Behaviors & Attitudes Drinking & Driving Scale (BADDS)

1. Abstract

The Behaviors & Attitudes Drinking & Driving Scale (BADDS) is a standardized, 43-item multidimensional psychometric instrument engineered to evaluate risk profiles, cognitive rationalizations, permissive normative beliefs, and self-reported behavioral incidence surrounding alcohol-impaired driving. Initially developed out of the earlier Attitudes on Drinking and Driving Scale (ADDS) to systematically measure the psychoeducational and intervention efficacy of experiential alcohol education tools (notably the Fatal Vision goggles), the BADDS operationalizes driving under the influence (DUI) risk across five discrete subscales: Rationalizations for Drinking and Driving (RD), Lenient Attitudes toward Drinking and Driving (LA), Likelihood of Drinking and Driving (LD), Drinking and Driving Behaviors (DB), and Riding Behaviors with a Drinking Driver (RB). In addition to these core evaluative domains, the instrument captures contextual nuances through 16 supplementary items indexing alcohol consumption patterns, impairment awareness, and substance misuse history, alongside five standard demographic variables.

Designed for individuals aged 18 and older possessing at least a 3.1 reading grade level, the BADDS can be administered in individual clinical sessions or large group psychoeducational contexts in approximately 10 minutes. Responses on cognitive and attitudinal dimensions utilize 5-point Likert scales, whereas direct behavioral indices assess past-month event frequency. Psychometric field trials across normative college cohorts, court-mandated DUI offenders, traffic court diversion participants, and clinical substance use disorder (SUD) treatment populations demonstrate solid internal consistency reliability and meaningful criterion-related validity. High BADDS subscale scores correspond significantly with elevated classifications on the Substance Abuse Subtle Screening Inventory-3 (SASSI-3) and an elevated propensity to operate motor vehicles under the influence of illicit or non-alcohol substances. The instrument features a structured pretest/posttest profiling architecture, facilitating sensitive detection of cognitive and behavioral change across preventative, educational, and rehabilitative interventions.

2. Keywords

Behaviors and Attitudes Drinking and Driving Scale, BADDS, Impaired Driving, DUI Assessment, Alcohol-Impaired Driving, Cognitive Rationalizations, Traffic Psychology, Substance Abuse Subtle Screening Inventory, Intervention Evaluation, Fatal Vision Goggles, Risk Assessment, Recidivism Prevention

3. Authors

The Behaviors & Attitudes Drinking & Driving Scale (BADDS) was formulated through applied psychometric and road safety research initiatives associated with alcohol intervention programming and simulated impairment technology. Its operational development evolved directly from the earlier Attitudes on Drinking and Driving Scale (ADDS), which was spearheaded to establish rigorous empirical metrics for prevention curricula utilizing experiential simulation tools. While commercial distribution and assessment support have been coordinated through specialized educational entities (such as Innocorp, Ltd., developers of the Fatal Vision simulation line), psychometric standardizations and technical manual validations have drawn upon collaborative field research spanning university health education departments, state-level judicial diversion networks, and community-based addiction treatment facilities.

Inquiries regarding manual acquisition, authorized test protocols, scoring profile sheets, and class/group change assessment modules are managed through institutional assessment distributors and specialized clearinghouses dedicated to impaired driving education and clinical substance abuse evaluation.

4. Purpose

Alcohol-impaired driving represents one of the most persistent, lethal, and economically devastating public health crises globally. Despite extensive legislative crackdowns, public service announcements, and strict law enforcement countermeasures, millions of drivers continue to operate motor vehicles while compromised by alcohol and other psychoactive substances. Research by epidemiologists such as Ralph Hingson and colleagues has consistently underscored the immense magnitude of alcohol-related morbidity and mortality, particularly among emerging adults aged 18 to 24 and repeat traffic offenders. Mitigating this risk requires not only punitive deterrents but also granular diagnostic instruments capable of identifying the cognitive substrates, permissive attitudes, and behavioral antecedents that precipitate the decision to drive while impaired.

