The amount and timing of parent-adolescent sexual communication as predictors of late adolescent sexual risk-taking behaviors.

This study examined the moderating role of timing of first discussion of sexual intercourse with mothers and fathers on the relationship between the amount of sexual communication and sexual risk-taking behaviors in late adolescence. Late adolescents (N = 214) completed questionnaire measures regarding the amount and timing of sexual communication with their mothers and fathers. In addition, the participants provided information about their various sexual risk-taking behaviors. In general, the results suggest that timing of first discussion of sexual intercourse contributes additional variance in several sexually risky behaviors beyond that contributed by the amount of communication with both fathers and mothers. The implications of these findings are discussed.

Preparing children for intimate relationships and providing them with an understanding of how to conduct themselves as sexual beings are crucial aspects of socialization and the family (Miller & Moore, 1990), especially in the context of a culture where sexuality is emphasized and the risks of engaging in sexual behavior are great lot adolescents (e.g., teenage pregnancies, HIV infection). There has been an increased interest in the role that parents play in influencing the sexual behavior of their adolescents (e.g., Jaccard & Dittus, 1991; Jaccard, Dittus, & Gordon, 1998: Kotva & Schneider, 1990). Parents can be viewed as socialization agents who provide direct information to their adolescents during discussions of various sexual topics. Theoretically, a socialization model provides a basis for the current examination of parent adolescent sexual communication in relation to adolescent sexual behavior. Socialization theory is based on the idea that children and adolescents learn certain attitudes and behaviors early in life from adult role models, such as parents. The attitudes and behaviors modeled by parents are learned by their children and portrayed in adolescence and adulthood (Philliber, 1980). Consequently, an adolescent's sexual behavior can, in part, be predicted from information on how the adolescent was socialized by his or her parents (Rodgers, 1996).

Parents are the first and most omnipresent force in the socialization of their children. Adolescents report that parents, along with their peers, are the main socialization agents who influence their sexual behavior (Miller & Fox, 1987). Research on parental sexual socialization has focused on the belief that particular parental behaviors such as the discussion of sexual topics will result in certain adolescent behaviors such as delay of the initiation of sexual behaviors (Miller & Fox, 1987). Typically, research on parental sexual socialization has focused on whether sexual communication has taken place between parents and adolescents and the amount of sexual communication that occurred. Amount of sexual communication can be viewed as the number or variety of sexual topics discussed (e.g., HIV/AIDS. contraception, pregnancy) and the extent to which these topics were discussed. Previous research on how the amount of sexual communication between parents and adolescents is related to adolescents' sexual behavior has been inconclusive. Some studies have found that more parent adolescent sexual communication is related to fewer sexual behaviors (Fox & Inazu, 1980; Pick & Palos, 1995), whereas other studies have found that more parent-adolescent sexual communication is related to more sexual behaviors (Pistella & Bonati, 1998; Somers & Paulson, 2000). One aspect of parent-adolescent sexual communication that has received little attention is the timing of such communication. An examination of timing of first discussion of sexual topics may offer more information on the relationship between parent adolescent sexual communication and adolescent sexual behaviors and provide insight into discrepancies reported in previous research on amount of parent adolescent sexual communication.

One aspect of adolescent sexual behavior that is of particular interest due to negative consequences for adolescents is sexual risk-taking behaviors. Sexual risk taking behaviors are those behaviors that contribute to the occurrence of unintended pregnancy and sexually transmitted infections (STIs), including such behaviors as multiple sexual partners, lack of condom use, and lack of birth control use (Centers for Disease Control [CDC], 1998). Early sexual debut can also be included under the category of sexual risk-taking behaviors due to its association with subsequent risky sexual behaviors such as having multiple sexual partners (Seidman, Mosher, & Aral, 1994). Early initiation of sexual intercourse also lengthens the time that adolescents who are not consistently using condoms and birth control are at risk for the contraction of STIs and for adolescent pregnancy (Miller et al., 1997).

Some previous research has indicated that parent-adolescent sexual communication has positive influences regarding several sexual risk-taking behaviors (Barnett, Papini, & Gbur, 1991; Handelsman, Cabral, & Weisfeld, 1987; Mueller & Powers, 1990; Pick & Palos, 1995). Adolescents who experience parent-adolescent sexual communication are more likely to delay sexual intercourse (Kahn, Smith, & Roberts, 1984: Pick & Palos, 1995), use contraceptives more frequently (Fisher, 1987), and report fewer sexual partners (Darling & Hicks, 1982; Kallen, Stephenson, & Doughty. 1983). Leland and Barth (1993) examined characteristics of adolescents who displayed HIV/AIDS avoidance behaviors, such as increasing condom use and decreasing number of sexual partners, and those who did not. Those adolescents who reported discussing a variety of sexual topics with parents were more likely to report greater condom use and fewer sexual partners. These findings suggest that the amount of parent-adolescent sexual communication may play a positive role in deterring adolescents from engaging in those sexual risk-taking behaviors (e.g., lack of contraceptive use and multiple sexual partners) that increase the chance of contracting STIs.

