Alternate Forms of HIV Prevention Attitude Scales for Teenagers

1. Abstract

The Alternate Forms of HIV Prevention Attitude Scales for Teenagers is an established psychometric assessment developed by Mohammad R. Torabi and William L. Yarber at Indiana University. The instrument was engineered to resolve critical methodological challenges in adolescent health education evaluation, specifically the lack of psychometrically sound, parallel-form survey instruments capable of minimizing pretest sensitization, practice effects, and memory bias in longitudinal and pretest-posttest intervention research designs. Grounded in the classical tripartite model of attitude structure—which delineates attitudes into cognitive (belief-based), affective (feeling- and emotion-based), and conative (action- or intentional-oriented) components—the battery comprises two mathematically equivalent, 15-item Likert-type scales designated as Form A and Form B.

Each form operationalizes adolescent attitudes across multiple behavioral and social domains relevant to the prevention of Human Immunodeficiency Virus (HIV) and Acquired Immunodeficiency Syndrome (AIDS), including condom negotiation and usage, injection drug avoidance, needle-sharing attitudes, social stigma, compassionate peer support, and educational receptivity. Administered on a 5-point Likert response continuum ranging from Strongly Agree (1 or 5) to Strongly Disagree (5 or 1), the scale yields continuous composite scores ranging from 15 to 75, wherein higher aggregate scores reflect systematically more favorable, prevention-oriented attitudes. Psychometric evaluation on a developmental cohort of 210 high school adolescents and an evaluation cohort of 600 secondary school students established robust psychometric integrity: alternate-form equivalence reliability between Form A and Form B was demonstrated at r = .82, with internal consistency coefficients (Cronbach’s alpha) of .78 (Form A) and .77 (Form B), alongside split-half reliability coefficients of .76 and .69. Factor analytic procedures affirmed comparable multidimensional structural invariance across both forms. This instrument serves as an indispensable tool for public health researchers, school health educators, adolescent psychologists, and program evaluators requiring reliable repeated-measures assessment.

2. Keywords

HIV prevention, adolescent attitudes, tripartite attitude model, alternate forms reliability, scale development, psychometrics, sexual health education, condom use intentions, health psychology, pretest-posttest evaluation, stigma reduction, conative intention

3. Authors

The Alternate Forms of HIV Prevention Attitude Scales for Teenagers were conceptualized, developed, and validated by:

• Mohammad R. Torabi, Ph.D., MPH, MSPH: Chancellor’s Professor Emeritus and Former Dean of the School of Public Health at Indiana University Bloomington. Dr. Torabi is an internationally recognized scholar in measurement theory, public health education, substance abuse prevention, and health behavior instrumentation.
• William L. Yarber, HSD, MS: Provost Professor in the Department of Applied Health Science, School of Public Health, and Senior Research Fellow at the Kinsey Institute for Research in Sex, Gender, and Reproduction at Indiana University Bloomington. Dr. Yarber is a foremost expert on sexual health behavior, sexually transmitted infection (STI) prevention curricula, and sex education assessment.

Institutional Address for Inquiries: William L. Yarber, Department of Applied Health Science, School of Public Health (formerly HPER Building), Indiana University, Bloomington, IN 47405; Email: [email protected].

4. Purpose

During the height of the global HIV/AIDS epidemic in the late 1980s and early 1990s, high rates of adolescent sexual experimentation, unprotected intercourse, and experimentation with illicit substances elevated youth into a critical epidemiological risk category. In response, national authorities—most prominently the National Research Council (Coyle, Boruch, & Turner, 1989)—issued urgent calls for rigorous empirical evaluation of school- and community-based HIV prevention programs. While researchers possessed standardized tools to measure factual epidemiological knowledge, empirical progress was severely impeded by the absence of valid, standardized psychometric devices capable of quantifying adolescents’ attitudes toward HIV and prevention practices.

