Latino sexual styles: developing a nuanced understanding of risk.

Within the next 50 years, Latinos are predicted to make up 25% of the U.S. population (U.S. Census, 2000a). It is estimated that Latino youth numbers are growing even faster, reaching 25% of the youth population in less than 25, not 50, years (U.S. Census, 2000b). Despite the increase in overall population percentage, Latinos still have disproportionately high incidence rates of human immunodeficiency virus infection (HIV) and sexually transmitted infections (STIs). Although the rates are not as high for Latinos as they are for African Americans, they are still significantly higher than rates for Whites (CDC, 2000; Driscoll, Biggs, Brindis, & Yankah, 2001). Many Latino youth have reported that they do not consider themselves at risk for HIV (Flores-Ortiz, 1994; Sneed, Morisky, Rotheram-Borus, et al., 2001) despite the use of contraception being rather low for Latino adolescents. Latino males were less likely to use condoms than their White or African American peers, and Latinas have declined in their use of birth control over time (Driscoll, Biggs, Brindis, & Yankah, 2001; Gurman & Borzekowski, 2004). Only half of Latino youth have reported using condoms (Taylor-Seehafer & Rew, 2000; Ventura, Mosher, Curtin, Abma, & Henshaw, 2001).

Indeed, Latinas have a high rate of teen pregnancy, lower than Blacks but higher than Whites, showing the slowest rate of decline for all groups over the past few years (Driscoll, Biggs, Brindis, & Yankah, 2001; Kenney, Reinholtz, & Angelini, 1997; Martin, Hamilton, & Ventura, 2001). Specifically, the risk seems concentrated among Latino groups: The rates for teen births are highest in Mexican Americans (Driscoll, Biggs, Brindis, & Yankah, 2001). Among ages 15-18 years, U. S.-born Latinas were most likely to have been pregnant, while among ages 19-24 years foreign-born Latinas were most likely to have been pregnant (Minnis & Padian, 2001). The study also found that the U. S.-born Latinas were more likely to have had an abortion or have contracted chlamydia than foreign-born Latinas. Taken together, the increase in the number of Latinos born and living in the United States and the high rates of teen pregnancy and STIs among this population indicate that there is a great need to understand the mechanisms underlying risk behavior to better inform prevention programming.

Latino Cultural Values

Crucial to our understanding of Latino sexual identity are several cultural values held by many Latino ethnic groups, including familism, marianismo, verguenza, and machismo.

Familism is the strong attachment and loyalty to immediate and extended family (Villarreal, Blozis, & Widaman, 2005) and is evidenced in adolescent sexuality in two ways: first, in the impact of families on adolescent sexuality, and second, in the expectations and anticipated life trajectories of adolescents. In Latino cultures, the family plays a strong role in sexual socialization and has consistent links with adolescent sexual behavior. As a result, there is an increased emphasis on respeto or respect for one's parents and other elders (Pavich, 1986). Girls in particular are influenced by their mothers and communities (East, Felice, & Morgan, 1993; Liebowitz, Castellano, & Cuellar, 1999), though there is evidence that these factors may also impact boys (Fasula & Miller, 2006). In addition, for girls, living with both parents, having recently immigrated, and living in an area with a high density of Hispanics decreased initiation into sexual activity (Upchurch, Aneshensel, Mudgal, & McNeely, 2001; Upchurch, Aneshensel, Sucoff, & Levy-Storms, 1999). Familism also distinguishes Latino youth and other groups in their expectation for life trajectory. Of all groups, Latino males had the highest desire for marriage (South, 1993). More often than Blacks or Anglos, Latino youth anticipate being married and/or pregnant at a younger age (East, 1998; South, 1993) and most Latinos hold views opposing abortion (Ellison, Echevarria-Cruz, & Smith, 2005). Generally, a lower percentage of Latinos say they are waiting until marriage to have sex than Anglos (Moran & Corley, 1991), and Latinos tend to have a large discrepancy in the ages of the partners with females being much younger than their male partners (Driscoll, Biggs, Brindis, & Yankah, 2001; Marin, Coyle, Gomez, Carvajal, & Kirby, 2000; Taylor, Chavez, Adams, Chabra, & Shah, 1999), resulting in a large relational power differential (Gowen, Feldman, Diaz, & Yisrael, 2004) and encouraging the traditional gender roles often promoted in Latino cultures.

The most prominent value for females is marianismo, which is embodied by women who are submissive, chaste, and relationship-oriented (Denner & Dunbar, 2004; Pavich, 1986). This value is associated with the attributed characteristics of the Virgin Mary (Wood & Price, 1997); morally superior and spiritually stronger than men, she is pure and asexual, ignorant of knowledge about sexuality or contraception. Paramount is virginity (Barkley & Mosher, 1995), while at the same time there is a reverence for motherhood, requiring a loss of virginity (Driscoll, Biggs, Brindis, & Yankah, 2001; Pavich, 1986; Villarruel, 1998). Interviews with Latinas aged 18 to 51 years (Davila, 2005) highlighted the additional role of verguenza, defined as sexual shame and lack of sexual knowledge (Sugar, 1995). Many women said that they were unable to speak about sexuality during their adolescence in their home and then later with their partners. These women "found it easier or less embarrassing to engage in an unwanted or unprotected sexual encounter than to openly discuss their sexual desires or preferences" (p. 364). Indeed, Hispanic female adolescents have been found to have lower sexually assertive beliefs than Anglos (Rickert, Sanghvi, & Wiemann, 2002), a factor which has been significantly predictive of condom use (Villarruel, Jemmott, Jemmott, & Ronis, 2004). However, some Latina mothers describe attempting to curb the influence of verguenza with open communication with their daughters (McKee & Karasz, 2006; Nadeem, Romo, & Sigman, 2006; Raffaelli & Green, 2003).

For males who are expected to act in traditionally "masculine" ways, the most frequently cited value is machismo, which has carried both positive and negative connotations among researchers (Cromwell & Ruiz, 1979; Zapata & Jaramillo, 1981). Initially conceptualized as exclusively negative, it was defined as a social behavior pattern in which the Latino male exhibits an overbearing attitude to anyone in a position inferior to his, demanding complete subservience, and characterized by unpredictable power (Wood & Price, 1997). Machismo has been linked to risk behaviors such as heavy drinking (Galanti, 2003) and risky sex (Castro-Vazquez, 2000; Locke, Newcomb, & Goodyear, 2005; Marin, 2003). Later scholars, in particular Latino scholars, fought the stereotype to focus on the positive attributes of machismo, such as honor, respect, and dignity (South, 1993).