Another area of concern in sexual risk taking is adolescent pregnancy. Similar to other positive effects of more parent-adolescent sexual communication, research has found that adolescents who experience more parent-adolescent sexual communication are less likely to have been pregnant or gotten a female partner pregnant during their adolescence (Leland & Barth, 1993; Pick & Palos, 1995). In addition, parent-adolescent sexual communication is related to more effective use of contraception (Casper, 1990; Leland & Barth, 1993; Pick & Palos, 1995). Overall, this research suggests that the amount of parent-adolescent sexual communication may have important influences on the probability of an adolescent becoming pregnant or getting an adolescent partner pregnant.

Taken together, the above-mentioned studies suggest that the amount of parent-adolescent communication is important to a number of sexual risk-taking behaviors and outcomes including early initiation of sexual intercourse, lack of condom and contraceptive use, multiple sexual partners, HIV/AIDS, and adolescent pregnancy. However, other studies have found contrasting results. Pistella and Bonati's (1998) examination of the amount of parent-adolescent sexual communication in a sexually active female adolescent population suggested that those female adolescents who had discussed sex-related topics of contraception and STIs with their parents were more likely to report a pregnancy. However, it was not clear when sexual communication occurred in relationship to the adolescent's pregnancy. It is possible that those parents who discussed sex-related topics with their daughters were responding to their daughter's sexual behavior. In addition, this sample is not representative of all adolescents because it only included female family planning patients.

Another study found similar results with a nonclinical population of female and male adolescents. Somers and Paulson (2000) found that more sexual communication with mothers and fathers was related to greater frequency of sexual behavior in adolescents. Their findings also revealed that age of the adolescent made a significant contribution to sexual behavior, indicating that age rather than amount of parent-adolescent communication may account for sexual behavior. Although age has previously been found to be related to sexual behavior (Newcomer & Baldwin, 1992: Santelli & Beilenson, 1992), parent-adolescent communication and age of first intercourse have not been examined together to determine if discussions occur before or after the initiation of sexual intercourse. An examination of the age of first discussion of sexual topics and the timing of sexual communication in relation to the onset of sexual activity would provide more information on the implications of parent-adolescent sexual communication.

One of the few studies that examined age of first discussion of sexual topics was Fox and Inazu's (1980) study of mother and daughter sexual communication. The researchers found that daughters who had sexual intercourse talked more frequently with their mothers about sex-related topics than those daughters who were virgins. In separate analyses, Fox and Inazu (1980) specified the time order of events such that the daugther's sexual behavior preceded recent mother-daughter sexual discussions. The authors restricted their analyses to virgins and those nonvirgins who had been sexually active for at least 6 months. They eliminated daughters who had initiated sexual activity within the preceding 6 months and assessed mother-daughter sexual discussions that occurred within the last 6 months. Fox and Inazu (1980) found that sexually experienced females had talked more frequently about sex with their mothers during the 6 months prior to the study than had girls who were still virgins. This suggests that recent sexual communication may be a response to a daughter's sexual behavior. Some mothers may become more involved in the sexual communication and socialization of their daughters when they suspect the importance of doing so, that is, when the daughter has become sexually active. This study did not, however, compare adolescents who had experienced mother-daughter sexual communication before the initiation of sexual activity (on-time discussions) with those who had experienced mother-daughter sexual communication after the initiation of sexual activity (off-time discussions).

A recent study looked more specifically at timing of mother-adolescent condom discussions and adolescent condom use (Miller, Levin, Whitaker, and Xu, 1998). Mother-adolescent discussions about condoms that occurred before the initiation of sexual intercourse were associated with greater condom use, while discussions after first intercourse were not related to greater condom use among male and female adolescents (Miller et al., 1998). These findings suggest that timing of parent-adolescent sexual communication is an important factor in adolescent sexual behavior.