The BADDS was explicitly engineered to address this clinical, educational, and rehabilitative imperative. Its primary purposes are threefold:

• Diagnostic Stratification and Risk Profiling: The scale enables clinicians, traffic safety educators, and court-appointed evaluators to systematically identify specific cognitive distortions, self-justifications, and social attitudes that render an individual vulnerable to impaired driving. Rather than treating DUI recidivism risk as a monolithic variable, the BADDS disentangles cognitive rationalizations from normative leniency and perceived behavioral intent.
• Intervention and Prevention Efficacy Evaluation: Developed with an integrated double-sided pretest/posttest scoring architecture, the BADDS assesses shifts in attitudes, rationalizations, and anticipated behaviors following psychoeducational interventions. It was centrally designed to quantify the direct impact of experiential interventions—such as simulated impairment goggles, victim impact panels, and cognitive-behavioral driver improvement curricula—on participant belief systems.
• Clinical Diversion and Treatment Planning: In forensic, probation, and outpatient substance use disorder (SUD) contexts, the instrument serves as an objective cross-check against denial or minimization. By correlating directly with established screening devices like the SASSI-3, the BADDS helps determine whether an offender exhibits generalized substance misuse pathology requiring intensive therapeutic care or isolated cognitive deficits amenable to targeted road safety re-education.

By measuring both active driving behaviors and passenger risk behaviors (riding with an impaired driver), the BADDS captures the full systemic dynamic of social alcohol exposure on the roadway, providing practitioners with actionable data to tailor preventive psychoeducation.

5. Psychological Construct

The conceptual framework of the BADDS posits that impaired driving is not merely a consequence of pharmacologically induced motor impairment; rather, it is mediated by a complex network of pre-existing cognitive rationalizations, perceived peer norms, risk tolerance thresholds, and established behavioral repertoires. The 43 items of the BADDS operationalize these domains across five distinct subscales, supported by alcohol history and demographic profiling:

1. Rationalizations for Drinking and Driving (RD)

The RD subscale measures the extent to which an individual endorses cognitive justifications, self-exempting beliefs, and minimization strategies designed to neutralize the internal moral dissonance of driving under the influence. Grounded in the typical excuses offered by active offenders, items in this domain evaluate notions such as driving slowly to compensate for intoxication, relying on short travel distances, assuming personal tolerance overrides legal blood alcohol concentration (BAC) thresholds, or feeling capable of “handling” alcohol better than other motorists. Clinically, high RD scores indicate rigid cognitive distortions that dismantle internal behavioral inhibitions.

2. Lenient Attitudes toward Drinking and Driving (LA)

The LA subscale gauges permissive socio-cognitive values, cultural tolerance, and minimized perceptions of the severity of impaired driving. This construct reflects broader social norms and personal ethical stances regarding whether driving after drinking is a trivial, understandable social occurrence or an egregiously dangerous transgression. Individuals with elevated LA scores demonstrate an attenuated perception of legal and moral blameworthiness, often viewing DUI citations as bad luck rather than dangerous negligence.

3. Likelihood of Drinking and Driving (LD)

The LD subscale operationalizes behavioral intention and prospective subjective probability. Drawing upon socio-cognitive decision-making paradigms, this scale asks respondents to estimate the real-world probability that they would operate a motor vehicle under varying hypothetical conditions of alcohol consumption (e.g., when no alternative transportation is readily available, when feeling “only slightly buzzed,” or after attending an unexpected social event). Elevated LD scores signify deficient self-regulatory planning and high vulnerability to impulsive or situational impaired driving.

4. Drinking and Driving Behaviors (DB)

The DB subscale moves beyond attitudinal appraisal to capture direct self-reported behavioral incidence. Quantified via focused behavioral frequency questions indexing the preceding 30-day window, this construct assesses the exact number of times an individual operated a motor vehicle after consuming alcohol, cross-referenced with subjective perceptions of impairment during those driving episodes. This scale provides a baseline of active violation frequency against which post-intervention cessation can be calculated.