Additional Considerations with Latino Youth

Preliminary research of these youth has shown that there are unique trends and factors. Latino adolescents vary from their Anglo and Black counterparts in terms of context and behavior (CDC, 2004). Specifically, two primary characteristics are noted here: country of origin and acculturation. Research has revealed a striking heterogeneity among Latino groups from varying countries of origin on sexuality related variables like menarche, sexual initiation age, and sexual education (DuRant, Pendergrast, & Seymore, 1990). In addition, differences in the reliability and validity of measures have been found between individuals from Colombia, Guatemala, Honduras, Mexico, Puerto Rico, and El Salvador (Umana-Taylor & Fine, 2001). The research thus far with "Latino" individuals has been too limited due to its lack of understanding of national origin (Barkley & Mosher, 1995; Kane, 2000; Torres, 2004).

Another consideration is the role of acculturation (Barkley & Mosher, 1995; Padilla, 1980; Pavich, 1986). Often less acculturated individuals, while they have fewer resources and less power, tend to have healthier outcomes than those who are more acculturated (e.g., Hayes-Bautista, Schink, & Chapa, 1988; Landale, Oropesa, & Llanes, 1998; Marin, Gomez, & Hearst, 1993; Sabogal & Catania, 1996). When applied to sexuality and sexual behavior, increased acculturation has been linked to less traditional gender roles (Phinney & Flores, 2002), younger ages for marriage and giving birth (East, 1998), increased risky sexual activity (Hines & Caetano, 1998; Kaplan, Erickson, & Juarez-Reyes, 2002), and increased alcohol use (Epstein, Botvin, & Diaz, 2000).

New Understandings of Sexuality

Various studies have looked to intrapsychic factors in understanding sexual behavior, such as expectancies (Bourdeau, Saltz, Bersamin, & Grube, 2007) and mood (Fortenberry, Temkit, Tu, Graham, Katz, & Orr, 2005), among others. New research is pointing to the notion of sexual identity (Buzwell & Rosenthal, 1996), noted as a "sense of self as a sexual being" (Graber, Brooks-Gunn, & Galen, 1998). Sexual identity has been further delineated as "the integration of psychological, psychical, societal, cultural, educational, economic, and spiritual factors" (Tolman, Striepe, & Harmon, 2003) that goes beyond individual attitudes and beliefs to an understanding of the interaction of attitudes and beliefs. Components of sexual identity have been found to be predictors of sexual and contraceptive behavior, each in their own right. These include sexual self-concept (Breakwell & Millward, 1997; Garcia & Carrigan, 1998; Murry, Brody, McNair, Luo, et al., 2005), body image (Halpern, King, & Oslak, 2005; Murry et al., 2005), sexual self-efficacy (Taris & Semin, 1999), sexual social comparison (Palmer & Murry, 1995), experiences of sexual desire or interest (Fortenberry et al. 2005; Welles, 2005), sexual self-perception (Garcia, Cieselka, & Fuchs, 1999), and sexual beliefs (Gowen, Feldman, Diaz, & Yisrael, 2004).

Recent research has begun the process of understanding the interactions of identity components to influence sexual behavior. To date, the most comprehensive work on sexual identities has been completed by researchers Buzwell, Rosenthal, and Smith (Buzwell & Rosenthal 1995, 1996; Smith & Rosenthal, 1998) with populations of Australian youth. These researchers chose three concepts with which to frame sexual identity: sexual self-esteem (perceptions of one's sexual activity, feelings of sexual appeal and attractiveness, adequacy of behavior in sexual situations, and body image), sexual self-beliefs (feelings of sexual energy, sexual adventurousness, anxiety in sexual situations, interest in commitment with sexual partner), and sexual self-efficacy (ability to say no to unwanted sexual behavior, ability to be assertive in achieving sexual satisfaction, ability to purchase and use condoms). Using cluster analytic techniques, they defined five sexual identities: sexually naive, sexually unassured, sexually competent, sexually adventurous, and sexually driven. Each identity was significantly different from the others. For example, the sexually naive cluster was low on self-efficacy, specifically assertiveness and using precautions, and was high on being able to say no to sex. By contrast, the sexually competent cluster had high self-efficacy on all items and had high sexual self-esteem. The sexually driven cluster had high self-esteem but low ability to say no to sex. The most important contribution of this research is that each cluster was found to differ on measures of sexual risk taking. The sexually competent and the sexually naive clusters were found to take significantly more risks (i.e., unprotected sex) with regular sex partners than the other groups. The sexually adventurous and the sexually driven clusters had more sex partners in the previous six months than the other three groups.

Understandings of Latino Sexuality

Studies of sexual identity show promise for untangling the complex relationships between individual concepts (e.g., body image and self-efficacy). No studies to date have taken this approach to understanding Latino adolescent sexual identity, however, despite urgent need to address the public health concerns of this population. Given the cultural values noted previously, we would expect that sexual identity would be heavily gendered for Latino youth. Specific to the scales developed by Buzwell and Rosenthal (2006), we would anticipate that girls would have low sexual self-esteem, lacking confidence in their sexual abilities or pride in their body image, given the implications of verguenza. Marianismo would also most likely suggest low sexual self-efficacy with regard to initiating sexual activities or knowing how to use birth control (as women are to be characterized by a lack of sexual knowledge). Conversely, they would be high in self-efficacy regarding saying no to sexual activity, fulfilling the value of chastity. In addition, they would value commitment for sexual relationships. In contrast, male youth would demonstrate the confidence of machismo, resulting in high interest in sexual exploration and expression of virility, confidence in sexual behaviors, and efficacy in initiating sexual encounters. It is unclear how values for adolescent males would impact efficacy with regard to birth control or condoms. The positive component of machismo would direct men to protect their loved ones, perhaps by wearing condoms, while familism and the orientation toward parenthood would indicate an openness to pregnancy and decrease the use of precautions.