Although the aforementioned research (Fox & Inazu, 1980; Miller et al., 1998) found that timing of discussion appears to be important for mothers, daughters, and sons, a limitation of these studies as well as other previous research is the lack of attention given to the fathers' role in sexual communication with their children. Research that has examined both mothers and fathers has found that mother-adolescent and father-adolescent communication can have different effects on adolescents' behavior. For example, the amount of communication between both adolescent males and females and their mothers was associated with less sexual activity (Kahn et al., 1984). while the amount of communication between adolescent males and their fathers was associated with increased sexual activity (Fisher, 1987). One explanation for the differences between mother-adolescent and father adolescent sexual communication may be that mothers are talking sooner to their adolescents about sex related topics. If this is true, mothers may be discussing sexual topics before they occur (on-time) while fathers may be discussing sexual topics after they occur (off-time). Further examination of the timing of first discussion may contribute to the understanding of previously reported gender differences.

The present study examined the timing of parent-adolescent sexual communication in relation to the onset of sexual intercourse to help determine the benefits of parents talking to their adolescents about sex before sexual activity begins. This was investigated by assessing late adolescents' perceptions of the amount of mother and father sexual communication that occurred, the timing of first discussion of sexual intercourse with both mothers and fathers, and the relationship of the communication variables to several adolescent sexual risk taking behaviors. Based on previous findings, we set forth several hypotheses.

HYPOTHESIS 1

We expected the amount of both mother-adolescent and father-adolescent sexual communication to be a significant predictor of sexual risk-taking behaviors. Specifically, we predicted that adolescents who reported more parent adolescent sexual communication would report fewer sexual risk taking behaviors, while those who reported less parent-adolescent sexual communication would report more sexual risk-taking behaviors. Although previous research in the area has found contradictory results, a large body of research supports the relationship between more parent-adolescent sexual communication and fewer sexual risk taking behaviors.

HYPOTHESIS 2

We predicted that timing of sexual discussions would be a significant predictor of sexual risk-taking behaviors, even when we controlled for amount of sexual communication. Sexual discussions that occurred prior to the initiation of sexual intercourse (on-time discussions) would predict fewer sexual risk-taking behaviors, while sexual discussions that occurred after the initiation of sexual intercourse (off-time discussion) would predict more sexual risk-taking behaviors in adolescents.

HYPOTHESIS 3

We expected an interaction between the amount of sexual communication and the timing of first discussion. We anticipated that more parent-adolescent sexual communication would be related to on-time discussions, and in turn, those adolescents would report less sexual risk taking behaviors. Likewise, less parent-adolescent sexual communication would be related to off time discussions, and those adolescents would report more sexual risk-taking behaviors.

METHOD

Sample and Procedures

Two-hundred fourteen late adolescents (101 males and 113 females) attending a large Midwestern university completed the study measures for extra course credit. The prerequisite for participation was that all participants had to be between 18 and 21 years of age. The mean age of the sample was 19.9 years. The majority of respondents were White (82.2%), while the remaining were African American (9.3%), Hispanic (5.6%), and Asian American (1.9%), Ten percent of the sample reported being virgins. The majority of the respondents reported being heterosexual (97.7%), while those remaining reported being bisexual (.9%), homosexual (.5%), or unsure (.5%). The majority of the respondents currently lived in a coed dormitory (37.4%) or an apartment with a same-sex roommate (33.6%). The remaining respondents lived in non-coed dormitories (18.2%), in apartments with opposite-sex roommates (7.5%), or with parents (2.3%). Regarding family composition, the majority of participants were from intact families (73.4%), while the remaining included 20.1% from divorced families, 2.8% from single-parent families, and 1.4% who lived with grandparents during their childhood and adolescence.

Measures

Amount of parent-adolescent sexual communication. Fisher (1987) developed the Weighted Topics Measure of Family Sexual Communication Scale to assess quickly and objectively the amount of communication about sexuality that has occurred between parents and their adolescent children. This scale, consisting of nine items, combines a relatively objective measure (the number of topics discussed) with a more subjective one (the extent of discussion). The instrument asks respondents to indicate on a scale of 1 to 5, with 1 indicating none and 5 indicating a lot, the extent to which nine specific sexual topics have been discussed (pregnancy, fertilization, intercourse, menstruation, sexually transmitted diseases, birth control abortion, prostitution, and homosexuality), first with their mothers and then with their fathers. Scores are computed by summing all items. Scores can range from 9 to 45 on both the mother and the father form, with higher scores indicating greater amounts of sexual communication. Cronbach's alpha reliability coefficient in the present study was .91 for discussion with father and .88 for discussion with mother.

