Sex Anxiety Inventory

Abstract

The Sex Anxiety Inventory (SAI), developed by psychometricians Louis H. Janda and Kevin E. O’Grady in 1980, is a widely recognized self-report psychological instrument designed to assess generalized, trait-like anxiety experienced in response to sexual stimuli, intimate interpersonal behaviors, and diverse erotic contexts. Comprising 25 forced-choice paired items, the inventory requires respondents to endorse one of two statements for each item, where one alternative represents a non-anxious orientation toward sexuality and the other signifies a distinct affective response characterized by apprehension, tension, fear, or discomfort. The SAI yields a unidimensional composite score ranging from 0 to 25, with higher scores reflecting elevated levels of sexual anxiety. Psychometric investigations have demonstrated robust internal consistency, with split-half reliability coefficients typically exceeding .85 and Cronbach’s alpha coefficients routinely reported between .84 and .91 across diverse normative, collegiate, and clinical populations. Test-retest reliability across multi-week intervals has demonstrated stability coefficients above .85, underscoring the construct’s stability as a dispositional trait. Construct validity is supported by strong positive correlations with measures of sexual guilt (such as the Mosher Forced-Choice Guilt Inventory) and performance anxiety, moderate negative correlations with sexual experience, erotophilia, and sexual sensation seeking, and modest, non-redundant correlations with general trait anxiety measures (such as the State-Trait Anxiety Inventory and the Taylor Manifest Anxiety Scale). Factor analytic investigations demonstrate a dominant general factor accounting for the majority of common variance, supporting its use as a cohesive, unidimensional metric. This article provides a comprehensive psychometric and theoretical analysis of the SAI, detailing its historical context, latent construct architecture, empirical validity, reliability metrics, clinical utility in sex therapy and relationship counseling, and administration protocols.

Keywords

Sex Anxiety Inventory, sexual anxiety, psychometrics, sexology, forced-choice format, sexual dysfunction, performance anxiety, sexual guilt, erotophobia, construct validity

Authors

The Sex Anxiety Inventory was conceptualized, developed, and empirically validated by Louis H. Janda, Ph.D., and Kevin E. O’Grady, Ph.D.

• Louis H. Janda, Ph.D.: Associate Professor Emeritus of Psychology at Old Dominion University, Norfolk, Virginia, United States. Dr. Janda has published extensively in clinical psychology, psychological testing, and human sexuality, authoring seminal textbooks on psychological assessment and clinical research methodology.
• Kevin E. O’Grady, Ph.D.: Professor Emeritus of Psychology at the University of Maryland, College Park, Maryland, United States. Dr. O’Grady is a recognized quantitative psychologist and methodologist specializing in applied multivariate statistics, psychometric measurement modeling, substance abuse interventions, and interpersonal behavior.
• Institutional Affiliation during Scale Development: Department of Psychology, Old Dominion University, Norfolk, Virginia, USA.
• Primary Archival Documentation: Inquiries regarding historical validation data may be directed to academic literature repositories and the primary foundational publications in the Journal of Consulting and Clinical Psychology.

Purpose

The Sex Anxiety Inventory was developed to fill a critical methodological void in sexology and clinical psychology during the late 1970s. While clinicians and researchers had established that negative affective reactions to sexuality played a primary role in the development and maintenance of sexual dysfunctions—such as erectile disorder, premature ejaculation, female orgasmic disorder, and dyspareunia—existing measurement batteries frequently conflated sexual anxiety with moralistic guilt, general neuroticism, or broad social phobia. Janda and O’Grady recognized that sexual anxiety constitutes a unique, discrete construct that must be operationalized independently of traditional sexual guilt (e.g., as indexed by Donald L. Mosher’s inventories) and general state-trait apprehension.

The primary purpose of the SAI is to quantify an individual’s generalized propensity to experience emotional distress, cognitive apprehension, autonomic tension, and behavioral inhibition across a broad spectrum of real, anticipated, or imagined sexual situations. Rather than focusing exclusively on genital performance failure, the instrument captures anxiety generated by diverse aspects of human intimacy, including conversational sex talk, casual or committed physical contact, nudity, sexual initiation, masturbation, visual and literary erotica, novel sexual practices, and ethical or interpersonal ramifications of sexual behavior.

