcoholism Screening Test (MAST) in a psychiatric outpatient population. Journal of Clinical Psychology, 36(4), 1024–1030. https://doi.org/10.1002/1097-4679(198010)36:4<1024::aid-jclp2270360443>3.0.co;2-y

13. Items of the Scale