Sexual Compulsivity Scale: Adaptation and Validation in the Spanish Population

Sexual compulsivity has been studied in relation to high-risk behavior for sexually transmitted infections. The aim of this study was the adaptation and validation of the Sexual Compulsivity Scale to a sample of Spanish young people. This scale was applied to 1,196 (891 female, 305 male) Spanish college students. The results of principal components factor analysis using a varimax rotation indicated a two-factor solution. The reliability of the Sexual Compulsivity Scale was found to be high. Moreover, the scale showed good temporal stability. External correlates were examined through Pearson correlations between the Sexual Compulsivity Scale and other constructs related with HIV prevention. The authors’ results suggest that the Sexual Compulsivity Scale is an appropriate measure for assessing sexual compulsivity, showing adequate psychometric properties in the Spanish population.

low-income groups. The scale showed high levels of internal consistency (α = .86 for gay men and α = .87 for African American men and women). In both samples, there are significant correlations between sexual compulsivity, high-risk sexual behavior, and drug use before sex. Although this scale has been used to evaluate sexual compulsivity, especially among individuals belonging to groups deemed at high risk or HIVpositive individuals, the few studies conducted in college students suggest that higher scores in sexual compulsivity are associated with a higher number of unprotected sexual encounters Gullete & Lions, 2005). To design optimal prevention campaigns, it is necessary to identify the characteristics of those most likely to engage in high-risk behavior (McBride, Reece, & Sanders, 2008).
The aim of our study was therefore to translate, adapt, and validate the SCS of Kalichman et al. (1994) in a Spanish population that has been little studied-young people-represented here by a sample of college students. Although college students have not been identified as a risk population, several studies in Spain and in other countries have emphasized the high prevalence of risky sexual behavior in this population (Ballester, Gil, Giménez, & Ruiz, 2009;McBride et al., 2008;Rolison, 2002).

Measures and Instruments
Participants responded individually to the SCS. To assess convergent and divergent validity, the participants also answered the Sexual Sensation Seeking Scale (Kalichman et al., 1994), the Spanish version of the Beck Depression Inventory (Beck, Rush, Shaw, & Emery, 1979[translated by Sanz & Vázquez, 1998) and The Brief Version of the Fear of Negative Evaluation Scale (Leary, 1983). Last, they responded to the Cuestionario de Prevención del Sida ("AIDS Prevention Questionnaire"; Ballester, Gil, Guirado, & Bravo, 2004), which was developed in Spanish. We subsequently explain these questionnaires: • SCS (Kalichman et al., 1994): This scale consists of 10 Likert-type items that assess sexual compulsivity, measured on a scale ranging from 1 (not at all like me) to 4 (very much like me), with a total score between 10 and 40. The internal consistency of the instrument applied to a sample of college students was .76 for men and .81 for women (Reece, Plate, & Daughtry, 2001). • Sexual Sensation Seeking Scale (Kalichman et al., 1994): This scale is designed to measure "the propensity to attain optimal levels of sexual excitement and to engage in novel sexual experiences." It consists of 9 Likert-type items measured on a 4-point scale ranging from 1 (not at all like me) to 4 (very much like me), with a total score range between 9 and 36. The instrument has shown an internal consistency of .81 in various samples, such as gay men (Kalichman & Rompa, 1995) and college students (Gray & Wilson, 2007). • Beck Depression Inventory in its Spanish Version (Sanz & Vázquez, 1998): This scale is a 21-question multiple-choice self-report inventory that reflects cognitive, affective, behavioral, and somatic symptoms of depression in relation to the 7 previous days. The items are answered according to a 4-point scale denoting different levels of severity and intensity. The Spanish version has shown good internal consistency in a sample of college students (.83). We present the SCS validated in this work for the Spanish population in Appendix A.

Procedure
We informed 1,196 participants of the purpose of the study. The research team told participants that they would be assessed on several sexual behaviors in young people. Then, they completed the SCS voluntarily, anonymously, and confidentially in one sitting during class time thanks to the collaboration of lecturers in different degree courses. In addition, the first 300 tested students also responded to questionnaires described earlier. After a week, 100 randomly selected participants answered the SCS a second time.

