Today, male and female adolescents in Taiwan are more sexually active and more prone to engage in high-risk sexual behaviors than at any previous time (Ball, 2009). Preventing risky sexual behavior in adolescents is necessary because their bodies are at a critical stage of physical and psychological maturity. Sexual communication is recognized as an important determinant factor of sexual behavior in adolescence (Atienzo, Walker, Campero, Lamadrid-Fiqueroa, & Gutirrez, 2009; Beckett et al., 2010). Ogle, Glasier, and Riley (2008) and Widman, Welsh, McNulty, and Little (2006) have indicated that communication about sex and sexuality play a major role in the attainment of sexual health, especially for young people. For parents, sexual parent-child communication requires suitable timing. Relevant topics include body image, birth control efficacy, declining to engage in sex, effective condom use, and partner condom refusal (Beckett et al., 2010). However, little is known about the sexual communication between adolescents and their parents or about how such correlates with demographics and sexual cognition.
Understanding the relationships between demographic data and sexual communication can be helpful in the development of suitable sexual health programs. Several studies have investigated the factors related to sexual communication. The study of Widman et al. (2006) indicated that adolescents who were engaged in satisfying boy or girlfriend relationships tended to have more sexual communication. Van Der Straten, Catania, and Pollack (1998) also indicated that sexual communication correlated positively with age, gender, education level, marital status, sexual assertiveness, condom comfort, and sexual and condom regulation skill. However, it was negatively correlated with perceived susceptibility to STDs. Parent-child sexual risk communication was found related to more responsible sexual attitudes and behavior (greater condom use self-efficacy and more sexual risk communication with their male partners) in sexually active adolescents (Hutchinson & Cooney, 1998). However, a longitudinal study of 141 parents and their children revealed that about 40% of adolescents started sexual communication with their parents only after engaging in sexual intercourse (Beckett et al., 2010).
A study by DiClemente, Wingood, Rose, Sales, and Crosby (2010) revealed that adolescents reported higher sexual communication frequency and consistent condom use after an intervention program designed to enhance sexual self-concepts and self-worth, safe sex practices, and HIV/STD prevention skills. Houlihan et al. (2008) indicated that sexual self-concepts were associated with risky sexual behavior and the consequences of unsafe sex among adolescents. On the basis of the unified theory of behavior, Guilamo-Ramos, Jaccard, Dittus, and Collins (2008) proposed that parent-adolescent sexual communication provided an effective way to discuss with children social norms, self-concepts, and self-efficacy related to sex. Although adolescents in Taiwan are increasingly willing to express their views on sex, the frequency in which adolescents talk about sexual activity and issues with their parents is less discussed. In this study, we investigated adolescent sexual communication behavior in Taiwan. Sexual communication was defined as the frequency with which adolescents discussed their sexual activity and talked about sexual issues with their parents.
Sexual risk cognition is an antecedent reciprocal factor as well as consequence of risky sexual behavior in adolescents (Houlihan et al., 2008). It has been confirmed that individual cognition of HIV risk behavior is correlated with sexual behavior (Shah, Thornton, & Burgess, 1997). A qualitative study of 22 adolescents conducted by Ott and Pfeiffer (2009) revealed that adolescent cognition regarding sexual abstinence varies, with cognition associated with the individual's age and progress along the transition to adulthood. However, they suggested that the transition period of early adolescence is an appropriate age to provide effective intervention programs to prevent high-risk sexual behavior.
Impett, Schooler, and Tolman (2006) defined sexual self-efficacy as individuals who make their own sexual decisions to have intercourse in a relationship. Childs, Moneyham, and Felton (2008) stated that sexual self-efficacy is influenced by antecedent factors (knowledge of HIV and spirituality) and psychosocial factors (perceived parental attitudes about premarital sex and attitudes toward sexual abstinence and condom use). They further found sexual self-efficacy to not be a significant predictor of sexual activity among adolescent women. A longitudinal cohort study by Woods, Hensel, and Fortenberry (2009) showed that the current trend away from the use of contraceptives may effectively raise sexual self-efficacy among adolescent women. In addition, sexual self-efficacy has been shown to be associated with first sexual intercourse. Adolescents in South Africa, for example, were identified as having a significantly higher likelihood of low sexual self-efficacy in their first sexual intercourse experience (Mathews et al., 2009).