The purpose of this study was to further the understanding of Latino adolescent sexual identity for males and females and elucidate its relationship to sexual risk taking behaviors that contribute to the high rates of pregnancy and STIs. Using a group of third-generation Mexican American male and female adolescents, this exploratory study investigated the following:

1. Validity of the Buzwell and Rosenthal scales with Latino adolescents;

2. Differences between male and female adolescents on measures of sexual identity;

3. Cluster group membership and characteristics; and

4. Relationship between cluster groups and risky sexual behaviors.

Method

Procedures

Participants were surveyed from five charter middle and high schools in the Southwest United States that received HIV and substance abuse prevention programs from a local AIDS service organization. School-wide surveys were conducted at the beginning and end of every school year as one method of evaluating the HIV and substance abuse prevention program. The schools varied greatly. One high school was college preparatory level and had a rigorous curriculum; its primary feeder middle school was also included. The third school was a combined middle and high school and was known as a "last-ditch" school for students who have been expelled from other schools; students in this school tend to have less educational attainment and are generally older than the average for their grade. The final two schools were both high schools known for their emphasis on Chicano studies and history and were run by a predominantly Latino social service agency. Data presented here were from the first school-wide survey.

Parents were contacted through the schools' orientation packets and during registration. Active consent was used. Students with parental consent completed paper and pencil measures during a designated class period. Participants who had limited literacy were read the items aloud by a program or research staff member as the students completed the survey. Students were informed that the surveys were available in Spanish, however, no student requested a Spanish version.

Participants

Slightly more than half the participants were female (59.4%). The majority identified as heterosexual (88.4%). Participants' ages ranged from 12.8 to 21.8 with a mean of 16.1, with no difference between genders (male M=15.98, female M = 16.22, t= -l.00, p = n.s.). Participants were in grades 7 to 12 grade, with a modal grade of 10 (26.5%). Just over half (56.8%) reported ever having sex. There was no significant difference between genders on sexual experience, with 50.0% of males and 43.8% of females reporting having had sex (Goodman and Kruskal's [tau] n.s.).

Measures

Country of origin. Country of origin was determined by responses to questions about the participants and the participants' parents. Participants were allowed to write in ethnicity for themselves and both parents. They were also asked to indicate a response for themselves and each parent that best fit from a list of choices. The list of choices included: Asian American, Black or African American, Hispanic or Latino, White or Caucasian not Hispanic, American Indian, Mixed (parents are from two groups), Other, Mexican National, Central American, and South American.

Acculturation. Participants were told to indicate the country of origin for themselves, their parents, and their grandparents.

Sexual self-esteem. Buzwell & Rosenthal's (1996) scales for sexual identity were used for this study. Sexual Self-Esteem has four subscales: Behavior (e.g., It is hard for me to know how to behave in a sexual situation, I feel good about my sexual behavior); Attractiveness (e.g., I frequently feel ugly or unattractive, People say I am good looking); Conduct (e.g., Partners have (would) find me sexually satisfying, I am confident about being able to get a girlfriend/boyfriend); and Body Perception (e.g., I like my body, When others look at me, they must think I am poorly developed). Responses ranged from 1 (Strongly Disagree) to 4 (Strongly Agree) with some items recoded such that higher responses indicated higher self-esteem.

Sexual self-beliefs. All four subscales were used: Arousal (e.g., I can feel frustrated if I don't have sex often, Sexual fulfillment is very important to me); Exploration (e.g., I would like to experiment when it comes to sex, When it comes to sex, I would try anything once); Anxiety (e.g., I would find it hard to relax while having sex, If I had sex, I would worry about someone finding out); and Commitment (e.g., It doesn't matter who you have sex with as long as you enjoy it, There needs to be commitment before I have sex). Responses ranged from 1 (Strongly Disagree) to 4 (Strongly Agree) with some items recoded such that higher responses indicated greater espousal of beliefs.

Sexual self-efficacy. Three subscales were used: Say No (e.g., Ask someone to wait for sex if no condoms available, Refuse to have sex with partner even when they really wanted to); Assertiveness (e.g., Be the one to start sexual activities, Tell partner how to treat you sexually); and Precautions (e.g., Discuss HIV protection even if contraception already used, Carry condoms just in case). The Sexual Self-Efficacy Scales had responses ranging from 0 (Cannot Do) to 5 (Can Do; Absolutely Certain).

Sexual risk-taking. Additional questions included intention items, assessing the likelihood of having sex in the next three months, the likelihood of having more than one partner in the next three months, and the likelihood of practicing safe sex in the next three months (scored 1 "Not at all likely" to 4 "very likely"). The likelihood of practicing safe sex was reverse coded. Questions also were more behavior-specific, including dichotomous "yes/no" items (ever having sex, having any sex in the last 30 days, using a condom at last sex, having a current sex partner, talking to that partner about safe sex in the past 30 days, having sex under the influence in the last three months or during the last sexual encounter, and having sex in exchange for drugs or money) and continuous items (the age of initiation into sexual activity, lifetime number of partners, and number of partners in the last three months) where participants were allowed to choose from a list of numbers.

Analyses

Cronbach's alpha and confirmatory factor analytic techniques were used to assess the suitability of these measures with the study sample. Independent t-tests were used to assess gender differences. Cluster analysis (Ward's method using squared Euclidean distance) was used to create cluster groups. Analyses of variance and Chi-square tests were used to test differences between the cluster groups on sexual risk behaviors.

Results

To control for issues related to country of origin and acculturation, only participants who indicated that either the participant or his/her parents were of Mexican (Mexican American) origin. When allowed to freely write in their own ethnicity, participants responded most frequently with "Hispanic" (30.1%), followed by Mexican (23.3%) and Mexican American (13.1%). The remaining write-in responses all had frequencies of 1 or 2 responses each. When forced to respond to a list of choices, the majority chose "Hispanic or Latino" (80.0%). The second most reported was "Mexican National" (11.0%). Finally, most reported their parents' ethnic identity as Hispanic or Latino (fathers: 60.6%; mothers: 62.6%) or Mexican National (fathers: 26.5%; mothers: 24.5%). Most responded that the previous three generations were all born in the United States (72.3%).

Analyses of missing data were conducted using Little's MCAR tests. All individuals with systematic missing data were deleted. Follow-up analyses indicated that younger students were more likely to have a larger percentage of items missing, including many who did not answer any of the scale questions. After deletion of these cases, the remaining missing items followed a random pattern. EM (expectation-maximization) estimation was used to impute missing values. The final sample size was 155.