Timing of parent-adolescent sexual discussions. We measured this variable by making an addition to Fisher's (1987) Weighted Topics Measure of Family Sexual Communication Scale. After indicating the extent to which the nine sexual topics had been discussed with parents. participants were asked an open ended question regarding the age at which their mothers and their fathers first discussed each sexual topic with them. We used the age of first discussion of sexual intercourse to create a timing variable by comparing the age of first discussion of sexual intercourse with mother and father to age of respondent's first intercourse. Sexual discussions that occurred before the initiation of sexual intercourse were coded as on time (l) while sexual discussions that occurred after initiation of sexual intercourse wine coded as off-time (0). We coded those adolescents who reported the same age of first discussion as age of first intercourse as off-time since discussions did not occur before but rather relatively at the same time as the initiation of sexual intercourse. Because timing was the focus of the study, not whether sexual discussion had occurred or not, we coded those who did not report sexual discussions as off-time because discussions occurring in the future would be considerably late in development. Virgins who reported sexual discussion were coded as on-time.

Adolescent sexual risk taking. Sexual risk taking was assessed from questions on the Center for Disease Control Sexual Risk Taking portion of the 1998 National College Health Risk Behavior Survey (CDC, 1998). Participants were asked to report the age when they had sexual intercourse for the first time and how many sexual partners they had had in their lifetime. To assess condom use in the past 30 days, we asked participants "During the past 30 days, how often did you or your partner use a condom?" Participants chose from the following responses: "I have not had sexual intercourse in the past 30 days," "never used a condom" "rarely used a condom," "sometimes used a condom," "most of the time used a condom," or "always used a condom." We asked respondents if they had ever been tested for the AIDS virus or HIV infection (yes = 1; no = 2). Respondents indicated the number of methods of birth control used the last time they had intercourse by circling the types of contraceptives from the following categories: condoms, spermicides, birth control pills. IUD, Depo-Provera, Norplant, none, "not sure," other, or "have not had sexual intercourse." We then summed the number of methods for a possible score of 6 total methods. The total score was used as an indicator of contraceptive use based oil the idea that an adolescent who is using multiple forms of birth control is engaging in less risky behavior than an adolescent who is only using one form. Additionally, it is important to recognize that some forms of contraceptives only prevent pregnancy and do not protect the adolescent against the contraction of STIs (e.g., birth control pills). We also asked respondents how many times they had been or had gotten someone pregnant. Adolescents could indicate never, once, two or ill more times, or "not sure." Because only a small number of adolescents reported a pregnancy, we coded the data as a dichotomous variable (1 = they had never been pregnant or gotten a female partner pregnant and 2 = they had been pregnant or gotten a female partner pregnant).

RESULTS

Descriptive Statistics

The mean scores and standard deviations for the Weighted Topics Measure of Family Sexual Communication Scale were M = 14.82, SD = 7.57 and M = 19.93, SD = 8.26 for sexual communication with fathers and mothers, respectively. The descriptive statistics for age of first discussion of sexual intercourse indicated a mean age of 13.4 years with a standard deviation of 2.79 for discussion with father and a mean age of 13.11 years with a standard deviation of 3.04 for discussion with mother. For those who were non-virgins, the mean age of first intercourse reported was 16.77 years, SD = 1.83 (n = 188), with males reporting a mean age of 16.8, SD=2.05 (n=90) and females reporting a mean age of 16.74, SD = 1.63 (n = 98). According to data on the age of first intercourse and age of firm sexual discussion with mother and father, respondents indicated that 36.9% had experienced on-time discussions with their fathers while 57.9% reported off-time discussions. The remaining 5.2% did not have a father figure to report on regarding timing of sexual communication. Respondents reported that 57.5% had experienced on-time discussions with their mothers while 41.6% reported off time discussions. Those who did not have a mother figure to report on made up the remaining .9%.

All respondents indicated their condom use in the last 30 days; 24.8% reported using condoms always, 11.2% most of the time, 4.7% sometimes, 6.1% rarely, and 22.0% never, while 31.3% reported not having sexual intercourse in the last 30 days. The average number of lifetime sexual partners was 2.86, SD = 2.14. Additionally, the majority of respondents (73.4%) indicated that they had not been tested for HIV/AIDS, while 26.6% indicated they had been tested and 3.3% reported they were not sure. The mean score and standard deviation for the number of different methods of birth control used were M = 1.3 and SD = .67, indicating that most respondents reported using between one and two methods of birth control to prevent pregnancy. The majority of the respondents reported that they had never been or never gotten someone pregnant (93.9%).