In this study, we propose that sexual cognitive variables can predict sexual communication. The conceptual framework is shown in Figure 1.
The following research questions were explored:
- Do sexual cognitive variables relate to sexual communication?
- Do sexual cognitive variables predict sexual communication among adolescents?
Participants and Procedures
This study was a cross-sectional survey of Taiwanese adolescents enrolled at a medical college in 2009. A total of 4,718 students, including 1,252 men (26.5%) and 3,466 women (73.5%), were currently studying at this medical college. Most students majored in one of three fields of study, including nursing, rehabilitation science, and medical technology. Subjects were selected for participation by a conventional sampling technique that used a list of all freshmen through junior (first to third year) classes. Twenty classes were selected at random and targeted by a total of 900 questionnaires. A total of 748 questionnaires were returned and accepted as valid (valid response rate = 83.1%).
Three structured, self-administered scales designed using reviews of relevant studies and consultations with experts were used to collect data. A pilot test was used to confirm the completeness and importance of each item in the instruments and eliminate logically duplicative items. The structured questionnaire included the following three sections:
1. Demographic data: This section included gender, age, whether subjects had heterosexual friends (boyfriends or girlfriends), the extent to which the subject felt satisfied with his or her heterosexual friends, the duration of relationships with heterosexual friends, and whether subjects had experienced sexual intercourse. Satisfaction with heterosexual friends was scored by only one item using a Likert-type 5-point scale graded from 1 (very unsatisfactory) to 5 (very satisfactory). Age and duration of relationships with heterosexual friends were reported in 1-year increments.
2. Sexual cognitive variables: Three structured instruments (the Sexual Self-Concept Inventory [SSCI], the Sexual Risk Cognitions Questionnaire [SRCQ], and the Sexual Self-efficacy Scale [SSE]) were used in this study. The SSCI used in this study was developed primarily on the basis of the SSCI (O'Sullivan, Meyer-Bahlburg, & McKeague, 2006) and revised on the basis of a review of relevant studies. This scale measured the current sexual self-concept of subjects. The original scale we developed included 16 items. Upon completion of a pilot test, 10 of the original SSCI items were retained by item analysis. The factor analysis divided the SSCI into three dimensions, namely, sexual arousability, which related to sexual responsiveness (four items), sexual agency, which related to sexual curiosity (three items), and negative sexual affect, which related to sexual anxiety and sexual monitoring (three items). A 5-point Likert-type scale in which 1 indicated strongly disagree and 5 strongly agree was designed to collect data. Higher scores indicated more positive sexual self-concepts. The content validity index (CVI) was .93. Cronbach's α ranged from .83 to .92 for subscales and was .93 for the overall SSCI. Confirmatory factor analysis (CFA) showed the model χ2 to be 53.50, df = 32, p = .010, the root mean square error of approximation (RMSEA) to be 0.050, the standardized root mean square residual (SRMR) to be 0.027, and the comparative fit index (CFI) as .98.
The SRCQ used in this study was developed mainly on the basis of the SRCQ (Shah et al., 1997) and a review of relevant studies. This scale was designed to measure the current sexual risk cognition of adolescents. The original scale included six items. Upon completion of the pilot test, four items were retained by item analysis as the SRCQ. The factor analysis divided the SRCQ into two dimensions: attitudes toward condom use (two items) and unsafe sexual behavior (two items). A 5-point Likert-type scale in which 1 indicated strongly disagree and 5 strongly agree was designed to collect data. Higher scores indicated high sexual risk cognitions. The CVI was .96, and Cronbach's α was .64 for the SRCQ and .72 and .80 for its two subscales, respectively. The CFA revealed the model χ2 as 0.39, df = 1, p = .53, the RMSEA < 0.001, the CFI as 1.00, and the SRMR as 0.006.
The SSE used in this study was developed primarily on the basis of the SSE (Libman, Rothenberg, Fichten, & Amsel, 1985). Incorporated revisions reflected suggestions gleaned from a review of relevant studies (Fisher, Fisher, Williams, & Malloy, 1994). This scale measured the current sexual self-efficacy of adolescents. The original scale included 12 items. Upon completion of the pilot test, 8 items were retained by item analysis in the SSE. Factor analysis divided the SSE into two dimensions, namely, personal efficacy (4 items) and behavioral skills (4 items). A 5-point Likert-type scale, in which 1 indicated strongly disagree and 5 strongly agree, was designed to collect data. Higher scores indicated higher self-efficacy. The CVI was .92, and Cronbach's α was .85 for the SSE and .82 and .87 for its two subscales, respectively. The CFA revealed the model χ2 to be 92.82, df = 19, p < .001, the RMSEA as 0.072, the SRMR as 0.050, and the CFI as .98.