Confirmatory factor analysis was used to assess the sexual identity measures' structure with this population. Factor structure of the sexual self-efficacy subscales was strong, with only one item dropped from each of the final scales used in this study. The analyses demonstrated a good fit for Precautions with 5 items ([chi square] [5] = 6.3, p = .27, Bentler-Bonett Normal Fit Index (BNFI) = .96, Comparative Fit Index (CFI) = .99, Root Mean-Square Error of Approximation (RMSEA) = .04), Say No with 8 items ([chi square] [19] = 19.9, p = .40, BNFI = .96, CFI = 1.0, RMSEA = .02), and Assertiveness with 4 items ([chi square] [6] = 8.7, p = .12, BNFI = .99, CFI = .97, RMSEA = .07). The other two scales had the best fit with a more reduced factor structure. Sexual self-beliefs had three factors, with items on arousal and exploration subscales falling into a combination subscale. The scales had good fit indices for Anxiety with 6 items ([chi square] [9]= 10.8, p=.29, BNFI = .96, CFI = .99, RMSEA = .04), Commitment with 8 items ([chi square] [20] = 25.4, p = .19, BNFI = .91, CFI = .98, RMSEA = .04), and Arousal/Exploration with 13 items ([chi square] [66] = 85.6, p = .04, BNFI = .85, CFI = .96, RMSEA = .05). Finally, rather than a four-factor solution, the best fit for Sexual Self-Esteem was a two-factor model with items from Body Perception and Attractiveness combined, as well as items from Conduct and Behavior combined. The resulting two scales fit well, with Body/Attractiveness at 5 items ([chi square] [5] = 7.46, p = .19, BNFI = .91, CFI = .97, RMSEA = .06) and Conduct/Behavior at 7 items ([chi square] [12] = 12.9, p .38, BNFI = .95, CFI = .99, RMSEA = .02). Reliability statistics were calculated for each subscale, ranging from .67 to .89 (see Table 1). The internal consistencies of the new scales were comparable or better than those reported in the original study.

Independent samples t-tests revealed statistically significant differences between genders on many of the subscale items (see Table 2). Females scored significantly higher on Anxiety, Commitment, Say No, and Assertiveness, while males scored higher on Arousal/ Exploration. No differences were found on Precautions, Behavior/Conduct, or Body/Attractiveness.

To determine if similar sexual identities result from a cluster analysis of Latino adolescent responses, methods were chosen based on the nature of the data (ordinal) and the desired results (a five-cluster solution as found in Buzwell and Rosenthal, 1996) from the array of procedures available for cluster analysis. First, scores were transformed into Z scores to standardize the measures. Hierarchical agglomerative methodology was chosen for determination of clusters; specifically, Ward's method was used for its efficiency as a space-dilating measure in strengthening the differences between the clusters (Aldenderfer & Blashfield, 1984). Squared Euclidean distance was used as the similarity measure given its superiority over correlation coefficients in sensitivity to the magnitude in differences between the variables. Results of the cluster analysis with five predetermined clusters were assessed for their fit to the data using several procedures. First, fusion coefficients were considered. A large increase in fusion coefficients indicates that two dissimilar clusters have been combined and that the optimal cluster solution is at the previous cluster level. The data indicated a large jump in fusion coefficients between the fourth and fifth clusters, indicating that a four-cluster solution might be a better fit. Second, a series of one-way ANOVAs were completed to test for similarity among the clusters on each of the subscale scores for both five-cluster and four-cluster solutions. Review of the posthoc analyses showed that two of the clusters were very similar on most subscale scores. The four-cluster solution was then considered the most appropriate.

[FIGURE 1 OMITTED]

The four clusters in these data matched those of the original sample quite closely, with one exception: Clusters labeled Unassured and Driven by Buzwell and Rosenthal (1996) were not found here; instead one cluster emerged that seemed to have some characteristics of the two. The four clusters were evenly matched, with two clusters having an average age of 15.8 and two having an average age of 16.6. In addition, one of each of these clusters was predominantly male and one predominantly female for both age categories, creating older female and male clusters and younger male and female clusters. Clusters varied in their patterns of subscale scores and are detailed below (see also Figure 1).

Cluster One (Aventurero, or adventurous, similar to Adventurous) was the smallest group with only 25 members (16% of sample), many of whom were male (72%). This group was older (M = 16.5 years) and very few were still virgins (28.0%). This group had the highest percentage of non-heterosexual members (12% bisexual and 8% gay/lesbian). Those who were sexually experienced had a young initiation age (M = 13.9 years) and many (87.5%) reported having a main partner. This group had the highest score on Arousal/Exploration and the lowest on Anxiety, and were also low on Commitment. They had high efficacy overall, with high Say No scores, and were highest on Precautions and Assertiveness. Finally, they had highest scores on Esteem items, Behavior/Conduct and Attractiveness/ Body Perception.

Cluster Two (Tradicional or traditional, similar to Naive) had 62 members (40% of sample), was a younger group (M = 15.8), mostly female (69.4%), more than half were virgins (58.3%), and most were heterosexual (93%). Of those who were sexually experienced, this group had the highest mean initiation age into sexual activity (14.8 years) and many had a main partner (66.7%). This group was the lowest on Arousal/ Exploration and highest on Commitment and Anxiety. Their scores were high on Saying No, low on Precautions, and low on Assertiveness. Finally, this group had the lowest scores on Attractiveness/Body Perception and the lowest scores on Behavior/Conduct.

Cluster Three (Discuidado, or Careless, similar to Unassured and Driven) had 27 members (17% of sample) who were the younger (M = 15.8 years) and more male (77.8%) cluster. Slightly less than half of this group were still virgins (48.1%) and their initiation age into sexual activity was the youngest (M = 13.2 years). This was also the only group to be exclusively heterosexual. Among those with sexual experience, fewer among this group than the other three groups reported having a main partner (50%). This group scored low on Anxiety, high on Arousal/Exploration, and lowest on Commitment. The hallmark of this group was their lowest average scores on all Efficacy subscales. They also were midrange on Behavior/Conduct and Attractiveness/ Body Perception.

The final group (Conciente, or conscientious, similar to Competent) had 41 members (26% of sample) and was the oldest of the group (M = 16.7). Most participants were female (88.8%) and few were virgins (22.5%). Most were heterosexual (90%), with 5% bisexual and 5% gay/lesbian. The sexually experienced individuals in this group had a high mean initiation age (14.4 years) and most (76.9%) reported having a main sexual partner. The group average score was high on Anxiety and highest on Commitment; they scored midrange on Arousal/Exploration. Self-Efficacy was the distinguishing factor with this group: they scored highest on Say No, high on Precautions, and highest on Assertiveness. They also scored high on Behavior/ Conduct and Attractiveness/Body Perception.