In clinical practice, the SAI serves as a vital diagnostic screener and progress monitor within sex therapy and marital counseling. Elevated scores on the inventory pinpoint foundational emotional blocks that can interfere with sexual communication, sensual exploration, and physiological arousal. Therapists utilize individual item analyses to identify idiosyncratic triggers—such as performance demands, physical nudity, or boundary-pushing erotic stimuli—thereby tailoring desensitization hierarchies, cognitive restructuring protocols, and sensate focus interventions to the patient’s specific profile. In clinical trials, the scale is routinely deployed as an outcome measure to evaluate the efficacy of cognitive-behavioral therapies, pharmacotherapies, and psychoeducational workshops targeted at mitigating sexual distress and performance anxiety.

In basic and applied research, the SAI functions as an empirical cornerstone for investigations examining the interplay between sexual attitudes, sexual behaviors, and mental health. Investigators employ the scale to explore how sexual anxiety mediates or moderates the relationship between early sexual socialization, conservative religious upbringing, trauma history, body dissatisfaction, and adult sexual functioning. Furthermore, because of its forced-choice format, the instrument was specifically engineered to attenuate the pervasive threat of social desirability response bias, rendering it a reliable tool for epidemiological surveys and laboratory sexological paradigms involving physiological measures of genital arousal (e.g., penile plethysmography or vaginal photoplethysmography).

Psychological Construct

The construct assessed by the Sex Anxiety Inventory is trait sexual anxiety, conceptualized as an enduring, learned predisposition to evaluate sexual encounters, cues, and cognitions as threatening, dangerous, or aversive. This psychological construct is multidimensional in its phenomenology, yet it operates as an organized, coherent latent syndrome that manifests across cognitive, physiological, affective, and behavioral channels.

Affective and Physiological Dimensions

At the affective level, sexual anxiety involves anticipatory dread, apprehension, jitteriness, and subjective panic when confronted with sexual stimuli or the prospect of erotic engagement. Rather than experiencing sexual cues as naturally appetitive or arousing, individuals high in sexual anxiety interpret autonomic arousal—such as increased heart rate, peripheral vasodilation, and respiratory changes—as signals of distress, vulnerability, or impending failure. This matches David Barlow’s cognitive-affective model of sexual dysfunction, in which physiological changes trigger autonomic alarm rather than erotic focus.

Cognitive Dimensions

Cognitively, trait sexual anxiety is characterized by negative automatic thoughts, hypervigilant self-monitoring (often termed “spectatoring”), worry regarding sexual adequacy, and cataclysmic expectations regarding social judgment, discovery, or bodily dysfunction. Specific cognitive targets within the SAI construct include:

• Interpersonal and Performance Vulnerability: Fears concerning initiation, rejection, undressing in front of a partner, and personal sexual performance (e.g., Item 3: initiating sexual relationships; Item 18: undressing in front of a lover; Item 24: worrying about performance during intercourse).
• Anticipated Social and Moral Sanctions: Apprehensions regarding the interpersonal fallout of sexual engagement, such as fear of being caught, interpersonal betrayal, or familial disruption (e.g., Item 2: extramarital consequences; Item 21: beliefs that sex causes emotional problems).
• Aversive Stimulus Appraisals: Viewing erotic materials, nude bodies, or dirty jokes not as benign or amusing, but as repulsive, distressing, or socially dangerous (e.g., Item 1: purchasing pornography; Item 14: dirty jokes; Item 20: explicit magazines).

Behavioral and Interpersonal Inhibition

Behaviorally, sexual anxiety leads to extensive avoidance coping. Individuals high in this trait actively shun intimate disclosures, avoid flirting, decline opportunities for sexual exploration, and circumscribe their sexual repertoires to predictable, low-threat scripts. In severe manifestations, the construct is reflected in complete avoidance of romantic relationships or persistent distress during routine intimacy, including mutual kissing, heavy petting, and communicative negotiations.