The primary purpose of developing the Alternate Forms of HIV Prevention Attitude Scales for Teenagers was to supply the health education, public health, and psychological research communities with a psychometrically rigorous instrument tailored specifically for longitudinal, quasi-experimental, and randomized controlled trial (RCT) designs. In standard educational evaluation designs utilizing a single-instrument pretest-posttest architecture, researchers frequently encounter profound methodological confounding due to testing effects, test-taking recall, reactive measurement, and item-specific practice sensitization. When adolescents encounter identical behavioral intention questions prior to and immediately following a curricular intervention, changes in posttest scores may reflect memory retrieval, cognitive priming, or defensive self-presentation rather than genuine shifts in underlying attitudinal disposition.

To overcome this measurement artifact, Torabi and Yarber (1992) developed parallel, interchangeable measurement instruments (Form A and Form B) characterized by equivalent statistical properties, identical length, matching item difficulties, comparable standard deviations, and equivalent information functions. The scales measure adolescents’ multi-faceted orientation toward HIV transmission risks, willingness to adopt barrier contraception, disposition toward individuals living with HIV, and intentions to avoid drug-related transmission pathways. Clinically and pedagogically, the scales serve three core functions:

1. Diagnostic Baseline Needs Assessment: Identifying preexisting misconceptions, social prejudices, moralizing attitudes, and behavioral resistance across student bodies prior to implementing health instruction.
2. Curricular Program Evaluation: Providing an unbiased, alternate-form metric to evaluate the empirical efficacy of comprehensive sexual health interventions, peer-led workshops, and public health risk-reduction programs.
3. Behavioral Health Research: Facilitating structural equation modeling, epidemiological tracking, and psychosocial studies exploring how affective reactions and cognitive beliefs mediate concrete adolescent health behaviors.

5. Psychological Construct

The conceptual architecture of the Alternate Forms of HIV Prevention Attitude Scales is rooted in the multidimensional conceptualization of an attitude. Far from being a monolithic evaluative judgment, an attitude toward HIV prevention operates as an organized constellation of psychological tendencies directed toward preventive behaviors, persons living with the disease, and institutional health policies. Torabi and Yarber operationalized the construct according to the classical three-component (tripartite) attitude paradigm, encompassing cognitive, affective, and conative domains.

The Cognitive Dimension (Beliefs and Knowledge Structures)

The cognitive component encompasses an individual’s propositional beliefs, factual perceptions, and intellectual judgments concerning HIV transmission, causality, institutional responses, and preventive efficacy. In the context of adolescent psychology, cognitive dimensions are heavily prone to superstition, peer-group folklore, and defensive rationalization. Specifically, the scale assesses:

• Transmission and Risk Perceptions: Rational recognition of the link between specific high-risk vectors (e.g., intravenous needle-sharing) and HIV acquisition, as captured by items assessing beliefs regarding needle hygiene (Form A, Item 9: “I believe that sharing IV drug needles has nothing to do with HIV”).
• Educational and Preventive Legitimacy: Intellectual valuation of systemic educational interventions versus cynical dismissals of preventive programming (Form A, Item 10: “HIV education in schools is a waste of time”).
• Perceptions of Prevention Feasibility: Cognitive appraisal of how easily barrier methods and harm reduction practices can be integrated into daily life (Form A, Item 7: “It is easy to use the prevention methods that reduce one’s chance of getting HIV”).

The Affective Dimension (Feelings, Stigma, and Emotional Reactions)

The affective component captures the emotional tone, visceral comfort or discomfort, empathy, revulsion, and moralistic sentiment aroused by HIV/AIDS and individuals diagnosed with the infection. Emotional barriers frequently override cognitive knowledge among adolescents, driving avoidance and discriminatory social behavior. The scale operationalizes this domain through:

• Social Discomfort and Proximity Anxiety: Anxious or visceral distress associated with interacting with seropositive individuals in social and educational environments (Form A, Item 1: “I would feel very uncomfortable being around someone with HIV”; Form A, Item 6: “It would be dangerous to permit a student with HIV to attend school”).
• Moralistic Condemnation vs. Compassion: Blaming orientations that view infection as retributive justice for behavioral non-conformity versus pro-social empathy and support (Form A, Item 2: “I feel that HIV is a punishment for immoral behavior”; Form A, Item 11: “I would be supportive of a person with HIV”).
• Partner-Related Emotional Barriers: Interpersonal discomfort, perceived insult, or affective resistance to introducing protective measures within romantic or sexual relationships (Form A, Item 3: “If I were having sex, it would be insulting if my partner insisted we use a condom”; Form A, Item 5: “I would dislike asking a possible sex partner to get the HIV antibody test”).

The Conative Dimension (Behavioral Intentions and Willingness to Act)

The conative component reflects an individual’s explicit behavioral commitment, stated intentions, and readiness to execute specific prophylactic actions under real-world temptation or pressure. Consistent with health behavior decision models, behavioural intention serves as the proximal psychological precursor to overt risk avoidance. The scale captures this dimension through direct action commitments:

• Communication and Negotiation Intentions: Stated resolve to initiate explicit risk-reduction dialogues prior to sexual involvement (Form A, Item 13: “I intend to talk about HIV prevention with a partner if we were to have sex”).
• Barrier Usage Implementation: Explicit commitment to utilize prophylactic barrier methods despite situational ambiguity or partner resistance (Form A, Item 12: “Even if a sex partner insisted, I would not use a condom”; Form A, Item 15: “I will use condoms when having sex if I’m not sure if my partner has HIV”).
• Substance Avoidance Intentions: Declarative commitment to abstain from high-risk substance use vectors known to transmit the virus (Form A, Item 14: “I intend not to use drugs so I can avoid HIV”).

6. Theoretical Framework

The theoretical foundations of the Alternate Forms of HIV Prevention Attitude Scales derive from two complementary paradigms: the Tripartite Model of Attitudes and the social-cognitive theories of behavioral decision-making, most notably the Theory of Reasoned Action (Ajzen & Fishbein, 1980) and its theoretical precursor studies (Kothandapani, 1971; Ostrom, 1969).

The Tripartite Architecture (ABC Model of Attitudes)

Pioneering attitude theorists such as Ostrom (1969) and Kothandapani (1971) empirically demonstrated that while cognitive beliefs, affective responses, and conative intentions correlate, they represent distinct, non-redundant structural components of human attitudes. Ostrom’s classical empirical investigations revealed that measuring all three channels accounts for substantially greater variance in social behavior than assessing any single dimension in isolation.

Prior to developing the adolescent HIV scale, Torabi and colleagues systematically tested and verified this three-component paradigm across diverse public health domains. They successfully developed the Three-Component Cancer Attitude Scale for college students (Torabi & Seffrin, 1986), an Alcohol Attitude Scale for Teenagers (Torabi & Veenker, 1986), and a Sexually Transmitted Diseases Attitude Scale for Young Adults (Yarber, Torabi, & Veenker, 1989). In each of these foundational studies, the tripartite structure provided superior psychometric differentiation, confirming that educational curricula must alter feeling states and behavioral intentions, rather than merely instilling factual knowledge.

Integration with Social-Cognitive Health Behavior Models

The scale integrates directly with Ajzen and Fishbein’s (1980) Theory of Reasoned Action (TRA), which posits that an individual’s overt behavior is immediately determined by their behavioral intention (the conative domain). This intention, in turn, is a joint function of their personal evaluation of the behavior (attitude toward the act, driven by cognitive beliefs and affective evaluations) and subjective norms (perceived peer and social pressure). In the context of HIV prevention:

• Beliefs regarding outcomes: Believing that condoms prevent transmission or that needle-sharing transmits pathogens constitutes the cognitive underpinning.
• Affective evaluation: Feeling comfortable with condom negotiation, lacking stigma toward HIV-positive individuals, and feeling that safer sex is desirable define the affective stance.
• Behavioral intention: The declared intention to carry out protective acts constitutes the conative bridge to real-world risk avoidance.