Statistical Analysis
We used SPSS to perform all of our analyses. We used a principal component factor analysis using varimax rotation to ascertain the factorial structure of the questionnaire. In addition, we performed a descriptive analysis of items and calculated internal consistency using Cronbach's alpha coefficient. We examined convergent validity through Pearson correlations between the SCS and other constructs related to HIV prevention (total score of Sexual Sensation Seeking Scale, Beck Depression Inventory, and an item from the Cuestionario de Prevención del Sida that requested information about the number of sexual partners). Selection was carried out according to the relationship between these elements and sexual compulsivity in the literature (Benostch et al., 1999;Gullete & Lions, 2005;Kalichman, Greenberg, & Abel, 1997;Kalichman & Rompa, 1995Raymond, Coleman, & Miner, 2003). We calculated test-retest reliability through Pearson's correlation coefficient.

Factorial Structure of the SCS
We obtained an index of .896 in the Kaiser-Meyer-Olkin test, and Bartlett test of sphericity results of χ 2 (45) = 3551.851, p < .001. We therefore concluded that factor analysis was appropriate.
We performed a factor analysis by extracting the principal components with varimax rotation, revealing a two-factor solution with eigenvalues higher than one, which explained 52.74% of total variance ( Table 2). The first factor, with an eigenvalue of 4.237, accounted for 42.37% of the variance and included five items (1,2,3,4, and 10) that represented an "interference of sexual behavior" whereby sexual compulsivity caused interpersonal relationship problems and social maladjustment. The second one, with an eigenvalue of 1.037, accounted for 10.37% of variance and represented a "failure to control sexual impulses" and entailed personal discomfort associated with sexual compulsivity (Items 5,6,7,8,and 9).

Descriptive Analysis of Items
In the total sample, the ratings given to the items ranged between 1.141 and 1.623 Table 3 shows that Items 5 and 6 had the highest scores and Items 2 and 4 had the lowest. In the men sample, the scores ranged between 1.24 and 1.84. Items 6 and 9 scored higher. In the female sample, Items 5 and 6 were rated higher. In both samples, Items 2 and 4 were rated lower.

Internal Consistency
To analyze internal consistency, we calculated Cronbach's alpha coefficient for total scale (α = .837). Correlations of individual items with corrected total score of the SCS, that is, the total score regardless of the item concerned,  Table 3). In the interference subscale, correlations ranged between 0.402 and 0.549. In the failure to control sexual impulses subscale, correlations ranged between 0.505 and 0.661.

Construct Validity
To examine convergent validity, we used data from 300 participants who answered the SCS, the Sexual Sensation Seeking Scale (Kalichman & Rompa, 1995), the Beck Depression Inventory (Beck et al., 1979), and the Cuestionario de Prevención del Sida (Ballester et al., 2004). Of total participants, 18 were removed because of a failure to respond to all items in the questionnaires. Therefore, we used data from 282 participants.

Convergent Validity
We calculated convergent validity using Pearson's correlation coefficient between the SCS, the Sexual Sensation Seeking Scale (Kalichman & Rompa, 1995), the Beck Depression Inventory (Beck et al., 1979) and the number of sexual partners, an item included in CPS (Ballester et al., 2004). Sexual compulsivity rarely-if ever-stands alone. There are often comorbid issues  Ballester-Arnal et al. in sexual compulsivity such as sexual dysfunction, physiological issues, other substances or addictions, depression and anxiety, or posttraumatic stress disorder (Delmonico & Griffin, 2011). Furthermore, there are several studies in literature that have concluded that sexual compulsivity and sexual sensation seeking had a significant contribution to realization of sexual risk behaviors (Beck, Thombs, Mahoney, & Fingar, 1995;Reece et al., 2001). As expected, our results show that the scale had a significant positive correlation with Sexual Sensation Seeking Scale (r = 0.444, p < .000), Beck Depression Inventory (r = 0.161, p < .007) and with the number of sexual partners, an item of Cuestionario de Prevención del Sida (r = .203, p < .001).

Test-Retest Reliability
Of the total sample (N = 1,196), 100 young people answered the SCS a week later, but 9 of these were excluded because of their failure to answer any of the items. We therefore analyzed test-retest reliability with 91 participants by calculating Pearson's correlation coefficient. There were significant correlations in both total scale and subscales between the two occasions of administration. Thus, correlation between total scale score in each pass was 0.725 (p < .01).