5. Riding Behaviors with a Drinking Driver (RB)

Impaired roadway mortality involves passengers as well as drivers. The RB subscale measures the frequency with which an individual placed their physical safety in jeopardy by riding as a passenger in a vehicle operated by someone who had been consuming alcohol over the past month. This construct identifies deficits in boundary setting, social assertiveness, risk recognition, and protective decision-making, highlighting vulnerability to alcohol-related harm even among individuals who may not personally hold a driver’s license.

Supplementary Diagnostic Domains

Complementing the five primary scales are 16 targeted items tracking personal alcohol use habits, awareness of legal/physiological intoxication markers, previous negative alcohol-related consequences (e.g., prior arrests, social conflicts, occupational impairment), and five demographic variables (gender, chronological age, race/ethnicity, highest completed educational grade, and marital status). These supplementary parameters allow clinicians to contextualize subscale performance within the individual’s broader psychosocial developmental profile.

6. Theoretical Framework

The BADDS is anchored in an integration of social cognitive theory, behavioral decision-making models, and criminological deterrence frameworks. Four primary theoretical pillars explain the conceptual architecture of the instrument:

1. The Theory of Planned Behavior (TPB)

Formulated by Icek Ajzen, the Theory of Planned Behavior asserts that human behavior is immediately guided by behavioral intentions, which are in turn shaped by three core components: attitudes toward the behavior, subjective norms, and perceived behavioral control. The BADDS maps directly onto the TPB construct matrix:

• The Lenient Attitudes (LA) and Rationalizations (RD) subscales reflect the cognitive evaluation and attitudinal valence toward the behavior (evaluating whether the act is acceptable, justified, or benign).
• The Likelihood of Drinking and Driving (LD) subscale captures the proximal behavioral intention, assessing how subjective norms and self-efficacy collapse under situational contingencies.
• The Drinking and Driving Behaviors (DB) subscale records the realization of that intention into overt action.

2. Cognitive Dissonance and Moral Disengagement

Operating a multi-ton motor vehicle while intoxicated is widely recognized across society as life-threatening and legally prohibited. To engage in such behavior without intolerable internal distress, individuals must deploy neutralizing rationalizations. Grounded in Leon Festinger’s Cognitive Dissonance Theory and Albert Bandura’s mechanics of moral disengagement, the Rationalizations for Drinking and Driving (RD) subscale measures the cognitive restructuring techniques—such as advantageous comparison, displacement of responsibility, and distortion of consequences—that permit individuals to violate social and personal ethical standards without experiencing guilt or moral culpability.

3. The Health Belief Model (HBM)

The Health Belief Model postulates that health-related decisions hinge upon perceived susceptibility to harm, perceived severity of the consequences, and perceived benefits versus barriers to adopting protective behaviors. The BADDS operationalizes this framework by identifying where the perceived susceptibility to accidents or arrest is artificially minimized by personal cognitive biases, and where the perceived barriers to arranging alternative sober transportation (e.g., inconvenience, expense, peer judgment) outweigh perceived vehicular risks.

4. Deterrence and Social Learning Paradigms

Within traffic psychology, classical deterrence theory posits that compliance with law relies on the certainty, celerity, and severity of punishment. However, social learning theory demonstrates that informal peer reinforcement, vicarious observation of peers drinking and driving without consequence, and baseline alcohol dependency significantly erode legal deterrence. The BADDS items capture these socio-environmental influences, elucidating why legal sanctions alone frequently fail to alter recidivism trajectories without direct cognitive-attitudinal restructuring.

7. Validity

Validation of the BADDS has encompassed multiple psychometric cohorts, evaluating construct, criterion-related, convergent, and discriminant validity across diverse legal, educational, and clinical settings.

Criterion-Related and Known-Groups Validity

The test manual provides validation through aggregated comparative analyses of known cohorts, particularly examining dichotomized contrast groups of confirmed “drinkers” versus “nondrinkers.” As theoretically hypothesized:

• Nondrinkers consistently exhibit markedly and statistically significantly lower scores across all BADDS subscales compared to individuals who consume alcohol.
• Items assessing subjective definitions and perceptions of what constitutes “safe” drinking and driving demonstrate profound score cleavages between regular consumers of alcohol and abstainers.
• Categorical classification analyses across the three primary scoring tiers (Low, Moderate, High risk) reveal strong predictive clustering. For example, among a validation sample of 325 participants categorizing within the High risk classification of the Rationalizations (RD) subscale, 87% were confirmed active drinkers, providing robust known-groups validation.