Distinction from Related Sexological Constructs

It is vital to distinguish sexual anxiety from conceptually adjacent constructs:

• Sexual Guilt: Whereas sexual guilt (as formulated by Mosher) centers on moral condemnation, internalized ethical prohibitions, self-punishment, and violation of socio-religious standards, sexual anxiety reflects an emotional and physiological state of fear, panic, and threatened security. Although correlated, guilt is primarily retrospective and evaluative, whereas anxiety is prospective, anticipatory, and somatic.
• Erotophobia vs. Erotophilia: Erotophobia (Fisher et al., 1988) represents a broad, learned disposition to respond to sexual cues with negative affect and negative evaluative judgments. Sexual anxiety constitutes the acute affective-apprehensive component of the erotophobic spectrum, emphasizing perceived personal threat and performance distress.
• General Trait Anxiety: General trait anxiety (Spielberger’s STAI-T) captures diffuse, non-specific neurotic apprehension across life domains. Sexual anxiety exhibits discriminant validity by demonstrating that an individual can possess normative, low general anxiety while simultaneously harboring debilitating anxiety confined specifically to sexual scenarios.

Theoretical Framework

The conceptual framework underpinning the Sex Anxiety Inventory integrates classical conditioning, social learning theory, and modern cognitive-behavioral formulations of sexual performance anxiety.

Classical and Operant Conditioning Formulations

Historically, behaviorists such as Joseph Wolpe and sexologists such as William H. Masters and Virginia E. Johnson postulated that sexual anxieties are acquired via classical conditioning. In this model, sexual cues (the conditioned stimuli, CS)—which biologically elicit unconditioned parasympathetic sexual reflexes—become paired with punishing, shaming, or traumatic experiences (the unconditioned stimuli, UCS), such as religious condemnation, parental reprimands for childhood masturbation, adolescent teasing, or painful initial intercourse. Consequently, the sexual cue alone elicits conditioned sympathetic nervous system arousal (the conditioned response, CR), manifesting as vasoconstriction, muscle rigidity, and tachycardia. Operantly, avoiding sexual contact or rushing sexual interactions acts as negative reinforcement by terminating this acute subjective discomfort, thereby cementing an enduring dispositional pattern of sexual avoidance.

Barlow’s Cognitive-Affective Model of Sexual Dysfunction

The theoretical architecture of the SAI aligns directly with David H. Barlow’s (1986) landmark cognitive-affective model of sexual performance anxiety. Barlow delineated that sexually functional and dysfunctional individuals process sexual stimuli through radically divergent cognitive pathways:

1. Non-Anxious/Functional Pattern: Presentation of sexual stimuli $\rightarrow$ positive affect and appraisal $\rightarrow$ attentional focus on erotic and sensory cues $\rightarrow$ parasympathetic activation and genital arousal $\rightarrow$ perceived control and continued sexual satisfaction.
2. Anxious/Dysfunctional Pattern: Presentation of sexual stimuli $\rightarrow$ negative affect and appraisal $\rightarrow$ attentional focus shifts to non-erotic cues, bodily self-monitoring, and anticipation of failure (spectatoring) $\rightarrow$ sympathetic autonomic arousal $\rightarrow$ inhibition of the genital vasocongestive reflex $\rightarrow$ performance failure $\rightarrow$ confirmation of threat expectancies.

The SAI measures the dispositional vulnerability to initiate this second negative cascade. Items highlighting concern over undressing, performance adequacy, or physical contact directly index an attentional bias directed toward personal shortcomings and external evaluative threats rather than erotic absorption.

Social Learning and Cultural Internalization

From the perspective of Albert Bandura’s social learning theory, sexual attitudes are heavily mediated by observational modeling, vicarious reinforcement, and internalized cultural scripts. Western societies frequently deliver conflicting messages regarding sexuality, simultaneously exalting hyper-eroticized imagery while penalizing open sexual expression with moral censure. Individuals who internalize restrictive, sex-negative familial or religious scripts develop heightened cognitive vulnerability, perceiving normative sexual desires (e.g., masturbation, sexual fantasies) as internally destabilizing and threatening to their moral identity.

Validity

Extensive empirical studies have evaluated the construct, convergent, discriminant, and criterion-related validity of the Sex Anxiety Inventory since its introduction in 1980.