Correlations between sexual communication and sexual risk-taking behaviors are reported in Table 1. Those who were virgins or did not report sexual communication with either mothers or fathers are not included in the analyses (N = 172). There was a positive relationship between amount of sexual communication reported with mother and father. That is, adolescents who reported more sexual communication with their fathers were also more likely to report more sexual communication with their mothers. Additionally, the amount of sexual communication with either parent was associated with on-time discussions of sexual topics with both parents. Specifically, adolescents who reported more sexual communication with either parent were more likely to report more on-time discussions with both parents.

Regression Analyses

We conducted 12 hierarchical regressions to test the study hypotheses using six outcome variables: (a) age of first sexual intercourse, (b) condom use in the last 30 days. (c) number of lifetime sexual partners, (d) number of methods of birth control, (e) being tested for HIV/AIDS, and (f) pregnancy, for both father and mother sexual communication, Hierarchical linear regression analyses were used for the continuous outcome variables (first intercourse, condom use, sexual partners, and birth control), while hierarchical binary logistic regression analyses were used for the dichotomous variables (being tested for HIV/AIDS and pregnancy).

In each regression, we entered the control variable of gender first, followed by amount of parent sexual communication (either father or mother) and the potential moderator variable of timing of first discussion of sexual intercourse (with father or mother). The amount of mother and father sexual communication variables, which are continuous, were centered around their means, while timing of discussion of sexual intercourse was dummy coded. We did this to facilitate interpretation of the interaction terms and to control for multicollinearity (Aiken & West, 1991). We entered the interaction term last to assess whether the impact of the amount of father-adolescent and mother-adolescent sexual communication varied as a function of the timing of first discussion of sexual intercourse with that parent. The hierarchical regressions are presented in Tables 2 and 3 for father-adolescent and Tables 4 and 5 for mother-adolescent. The change in as reported in Tables 2, 3, 4, and 5 reflect missing data reported for either mother-or father-adolescent communication and the exclusion of virgins in the analyses with the outcome variables of age of first sexual intercourse and number of birth control methods used.

H1: The amount of both mother-adolescent and father-adolescent sexual communication will be a significant predictor of sexual risk-taking behaviors. Consistent with the zero-order correlations, amount of sexual communication with father made a significant contribution in predicting age of first intercourse, number of lifetime sexual partners, and being tested for HIV/AIDS. Adolescents who reported a greater amount of sexual communication with their fathers were younger at first intercourse, reported more lifetime sexual partners, and were more likely to have been tested for HIV/AIDS. Similar results were found with mother-adolescent sexual communication. The amount of sexual communication with mother made a significant contribution in predicting age of first intercourse, number of lifetime sexual partners, number of methods of birth control, being tested for HIV/AiDS, and having been or gotten someone pregnant. Adolescents who reported more sexual communication with their mothers also reported a younger age of first intercourse, more lifetime sexual partners, and more methods of birth control used, and were more likely to have been tested for HIV/AIDS and to have been or gotten someone pregnant, in general, the results did not support the hypothesis that more sexual communication was related to fewer sexual risk-taking behaviors.

H2: Timing of sexual discussions will be a significant predictor of sexual risk-taking behaviors, even when we control for amount of sexual communication. For fathers, timing of discussion of sexual intercourse made a significant contribution in predicting age of first intercourse, number of lifetime sexual partners, and pregnancy. Adolescents who reported on-time discussions with their fathers were older at first intercourse and had fewer life time sexual partners, but were more likely to have been or gotten someone pregnant. For mothers, tinting of discussion of sexual intercourse made a significant contribution in predicting age of first intercourse, number of lifetime sexual partners, number of methods of birth control used, and pregnancy, Adolescents who experienced on-time discussions with their mothers were older at the time of first intercourse, had fewer lifetime sexual partners, and used more methods of birth control but were more likely to have reported that they had been or gotten someone pregnant.

H3: An interaction between the amount of sexual communication and timing of first discussion is expected, The interaction between father sexual communication and timing of discussions of sexual intercourse was significant

ill predicting age of first intercourse and pregnancy. The plot of the Sexual Communication with Father X Timing of Sexual Communication with Father interaction revealed a strong relation between sexual communication with father and age or first intercourse for those adolescents who had off-time discussions of sexual intercourse. Although the direction of effect for the on time group is similar to the off-time group, the effect ix less dramatic (See Figure 1). This suggests that timing of sexual discussions with a father moderated the effects of the amount of lather-adolescent sexual communication on the age of first intercourse. The interaction indicates that regardless of the amount of sexual communication, oil-time discussions with fathers resulted in an older age of first intercourse.