3. Sexual Communication Scale (SCS): The scale used in this study was developed mainly in reference to the SCS (Somer & Ganivez, 2003) and revised on the basis of a review of relevant studies. This scale measured the frequency with which adolescents discussed sexual activity and sexual issues with their parents. The original scale included 12 items. Upon completion of a pilot test, 6 items on the original SCS were retained after item analysis. Factor analysis divided the SCS into two dimensions, namely, sexual activity (3 items) and sexual issues (3 items). A 5-point Likert-type scale that ranged from 1 (never) to 5 (very often) was designed to collect the data, with higher scores indicating adolescents discussed sexual activity or sexual issues with their parents more frequently. The CVI for the SCS was .92, and Cronbach's α was .91 for the SCS and .94 and .82 for its two subscales, respectively. The CFA showed the model χ2 to be 45.60, df = 8, p < .001, the RMSEA as 0.133, the SRMR as 0.046, and the CFI as .98. All evaluation results, with the exception of the higher RMSEA value, were satisfactory for this scale.
The study protocol and the informed consent form were reviewed and approved by the institutional review board with regard to ethical considerations and procedures. All subjects included in this study were volunteers and received assurances of confidentiality. Informed consent was obtained from all study participants.
SPSS Version 13.0 (SPSS Inc., Chicago, IL) and LISREL 8.8 were used to manage all data entry and analysis work. Confirmatory factor analyses were used to set up the number of factors as suggested for each instrument. First, descriptive statistics were used to determine the sexual communication situation in relation to demographic data and to identify the distribution of adolescent sexual cognitive and sexual communication variables. Second, the Pearson product-moment correlation was adopted to explore relationships between sexual cognitive variables and the sexual communication variable. Finally, multiple regression analysis was applied to analyze the influence of adolescent demographic data and sexual cognitive variables on the sexual communication variable.
Descriptive Statistics of Demographic Data on Sexual Communication
Of the 748 participants, 355 were women and 393 were men (47.5% and 52.5%, respectively). The mean and the standard deviations of age within the sample were 16.70 and 0.69 years, respectively. More than half of the participants had heterosexual friends (53.6%), and a similar percentage had prior sexual experience (52.1%). Subjects expressed feeling moderately satisfied with their current heterosexual relationship (mean = 3.38, SD = 1.40). The duration of relationships with heterosexual friends lasted, on average, 1.35 years (Table 1).
Adolescent Sexual Cognitive Variables and Sexual Communication
As shown in Table 2, the mean for overall sexual self-concept was 3.22 (SD = 0.78). Adolescents had similar scores, on average, in the three subdimensions of sexual arousability, sexual agency, and negative sexual affect. With regard to sexual risk cognitions, the mean overall sexual risk cognition was 3.21 (SD = 1.07). Adolescents had more risk awareness in terms of their attitudes toward condom use (mean = 3.36) than in appreciating the consequences of engaging in sexual relations (mean = 3.06). In respect to sexual self-efficacy, the mean overall sexual self-efficacy was 3.43 (SD = 0.78). Adolescents had relatively better behavioral skills (mean = 3.70) in comparison with personal efficacy (mean = 3.17). As for sexual communication, the frequency of discussing sexual activity and sexual issues with their parents was moderate (mean = 2.52, SD = 1.24). Moreover, as shown in Table 3, all sexual cognitive variables and their subdimensions exhibited positive correlation coefficients with sexual communication and reached significance levels of at least .001.
Predictors of Sexual Communication
Multiple regression analysis using the "enter method" in SPSS software was conducted to identify predictors of sexual communication. Demographic data, reference values, and sexual cognitive variables with their subdimensions were entered into the regression model. Some demographic data consisting of categorical variables were coded as dummy variables when conducting regression analysis. Results in Table 4 indicate that demographic data and the relevance of each variable (having heterosexual friends, satisfaction with heterosexual friends, and duration of relationships with heterosexual friends) were the main predictors of sexual communication reaching significant levels of at least .05. As for sexual cognitive variables, all subdimensions, with the exception of sexual self-efficacy behavioral skills, were predictive of sexual communication, all reaching significant levels of at least .001. These variables accounted for about 62.0% of variance in sexual communication. Variance inflation factors were all lower than the common standard (10), which indicated this multiple regression analysis was free of noteworthy multicollinearity problems (Hair, Black, Babin, Anderson, & Tatham, 2006).