Behaviors varied among clusters. The Tradicional were lowest on their likelihood of being sexually active and low on the likelihood of having more than one partner in the next three months. They had the lowest average number of lifetime partners (1.43) and the lowest average number of three-month partners (0.75). Of those who were sexually experienced, very few had had sex in the past 30 days (18.2%) and few reported being under the influence during sex, either the last time (14.3%) or sometime in the last three months (15.0%). Slightly less than half had used a condom the last time they had sex (40%), and less than half had talked with their partner about condoms (40%).

The Conciente cluster was high on their expectation for sex in the next three months, also low on having more than one partner, and highest on the likelihood of practicing safe sex. This group had a low lifetime number of partners (approximately 2) and three-month partners (approximately 1). Over half (61.5%) had sex within the past 30 days. One-third reported having sex under the influence in the last three months and a small percentage had sex under the influence the last time they had sex (10.3%). Half of those who were sexually experienced reported using a condom during their last sexual encounter and the same percentage (56%) had talked with their partner about condoms.

Aventurero had the highest expectation of being sexually active, low expectation of having more than one partner, and a midrange expectation for practicing safe sex. This group had a high number of lifetime partners (2.7) and midrange number of three-month partners (1.3), with most reporting a main sex partner (87.5%). About two-thirds of this group had had sex in the last 30 days. This group had few who reported being the influence of alcohol or drugs the last sexual encounter (11.8%) or in the past three months (17.6%). More than half reported using a condom the last time they had sex (60%) and more than half (62.5%) had talked with a partner about condoms.

The Discuidado had a high expectation for being sexually active, the highest expectation for more than one partner, and the lowest average score for practicing safe sex. Although only half of this group reported being sexually experienced (vaginal intercourse), those who were not virgins had a high number of lifetime partners (M = 4.3) and the highest number of partners in the last three months (M = 2.2 partners). About half had had sex in the last 30 days (41.7%), and few reported being under the influence at the last time (7.1%), although more reported doing so in the last three months (30.8%). Less than half reported using a condom during the last sexual encounter (40%) and very few reported talking with a partner about condoms (16.7%).

These clusters were also found to be significantly different from each other in a number of ways. ANOVAs and Chi-square analyses were completed to assess for differences between the clusters on the behavior items. Both the Aventurero (72%) and the Conciente (77.5%) had the highest percentages of individuals to have initiated sexual activity (intercourse) and the Tradicional was the lowest with 41.7% ([chi square] [3, n = 152] = 15.23, p = .00). Tradicional were found to be significantly lower and Aventurero significantly higher than the other clusters (F[3,194] = 8.29, p = .00) for likelihood of sexual activity in the next three months. The Discuidado reported significantly more intention to have more than one sexual partner than any other group (F[3,129] = 12.28, p = .00). Differences were also found in terms of the likelihood for practicing safe sex (F[3,103] = 2.51, p <. 10), with the Discuidado less likely than the Conciente. Of those who reported being sexually experienced, Discuidado were significantly younger than the Tradicional for the age of initiation into sexual activity (F[3,81] = 2.25, p < .10) and Discuidado were much higher than all others in their lifetime number of sexual partners (F[3,70] = 3.97, p = .00). As noted previously, there were strong differences in the gender make-up of each cluster ([chi square] [3, n= 155] = 41.95, p = .00) and a significant difference in virginity status ([chi square] [3, n = 155] = 15.23, p = .00), like the original clusters. Having talked with a main partner was significant between the groups, with fewer Discuidado reporting doing so ([chi square] [3] = 7.19, p < .10).

Discussion

The scales seemed easily adaptable for this population. The self-efficacy scales underwent little modification and had strong reliabilities. As for the remaining modifications, many seemed intuitive, as the distinction between conduct and behavior, attractiveness and body image, as well as arousal and exploration, appeared slight. Modifications to these scales reflected a lack of face validity with the original scales more than cultural differences.

In general, average scores on the subscales for each gender supported the cultural values for gender among Latino youth. Adolescent girls were more likely to endorse items relating to needing a commitment to have sex, expressing anxiety about being in sexual situations, and feeling confident in their ability to refuse sex. Likewise, males scored higher on items related to having a high sex drive and expressing interest in sexual novelty. Unexpected was the significantly higher average score for females on the Assertiveness subscale. The items included statements noting the ability to initiate sexual activities and insisting that a partner respect the respondent's sexual needs. The finding is contrary to the cultural value of verguenza and not knowing about sex or being self-confident in sexual relationships. Unfortunately, data was not collected on the age of the sexual partner, perhaps indicating these Latinas were dating same-age or younger males and felt more comfortable speaking up (Driscoll, Biggs, Brindis, & Yankah, 2001; Galanti, 2003). Qualitative interviewing could shed some light on how Latinas express assertiveness in sexual relationships.

For the third research question, regarding the duplication of the original clusters, the data strongly support a similar typology to that which was already found. Initial reliabilities of the measures and their correlations with gender, age, virginity status, and behaviors, all indicated that the measures would operate in a manner similar to that of Buzwell and Rosenthal (1996). Again, the difference between the original sample and the current group of Latino adolescents seems to be one of a degree of difference, rather than a qualitative divergence. The original scales clustered in a manner very similar to that of the original and replication studies.

However, the most striking finding is the four-cluster solution, resulting in no clusters that distinguished between the Driven and Unassured clusters. It would seem that the rigidly prescribed sex roles typical of traditional Latino culture would foster the style of young males that would encourage hyper-sexuality, which was indeed found in the cluster called Discuidado. However, this group also shared characteristics with the Unassured, in their low scores on sexual self-efficacy. It would seem, then, that low sexual self-efficacy doesn't inhibit risky sexual activity among those who reported being sexually experienced. Granted, no standard significance tests exist for cluster analytic procedures and interpretation of the results rely on heuristics.

Also of interest is the presence of Conciente as primarily (88.8%) female. What is exceptional about this group is that they seem to have defied the Latino stereotype by having high efficacy overall, especially efficacy around condoms. Of note is that this group not only reported that they had high efficacy but reported high condom usage and frequent conversations with sexual partners about condoms. This group is unlike the less acculturated Latinas, who might subscribe more to traditional beliefs about women, and yet unlike more acculturated Latinas, who report infrequent condom use and have a higher rate of teen pregnancy.

For the fourth research question, regarding the differences between the clusters on sexual behaviors, the analyses partially supported the original cluster structure. Groups were found to be different in terms of their age, gender composition, virginity status, and importantly, their sexual activity under the influence of drugs or alcohol.