[FIGURE 1 OMITTED]

We found a similar relationship with the dependent variable of pregnancy. The plot of the Sexual Communication with Father X Timing of Sexual Communication with Father interaction also revealed a relationship between sexual communication with rather and pregnancy for those adolescents who had off-time discussions of sexual intercourse. A relationship did not exist for those adolescents who experienced on-time discussions of sexual intercourse (See Figure 2). This suggests that timing of sexual discussions with a lather moderated the effects between amount of father-adolescent sexual communication and pregnancy. However, due to the small number of pregnancies reported in the sample (n = 121), the results regarding pregnancies should be interpreted with caution. None of the other interactions were significant. indicating that the timing of discussion of sexual intercourse did not moderate the effects of amount of sexual communication with father on condom use in last 30 days, number of lifetime sexual partners, being tested for HIV/AIDS, or number of methods of birth control.

[FIGURE 2 OMITTED]

Regarding mother adolescent sexual communication, none of the interactions were significant, indicating that the timing of discussion of sexual intercourse did not rood crate the effects of amount of sexual communication with mother on the outcome variables.

Discussion

This study was designed to investigate the associations of parent-adolescent sexual communication and sexual risk-taking behaviors. Previous research examining parent-adolescent sexual communication and adolescent sexual behavior has not been conclusive. Some studies have suggested that sexual communication is not related to adolescent sexuality (e.g., Miller & Fox, 1987; Newcomer & Udry, 1985) while others have found that high levels of parent-adolescent communication are related to low levels of sexual risk taking in late adolescence (Barnett et al., 1991; Pick & Palos, 1995). While previous research has neglected the timing of sexual communication, the current study assessed how timing of parent adolescent sexual communication moderates the relationship between the amount of parent-adolescent sexual communication and sexual risk-taking behaviors.

Amount of parent-adolescent sexual communication made a significant contribution in predicting sexual risk-taking behaviors in this study, but in the opposite of the hypothesized direction. Although adolescents who reported more sexual communication with their mothers used more methods of birth control, more sexual communication with parents was generally related to more sexually risky behaviors. Adolescents who reported more sexual communication with their mothers and fathers reported a younger age of first intercourse and more lifetime sexual partners and were more likely to have been tested for HIV/AIDS. In addition, more mother-adolescent communication was related to pregnancy. This is consistent with previous research that has indicated that more sexual communication with mothers and fathers is related to greater frequency of sexual behavior in adolescents (Somers & Paulson. 2000), and that female adolescents who have discussed sex-related topics with a parent are more likely to report a pregnancy (Pistella & Bonati, 1998). These results suggest that both parents may influence late adolescents' sexual risk-taking behaviors.

Along with the amount of parent-adolescent sexual communication, the timing of sexual discussions also made a significant contribution in predicting sexual risktaking behaviors. On-time discussions with fathers and mothers predicted an older age at the time of first intercourse and fewer lifetime sexual partners, but also predicted a greater likelihood of having been or gotten someone pregnant. Mother-adolescent sexual discussions that were on-time also predicted more methods of birth control used. These data support the importance of examining the timing of sexual discussions when investigating parent-adolescent sexual communication and adolescent sexual behavior. Although our results suggest that timing of first sexual discussion is important in predicting some adolescent sexual behaviors, it is important to note that the first sexual discussions that occur may not be as explicit or detailed as subsequent discussions. In fact, later discussions may include more information regarding contraception or STIs and be more interactive than the first discussion. Additionally, it is possible that adolescents who reported on-time discussion were older at first intercourse because adolescents who delayed sexual intercourse provided parents with more time to have discussions that would be identified as on-time, or prior to sexual intercourse. Therefore, subsequent studies may need to control for the age of discussion and/or the depth of content in those discussions.

Regression analyses supported the moderating relationship of timing of sexual discussions for father-adolescent sexual communication and the outcome variables of age of first intercourse and pregnancy. The findings indicated that discussions with fathers that were on time, regardless of the amount of sexual communication, resulted in an older age of first intercourse. Off-time discussions of sexual intercourse with fathers coupled with more reported father-adolescent sexual communication predicted an earlier age of first intercourse, while off-time discussions coupled with less reported father-adolescent sexual communication predicted an older age of first intercourse. One explanation for these findings may be that fathers are responding to their adolescents' behavior and initiation of sexual intercourse. When fathers discover that their adolescents have had sexual intercourse for the first time at a younger age, fathers may be more concerned with the developmental implications of this activity and discuss sexual topics more with their children. When the adolescent is older, however, the father may feel that the activity is more developmentally appropriate and that the adolescent has already acquired the sexual knowledge needed to make sexual decisions. Therefore. lathers may not see a need for more sexual communication with their older adolescents.