Psychometric Equivalence and Classical Test Theory

Under Classical Test Theory (CTT), two test forms are defined as strictly parallel if they have identical true score variances, identical error score variances, identical means, and identical correlations with any external criterion. Timminga (1990) further extended this definition within Item Response Theory (IRT), establishing that parallel tests must possess congruent information functions across the latent trait continuum ($ heta$). Torabi and Yarber applied these psychometric tenets to construct alternate forms where items in Form A precisely match the conceptual domain, directionality, and discriminatory power of items in Form B, thereby eliminating testing bias across pretest-posttest measurement intervals.

7. Validity

The validation process for the Alternate Forms of HIV Prevention Attitude Scales followed rigorous multi-stage procedures to establish content, construct, and structural validity.

Content and Curricular Validity

Content validity was established through structured item generation governed by a formal Table of Specifications. This blueprint balanced items evenly across the three attitude dimensions (Cognitive, Affective, Conative) and across critical thematic domains of HIV/AIDS prevention (e.g., sexual transmission, barrier methods, needle-sharing, stigma, and school attendance policies). A comprehensive pool of 50 candidate Likert-type items was constructed and submitted to an independent jury of national experts in health education, adolescent psychology, and infectious disease epidemiology. The panel scrutinized each statement for semantic clarity, readability at the secondary-school reading level, developmental appropriateness, and conceptual alignment with HIV prevention priorities. Items exhibiting ambiguity, multidimensional phrasing, or lack of consensus were eliminated or revised.

Construct and Discriminant Validity

Construct validity was evaluated across successive pilot and field testing administrations:

• Preliminary Cohort Testing: The revised 50-item scale was initially administered to an urban/suburban cohort of 210 Midwestern high school students. Comprehensive item analyses, including item-to-total score correlations and upper-versus-lower quartile discrimination indices ($D$), were computed. Items with weak discriminating capability or negligible variance were excised, distilling the pool into two 15-item subsets displaying maximal discriminative efficiency.
• Field Cohort Validation: The resultant 15-item parallel forms (Form A and Form B) were administered simultaneously to a representative sample of 600 secondary school students in a Midwestern high school. Corrected item-total correlations across all items in both forms were uniformly positive, substantial, and statistically significant ($p < .001$), demonstrating that each item actively measures the shared overarching construct of HIV prevention attitude.
• Comparative Normative Equivalence: Mean total scores, item variance distributions, and scale standard deviations between Form A and Form B exhibited exceptional statistical concordance, demonstrating that both forms capture identical construct variance.

Criterion-Related Validity Challenges and Methodological Constraints

A recognized empirical limitation documented by Torabi and Yarber (1992) pertains to concurrent and predictive criterion-related validity. In ideal psychometric validation, scale scores are directly correlated against objective external criteria, such as verified biomarker testing, clinically documented STI incidence, or diary-based coital diaries. However, during the validation epoch, school district institutional review boards (IRBs) and parental consent protocols strictly prohibited surveying minor adolescents regarding their active illegal intravenous drug use or private sexual practices. Consequently, formal criterion-related validity was not directly observed through behavioral auditing; instead, conative intention items served as the theoretically validated proxy for future behavior.

8. Reliability

The reliability of the Alternate Forms of HIV Prevention Attitude Scales for Teenagers was established using multiple classical psychometric metrics, proving high internal consistency and exceptional alternate-form equivalence.

Alternate-Forms (Equivalent-Forms) Reliability

The defining psychometric feature of this instrument is its parallel-form reliability. Administered concurrently to the target evaluation sample of 600 high school students, the correlation between total scores on Form A and Form B yielded an alternate-form reliability coefficient of:

$$r_{AB} = .82$$

This high coefficient ($p < .001$) demonstrates that Form A and Form B are interchangeable assessments. Researchers can administer Form A as an initial baseline pretest and Form B as a post-intervention posttest (or vice versa, counterbalanced across cohorts) without introducing systematic measurement error or score discrepancies attributable to test form variation.