Limitations

Several important limitations of this study should be noted. First, the inclusion of the measures in a program evaluation packet most likely contributed to survey fatigue. Many students took almost an hour and a half to complete all the measures. Some never finished or never got to the section of the survey containing the measures, accounting for almost all of the missing data. The potential impact is hard to estimate for those who completed the measures but did so while fatigued.

The other limitation is the lack of generalizability to the Latino population of adolescents. While the inclusion of significantly different schools assisted in creating heterogeneity of the sample, nonetheless, all participants were in school and thus cannot be assumed to reflect the identities of those who may have dropped out. Given the high rates of school drop-out among this population, more than double that of other ethnic groups (U.S. Department of Education, 2005), any school-based survey of Latino adolescents will be missing information on much of the population. Of particular interest are those not in school due to changes in their family structure, namely those who have become parents themselves.

Third, the sample size for this study is smaller than the original studies. While the current sample size is adequate for cluster analysis, there may be additional patterns that are not captured in this data set because it was smaller (N = 155) than both the original (N = 470) and the replication (N = 522) samples. Differences in the results that were found could be due to the smaller sample size.

Implications

The presence of sexual styles has important implications for group-based sexuality education programs. As indicated by the strong differences in scores of efficacy, esteem, and beliefs, adolescents can have very different prevention needs, regardless of their virginity status, gender, or sexual orientation. The Conciente group already seems strong on efficacy but had a surprising number who had sex under the influence of alcohol or drugs, given their trend of otherwise responsible sexual behavior. Perhaps their high self-perceptions of efficacy lead them to engage in sex under the influence because they thought they could handle it. This group could potentially benefit from improved understanding related to the increased risk of having sex while under the influence of alcohol or other drugs.

Of particular concern is the Discuidado, with their high number of sexual partners, low efficacy for precautions, and low intention to practice safe sex. With the high self esteem of this group, perhaps this is indicative of an inflated ego, or identification with machismo, and is related to the high number of sexual partners. An intervention for these youth should be tailored to cultural expectations around masculinity and increasing efficacy particularly with regard to condom use.

Finally, it is important to note that all clusters had members younger and older, male and female, virgin and sexually experienced, heterosexual and sexual minority. The preponderance of adolescent sexual literature often divides adolescents into groups based on age, gender, sexual experience, or sexual orientation. However, the results of Buzwell and Rosenthal (1996) and this study indicate that there are more complex relationships between sexual identity and sexual behavior than is captured by these typical divisions. We hope that this work, and work like it, will help bring a more nuanced understanding to sexual risk behavior specifically and sexuality in general.

References

Aldenderfer, M. S. & Blashfield, R. K. (1984). Cluster analysis. Newbury Park, CA: Sage Publications.

Barkley, B. H. & Mosher, E. S. (1995). Sexuality and Hispanic culture: Counseling children with their parents. Journal of Sex Education and Therapy, 21, 255-267.

Bourdeau, B., Saltz, R. F., Bersamin, M. M., & Grube, J. W. (2007). Understanding the relationship between alcohol and sex: Latino and white college students and problematic sexual experiences while drinking. Journal of American College Health, 56, 299-306.

Breakwell, G. M. & Millward, L. J. (1997). Sexual self-concept and sexual risk-taking. Journal of Adolescence, 20(1), 29-41.

Buzwell, S. & Rosenthal, D. (1995). Exploring the sexual world of the unemployed adolescent. Journal of Community and Applied Social Psychology, 5(3), 161-166.

Buzwell, S. & Rosenthal, D. (1996). Constructing a sexual self: Adolescents' sexual self-perceptions and sexual risk-taking. Journal of Research on Adolescence, 6, 489-513.

Castro-Vazquez, G. (2000). Masculinity and condom use among Mexican teenagers: The escuela nacional preparatoria's No. 1 case. Gender and Education, 12(4), 479-492.

Centers for Disease Control and Prevention (2000). U.S. HIV and AIDS cases reported through June 2000. HIV/AIDS Surveillance Report, 12(1), 1-41.

Cromwell, R. E. & Ruiz, R. A. (1979). The myth of macho dominance in decision making within Mexican and Chicano families. Hispanic Journal of Behavioral Sciences, 1, 355-373.

Davila, Y. R. (2005). The social construction and conceptualization of sexual health among Mexican American women. Research and Theory for Nursing Practice, 19(4), 357-368.

Denner, J. & Dunbar, N. (2004). Negotiating femininity: Power and strategies of Mexican American girls. Sex Roles, 50(5-6), 301-314.

Driscoll, A. K., Biggs, M. A., Brindis, C. D., & Yankah, E. (2001). Adolescent Latino reproductive health: A review of the literature. Hispanic Journal of Behavioral Sciences, 23, 255-326.

DuRant, R. H., Pendergrast, R., & Seymore, C. (1990). Sexual behavior among Hispanic female adolescents in the United States. Pediatrics, 85, 1051-1058.

East, P. L. (1998). Racial and ethnic differences in girls' sexual, marital, and birth expectations. Journal of Marriage and the Family, 60, 150-162.

East, P. L., Felice, M. E., & Morgan, M. C. (1993). Sisters' and girlfriends' sexual and childbearing behavior: Effects on early adolescent girls' sexual outcomes. Journal of Marriage and the Family, 55, 953-963.

Ellison, C. G., Echevarria-Cruz, S., & Smith, B. (2005). Religion and abortion attitudes among US Hispanics: Findings from the 1990 Latino National Political Survey. Social Science Quarterly, 86, 192-208.

Epstein, J. A., Botvin, G. J., & Diaz, T. (2000). Alcohol use among Hispanic adolescents: Role of linguistic acculturation and gender. Journal of Alcohol and Drug Education, 45(3), 18-32.

Fasula, A. M. & Miller, K. S. (2006). African-American and Hispanic adolescents' intentions to delay first intercourse: parental communication as a buffer for sexually active peers. Journal of Adolescent Health, 38, 193-200.

Flores-Ortiz, Y. G. (1994). The role of cultural and gender values in alcohol use patterns among Chicana/Latina high school and university students: Implications for AIDS prevention. The International Journal of the Additions, 29, 1149-1171.

Fortenberry, J. D., Temkit, M., Tu, W., Graham, C. A., Katz, B. P., & Orr, D. P. (2005). Daily mood, partner support, sexual interest, and sexual activity among adolescent women. Health Psychology, 24(3), 252-257.

