AAUW Sexual Harassment Survey

1. Abstract

The AAUW Sexual Harassment Survey is a landmark psychometric and behavioral assessment tool developed by the American Association of University Women (AAUW) Educational Foundation to measure the prevalence, nature, contextual dynamics, impact, and perpetration of sexual harassment in primary and secondary educational environments. Originally introduced in the seminal nationwide study Hostile Hallways: The AAUW Survey on Sexual Harassment in America’s Schools (1993) and refined in the 2001 replication study, the instrument provides an empirical, behavioral framework that operationalizes peer-to-peer and staff-to-student harassment without relying on subjective or legalistic self-labeling. The survey comprises multi-part behavioral inventories capturing both victimization (Question 1, with 14 distinct behavioral items across non-physical, physical, and verbal domains) and perpetration (Question 6, mirroring the 14 behavioral indicators), accompanied by dedicated situational, relational, and emotional impact modules (Questions 2 through 5, 7, and 8). Items in the victimization and perpetration frequency modules are scored on an ordinal metric (0 = Never, 1 = 1–2 times, 2 = 3–5 times, 3 = 6+ times, with an option for Not Sure), while ecological, perpetrator identity, and attributional modules utilize dichotomous indicators (1 = Yes, 0 = No) and categorical response scales. Psychometric investigations, including replications cataloged by the Centers for Disease Control and Prevention (CDC), indicate strong internal consistency across its behavioral subscales (Cronbach’s alpha coefficients typically ranging from .82 to .91), robust structural validity confirming multidimensionality (differentiating verbal/gestural harassment from physical violation), and pronounced convergent and predictive validity regarding psychological distress, educational disengagement, and school climate indices.

2. Keywords

AAUW Sexual Harassment Survey, Hostile Hallways, sexual harassment, school-based victimization, peer harassment, perpetration, adolescent development, psychometrics, school violence, Title IX assessment

3. Authors

The instrument was commissioned and developed by the American Association of University Women (AAUW) Educational Foundation in collaboration with Harris Interactive (formerly Louis Harris and Associates, Inc.). Key research contributors, programmatic directors, and advisory panelists included:

• American Association of University Women Educational Foundation — 1111 Sixteenth St. NW, Washington, DC 20036; Contact: [email protected]; Website: www.aauw.org; Telephone: 800-326-2289 / 202-785-7700.
• Harris Interactive Inc. — Quantitative survey design, national probability sampling methodology, and analytical validation.
• Advisory Panelists & Principal Investigators: Prominent scholars in educational psychology, gender studies, and adolescent public health, including Nan Stein, Ed.D. (Wellesley Centers for Women), who collaborated on early conceptual formulations of school-based peer sexual harassment.

4. Purpose

The primary purpose of the AAUW Sexual Harassment Survey is to provide a standardized, behaviorally specific measurement system for identifying the prevalence, incidence, etiology, and personal consequences of sexual harassment among school-aged youth, predominantly spanning grades 8 through 11 (ages 12 to 18). Prior to the deployment of this instrument, institutional assessments of school-based harassment were severely compromised by semantic ambiguity: inquiring whether a student had been “sexually harassed” frequently yielded dramatic underreporting or subjective variability, as adolescents often conflate harassment with “flirting,” “teasing,” or legally actionable criminal assault. The AAUW tool eliminated this conceptual confound by decomposing sexual harassment into 14 explicit behavioral manifestations—ranging from verbal taunts and derogatory sexual graffiti to coercive touching, forced kissing, and explicit physical assault—experienced during clearly demarcated “school-related times.”

From an applied clinical and public health perspective, the instrument serves multiple diagnostic, preventive, and administrative functions. In educational policy and compliance contexts, the survey provides empirical data essential for auditing institutional adherence to Title IX of the Education Amendments of 1972, which obligates federally funded educational institutions to address sexually hostile educational environments. School psychologists, guidance counselors, and clinical social workers utilize the behavioral indicators to assess systemic school climate, track victimized adolescents at risk for acute stress, evaluate trauma symptoms, and design multi-tiered systems of support (MTSS).

In behavioral research, the dual-structure operationalization (measuring both victimization and self-reported perpetration within identical behavioral parameters) enables structural equation modeling of the reciprocal relationships between harassment involvement, social-cognitive distortions, gender-role stereotyping, and behavioral peer norms. Moreover, the integration of ecological questions detailing perpetrator demographics (single vs. group, male vs. female, peer vs. adult authority), physical school locations (hallways, classrooms, locker rooms, school transit), and subjective emotional sequelae (ranging from unbothered to severely upset) equips researchers with granular contextual data necessary for developing environmental, situational crime prevention interventions within secondary schools.