Internal Consistency Reliability

Internal consistency analyses conducted on both forms confirmed that individual items within each form measure a coherent psychological construct:

• Cronbach’s Coefficient Alpha ($lpha$): Form A achieved an alpha coefficient of .78, while Form B achieved an alpha coefficient of .77. In adolescent attitudinal measurement—where diverse topics such as drug use, condom purchase, and interpersonal stigma are combined—coefficients approaching .80 reflect optimal internal consistency without redundancy.
• Split-Half Reliability: Using the odd-even split-half reliability method corrected by the Spearman-Brown prophecy formula, Form A demonstrated a split-half coefficient of .76, while Form B demonstrated a split-half coefficient of .69.

The standard error of measurement (SEM) for both forms remained low and uniform across the distribution of raw scores, confirming stable measurement precision for group-level curricular evaluations.

9. Factor Analysis

To confirm structural integrity, construct dimensionality, and cross-form factor comparability, the validation data gathered from the 600 adolescent participants were subjected to extensive exploratory and confirmatory factor analytic procedures.

Factor Extraction and Rotation Procedures

Principal Axis Factoring (PAF) and Principal Component Analysis (PCA) accompanied by orthogonal (Varimax) and oblique (Promax) rotations were performed separately on the correlation matrices of Form A and Form B. The empirical objectives were twofold: (1) to determine whether the empirical latent dimensions aligned with the theoretical tripartite model (Cognitive, Affective, Conative), and (2) to verify factorial invariance and structural equivalence across both forms.

Empirical Factor Structure

The factor solutions revealed stable, highly comparable multidimensional configurations across both instruments:

• Factor 1: Personal Intentions and Condom Self-Efficacy (Conative Domain): Highly saturated with items reflecting direct personal agency and behavioral commitments regarding safer sexual practices, condom negotiation, and drug abstinence (e.g., Form A Items 12, 13, 14, 15; Form B Items 3, 12, 13, 14). Factor loadings for these items consistently exceeded .55 to .75.
• Factor 2: Interpersonal Stigma, Empathy, and Social Interaction (Affective Domain): Composed of items addressing emotional comfort around persons living with HIV, social support, and moralistic judgment (e.g., Form A Items 1, 2, 6, 11; Form B Items 1, 2, 11). Primary loadings on this factor ranged from .50 to .72.
• Factor 3: Knowledge Beliefs and Transmission Perceptions (Cognitive Domain): Anchored by items evaluating cognitive acceptance of prevention education, needle hygiene, and risk reduction feasibility (e.g., Form A Items 7, 8, 9, 10; Form B Items 8, 9, 10). Factor loadings clustered between .45 and .68.

Cross-Form Comparability and Factorial Invariance

Crucially, factor analysis of Form B identified an essentially identical latent structure to that of Form A. Item loadings, scree plot trajectory breaks, and eigenvalues accounted for comparable proportions of total variance in both forms. This empirical congruence supported the hypothesis that both scales share identical latent psychometric spaces, confirming the validity of using either form interchangeably in longitudinal intervention designs.