Galanti, G. (2003). The Hispanic family and male-female relationships: An overview. Journal of Transcultural Nursing, 14(3), 180-185.

Garcia, L. T. & Carrigan, D. (1998). Individual and gender differences in sexual self-concept. Journal of Psychology & Human Sexuality, 10(2), 59-70.

Garcia, L. T., Cieselka, C., & Fuchs, D. (1999). Social comparison processes in sexual self-perceptions. Journal of Psychology & Human Sexuality, 11(2), 35-42.

Gowen, L. K., Feldman, S. S., Diaz, R., & Yisrael, D. S. (2004). A comparison of the sexual behaviors and attitudes of adolescent girls with older vs. similar-aged boyfriends. Journal of Youth and Adolescence, 33(2), 167-175.

Graber, J. A., Brooks-Gunn, J., & Galen, B. R. (1998). Betwixt and between: Sexuality in the context of adolescent transitions. In R. Jessor (Ed.), New perspectives on adolescent risk behavior (pp. 270-316). Cambridge, England: Cambridge University Press.

Gurman, T. & Borzekowski, D. L. G. (2004). Condom use among Latino college students. Journal of American College Health, 52, 169-178.

Halpern, C. T., King, R. B., & Oslak, S. G. (2005). Body mass index, dieting, romance, and sexual activity in adolescent girls: Relationships over time. Journal of Research on Adolescence, 15(4), 535-559.

Hayes-Bautista, D. E., Schink, W. O., & Chapa, J. (1988). The burden of support: Young Latinos in an aging society. Stanford, CA: Stanford University Press.

Hines, A. M. & Caetano, R. (1998). Alcohol and AIDS-related sexual behavior among Hispanics: Acculturation and gender differences. AIDS Education and Prevention, 10, 533-547.

Kane, E. W. (2000). Racial and ethnic variations in gender-related attitudes. Annual Review of Sociology, 26, 419-439.

Kaplan, C. P., Erickson, P. I., & Juarez-Reyes, M. (2002). Acculturation, gender role orientation, and reproductive risk-taking behavior among Latina adolescent family planning clients. Journal of Adolescent Research, 17(2), 103-121.

Kenney, J. W., Reinholtz, C., & Angelini, P. J. (1997). Ethnic differences in childhood and adolescent sexual abuse and teenage pregnancy. Journal of Adolescent Health, 21, 3-10.

Landale, N. S., Oropesa, R. S., & Llanes, D. (1998). Schooling, work, and idleness among Mexican and non-Latino White adolescents. Social Science Research, 27, 457-480.

Liebowitz, S. W., Castellano, D. C., & Cuellar, I. (1999). Factors that predict sexual behaviors among young Mexican American adolescents: An exploratory study. Hispanic Journal of Behavioral Sciences, 21(4), 470-479.

Locke, T. F., Newcomb, M. D., & Goodyear, R. K. (2005). Childhood experiences and psychosocial influences on risky sexual behavior, condom use, and HIV attitudes/behaviors among Latino males. Psychology of Men and Masculinity, 6(1), 25-38.

Marin, B. V. (2003). HIV prevention in the Hispanic community: Sex culture, and empowerment. Journal of Transcultural Nursing, 14(3), 186-192.

Marin, B. V., Coyle, K. K., Gomez, C. A., Carvajal, S. C., & Kirby, D. B. (2000). Older boyfriends and girlfriends increase risk of sexual initiation in young adolescents. Journal of Adolescent Health, 27(6), 409-418.

Marin, B. V., Gomez, C., & Hearst, N. (1993). Multiple heterosexual partners and condom use among Hispanics and non-Hispanic Whites. Family Planning Perspectives, 25, 170-174.

Martin, J. A., Hamilton, B. E., & Ventura, S. J. (2001). Births: preliminary data for 2000. National Vital Statistics Reports, 49(5), 1-20.

McKee, M. D. & Karasz, A. (2006). You have to give her that confidence: Conversations about sex in Hispanic mother--daughter dyads. Journal of Adolescent Research, 21(2), 158-184.

Minnis, A. M. & Padian, N. S. (2001). Reproductive health differences among Latin American- and US-born young women. Journal of Urban Health, 78, 624-637.

Moran, J. R. & Corley, M. D. (1991). Sources of sexual information and sexual attitudes and behaviors of Anglo and Hispanic adolescent males. Adolescence, 26, 857-864.

Murry, V. M., Brody, G. H., McNair, L. D., Luo, Z., Gibbons, F. X., Gerrard, M., & Wills, T. A. (2005). Parental involvement promotes rural African American youths' self-pride and sexual self-concepts. Journal of Marriage and the Family, 67, 627-642.

Nadeem, E., Romo, L. F., & Sigman, M. (2006). Knowledge about condoms among low-income pregnant Latina adolescents in relation to explicit maternal discussion of contraceptives. Journal of Adolescent Health, 39, 19.e9-1-19.e15.

Padilla, A. M. (1980). Notes on the history of Hispanic psychology. Hispanic Journal of Behavioral Sciences, 2, 109-128.

Palmer, J. T. & Murry, V. M. (1995). Sexual self-concept inventory: A preliminary investigation of reliability and construct validity. Unpublished Master's Thesis; University of Georgia, Athens.

Pavich, E. G. (1986). A Chicana perspective on Mexican culture and sexuality. Journal of Social Work and Human Sexuality, 4(3), 47-65.

Phinney, J. S. & Flores, J. (2002). "Unpackaging" acculturation: Aspects of acculturation as predictors of traditional sex role attitudes. Journal of Cross-Cultural Psychology, 33, 320-331.

Raffaelli, M. & Green, S. (2003). Parent-adolescent communication about sex: Retrospective reports by Latino college students. Journal of Marriage and Family, 65(2), 474-481.

Rickert, V. I., Sanghvi, R., & Wiemann, C. M. (2002). Is lack of sexual assertiveness among adolescent and young adult women a cause for concern? Perspectives on Sexual and Reproductive Health, 34, 178-183.

Sabogal, F. & Catania, J. A. (1996). HIV risk factors, condom use, and HIV antibody testing among heterosexual Hispanics: The National AIDS Behavioral Surveys (NABS). Hispanic Journal of Behavioral Sciences, 18, 367-391.

Smith, A. M. A. & Rosenthal, D. A. (1998). Revisiting sexual styles: Patterns of adolescents' self-perceptions. Psychological Reports, 83, 65-66.