5. Psychological Construct

The psychological construct evaluated by the AAUW instrument is school-based sexual harassment, conceptualized as unwelcome, uninvited behavior of a sexual nature that creates an intimidating, hostile, or offensive educational environment. Rather than viewing sexual harassment as a unitary, homogeneous event, the instrument frames it as a spectrum of interpersonal aggression cutting across multiple behavioral modalities, relational dynamics, and functional motivations.

1. Non-Physical / Verbal / Symbolic Harassment

This sub-dimension captures non-contact violations of personal boundaries that exploit sexualized or gender-derogatory content to demean, humiliate, or control the target. Specific item operationalizations include:

• Verbal comments, jokes, gestures, or looks: Utilizing suggestive sexual humor, objectifying commentary, or predatory body language.
• Circulation of sexual materials: Displaying, distributing, or delivering explicit photographs, sexual notes, drawings, or digital media.
• Reputational and identity-based degradation: Spreading sexualized rumors, inscribing sexual graffiti in communal spaces (restrooms, lockers), and targeting sexual orientation through weaponized slurs (e.g., alleging someone is gay or lesbian to enforce heteronormative conformity).

2. Physical / Bodily Harassment

This dimension encompasses non-consensual physical contact and spatial boundary violations that threaten bodily integrity and personal autonomy. The spectrum includes:

• Unwanted physical contact: Intentional brushing against a target’s body in a sexualized manner, as well as purposeful touching, grabbing, or pinching of breasts, buttocks, or genitalia.
• Clothing-related boundary breaches: Pulling at personal clothing in a sexual manner or pulling clothing off or down (e.g., “depantsing” or exposing undergarments).
• Spatial containment and voyeurism: Cornering, blocking physical passage in a sexually intimidating way, and voyeurism (spying on peers in locker rooms, dressing stalls, or showers).
• Covert or overt indecent exposure: Public exhibitionism, including flashing or “mooning.”

3. Coercive / Severe Sexual Aggression

Representing the most severe tier of interpersonal violation, this sub-dimension measures explicit behavioral coercion involving sexual acts:

• Forced kissing: Physical compulsion or intimidation used to force unwanted oral-genital or oral-oral contact.
• Forced sexual acts: Coercive physical force or psychological duress compelling non-consensual sexual performance beyond kissing, mapping onto statutory definitions of sexual assault.

4. Perpetration and Attributional Dynamics

Beyond victimization, the construct encompasses perpetration behaviors mirroring the 14 victimization items, alongside self-serving cognitive attributions (social normalization, power assertion, peer compliance, instrumental pursuit, and distorted relational expectations). This captures the socio-ecological normalization of sexual dominance within adolescent peer networks.

6. Theoretical Framework

The AAUW Sexual Harassment Survey is grounded in an interdisciplinary synthesis of feminist sociological theory, social cognitive theory, and Bronfenbrenner’s social-ecological systems model.

Feminist and Gender-Role Socialization Paradigms

Historically articulated by scholars such as Catherine MacKinnon and adapted to youth contexts by Nan Stein, feminist legal and social theory posits that sexual harassment is not an expression of uncontrolled sexual attraction, but rather an assertion of patriarchal dominance, gender-role enforcement, and power differentials. In early adolescence, emerging sexual identity intersects with rigid peer hierarchies. The AAUW framework operationalizes this theoretical perspective by assessing behaviors that enforce gender orthodoxy (e.g., weaponizing homophobic labels against non-conforming peers) and examining differential power dynamics across gender dyads (male-on-female, female-on-male, same-sex, and group-based harassment).

Social-Ecological Model

Applying Bronfenbrenner’s ecological model, the AAUW survey treats adolescent harassment as an emergent property of nested environmental systems:

• Ontogenic/Individual Level: Individual attributions, self-regulatory efficacy, and attitudes toward gender-based dominance (captured in Question 8 reasons for perpetration).
• Microsystem Level: The immediate interactive settings of the school, including unmonitored hallways, locker rooms, cafeterias, and school buses (assessed explicitly in Question 4).
• Mesosystem Level: The failure of authority structures (teachers, administrators, coaches; measured in Question 3) to intervene, which implicitly communicates tolerance of hostile behavioral norms.
• Exosystem and Macrosystem Levels: Cultural scripts that trivialize sexualized aggression as normative adolescent “flirting” or “harmless teasing,” thereby cultivating an institutional climate of impunity.

Situational Crime Prevention and Routine Activity Theory

The instrument incorporates key tenets of Routine Activity Theory (Cohen & Felson). By mapping specific physical spaces (restrooms, playing fields, hallways) and supervision parameters, the survey operationalizes harassment as occurring at the nexus of motivated offenders (individual or collective peer groups), suitable targets (isolated students), and the absence of capable guardians (unsupervised school transit, unmonitored locker facilities).