10. Instrument / Measurement Tool

• Name: Alternate Forms of HIV Prevention Attitude Scales for Teenagers
• Authors: Mohammad R. Torabi, Ph.D., and William L. Yarber, HSD
• Target Population: Adolescents and teenagers enrolled in secondary education (high school students, typically aged 14–18 years).
• Administration Format: Standardized self-report paper-and-pencil or digital survey questionnaire.
• Administration Modality: Group or individual administration in classroom, laboratory, or clinical evaluation settings.
• Administration Time: Approximately 8 to 10 minutes per form.
• Number of Items: 15 items per alternate form (Form A = 15 items; Form B = 15 items).
• Response Scale: 5-point Likert-type response continuum:
  • A = Strongly agree
  • B = Agree
  • C = Undecided
  • D = Disagree
  • E = Strongly disagree
• Scoring System:
  • Score Range: 15 to 75 points per form. Higher scores denote more positive, compassionate, and prevention-oriented attitudes.
  • Form A Scoring Key:
    • Positively Phrased Items (7, 8, 11, 13, 15): Strongly agree = 5; Agree = 4; Undecided = 3; Disagree = 2; Strongly disagree = 1.
    • Negatively Phrased / Reverse-Scored Items (1, 2, 3, 4, 5, 6, 9, 10, 12, 14*): Strongly agree = 1; Agree = 2; Undecided = 3; Disagree = 4; Strongly disagree = 5. (Note: As specified in original publication instructions, items not listed as directly scored are reverse-scored so that prevention-positive attitudes receive 5 points).
  • Form B Scoring Key:
    • Positively Phrased Items (1, 3, 8, 9, 10, 11, 12, 13, 14, 15): Strongly agree = 5; Agree = 4; Undecided = 3; Disagree = 2; Strongly disagree = 1.
    • Negatively Phrased / Reverse-Scored Items (2, 4, 5, 6, 7): Strongly agree = 1; Agree = 2; Undecided = 3; Disagree = 4; Strongly disagree = 5.

11. Permissions & Fee and Test Year

The Alternate Forms of HIV Prevention Attitude Scales for Teenagers were published in 1992 in the peer-reviewed journal AIDS Education and Prevention (Guilford Press). The instrument was developed under academic research auspices at Indiana University to advance adolescent public health education.

• Commercial Fees: The scale is non-commercial. No licensing fees or royalty payments are required for academic research, non-profit institutional evaluations, or school-based health educational assessment.
• Permissions & Reproduction: Researchers, health educators, and clinicians are permitted to utilize and reproduce the scales for educational and scientific research purposes, provided proper scholarly citation is attributed to Torabi & Yarber (1992). For commercial packaging, digital distribution platforms, or inclusion in commercial testing software, formal written permission should be requested from the authors or the publisher (Guilford Publications).
• Contact: Dr. William L. Yarber, School of Public Health, Indiana University Bloomington ([email protected]).

12. References

• Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Prentice-Hall.
• Coyle, S. L., Boruch, R. F., & Turner, C. F. (Eds.). (1989). Evaluating AIDS prevention programs. National Academy Press. https://doi.org/10.17226/1458
• Kothandapani, V. (1971). A psychological approach to the prediction of contraceptive behavior. Carolina Population Center, University of North Carolina, Monograph 15.
• Ostrom, T. M. (1969). The relationship between the affective, behavioral, and cognitive components of attitude. Journal of Experimental Social Psychology, 5(1), 12–30. https://doi.org/10.1016/0022-1031(69)90003-1
• Timminga, E. B. (1990). The construction of parallel tests from IRT-based item banks. Journal of Educational Statistics, 15(2), 129–145. https://doi.org/10.3102/10769986015002129
• Torabi, M. R., & Seffrin, J. R. (1986). A three component cancer attitude scale. Journal of School Health, 56(5), 170–174. https://doi.org/10.1111/j.1746-1561.1986.tb05720.x
• Torabi, M. R., & Veenker, C. H. (1986). An alcohol attitude scale for teenagers. Journal of School Health, 56(3), 96–100. https://doi.org/10.1111/j.1746-1561.1986.tb05697.x
• Torabi, M. R., & Yarber, W. L. (1992). Alternate forms of the HIV prevention attitude scales for teenagers. AIDS Education and Prevention, 4(2), 172–182. https://pubmed.ncbi.nlm.nih.gov/1606019/
• Yarber, W. L., Torabi, M. R., & Veenker, C. H. (1989). Development of a three-component sexually transmitted diseases attitude scale for young adults. Journal of Sex Education and Therapy, 15(1), 36–49. https://doi.org/10.1080/01614576.1989.11074945

13. Items of the Scale