This study examined student sexual communication by exploring sexual cognitive variables in Taiwan adolescents. Findings showed that the frequency of sexual communication between adolescents and their parents was, on average, "moderate." This finding was consistent with that of Liao and Hong (2006), who also found infrequent parent-adolescent sexual communication in Taiwan. Parent-adolescent sexual communication is an effective way to discuss social norms, personal values, social values, and self-sexual concepts (Guilamo-Ramos et al., 2008). The issue of heterosexual sexual communications has been emphasized in previous sexual health programs. As sexual communication between adolescents and their parents is relatively low in Taiwan, we suggest that future sexual health programs may help increase communication frequencies on this issue. More encouragement and promotion of sexual communication among adolescents and their parents are needed.
Study results showed sexual cognitive variables to be predictors of sexual communication. Sexual self-concept was correlated most strongly with sexual communication. Results also indicate that sexual self-concept has a positive influence on sexual communication. The results of this study are consistent with Tasic's (1997) conclusion that teens with more positive self-concepts have more communication with parents. In addition, teens' self-concepts are significantly related to parent-adolescent communication. The study of Salazar et al. (2005) revealed that female African American teens' self-concepts correlate strongly with their parents' sexual communication.
The findings in our study indicate that sexual risk cognitions are correlated with communication and that these cognitions have a positive influence on sexual communication. Previous studies have shown that children's sexual cognitions are significantly associated with sexual communication with their mothers on the issues of abstinence and reproduction (Miller et al., 2009). Similarly, sexual cognition is positively related to the sexual communication of female teens. Moreover, most adolescents perceive more sexual communication coming from their mothers than from their fathers (Barone & Wiederman, 1997).
Our study also shows that sexual self-efficacy correlates with sexual communication. Cha, Kim, and Patrick (2008) established that sexual self-efficacy with regard to consistent and correct condom use is significantly correlated with mother-adolescent sexual communication. Moreover, higher sexual self-efficacy has been shown to be related to higher levels of sexual communication, and both have been proven to be effective health-protective strategies to increase condom self-efficacy and sexual comfort for women (Van Der Straten et al., 1998).
As for the relationship between demographic data and variables and sexual communication, results indicate that demographic data and variables (i.e., having heterosexual friends, satisfaction with heterosexual friends, and duration of relationships with heterosexual friends) impact upon sexual communication. These findings are consistent with the study of Widman et al. (2006), which revealed that adolescents who reported satisfaction in their relationships with heterosexual friends had higher levels of sexual communication. Furthermore, adolescents having heterosexual friends and long relationship durations with heterosexual friends tended to have more frequent sexual communication with their parents. We suggest that parents in Taiwan must maintain an open attitude regarding their children's heterosexual relationships and discuss sexual issue and activity with their child frequently.
Moreover, in this study, gender, age, and sexual experience were not significantly predictive of sexual communication. These findings are inconsistent with the study of Van Der Straten et al. (1998), which showed gender and age to be associated with sexual communication. Their study explained that respondents who were younger and/or male were more likely to have higher levels sexual communication. Furthermore, Swain, Ackerman, and Ackerman (2006) showed being in the teenage age bracket correlating positively to sexual communication. A reasonable explanation for this difference may be that our participants were all 5-year junior college students in Taiwan and thus had a more limited age range than comparable studies done overseas. Future studies may investigate adolescents of different age groups.
Safe sexual behavior and heterosexual sexual communication issues have been emphasized in previous studies and health education programs (Bruhin, 2003; Kirby et al., 2004; Noar, Zimmerman, Palmgreen, Lustria, & Horosewski, 2006). However, sexual communications between adolescents and their parents have been largely overlooked. We suggest that future sexual health programs may introduce and discuss sexual communication issues between parents and adolescents to help increase the general level of sexual communication between parents and their adolescent children. More sexual self-concept, sexual risk cognitions, and sexual self-efficacy discussion with adolescents should be encouraged in sexual health education programs as such can help adolescents better understand and appreciate self-sexual concepts. Finally, we suggest introducing intervention studies or sexual health education programs in the future as well as the use of various methods, for example, role playing, practicing skills, and increasing assertiveness (Mitchell, Kaufman, & the Pathways of Choice and Healthy Ways Project Team, 2002; Shercliffe et al., 2007), to increase safe sexual activities and concepts among young students.
