Garos Sexual Behavior Inventory

Garos Sexual Behavior Inventory

SHEILA GAROS Texas Tech University

The Garos Sexual Behavior Inventory (GSBI; Garos, 2009; Garos & Stock, 1998a, 1998b) is designed to as- sist forensic specialists and mental health professionals in making assessments and treatment decisions about individuals with problems related to sexuality and sexual behavior. The GSBI’s subscale scores can be used to evaluate the cognitive, affective, and behavioral dimensions of an individual’s overall sexual adjustment. The GSBI has been used effectively with clinical populations known to have difficulty with sexuality and sexual behavior, such as victims of sexual abuse, individuals with substance abuse problems, those with compulsive sexual behaviors, individuals with paraphilic interests (Garos & Stock, 1998b), and forensic populations (Garos, Bleckley, Beggan, & Frizzell, 2004). Yet the GSBI is sufficiently versatile that it can also be used for couples counseling to uncover differences in sexual values, feelings, and comfort levels.


The first objective in development of the GSBI was to develop a conceptual model that would act as the basis for the investigation of constructs underlying disorders of sexual frequency and control. With consultation from experts in the field, a series of rationally derived item elimination procedures was instituted to identify items that best described dimensions of the constructs of interest. To evaluate the adequacy of the conceptually derived model, initial items on the GSBI were subjected to a principle components factor analysis using the varimax procedure.

In a second study, the 70-item GSBI was subjected to a maximum likelihood factor analysis with oblique rotation of the factor structure to confirm the conceptual dimen- sions believed to underlie the GSBI. The GSBI consists of four main scales, and three masking scales designed to reduce the potential for defensive responding, as well as an Inconsistent Responding index.

Main scales

  • Discordance: Overall sexual adjustment; shame, fear, internal conflict about sexual behavior and interests
  • Sexual Obsession: Preoccupation with sex and sexual stimuli
  • Permissiveness: Attitudes and values about sex—con- servative or unconventional
  • Sexual Stimulation: Level of comfort with sexual arousal

    Masking Scales

  • Sexual Control Difficulties
  • Sexual Excitability
  • Sexual Insecurity

Response Mode and Timing

The GSBI can be completed in just 20 to 30 minutes. Items are written at a sixth-grade reading level and are easy to understand and answer. Respondents indicate the extent of their agreement on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). An AutoScore Form is designed to protect the privacy of respondents. Test items do not appear on the AutoScore Form. Instead, they are printed on a separate reusable Administration Card, thereby ensuring that sensitive information remains confidential.


The GSBI generates normalized T-scores for the main and masking scales, making it easy to compare scores across scales and across tests. Normative data are based on a nonclinical group, spanning ages 17 to 104, with roughly equal numbers of males and females. In addition, the test manual provides average scores for males, females, and clinical and correctional subgroups (sex offenders, sub- stance abusers, and sexual addicts). It has not been found useful to obtain a total score on the GSBI, given that the subscales are relatively independent, representing differ- ent constructs of interest.


In Study 1 (= 500; Garos & Stock, 1998b), internal re- liability coefficients for the GSBI subscales for females were .75 for Sexual Obsession, .73 for Permissiveness, .74 for Sexual Stimulation, and .80 for Discordance. For males the coefficients were .72 for Sexual Obsession, .66 for Permissiveness, .67 for Sexual Stimulation, and .85 for Discordance. In Study 2 (= 1,000; Garos & Stock, 1998a), internal reliability coefficients for the total sample were .80 Sexual Obsession, .70 Permissiveness, .72 Sexu- al Stimulation, and .82 Discordance.

The three masking scales have reported alpha reliabilities of Control (8 items, alpha = .85), Excitement (7 items, alpha = .82), and Insecurity (6 items, alpha = .74; Garos, 2009).

In Study 2 (= 1,000; Garos & Stock, 1998b) the fol- lowing test-retest reliability coefficients for a subset of participants were reported: .62 on the Permissiveness sub- scale (= 53), .84 on the Sexual Obsession subscale (= 53), .72 for the Discordance subscale (= 54) and .79 on the Sexual Stimulation (= 52) subscale. All results were significant at the < .001 level. The time between administrations was 19 days.


Factorial validity for the GSBI was established through equivalence of factor structures and item retention in Studies 1 and 2 (Garos & Stock, 1998a, 1998b). The final factor structure obtained in Study 1 was replicated for males and females, and for the total sample in Study 2. To evaluate the discriminant and convergent validity of the GSBI, correlations were computed to determine the independence of GSBI subscales with related but conceptually distinct subscales of other behavioral and psycho- logical measures (Garos & Stock, 1998a). These measures included the Sexual Addiction Screening Test (Carnes, 1992), the Sexuality Scale (Snell & Papini, 1989), the Zuckerman-Kuhlman Personality Questionnaire (Zucker- man, Kuhlman, Joireman, Teta, & Kraft, 1993), the Sexual Opinion Survey (Fisher, Byrne, White, & Kelley, 1988), the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), and the Rosenberg Self-Esteem Scale (Rosenberg, 1962). All scales correlated as predict- ed ranging from .05 to .52.