Sneed, C. D., Morisky, D. E., Rotheram-Borus, M. J., Ebin, V., Malotte, C. K., Lyde, M., & Gill, J. K. (2001). "Don't know" and "didn't think of it": Condom use at first intercourse by Latino adolescents. AIDS Care, 13(3), 303-308.

South, S. J. (1993). Racial and ethnic differences in the desire to marry. Journal of Marriage and the Family, 55, 357-370.

Sugar, M. (1995). Facets of adolescent sexuality. In R. C. Mahron & S. C. Feinstein, et al. (Eds.), Adolescent psychiatry: Developmental and clinical studies, Volume 20 (pp. 139-161). Hillsdale, NJ: Analytic Press, Inc.

Taris, T. W. & Semin, G. R. (1999). Sexual efficacy of adolescents, permissiveness, and the transition toward nonvirginity: From intention to behavior. Journal of Applied Social Psychology, 29(5), 903-917.

Taylor, D. J., Chavez, G. F., Adams, E. J., Chabra, A., & Shah, R. S. (1999). Demographic characteristics in adult paternity for first births to adolescents under 15 years of age. Journal of Adolescent Health, 24, 251-258.

Taylor-Seehafer, M. & Rew, L. (2000). Risky sexual behavior among adolescent women. Journal of the Society of Pediatric Nurses, 5, 15-25.

Tolman, D. L., Striepe, M. I., & Harmon, T. (2003). Gender matters: Constructing a model of adolescent sexual health. Journal of Sex Research, 40, 4-12.

Torres, V. (2004). The diversity among us: Puerto Ricans, Cuban Americans, Caribbean Americans, and Central and South Americans. New Directions for Student Services, 105, 5-16.

Umana-Taylor, A. J. & Fine, M. A. (2001). Methodological implications of grouping Latino adolescents into one collective ethnic group. Hispanic Journal of Behavioral Sciences, 23, 347-362.

Upchurch, D. M., Aneshensel, C. S., Mudgal, J., & McNeely, C. S. (2001). Sociocultural contexts of time to first sex among Hispanic adolescents. Journal of Marriage and the Family, 63, 1158-1169.

Upchurch, D. M., Aneshensel, C. S., Sucoff, C. A., & Levy-Storms, L. (1999). Neighborhood and family contexts of adolescent sexual activity. Journal of Marriage and the Family, 61, 920-933.

U.S. Census. (2000a). Statistic abstract of the United States, 1999. (Table 12, Resident population-selected characteristics, 1790 to 1998, and projections, 2000 to 2050, p. 14.) Washington, D.C.: Government Printing Office.

U.S. Census. (2000b). Statistic abstract of the United States, 1999. (Table 24, Projections of resident population, by age, sex and race: 2000 to 2025, p. 25.) Washington, D.C.: Government Printing Office.

U.S. Department of Education, National Center for Education Statistics. (2005). The Condition of Education 2005 (NCES 2005-094). Washington, D.C.: U.S. Government Printing Office.

Ventura, S. J., Mosher, W. D., Curtin, S. C., Abma, A. C. & Henshaw, S. K. (2001). Trends in Pregnancies and Pregnancy Rates for the United States, 1976-97: An Update. National Vital Statistics Reports, 49(4). Hyattsville, MD: National Center for Health Statistics.

Villarruel, A. M. (1998). Cultural influence on the sexual attitudes, beliefs, and norms of young Latina adolescents. Journal of the Society of Pediatric Nurses, 3(2), 69-79.

Villarreal, R., Blozis, S. A., & Widaman, K. F. (2005). Factorial Invariance of a Pan-Hispanic Familism Scale. Hispanic Journal of Behavioral Sciences, 27(4), 409-425.

Villarruel, A. M., Gallegos, E. C., Cherry, C. J. L., & de Duran, M. R. (2003). La uniendo de fronteras: Collaboration to develop HIV prevention strategies for Mexican and Latino youth. Journal of Transcultural Nursing, 14, 193-206.

Villarruel, A. M., Jemmott, J. B. III, Jemmott, L. S., & Ronis, D. L. (2004). Predictors of sexual intercourse and condom use intentions among Spanish-dominant Latino youth: A test of the planned behavior theory. Nursing Research, 53(3), 172-181.

Welles, C. E. (2005). Breaking the silence surrounding female adolescent sexual desire. Women and Therapy, 28(2), 31-45.

Wood, M. L. & Price, P. (1997). Machismo and marianismo: Implications for HIV/AIDS risk reduction and education. American Journal of Health Studies, 13, 44-52.

Zapata, J. T. & Jaramillo, P. T. (1981). Research on the Mexican-American family. Journal of Individual Psychology, 37, 72-85.

Beth Bourdeau

Prevention Research Center, Pacific Institute for Research and Evaluation

Volker K. Thomas

Child Development and Family Studies, Purdue University

Janie K. Long

Center for Lesbian, Gay, Bisexual, & Transgender Life, Duke University

Correspondence should be addressed to Beth Bourdeau, Ph.D., Prevention Research Center, Pacific Institute for Research and Evaluation, 1995 University Ave., Suite 450, Berkeley, CA 94706. E-mail: [e-mail].

Table 1. Original and Current Scale Reliabilities (N = 155)

                       Original   Original   Current   Current
                         N of      Study      N of      Study
Scale                   items     [alpha]     items    [alpha]

Sexual Self-Beliefs
  Anxiety                 11        .75         6        .81
  Arousal                 10        .87                  .87
  Exploration              9        .81        13
  Commitment              10        .84         8        .82
Sexual Self-Efficacy
  Say No                   9        .81         8        .89
  Precautions              6        .74         5        .76
  Assertiveness            5        .75         4        .77
Sexual Self-Esteem
  Body Perception          3        .64                  .67
  Attractiveness           5        .81         5
  Conduct                  5        .81                  .83
  Behavior                 6        .80         7

Table 2. Gender Differences on Subscale Scores

Scale                               Male M   Female M       t

Sexual Self-Beliefs
  Anxiety                            13.3      16.4     -5.14 **
  Arousal/Exploration                32.0      26.6      5.34 **
  Commitment                         21.8      26.8     -7.42 **
Sexual Self-Efficacy
  Say No                             22.2      30.5     -5.34 **
  Precautions                        16.2      17.4     -1.08
  Assertiveness                      11.6      13.9     -2.56 *
Sexual Self-Esteem
  Body Perception /Attractiveness    14.1      13.3      1.86
  Conduct/ Behavior                  20.7      19.8      1.50