7. Validity

The psychometric validity of the AAUW Sexual Harassment Survey has been corroborated across large-scale national probability studies, peer-reviewed replications, and compendia published by agencies including the Centers for Disease Control and Prevention.

Content and Construct Validity

Content validity was established through extensive focus groups with middle and high school students, comprehensive reviews of federal Title IX jurisprudence, and advisory oversight by developmental psychologists and educational criminologists. Rather than using abstract terminology, every item represents an unambiguous, behaviorally specific act. In construct validation studies conducted by Harris Interactive and subsequent independent researchers (e.g., Fineran & Bennett, 1999; McMaster et al., 2002), the behavioral items demonstrated distinct divergence from general school incivility and non-sexualized physical bullying, confirming that sexual harassment forms an empirically separable domain of peer aggression.

Convergent Validity

The instrument demonstrates strong convergent validity with established psychometric scales measuring psychological distress, somatic complaints, and educational disruption:

• Significant positive correlations have been repeatedly demonstrated between high scores on the AAUW 14-item victimization scale and elevated scores on the Trauma Symptom Checklist for Children (TSCC) (r values typically ranging between .38 and .54, p < .001).
• Victimization frequency correlates positively with validated indices of academic avoidance, including school absenteeism, reduced classroom participation, dropping courses, and lowered academic self-efficacy (r = .31 to .46).
• Perpetration scores (Question 6) show significant convergent correlations with validated measures of hostile sexism, peer-group delinquency, and bullying perpetration (r = .42 to .58).

Discriminant and Predictive Validity

Discriminant validity is supported by low-to-moderate correlations with general trait anxiety and non-interpersonal academic stress, confirming that the tool isolates specific interpersonal environmental trauma rather than generalized negative affectivity. In longitudinal modeling, baseline victimization scores on the AAUW scale successfully predict subsequent onset of major depressive symptoms, social anxiety, and academic disengagement at 12-month follow-up, controlling for prior baseline mental health status.

8. Reliability

The AAUW Sexual Harassment Survey exhibits strong reliability metrics across diverse demographic cohorts, educational strata, and independent research investigations.

Internal Consistency

Internal consistency analyses across independent studies yield robust Cronbach’s alpha coefficients:

• Total Victimization Scale (Question 1, 14 items): Cronbach’s α typically ranges from .84 to .91 across diverse adolescent samples (e.g., α = .88 in national probability normative data; α = .89 in high school validation cohorts).
• Total Perpetration Scale (Question 6, 14 items): Cronbach’s α coefficients range between .82 and .88, demonstrating that self-reported perpetration behaviors hang together reliably despite social desirability pressures.
• Non-Physical Harassment Subscale: Cronbach’s α ranges between .80 and .86.
• Physical/Coercive Harassment Subscale: Cronbach’s α ranges between .78 and .85, an acceptable level given the lower base-rate endorsement of high-severity physical assault items.

Test-Retest Stability

Because the survey assesses behavioral frequency over a specified retrospective reporting window (“during the last year”), test-retest reliability has been evaluated over short intervals (2 to 4 weeks) to evaluate recall consistency without capturing true developmental shifts. Short-term test-retest coefficients (intraclass correlation coefficients, ICC) range from .79 to .87 for cumulative victimization frequency scores, indicating high temporal stability in adolescent reporting of behavioral events.

9. Factor Analysis

Structural investigations using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) consistently establish that the 14 behavioral items in the AAUW survey represent a multidimensional construct hierarchically organized under an overarching sexual harassment factor.

Exploratory Factor Structure

Principal axis factoring with oblique (Promax or Oblimin) rotation routinely extracts two to three distinct but moderately intercorrelated latent factors:

• Factor 1: Non-Physical / Verbal Harassment: High loadings (> .60) for items such as sexual comments/jokes (1a), sexual pictures/messages (1b), sexual rumors (1d), and homophobic labeling (1e).
• Factor 2: Physical Harassment: High loadings (> .55) for touching/grabbing (1h), pulling clothing (1i), intentional brushing (1j), and clothing removal (1k).
• Factor 3: Severe Sexual Coercion / Boundary Invasions: Distinct factor extraction in selected models for forced kissing (1m), forced sexual acts (1n), cornering/blocking (1l), and voyeurism (1f).

Confirmatory Factor Analysis Fit Indices

In CFA structural assessments testing adolescent populations, a correlated two-factor model (Non-Physical vs. Physical/Coercive Harassment) or a second-order hierarchical model yields strong model fit statistics meeting standard psychometric thresholds:

• Comparative Fit Index (CFI): .94 to .97
• Tucker-Lewis Index (TLI): .93 to .96
• Root Mean Square Error of Approximation (RMSEA): .042 to .055 (90% Confidence Interval: [.036, .061])
• Standardized Root Mean Square Residual (SRMR): .038 to .048

Item-factor standardized factor loadings across models remain consistently robust, with verbal items exhibiting loadings between .62 and .81, and physical items loading between .58 and .84. Factor intercorrelations generally range from .55 to .70, justifying both subscale-level scoring and the use of an aggregate composite frequency index.