There were several limitations to this study. Although the sample was collected from one medical college, inferential validity was not sufficient to represent other adolescents. We hope that more sample sources and probability sampling methods may be used to enhance the representative nature of research on this subject. Also, as this was a cross-sectional study, causality cannot be inferred directly from results. Data collection occurred only once, and thus study findings can only infer that one variable predicts the dependent variable, as derived from statistical analysis. Moreover, demographic data on subject parents were not collected. Future studies may consider data on parents as well. Another possible limitation of this study relates to the sexual issues faced by adolescents. Such may introduce bias into our findings because some respondents may report what they think is ideal sexual knowledge, behavior, and cognition instead of reporting their actual sexual patterns. However, the bias effect was minimized through, ensuring respondents of the absence of identifiers that would link them with their survey answers. In future studies, interviews could be used to collect more detailed information about student sexual communication.
This study used demographic data and variables as well as sexual cognitive variables to assess sexual communication. Such were used as independent constructs. Because of the significant associations among the three sexual cognitive variables and sexual communication subdimensions, sexual cognitive variables appear to be important with regard to identifying adolescent sexual communication. Results also indicate that demographic data and variables (having heterosexual friends, satisfaction with heterosexual friends, and duration of relationships with heterosexual friends) were the important predictors of sexual communication. This study suggests that increasing adolescent sexual communication could be a way to decrease sexually transmitted infections among youths.
Atienzo, E. E., Walker, D. M., Campero, L., Lamadrid-Fiqueroa, H., & Gutirrez, J. P. (2009). Parent-adolescent communication about sex in Morelos, Mexico: Does it impact sexual behaviour. European Journal of Contraception and Reproductive Health Care
Ball, H. (2009). Prevalence of adolescent sexual behaviours may be rising in Asian cities. International Perspectives on Sexual and Reproductive Health
Barone, N. M., & Wiederman, M. W. (1997). Young women's sexuality as a function of perceptions of maternal sexual communication
during childhood. Journal of Sex Education and Therapy
Beckett, M. K., Elliott, M. N., Martino, S., Kanouse, D. E., Corona, R., Klein, D. J., et al. (2010). Timing of parent and child communication about sexuality relative to children's sexual behaviors. Pediatrics
Bruhin, E. (2003). Power, communication and condom use: Patterns of HIV-relevant sexual risk management in heterosexual relationships. AIDS Care
Cha, E. S., Kim, K. H., & Patrick, T. E. (2008). Predictors of intention to practice safer sex among Korean college students. Archives of Sexual Behavior
Childs, G., Moneyham, L., & Felton, G. (2008). Correlates of sexual abstinence and sexual activity of low-income African American adolescent. Journal of the Association of Nurses in AIDS Care
DiClemente, R. J., Wingood, G. M., Rose, E., Sales, J. M., & Crosby, R. A. (2010). Evaluation of an HIV/STD sexual risk-reduction intervention for pregnant African American adolescents attending a prenatal clinic in an urban public hospital: Preliminary evidence of efficacy. Journal of Pediatric & Adolescent Gynecology
Fisher, J. D., Fisher, W. A., Williams, S. S., & Malloy, T. E. (1994). Empirical tests of an information-motivation-behavioral skills model of AIDS-preventive behavior with gay men and heterosexual university students. Health Psychology
Guilamo-Ramos, V., Jaccard, J., Dittus, P., & Collins, S. (2008). Parent-adolescent communication about sexual intercourse: An analysis of maternal reluctance to communicate. Health Psychology
Hair, J. F., Black, B., Babin, B., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis
(6th ed.). Upper Saddle River, NJ: Prentice Hall.