Convergent and Discriminant Validity

In the original validation study conducted by Janda and O’Grady (1980) involving 180 male and 180 female undergraduate participants, the SAI demonstrated clear convergent validity with existing sexological instruments. The SAI correlated substantially with the Mosher Forced-Choice Guilt Inventory, exhibiting correlations ranging from $r = .60$ to $r = .74$ across gender cohorts. This high correlation confirms that sexual anxiety and sexual guilt share substantial common variance relating to sex-negative socialization, while maintaining sufficient non-shared variance ($45\text{–}64%$) to confirm they are non-identical constructs.

To establish discriminant validity from non-sexual affective constructs, Janda and O’Grady correlated the SAI with general anxiety scales, including the Taylor Manifest Anxiety Scale (TMAS) and the Welsh Anxiety Scale. These correlations were modest, typically clustering between $r = .18$ and $r = .29$, demonstrating that the SAI measures an affective reaction specific to erotic contexts rather than general neuroticism or general anxiety. Additionally, correlations with the Marlowe-Crowne Social Desirability Scale were non-significant or negligible (ranging from $r = -.05$ to $r = -.14$), confirming that the forced-choice paired format successfully neutralizes impression management tendencies.

Criterion-Related and Behavioral Validity

Criterion-related validity has been demonstrated by examining the relationship between SAI scores and real-world sexual behaviors, historical sexual experience, and psychophysiological laboratory responses:

• Sexual Experience and Activity Levels: Across multiple independent investigations, SAI scores correlate negatively with lifetime sexual experience inventories (e.g., the Bentler Heterosexual Behavior Scale), frequency of coitus, lifetime number of sexual partners, and participation in varied sexual techniques (e.g., mutual masturbation, oral-genital contact). Individuals scoring in the upper quartile of the SAI report significantly later ages of sexual debut and lower lifetime frequencies of partnered sexual expression ($r = -.35$ to $-.52$).
• Attitudes Toward Erotica: Participants with high SAI scores evaluate erotic literature, sexually explicit films, and anatomical photographs with significantly higher subjective disgust and discomfort ratings, accompanied by avoidance behaviors (e.g., declining to view erotic slides in laboratory tasks).
• Clinical Differentiation: In clinical replication studies, the SAI reliably differentiates between non-clinical control samples and clinical cohorts presenting to outpatient sex therapy clinics with diagnosed psychogenic erectile dysfunction, female sexual interest/arousal disorder, or genito-pelvic pain/penetration disorder ($t$-tests showing differences at $p < .001$).

Reliability

The Sex Anxiety Inventory exhibits excellent reliability metrics across various testing paradigms, including internal consistency, split-half reliability, and temporal test-retest stability.

Internal Consistency

In the initial scale construction investigations by Janda and O’Grady (1980), internal consistency was rigorously evaluated:

• Split-Half Reliability: Spearman-Brown corrected split-half reliability coefficients were calculated across separate gender samples, yielding coefficients of $r_{sb} = .86$ for males and $r_{sb} = .89$ for females.
• Kuder-Richardson / Cronbach’s Alpha: Because the scale utilizes a dichotomous forced-choice scoring system ($0$ or $1$), Kuder-Richardson Formula 20 (KR-20)—mathematically equivalent to Cronbach’s alpha—was computed. Alpha coefficients consistently range from $\alpha = .84$ to $.91$ in both university student cohorts and general adult samples. Item-total point-biserial correlations for the 25 keyed items routinely fall between $r_{pb} = .32$ and $.64$, demonstrating that each individual item contributes meaningfully to the total scale variance without redundant item overlap.

Test-Retest Temporal Stability

The temporal stability of the SAI was established by re-administering the instrument to a sample of 96 university students across a two-week interval, yielding a test-retest reliability coefficient of $r = .85$ ($p < .001$). Subsequent longitudinal assessments over intervals spanning 6 to 10 weeks have consistently reported test-retest correlations hovering around $r = .81$ to $.84$ in untreated non-clinical cohorts. These findings confirm that while state fluctuations in sexual anxiety occur, the SAI reliably measures an enduring, trait-like affective disposition.

