Area Specific Self-Esteem, Values, and Adolescent Sexual Behavior

This study examined area-specific self-esteem scores by sexual behavior relative to adolescents’ values concerning participation in sexual intercourse as an unmarried teenager. The sample consisted of 332 students in grades 7–12 from a Southern rural school district. Students were asked if they had ever had sexual intercourse (yes/no) and if they had participated in sexual intercourse in the last month (yes/no). Respondents also indicated on a 4-point scale their response to the statement “It is against my values to have sex as an unmarried teenager.” Data were analyzed using a 2 x 4 (behavior x values) analysis of variance for each of the three area-specific self-esteem scores (peer, school, and home). Results indicated that students who had participated in sexual intercourse had significantly lower scores in school and home self-esteem than those who had not participated. In addition, those who “strongly agreed” with the values statement and indicated they had not had intercourse had the highest school and home self-esteem scores. Those who strongly agreed with the values statement but indicated they had participated in sexual intercourse had the lowest school and home self-esteem scores. This behavior x values interaction was significant for sexual intercourse–ever, and for school self-esteem and sexual intercourse in the last month. No difference was seen in peer self-esteem scores by behavior nor were there behavior x values interactions. Michael Young, PhD, FAAHB, is a university professor with the Program in Health Science at the University of Arkansas, HP 326A, Fayetteville, AR 72701; E-mail: meyoung@ uark.edu. Joseph Donnelly, PhD, is a professor in the Department of HPPERLS at Montclair State University in Upper Montclair, NJ. George Denny, PhD, is an associate professor with the Education Foundation at the University of Arkansas. Michael Young, Joseph Donnelly, and George Denny American Journal of Health Education — September/October 2004, Volume 35, No. 5 283 areas of self-esteem according to the values held by adolescents. Values may be implicated both as a predictor of sexual behavior in adolescents and as a determinant of self-esteem after sexual behavior has been initiated. For example, Miller and coworkers (1987) found that self-esteem was positively related to sexual intercourse for adolescents who believed that intercourse was always right, but negatively related for those who believed it was wrong. In this study we hoped to find evidence that would enrich the understanding of how particular areas of self-esteem and sexual behavior interact, and how personal values related to sex might mediate their relationship. SELF-ESTEEM: DEFINITION AND MEASUREMENT Many studies have conceptualized selfesteem or provided some sort of definition of self-esteem. There has, however, been no consensus among these researchers as to what self-esteem actually is. Few researchers who have examined the relationship between self-esteem and sexual behaviors have offered a concrete operational definition for self-esteem. Wells (1976) analyzed the situation relative to self-esteem as follows: Because self-esteem seems to many an intuitive or common sense idea, there has been no need to spell out its nature or the processes by which it operates. Though this comment was made nearly 30 years ago, it continues to be true today. Rosenberg (1965) provided one of the broadest and most frequently cited definitions of selfesteem within psychology, describing it as a favorable or unfavorable attitude toward the self. Although the term self-esteem usually refers to a global sense of self-worth, narrower concepts such as self-confidence or body-esteem are often used to describe a more specific area of self-esteem. Selfesteem is considered the evaluative component of self-concept, which more broadly includes cognitive and behavioral aspects of the self as well as evaluative or affective ones (Blascovitch & Tomaka, 1991). Research has found that self-esteem significantly affects general health behavior for adolescents of all ages (Rivas Torres & Fernandez Fernandez, 1995). Most research investigating self-esteem and adolescent behavior relative to such activities as sexual activity and drug use have used a generalized, or global, selfesteem scale to determine participants’ self-esteem (Cole & Slocumb, 1995; Hally & Pollack, 1993; Hollar & Snizek, 1996; Orr, Wilbrandt, Brack, Rauch, & Ingersoll, 1989; Pearlman, 1974; Robinson & Frank, 1994; Stratton & Spitzer, 1967; Walsh, 1991). Several researchers have used area-specific self-esteem scales (Emery, McDermott, Holcomb & Marty, 1993; Young, 1989; Young, Denny & Spear, 2000) that measure self-esteem in specific arenas of participants’ lives. In this study, self-esteem was defined as one’s self-evaluation within the context of three particular areas of experience—peers, family, and school. FINDINGS RELATIVE TO SELF-ESTEEM AND ADOLESCENT SEXUAL BEHAVIOR Studies have shown sometimes contradictory results when assessing the relationship between self-esteem and sexual behavior. Most of the research has utilized a global measure of self-esteem, although a few studies have used area-specific measures of selfesteem. This review of findings focuses on studies that have involved high school or middle/junior high school students. Miller and colleagues (1987) used the Rosenberg self-esteem scale and found that in a sample of 2,423 high school students across the West, self-esteem was negatively correlated with sexual attitudes and behavior. A recent study (Spencer et al., 2002) also used the Rosenberg scale to examine the role of self-esteem in predicting coitus initiation in a population of early adolescents. Data collected longitudinally while participants were in seventh and ninth grades found that boys with higher self-esteem rating in seventh grade were more likely to initiate intercourse by ninth grade. An opposite relationship was seen in girls, with those starting out with higher self-esteem more likely to remain virgins than girls with lower self-esteem. These results highlight the need to examine gender as a factor in the relationship between self-esteem and sexual behavior. Several studies contradict the results above, finding no difference in self-esteem between sexually active or inactive adolescents, or between virgins and nonvirgins. Robinson and Frank (1994), using the Coopersmith