Houlihan, A. E., Gibbons, F. X., Gerrard, M., Yeh, H. C., Reimer, R. A., & Murry, V. M. (2008). Sex and the self: The impact of early sexual onset on the self-concept and subsequent risk behaviour of African American adolescents. The Journal of Early Adolescence
Hutchinson, M. K., & Cooney, T. M. (1998). Parent-teen risk communication: Implications for intervention. Family Research
Impett, E. A., Schooler, D., & Tolman, D. L. (2006). To be seen and not heard: Femininity ideology and adolescent girls' sexual health. Archives of Sexual Behavior
Kirby, D. B, Baumler, E., Coyle, K. K., Basen-Engquist, K., Parcel, G. S., Harrist, R., et al. (2004). The "safer choices" intervention: Its impact on the sexual behaviors of different subgroups of high school students. Journal of Adolescent Health
Liao, L. L., & Hong, X. L. (2006). The communication practices of sex-related issues between adolescents and their caregivers: Elementary and junior high school students in Changhua County, Taiwan as an example. Formosan Journal of Sexology
(2), 1-12. (Original work published in Chinese)
Libman, E., Rothenberg, I., Fichten, C. S., & Amsel, R. (1985). The SSES-E: A measure of sexual self-efficacy in erectile functioning. Journal of Sex & Marital Therapy
Mathews, C., Aaro, L. E., flisher, A. J., Mukoma, W., Wubs, A. G., & Schaalma, H. (2009). Predictors of early first sexual intercourse among adolescents in Cape Town, South Africa. Health Education Research
Miller, K. S., Fasula, A. M., Dittus, P., Wiegand, R. E., Wyckoff, S. C., & McNair, L. (2009). Barriers and facilitators to maternal communication with preadolescents about age-relevant sexual topics. AIDS and Behavior
Mitchell, C. M., Kaufman, C. E., & Pathways of Choice and Healthy Ways Project Team. (2002). Structure of HIV knowledge attitudes and behaviors among American Indian youth adults. AIDS Education and Prevention
Noar, S. M., Zimmerman, R. S., Palmgreen, P., Lustria, M., & Horosewski, M. L. (2006). Integrating personality and psychosocial theoretical approaches to understanding safer sexual behaviour: Implications for message design. Health Communication
Ogle, S., Glasier, A., & Riley, S. C. (2008). Communication between parents and their children about sexual health. Contraception
O'Sullivan, L. F., Meyer-Bahlburg, H. F. L., & McKeague, I. W. (2006). The development of the Sexual Self-concept Inventory for early adolescent girls. Psychology of Women Quarterly
Ott, M. A., & Pfeiffer, E. J. (2009). "That's nasty" to curiosity: Early adolescent cognitions about sexual abstinence. Journal of Adolescent Health
Salazar, L. F., Crosby, R. A., DiClement, R. J., Wingood, G. M., Lescano, C. M., Brown, L. K., et al. (2005). Self-esteem and theoretical mediators of safer sex among African American female adolescents: Implications for sexual risk reduction interventions. Health Education and Behavior
Shah, D., Thronton, S., & Burgess, A. P. (1997). Sexual Risk Cognitions Questionnaire: A reliability and validity study. AIDS Care
Shercliffe, R. J., Hampton, M., McKay-McNabb, K., Jeffery, B., Beattie, P., & McWatters, B. (2007). Cognitive and demographic factors that predict self-efficacy to use condoms in vulnerable and marginalized aboriginal youth. The Canadian Journal of Human Sexuality
Somer, C. L., & Ganivez, G. L. (2003). The Sexual Communication
Scale: A measure of frequency of sexual communication
between parents and adolescents. Adolescence
Swain, C. R., Ackerman, L. K., & Ackerman, M. A. (2006). The influence of individual characteristics and contraceptive beliefs on parent-teen sexual communications: A structural model. Journal of Adolescent Health
Tasic, D. (1997). Adolescents' self-concept and their communication with parents. Croatian Medical Journal
Van Der Straten, A., Catania, J. A., & Pollack, L. (1998). Psychosocial correlates of health-protective sexual communication
with new sexual partners: The national AIDS behavioural survey. AIDS and Behavior
Widman, L., Welsh, D. P., McNulty, J. K., & Little, K. (2006). Sexual communication
and contraceptive use in adolescent dating couples. Journal of Adolescent Health
Woods, J. L., Hensel, D. J., & Fortenberry, J. D. (2009). Contraceptive withdrawal in adolescents: A complex picture of usage. Journal of Pediatric & Adolescent Gynecology