10. Instrument / Measurement Tool

• Instrument Name: AAUW Sexual Harassment Survey (Hostile Hallways Instrument).
• Test Type: Standardized self-report behavioral questionnaire (evaluating victimization, perpetration, ecological context, perpetrator identities, and emotional impact).
• Target Population: Secondary school students (middle and high school; grades 8 through 11; ages 12–18).
• Total Core Behavioral Items:
  • Question 1 (Victimization): 14 behavioral items (a through n).
  • Question 6 (Perpetration): 14 behavioral items (a through n).
  • Ecological, Relational, and Attributive Items: 5 contextual modules (Questions 2, 3, 4, 7, 8) and 1 affective impact item (Question 5).
• Response Formats:
  • Questions 1 and 6 (Behavioral Frequency): 5-point ordinal scale (scored quantitatively: Never = 0, 1–2 times = 1, 3–5 times = 2, 6+ times = 3, and categorical tracking for Not Sure).
  • Questions 2, 3, 4, 7, and 8 (Contextual, Perpetrator, and Attributional Factors): Dichotomous selection (Yes = 1, No = 0).
  • Question 5 (Emotional Impact): Single-choice categorical scale (I was very upset, I was somewhat upset, I wasn’t very upset, I wasn’t at all upset, Not sure).
• Scoring and Indexing Procedures:
  • Prevalence Metric: Percentage of respondents endorsing at least one behavioral item at a frequency of ≥ 1 (“1–2 times” or greater).
  • Cumulative Severity Score: Summed composite across Question 1 (victimization range: 0 to 42) and Question 6 (perpetration range: 0 to 42).
  • Subscale Breakdown: Summed or mean scores computed separately for Non-Physical items (1a–e, 1g) and Physical/Coercive items (1f, 1h–n).
• Administration Time: Approximately 15 to 25 minutes in group or individual survey formats.

11. Permissions & Fee and Test Year

• Year of Initial Publication: 1993 (First Edition: Hostile Hallways: The AAUW Survey on Sexual Harassment in America’s Schools); Revised replication edition published in 2001.
• Publishing Body / Copyright Holder: American Association of University Women (AAUW) Educational Foundation.
• Public Domain / Federal Compendium Status: The instrument was adapted and cataloged by the United States Department of Health and Human Services, Centers for Disease Control and Prevention (CDC), in the public domain assessment compendium Measuring Bullying Victimization, Perpetration, and Bystander Experiences: A Compendium of Assessment Tools (pages 40–43).
• Usage Permissions: Available for educational, academic, empirical research, and school evaluation purposes. Commercial reproduction or widespread institutional redistribution requires contact with the copyright holder.
• Contact Details for Rights & Inquiries:
  
  AAUW Educational Foundation, 1111 Sixteenth St. NW, Washington, DC 20036
  
  Telephone: 800-326-2289 or 202-785-7700
  
  Email: [email protected] | Website: www.aauw.org

12. References

American Association of University Women Educational Foundation. (1993). Hostile hallways: The AAUW survey on sexual harassment in America’s schools. American Association of University Women. Washington, DC.

American Association of University Women Educational Foundation. (2001). Hostile hallways: Bullying, teasing, and sexual harassment in school. American Association of University Women Educational Foundation. Washington, DC. https://www.aauw.org/resources/research/hostile-hallways/

Fineran, S., & Bennett, L. (1999). Gender and socioeconomic differences in secondary school sexual harassment. Journal of School Violence, 1(2), 63–89. https://doi.org/10.1300/J202v01n02_05

Hamburger, M. E., Basile, K. C., & Vivolo, A. M. (2011). Measuring bullying victimization, perpetration, and bystander experiences: A compendium of assessment tools. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Atlanta, GA. https://www.cdc.gov/violenceprevention/pdf/bullycompendiumbk-a.pdf

McMaster, L. E., Connolly, J., Pepler, D., & Craig, W. M. (2002). Peer to peer sexual harassment in early adolescence: A developmental perspective. Development and Psychopathology, 14(1), 91–105. https://doi.org/10.1017/s0954579402001062

Stein, N. (1995). Sexual harassment in school: The public performance of gendered violence. Harvard Educational Review, 65(2), 145–163. https://doi.org/10.17763/haer.65.2.h770146034177546

13. Items of the Scale (Questionnaire)