self-esteem scale, found no significant differences in self-esteem between sexually active males and sexually inactive males, or between sexually active females or sexually inactive females. Nor did they find significant self-esteem differences between virgins and nonvirgins of either sex. In a study of junior high students from blue collar homes Orr and his coworkers (1989) found no overall statistical difference in self-esteem (apparently a global measure) of sexually experienced and virgin adolescents. However, they did find that the selfesteem of sexually active girls was significantly lower than that of virgin girls (perhaps reflecting the continued existence of a double standard). Benson and Torpy (1995) examined the relationship of selfesteem and other variables in self-reported virginity loss among junior high students. They found that when considered in the context of a logistic regression analysis, selfesteem was not associated with age at first sexual intercourse. Some studies have used area-specific measures of self-esteem. Young (1989) found that among early adolescents virgins displayed higher school self-esteem than nonvirgins. Home self-esteem and peer selfesteem were not related to participation in sexual intercourse. Young and colleagues (2000) used the Kelley short form of the Hare Self-Esteem scale (an area-specific scale) to study the relationship of self-esteem to sexual behavior among 1,659 junior and senior high school students in rural Arkansas. Results indicated that higher peer self-esteem was significantly associated with increased likelihood of past participation in sexual intercourse (both ever and recent participation) and greater intent to participate in the future. Higher home and school Michael Young, Joseph Donnelly, and George Denny 284 American Journal of Health Education — September/October 2004, Volume 35, No. 5 self-esteem were significantly associated with decreased likelihood of past participation (both ever and recent) and less intent to participate in the future. FINDINGS RELATIVE TO VALUES AND ADOLESCENT SEXUAL BEHAVIOR Although self-esteem might be considered an individual adolescent characteristic that acts as a predictor variable for sexual behavior, family characteristics such as parental values and attitudes toward sexual behavior also need to be considered (Miller & Fox, 1987; Werner-Wilson, 1998). Miller (2002) recently found that parent–child closeness and parental supervision in combination with parents’ values against teen sexual intercourse decreased the risk of adolescent pregnancy. A sample of 697 abstinent adolescents in Missouri revealed three factors in common contributing toward their abstinence. These were labeled by researchers “fear-based postponement,” “emotionality and confusion,” and “conservative values” (Blinn-Pike, 1999), which can be assumed to be at least in part reflective of conservative family values. Miller, McCoy, Olson, and Wallace (1986) found that sexual permissiveness and intercourse experience were highest among adolescents who viewed their parents as not being strict at all or not having any rules. Parental strictness and discipline can be viewed as an enactment or expression of parental values. Values and attitudes about sex have been shown to be related to behavior (Glass, 1972; Miller et al., 1987; Miller & Olsen, 1988; Thomson, 1982). Identification with a church or synagogue and religious identification in general have been seen to be protective factors against early sexual behavior (Brown, 1985; Glass, 1972). This again at least points to the idea of personal values. Indeed, sexual behavior that contradicts personal values is associated with emotional distress and lower self-esteem. These values likely correspond to social norms (Miller et al., 1987), which necessarily implicate gender and ethnicity as variables. Both gender and ethnicity affect the risk of first sexual intercourse through differences in sexual norms, attitudes, and values (Upchurch, Levy-Storms, Sucoff, & Aneshensel, 1998). In a sample of 877 ethnically diverse youths in Los Angeles County, the median age at first sex was found to differ by gender and ethnicity group. Socioeconomic conditions were thought to account for ethnic differences among girls in the age at first sex, whereas cultural influences may have contributed to the differences between Hispanic boys and girls. The social and cultural meanings of sexual behavior, as well as the norms of when first sex should occur, seem to vary according to adolescents’ gender and ethnicity (Upchurch et al., 1998). Liebowitz, Castellano, and Cuellar (1999) used a multiple regression model to predict the absence of sexual activity among young Mexican American adolescents. They found that among the predictors of absence of sexual activity the most important was the child’s perception of the congruency of parent–child sexual values. Whether this finding is valid across ethnic groups remains to be studied. Values of adolescents regarding sexual behavior have been seen to differ greatly by gender. Girls have indicated greater commitment to abstinence and less permissive sexual attitudes than boys in general (De Gaston & Weed, 1996). Miller (1986) and Gilligan (1982) suggested that the culture into which most girls are socialized perceives, values, and interprets sex quite differently than it is seen and valued in the dominant masculine culture of our society. Sex roles and sexual identity seem to be major factors in the development and maintenance of sexual attitudes (DeGaston & Weed, 1996). These values are implicated both as predictors of sexual behavior in adolescents of both genders and varying ethnicities and as a determinant of self-esteem after sexual behavior has been initiated. In this study the interaction of values and behaviors on specific areas of self-esteem was explored in a sample of adolescents attending a public school in the rural south. We hypothesized that students who reported sexual behavior consistent with their stated values regarding sex before marriage (i.e., strongly agreed that it was against their personal values to have sex before marriage and had not yet experienced sexual intercourse) would score higher on measures of self-esteem than students who reported behavior that was inconsistent with their stated values (i.e., strongly agreed that it was against personal values to have sex before marriage but had experienced sexual intercourse).