Convergent Validity with Clinical Substance Misuse Metrics

To assess convergence with established clinical diagnostic criteria, BADDS subscale scores were systematically correlated against the Substance Abuse Subtle Screening Inventory-3 (SASSI-3), an internationally utilized instrument for detecting substance dependence and chemical dependency profiles:

• Statistically significant correlations in theoretically predicted directions were identified between the three BADDS risk classifications and positive screening diagnoses for substance use disorders on the SASSI-3.
• Higher dimensional scores across the RD, LA, and LD subscales demonstrated positive correlations with three independent clinical scales of substance misuse embedded within the SASSI-3 protocol.
• Crucially, convergent behavioral validity was demonstrated through cross-substance risk generalization: individuals classified into the “High” risk categories on BADDS subscales demonstrated a significantly heightened statistical likelihood of having driven under the influence of illicit or prescription drugs other than alcohol.

Methodological Caveats and Sample Size Dynamics

While correlational studies reported in the BADDS technical documentation substantiate statistically significant relationships ($p < .05$ to $p < .001$), psychometric reviewers (e.g., referencing principles by Jum C. Nunnally and I. H. Bernstein) emphasize caution regarding sample size artifacts. Many BADDS validation cohorts comprise substantial samples ranging from $N = 572$ to $N = 1,850$. In samples of this magnitude, statistical significance can be achieved even when the magnitude of shared variance ($R^2$) is as low as 2% to 4%. Similarly, although chi-square tests comparing offender classifications to non-offenders are significant, certain comparisons reflect modest effect sizes. Clinical evaluators are therefore advised to interpret subscale classifications as continuous indicators of risk rather than definitive, isolated diagnostic thresholds.

8. Reliability

The reliability of the BADDS has been investigated through internal consistency analyses and temporal stability assessments across several target populations, including college undergraduates, adult DUI offenders, municipal traffic court participants, and clinical substance use disorder patients.

Internal Consistency Reliability

The attitudinal and intentional dimensions of the BADDS—specifically the Rationalizations for Drinking and Driving (RD), Lenient Attitudes (LA), and Likelihood of Drinking and Driving (LD) scales—demonstrate robust internal consistency. Across standard university and court-mandated standardization samples:

• Cronbach’s alpha ($lpha$) coefficients for the RD subscale typically range between .84 and .91, indicating high item homogeneity and coherent measurement of cognitive rationalization constructs.
• The LD subscale exhibits alpha coefficients generally falling between .82 and .88, reflecting strong internal reliability for behavioral intention items.
• The LA subscale yields alpha coefficients generally observed between .76 and .84. (Independent psychometric reviews note that in certain early raw technical reports, the LA scale was omitted from specific sub-analyses, but subsequent aggregate testing has confirmed satisfactory internal consistency).

Test-Retest Stability

Temporal stability estimates have been compiled across pre- and post-intervention evaluation windows ranging from immediate post-test intervals to 30-day follow-ups. In control conditions receiving no active psychoeducational intervention, the BADDS subscales demonstrate acceptable stability coefficients, generally ranging from $r = .74$ to $r = .86$. This temporal consistency confirms that the instrument measures relatively enduring cognitive-attitudinal traits rather than transient emotional states, while remaining sufficiently sensitive to detect authentic cognitive shifts following targeted didactic or experiential interventions (such as Fatal Vision simulations).

Limitations of Available Reliability Documentation

Psychometric evaluation standards (e.g., Foster & Cone, 1995) require scrutiny of published documentation. A notable limitation of the BADDS literature is that out of seven core reliability references cited within its primary technical manual, four correspond to unpublished raw corporate data reports and one represents an unpublished master’s thesis. Peer-reviewed, independent replications of its internal consistency across ethnically and socioeconomically diverse demographics remain an ongoing clinical need.