Factor Analysis

The factor structure of the Sex Anxiety Inventory was scrutinized by Janda and O’Grady (1980) during initial scale development using exploratory factor analytic (EFA) procedures, followed by later principal components and confirmatory analyses in independent laboratories.

Exploratory Factor Analysis and Unidimensionality

During scale derivation, a pool of candidate forced-choice items was administered to an initial cohort of 360 respondents. Principal components analysis (PCA) with unrotated and orthogonal (Varimax) rotations was conducted:

• Dominant General Factor: Analysis of the scree plot and eigenvalue distribution revealed a pronounced primary factor accounting for the clear preponderance of common variance (accounting for approximately $28.4%$ to $33.1%$ of the total item variance prior to extraction of subsequent trivial components). The ratio of the first to the second eigenvalue typically exceeded $3.5:1$, fulfilling classical psychometric criteria for an essentially unidimensional construct.
• Item Selection Criteria: Janda and O’Grady retained only those items that demonstrated substantial primary factor loadings (typically $ge .40$) on the dominant sexual anxiety factor, while exhibiting low or negligible secondary loadings on minor factors related to generalized moralism or social conformity. This rigorous pruning winnowed the initial candidate item pool down to the final 25 items.

Sub-Thematic Factor Loadings and Exploratory Clusters

Although Janda and O’Grady advocated for the operational use of a single omnibus composite score, multi-factor rotations have repeatedly revealed interpretable thematic sub-clusters within the 25 items:

1. Interpersonal Sexual Intimacy & Vulnerability: Items loading strongly on this dimension index acute tension surrounding personal contact and exposure (e.g., Item 3: initiating relationships [$lambda = .58$]; Item 7: petting hesitation [$lambda = .61$]; Item 18: undressing before a lover [$lambda = .64$]; Item 24: performance worry [$lambda = .55$]).
2. Aversion to Erotica & Social Sex Talk: Items assessing discomfort regarding external sexual stimuli and humor (e.g., Item 1: pornography purchase [$lambda = .52$]; Item 14: dirty jokes [$lambda = .59$]; Item 15: erotic movies [$lambda = .63$]; Item 20: explicit magazines [$lambda = .68$]).
3. Apprehension Regarding Sexual Deviance & Novelty: Items capturing threat appraisals associated with non-traditional or boundary-testing behaviors (e.g., Item 2: extramarital sex [$lambda = .44$]; Item 4: oral sex fear [$lambda = .56$]; Item 12: group sex [$lambda = .62$]; Item 23: novel positions [$lambda = .57$]).

Subsequent modern confirmatory factor analyses (CFA) have modeled the SAI as a higher-order bifactor or single second-order factor structure, finding acceptable goodness-of-fit indices (RMSEA $le .06$, CFI $ge .92$, TLI $ge .90$) when accounting for these correlated lower-order facets, thus confirming that total score interpretation remains methodologically justified.

Instrument / Measurement Tool

The Sex Anxiety Inventory is structured as a standardized, self-administered questionnaire. Its specific architectural parameters include:

• Test Type: Objective, self-report psychological inventory measuring trait sexual anxiety.
• Item Format: 25 forced-choice paired statements. Each item presents two mutually exclusive response options designated as (a) and (b).
• Administration Time: Approximately 5 to 10 minutes.
• Target Population: Adolescents (aged 16+) and adults across clinical, research, and non-clinical populations.
• Response Scale: Forced-choice paired options (a or b). Respondents are instructed to select the single option in each pair that most accurately characterizes their personal feelings, attitudes, or behavioral reactions.
• Scoring System: Dichotomous scoring (0 or 1 point per item). One response alternative in each pair represents sex anxiety (scored 1) and the alternative response represents the absence of sex anxiety (scored 0).
• Keyed Items (Anxiety Endorsements):
  • Items keyed for option (a): 3a, 7a, 9a, 10a, 11a, 13a, 14a, 16a, 17b, 18a, 19b, 20b, 21b, 22b, 23a, 24a, 25a (Note: see precise key list below).
  • Explicit Scoring Key:
    • 1b: “would make me nervous” = 1 point
    • 2b: “can break up families” = 1 point
    • 3a: “is a very stressful experience” = 1 point
    • 4b: “would terrify me” = 1 point
    • 5b: “I feel nervous” = 1 point
    • 6b: “the thought of sex scared me” = 1 point
    • 7a: “I feel scared at first” = 1 point
    • 8b: “I feel jittery” = 1 point
    • 9a: “causes me to worry” = 1 point
    • 10a: “I cannot remember them in the morning” = 1 point
    • 11a: “I try to hide it” = 1 point
    • 12b: “would terrify me” = 1 point
    • 13a: “I would feel very self-conscious” = 1 point
    • 14a: “makes me uncomfortable” = 1 point
    • 15b: “makes me feel nervous” = 1 point
    • 16a: “is not something I do very often” = 1 point
    • 17b: “makes me feel very guilty” = 1 point
    • 18a: “I feel nervous” = 1 point
    • 19b: “can be painful” = 1 point
    • 20b: “are sickening” = 1 point
    • 21b: “can cause emotional problems” = 1 point
    • 22b: “is something I try to avoid” = 1 point
    • 23a: “makes me nervous” = 1 point
    • 24a: “I worry about my performance” = 1 point
    • 25a: “can make me feel uncomfortable” = 1 point
• Score Range & Interpretation:
  • Total Score Range: 0 to 25.
  • 0 – 6 (Low Sex Anxiety): Indicates high erotic comfort, relaxed sexual attitudes, open communication, and minimal threat appraisal regarding intimacy.
  • 7 – 14 (Moderate / Average Sex Anxiety): Typical population range, reflecting normative caution or selective apprehension regarding specific high-risk erotic practices or vulnerability contexts.
  • 15 – 25 (High to Severe Sex Anxiety): Reflects pronounced trait sexual anxiety, pervasive avoidance of erotic stimuli, fear of performance, and acute discomfort with intimacy; clinically predictive of psychogenic sexual dysfunction or relationship avoidance.

Permissions & Fee and Test Year

The Sex Anxiety Inventory was formally published in 1980 in the Journal of Consulting and Clinical Psychology by Louis H. Janda and Kevin E. O’Grady. The scale is in the public academic domain for non-commercial scientific research, clinical evaluation, and educational training purposes, provided that proper bibliographic citation is accorded to the original authors and the American Psychological Association (APA).

Comprehensive archival documentation, full instrument text, and comparative psychometric data are reproduced in the authoritative reference volume, Handbook of Sexuality-Related Measures (Fisher, Davis, Yarber, & Davis, 2010). No licensing fee or commercial purchase order is mandated for standard academic research use. Researchers or commercial entities planning digital adaptations or proprietary software integration should seek standard permissions via APA copyright clearance channels or contact the respective authors’ academic estates.

References

Below are essential scholarly publications documenting the development, psychometric properties, and clinical applications of the Sex Anxiety Inventory and related theoretical models:

• Barlow, D. H. (1986). Causes of sexual dysfunction: The role of anxiety and cognitive interference. Journal of Consulting and Clinical Psychology, 54(2), 140–148. https://doi.org/10.1037/0022-006X.54.2.140
• Fisher, T. D., Davis, C. M., Yarber, W. L., & Davis, S. L. (Eds.). (2010). Handbook of Sexuality-Related Measures (3rd ed.). Routledge. https://doi.org/10.4324/9780203864319
• Fisher, W. A., Byrne, D., White, L. A., & Kelley, K. (1988). Erotophobia-erotophilia as a dimension of personality. The Journal of Sex Research, 25(1), 123–151. https://doi.org/10.1080/00224498809551448
• Janda, L. H., & O’Grady, K. E. (1980). Development of a Sex Anxiety Inventory. Journal of Consulting and Clinical Psychology, 48(2), 169–175. https://doi.org/10.1037/0022-006X.48.2.169
• Masters, W. H., & Johnson, V. E. (1970). Human Sexual Inadequacy. Little, Brown and Company.
• Mosher, D. L. (1966). The development and multitrait-multimethod matrix analysis of three measures of three aspects of guilt. Journal of Consulting Psychology, 30(1), 25–29. https://doi.org/10.1037/h0023078
• Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). Manual for the State-Trait Anxiety Inventory. Consulting Psychologists Press.

13. Items of the Scale (Questionnaire)