DISCUSSION
In previous studies, Kalichman and Cain (2004) and McBride et al. (2008) found a two-factor solution for the SCS. Both called these factors in the same way: Factor 1 was called social disruptiveness and Factor 2 was called personal discomfort. However, some items (2, 5, 6, and 7) are grouped on opposite factors on these studies. In our study, there is also a twofactor solution, but we have named them differently. Factor 1 was labeled as interference of sexual behavior, while Factor 2 was called failure to control sexual impulses. When analyzing the two factors obtained, it is apparent that they both reflect variations in factor loadings of individual items compared to those obtained in the study by McBride et al. (2008), which was also conducted with a sample of young people. However, our results are similar to those obtained by Kalichman and Cain (2004) in a sample of men and women receiving services from sexually transmitted infection clinic with different ages to those in our study. There is a difference in factor loading of Item 10, "It has been difficult for me to find sex partners who desire having sex as much as I want to." The item belongs to the personal discomfort factor in the study of Kalichman and Cain (2004), while it belongs to the interference of sexual behavior factor in our sample.
Relating to descriptive analysis of the items, the sample of the present study has the lowest scores of sexual compulsivity compared with other samples in other studies.
Because sexual compulsivity has been associated to risky sexual behaviors and increased sexual frequency, it is reasonable that means obtained in the college students sample were lower than those of other that assessed men who have sex with men campus cruisers (MSM) and men who have sex with men and women campus cruisers (MSMW) campus cruisers , people living with HIV (Benotsch, Kalichman, & Pinkerton, 2001;Kalichman & Rompa, 2001;Reece, 2003) or gay/bisexual male sex workers (Parson et al., 2001). In contrast, it is important to note that in all studies the mean score of sexual compulsivity is higher in men than in women; educational and social factors may possibly account for these differences.
In terms of reliability, SCS has shown good internal consistency in total scale (α = .837) and in subscales. Internal consistency of the scale obtained in this study is slightly higher than that found by Dodge et al. (2004) among students aged 18 to 25 years old (α = .82). However, our result is slightly lower than the pilot study of the scale undertaken by Kalichman et al. (1994) in which participants were sexually active men who considered themselves homosexual (α = .89). It was also been slightly lower than the original study by Kalichman and Rompa (1995) of gay men (α = .86) and of primarily African American men and women in inner-city areas on low incomes (α = .88). In any case, few differences are mere hundredths, that is, the results are virtually identical.
In relation to convergent validity, this study found significant relationships between SCS and another measures. As expected, the Sexual Sensation Seeking Scale, the number of sexual partners and the participation in risky behavior correlated positively and significantly with SCS. Our result is consistent with those obtained in previous studies, where it was concluded that the Sensation seeking and sexual compulsivity had a significant contribution to realization of risky sexual behaviors (Beck et al, 1995;Reece et al., 2001). In another study, age, sexual sensation seeking, and sexual compulsivity were shown as predictors of risk behaviors (Gullette & Lyons, 2005). Referring to depression, our results show significant positive correlations between this variable and sexual compulsivity. This is also in line with the literature reviewed. Across studies, people with high scores on compulsivity have extremely high rates of psychological disorders, in particular depression, anxiety disorders and substance use disorders (Black et al., 1997;Raymond et al., 2003).
As for the scale's temporal stability, the results of this study show that it is high for the scale. That value is similar to previous studies (Kalichman & Rompa, 1995). The results obtained suggest that the SCS built by Kalichman et al. (1994) and translated by our team for Spanish population, is an appropriate measure for evaluating sexual compulsivity related with HIV prevention in young people. The scale shows adequate psychometric properties.
Our study has some limitations that should be considered in future research, such as number of participants of each sex-there were many more women than men in this study-and sexual orientation-most of the individuals in this sample were heterosexual students.
It is highly desirable to complete the sample with more men and people with different sexual orientations. Furthermore, it would be interesting to use a clinical group of patients suffering from sexual compulsivity to compare their scores with those of the general population and thus to establish a cutoff point with greater clinical significance.