9. Factor Analysis

The structural dimensionality of the BADDS was derived through classical exploratory factor analysis (EFA) during its transition from the original ADDS, followed by confirmatory empirical modeling to verify its five-subscale partition.

Exploratory Factor Structure

Initial factor analytic investigations conducted during instrument development targeted the cognitive justification items, ultimately yielding the robust Rationalizations for Drinking and Driving (RD) factor. When items representing general social leniency, perceived personal risk, and hypothetical future actions were introduced, multi-factor extraction (using principal component analysis with varimax and oblique rotations) confirmed three distinct attitudinal-intentional factors:

1. Factor 1: Cognitive Justification / Rationalization (RD): High primary loadings (predominantly ranging from .55 to .78) on items that minimize personal impairment, assert exceptional driving competence while intoxicated, or justify DUI on the grounds of perceived necessity.
2. Factor 2: Normative Leniency and Societal Tolerance (LA): Loadings clustering between .48 and .71 on items capturing permissive societal perspectives, minimization of DUI as a criminal infraction, and tolerance for moderate intoxication prior to driving.
3. Factor 3: Anticipated Behavioral Enactment (LD): Substantial loadings (ranging from .58 to .82) on situational likelihood scenarios assessing prospective self-reports of driving under the influence.

Behavioral Factor Independence

The behavioral metrics—Drinking and Driving Behaviors (DB) and Riding Behaviors with a Drinking Driver (RB)—function as distinct, single-construct behavioral indices based on explicit count data rather than multi-item latent factor dimensions. Structural equation models that treat DB and RB as criterion outcome variables predicted by the latent factors of RD, LA, and LD have consistently demonstrated good model fit, validating the theoretical premise that rationalizations and lenient attitudes directly drive behavioral violation rates.

Demographic Invariance and Covariance

Factorial invariance testing across demographic strata has yielded important structural findings:

• Gender Covariance: Male respondents consistently score significantly higher than female respondents across the latent RD, LA, and LD dimensions, mirroring established epidemiological disparities in DUI arrest rates. However, multi-group confirmatory factor analyses demonstrate that the factor structure itself remains structurally invariant across genders; the underlying constructs operate identically even though baseline latent means diverge.
• Racial and Educational Invariance: Standardization samples reported in the BADDS manual were predominantly Caucasian (87% to 94% White across various sample waves). Although factor loadings remain coherent across educational levels (from high school students to college graduates and court diversion groups), comprehensive measurement invariance across more ethnically diverse populations has not yet been definitively established.

10. Instrument / Measurement Tool

The BADDS is structured as a streamlined, paper-and-pencil inventory engineered for rapid clinical scoring and direct pretest-to-posttest comparison. Below is the comprehensive technical breakdown of its architecture:

• Instrument Structure & Administration:
  • Total Item Count: 43 items organized systematically across cognitive, behavioral, diagnostic, and demographic sections.
  • Administration Mode: Paper-and-pencil questionnaire administered individually or in large group psychoeducational classrooms.
  • Administration Time: Approximately 10 minutes.
  • Reading Level Requirement: Standardized at a 3.1 reading grade level, ensuring comprehension across varied educational backgrounds.
  • Target Population: Adults aged 18 years and older (with specialized high school reference groups available).
• Subscale Breakdown:
  • Rationalizations for Drinking and Driving (RD): Multi-item scale assessing cognitive excuses, justifications, and self-exempting beliefs.
  • Lenient Attitudes toward Drinking and Driving (LA): Multi-item scale evaluating permissive social norms and tolerance of impaired driving.
  • Likelihood of Drinking and Driving (LD): Multi-item scale capturing self-estimated prospective likelihood of driving while intoxicated under various real-world scenarios.
  • Drinking and Driving Behaviors (DB): Calculated from one core question with two sub-questions recording past-month frequency of driving after drinking.
  • Riding Behaviors with a Drinking Driver (RB): Calculated from one core question with two sub-questions recording past-month frequency of riding as a passenger with an impaired driver.
  • Alcohol Awareness, Consumption, and History: 16 supplementary items indexing personal consumption volume, intoxication awareness, and historical legal/social consequences.
  • Demographic Metrics: 5 items recording gender, age, race/ethnicity, highest grade completed, and marital status.
• Response Formats & Scaling:
  • The RD, LA, and LD scales employ a standard 5-point Likert rating scale (typically anchored from Strongly Disagree [1] to Strongly Agree [5], or Very Unlikely [1] to Very Likely [5]).
  • The DB and RB scales utilize open-ended numerical frequency counts indexing specific behavioral events occurring over the past 30 days.
  • Supplementary alcohol history items use mixed categorical, dichotomous (Yes/No), and ordinal frequency metrics.
• Scoring and Profile Documentation:
  • Double-Sided Profile Sheet: Responses are transferred to a standardized profile sheet featuring a Pretest Profile on one side and a Posttest Profile on the reverse.
  • Risk Classification: Subscale raw scores are summed and categorized into three standardized interpretive risk bands: Low, Moderate, and High.
  • Score Difference Calculation: The posttest profile features dedicated calculation fields to compute direct numerical score differences ($\Delta = \text{Posttest} – \text{Pretest}$) across each subscale.
  • Class/Group Change Assessment Sheet: For instructors and group facilitators, a companion worksheet allows entry of aggregate pretest and posttest averages, determination of group-level average change, and calculation of an overall percent improvement metric.

11. Permissions & Fee and Test Year

The Behaviors & Attitudes Drinking & Driving Scale (BADDS) was copyrighted and formally published in its standardized format in 2003, emerging from initial research and development phases on the ADDS conducted during the late 1990s. The test materials, scoring templates, class assessment sheets, and technical manual are proprietary and commercialized products.

The instrument is protected by international copyright law and is distributed primarily through educational and traffic safety organizations, notably Innocorp, Ltd. (P.O. Box 401, Oregon, WI 53575, USA; website: fatalvision.com), in connection with alcohol awareness and impaired driving intervention programs. Test packages—comprising test booklets, double-sided profile scoring forms, group assessment sheets, and the comprehensive administrative manual—are sold for a fee under commercial licenses. Academic researchers seeking to utilize the scale for non-commercial scientific investigations, dissertations, or grant-funded research must obtain formal written licensing permissions or purchase authorized assessment kits from the publisher.

12. References

Engs, R. C. (1975). The Student Alcohol Questionnaire. Bloomington, IN: Department of Health and Safety Education, Indiana University. https://hdl.handle.net/2022/17484

Festinger, L. (1957). A theory of cognitive dissonance. Stanford University Press. https://doi.org/10.1515/9781503620766

Foster, S. L., & Cone, J. D. (1995). Validity issues in clinical assessment. Psychological Assessment, 7(3), 248–260. https://doi.org/10.1037/1040-3590.7.3.248

Hingson, R., Heeren, T., Winter, M., & Wechsler, H. (2005). Magnitude of alcohol-related mortality and morbidity among U.S. college students ages 18–24: Changes from 1998 to 2001. Annual Review of Public Health, 26, 259–279. https://doi.org/10.1146/annurev.publhealth.26.021304.144652

Hingson, R., & Winter, M. (2003). Epidemiology and consequences of drinking and driving. Alcohol Research & Health, 27(1), 63–78.

Jacobson, N. S., Roberts, L. J., Berns, S. B., & McGlinchey, J. B. (1999). Methods of defining and determining clinical significance of treatment effects: Description, application, and alternatives. Journal of Consulting and Clinical Psychology, 67(3), 300–309. https://doi.org/10.1037/0022-006X.67.3.300

Miller, G. A. (1999). The Substance Abuse Subtle Screening Inventory-3 (SASSI-3) manual. The SASSI Institute.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.

Vogt, D. S., King, D. W., & King, L. A. (2004). Focus groups in psychological assessment: Enhancing content validity by consulting members of the target population. Psychological Assessment, 16(3), 231–243. https://doi.org/10.1037/1040-3590.16.3.231

13. Items of the Scale