A similar interaction between father-adolescent sexual communication and timing of first discussion of sexual intercourse emerged regarding whether participants had ever been or gotten someone else pregnant. It should be noted that this sample reported a small number of pregnancies and results should be interpreted with caution. These results again suggest that fathers may be responding to their adolescents' behavior. Adolescents who engage in more sexually risky behaviors are more likely to become or get someone pregnant (Poppen, 1995). A likely explanation for the interaction we found may he that fathers who perceive that their adolescents are becoming interested in sex or are aware that their adolescents are engaging in sexual behaviors are prompted to discuss more sexual topics and discuss them more frequently with their adolescents (Somers & Paulson, 2000).

Although we found interactions for two dependent variables, the interaction models between liming of sexual discussions and amount of sexual communication were not supported overall. The findings of the present study suggest that timing of sexual discussions interacts with the level of father-adolescent sexual communication to influence the age of first intercourse and pregnancy. Regression analyses did not support the moderating relationship between amount of father-adolescent sexual communication and timing of first discussion with the dependent variables of condom use, nun/bet of lifetime sexual partners. being tested for HIV/AIDS, and number of methods of birth control used. We found main effects for mother-adolescent sexual communication: however, regression analyses did not support the moderating relationship between amount of parent-adolescent sexual communication and timing of first discussion with any dependent variables. One explanation for these nonsignificant results may be that these are sexual behaviors that may be easier for an adolescent to hide, Parents may suspect that their adolescents have become sexually active due to long-term dating relationships or the adolescents" requests for birth control pills, and may therefore discuss sexual topics surrounding pregnancy and first intercourse. However, adolescents may better hide their sexual activity with numerous partners and concerns with HIV/AIDS.

Although a number of interpretations are possible for the present findings, limitations of the study must be taken into account when examining the results. With respect to the measures used, the classification of on-time versus off-time sexual discussions with parents presents a problem in its relationship to age. The older an adolescent is at first intercourse, the more likely that parent-adolescent sexual discussions will occur belt, re initiation of sexual intercourse. Therefore, a parent-adolescent sexual discussion may be interpreted as on-time even though it may have occurred relatively late in the adolescent's development. Future research needs to examine this relationship more closely and perhaps investigate this phenomenon in younger adolescent populations.

The present study also only considered adolescent Respondents' perceptions of the amount sexual discussion and age of first discussion of sexual topics from a retrospective view. Thus, their perceptions may not have been as accurate as those of younger adolescents who are currently in junior high and high school and discussing sexual topics with their parents for the first time. Future research would benefit from a longitudinal examination of adolescents as they begin parent-adolescent sexual communication, with follow-ups to determine subsequent sexual behavior to examine the effects of timing of first discussion of sexual topics on adolescent sexual activity and risk taking.

Another limitation of the present study was the respondents' lack of diversity. A sample with more diversity may yield different results and would allow for more generalizability to a larger adolescent population. For example, on average, Black adolescents are more likely to be sexually active than Whites, even when other factors are controlled (Casper. 1990). Additionally. Latino adolescents report significantly less sexual information from parents than do non-Latino White adolescents (Davis & Harris, 1982). Furthermore, this study only examined late adolescents in a college population. When adolescents go to college they are often leaving their parents for the first time. This new independence can result ill more opportunities for sexual behavior to occur. It can also increase the amount of communication that adolescents experience with their parents on all topics including sexuality (Barnes & Olson. 1985). Therefore, an examination of a population of late adolescents who are not experiencing increased independence, such as college students living at home or late adolescents who are not attending college and still reside with their parents, is important to offer more generalizability of the findings.

In the present study, the amount of sexual communication was relatively low. It would be advantageous to explore the variable of timing in a sample that has more variability in the amount of sexual communication to better understand the interaction between timing of sexual discussions and amount of sexual communication. The relatively low reported levels of sexual communication may be due to the topics included in the measure. Some topics may be discussed more often or in more detail (e.g., pregnancy) whereas others may not be discussed at all (e.g., prostitution). Future studies should examine whether discussions of certain topics more strongly predict sexual risk-taking behaviors. For example, researchers could focus on selected topics (e.g., intercourse, pregnancy, sexually transmitted infections) to determine if these topics are more strongly related to sexual risk-taking as compared to other topics (e.g., fertilization or prostitution). Although the current study includes this data, the purpose of the current study was to examine the specific relationship between sexual communication about intercourse and sexual risk-taking behaviors: therefore these other topics were not examined.