A validity study was conducted with prisoners to determine how well the associated affective and cognitive characteristics of intrapsychic conflict could distinguish between incarcerated sex offenders and incarcerated nonsex offenders (Garos et al., 2004). The Discordance subscale emerged as the most significant predictor variable between sex offenders and nonoffenders (Wilks’ λ = .553) F(1,184) = 148.63, < .001. A discriminant function analysis was conducted to establish the GSBI’s ability to correctly predict classification of group membership. For sex offenders, 80.6% were correctly classified, as were 88.3% of nonsex offenders. Overall, 84.2% of cases were correctly classified. In the cross-validation sample, results for placement in group membership for sex offenders remained the same (80.6%), whereas 87.2% of nonsex of- fenders were correctly classified. For this sample, an over- all 83.7% of cases were correctly classified. These results indicated significant improvement over base rates, with the Discordance subscale emerging as the strongest of the four predictors. In the same study, a logistic regression was conducted as a second index of the GSBI’s ability to predict group membership. A positive predictive value of 82.8% was obtained, which was a significant improvement relative to the base rate of cases in the sex offender category (53%), again with Discordance emerging as the most significant predictor. To test whether the Permissive- ness, Sexual Stimulation, and Sexual Obsession subscales contributed significantly to the predictive ability of the GSBI, a logistic regression was performed with the omission of the Discordance subscale. Although the percentage of correctly classified cases decreased with the omission of Discordance relative to the base rate of cases in the sex offender category (52%), a positive predictive value of 68.4% represented a 17.6% improvement over chance.

In a study of inpatients receiving experiential therapy for sexual addiction (Klontz, Garos & Klontz, 2005), two subscales of the GSBI were sensitive to a main effect of time. Male and female patients reported less sexual preoccupation and less discordance from pretreatment to posttreatment. Changes were stable at 6-month follow-up.

Other Information

Copyright: Western Psychological Services. Testing materials are available from Western Psychological Services,

12031 Wilshire Blvd., Los Angeles, CA 90025-1251; Telephone: (800) 648–8857;

Address correspondence to Sheila Garos, Psychology Department, Texas Tech University, Lubbock, TX 79409-2051; e-mail: [email protected]

Garos Sexual Behavior Inventory (Sample Items)

Main Scales

Discordance Subscale

I feel uncomfortable about my sexual practices. I am afraid I am not sexually attractive enough.

Sexual Obsession

It feels impossible to stop masturbating.

Once I am aroused it is difficult not to masturbate or have sex.


Birth control should be readily available to young people. Women’s sexual advances should be subtle.

Sexual Stimulation

I like not wearing underwear.

I enjoy being aroused by my sexual thoughts and feelings.

Masking Scales


When I wake up from sexual dreams, I like to masturbate. Sexual thoughts overpower my better judgment.


As a child, I was more interested in sex than my friends were.

The smell of men’s cologne/ladies’ perfume makes me think of sex.


It is easier to have sex if my partner is the aggressor.

I feel like a failure if I cannot keep someone sexually excited.

Sample items from the GSBI copyright © 2008 by Western Psychological Services. Reprinted for sole purpose of scholarly reference by permission of WPS, 12031 Wilshire Boulevard, Los Angeles, California, 90025, U.S.A., Not to be reprinted in whole or in part for any additional purpose without the expressed, written permission of the publisher ([email protected]). All rights reserved.


Beck, A. T., Ward, A. T., Mendelson, C. H., Mock, M., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 53–63.

Carnes, P. (1992). Out of the shadows: Understanding sexual addiction (2nd ed.). Center City, MN: Hazelden.

Fisher, W. A., Byrne, D., White, L. A., & Kelley, K. (1988). Erotopho- bia-erotophilia as a dimension of personality. The Journal of Sex Re- search, 25, 123–151.

Garos, S. (2009). Garos Sexual Behavior Inventory test manual. Los An- geles: Western Psychological Services.

Garos, S., Bleckley, K. M., Beggan, J. K., & Frizzell, J. (2004). The rela- tionship between intrapsychic conflict and deviant sexual behavior in sex offenders. Journal of Offender Rehabilitation, 40, 23–38.

Garos, S., & Stock, W. A. (1998a). Investigating the discriminant validity and differentiating capability of the Garos Sexual Behavior Index. Sexual Addiction and Compulsivity: The Journal of Treatment and Prevention, 5, 251–267.

Garos, S., & Stock, W. A. (1998b). Measuring disorders of sexual fre- quency and control: The Garos Sexual Behavior Index. Sexual Ad- diction and Compulsivity: The Journal of Treatment and Prevention, 5, 159–177.

Klontz, B., Garos, S., & Klontz, P. (2005). The effectiveness of mul- timodal experiential therapy in the treatment of sexual addiction. Sexual Addiction and Compulsivity: The Journal of Treatment and Prevention, 12, 275–294.

Rosenberg, M. (1962). The Rosenberg Self-Esteem Scale. Department of Sociology, University of Maryland, College Park.

Snell, W., & Papini, D. (1989). The sexuality scale: An instrument to measure sexual esteem, sexual depression and sexual preoccupation. The Journal of Sex Research, 26, 756–763.

Zuckerman, M., Kuhlman, M., Joireman, J., Teta, P., & Kraft, M. (1993). A comparison of three structural models for personality: The Big Three, the Big Five, and the Alternative Five. Journal of Personality and Social Psychology, 65, 757–768.