Finally, the present study did not examine the attitudinal tones of the communication, the specific content of the discussions, or who initiated the discussions. An examination of these factors may help explain why father-adolescent and mother-adolescent communication are associated with different sexual outcomes for their adolescents. Research that examines parental attitudes and their relationship to discussing sexual topics as well as their influence on how the timing of sexual discussions affects sexual behavior is needed to further understand the relationships between parent-adolescent sexual communication and adolescent sexual behavior.

Given the findings that timing of sexual discussions does contribute to adolescents' sexual risk-taking behaviors. future research is needed that examines why some parents are more likely to have on-time discussions with their children as compared to parents whose discussions are off-time. One hypothesis is that other parenting processes may covary with amount of sexual communication and timing of sexual discussions (Christopher, 2000). For example, high levels of parental monitoring and support may predict which parents are likely to provide on-time discussions with their adolescents. Research examining the processes that may lead to more on time sexual discussions between parents and adolescents, and hence fewer sexual risk-taking behaviors, is needed in this area.

In conclusion, parent-adolescent communication and sexual activity is an important topic because of the possible negative consequences of sexual activity such as unwanted pregnancies and STIs. Little research has been done regarding the timing of parent-adolescent sexual communication. The present study examined the relationship between the timing of parent adolescent sexual communication and sexual risk taking behaviors. Based upon the present results, timing is an important factor in predicting sexual risk-taking behaviors. For most sexual behaviors, timing of first discussion of sexual intercourse contributed additional variance beyond that contributed by amount of communication in several sexually risky behaviors for both fathers and mothers. Therefore, emphasis should be placed on encouraging parents to discuss issues with their adolescents early in their adolescents' development.

Table 1. Correlation coefficients (N=172) for sexual Communication and
Sexual Risk-Taking Behaviors

                                1        2        3        4       5

1. Dad sexual communication   1.00
2. Mom sexual communication    .61**   1.00
3. Dad timing                  .70**    .38**   1.00
4. Mom timing                  .38**    .59**    .51**   1.00
5. First intercourse          -.23**    -.19*    .03      .15    1.00
6. Condom use                  .05      .08      .07      .11     .02
7. Number of partners          .22**    .20**   -.01     -.09    -.42**
8. HIV/AIDS testing           -.12      -.21**  -.06     -.15     .13
9. Birth control               .10      .27**    .14      .29**  -.04
10. Number of pregnancies      .15*     .16*    -.02     -.05    -.23**

                                6        7        8        9       10

1. Dad sexual communication
2. Mom sexual communication
3. Dad timing
4. Mom timing
5. First intercourse
6. Condom use                 1.00
7. Number of partners         -.07     1.00
8. HIV/AIDS testing           -.05     -.14     1.00
9. Birth control               .06     -.07     -.13     1.00
10. Number of pregnancies     -.13      .23**   -.20**   -.10    1.00

* p < .05. ** p < .01.

Table 2. Hierarchical Regression of Sexual Risk-Taking Behaviors
Regressed on Sexual Communication and Timing of Sexual Discussions With
Fathers

Step/Variable                               R(2)            Change R(2)

First sexual intercourse (n = 176)
  Gender                                    .00               .00
  Father sexual communication               .04               .04**
  Father timing of discussion               .11               .07**
  Sexual communication X Timing             .13               .03*

Condom use in last 30 days (n = 201)
  Gender                                    .01               .00
  Father sexual communication               .01               .00
  Father timing of discussion               .02               .01
  Sexual communication X Timing             .02               .00

Number of lifetime sexual partners (n = 201)
  Gender                                    .01               .01
  Father sexual communication               .04               .03*
  Father timing of discussion               .09               .05**
  Sexual communication X Timing             .10               .01

Number of birth control methods (n = 176)
  Gender                                    .00               .00
  Father sexual communication               .01               .01
  Father timing of discussion               .02               .01
  Sexual communication X Timing             .03               .01

Step/Variable                              Beta             Final model
                                                              F value

First sexual intercourse (n = 176)
  Gender                                    .01
  Father sexual communication              -.21**
  Father timing of discussion               .37**
  Sexual communication X Timing             .41               6.65**

Condom use in last 30 days (n = 201)
  Gender                                   -.09
  Father sexual communication               .02
  Father timing of discussion               .12
  Sexual communication X Timing             .18                .99

Number of lifetime sexual partners (n = 201)
  Gender                                   -.11
  Father sexual communication               .18*
  Father timing of discussion              -.32**
  Sexual communication X Timing            -.26               5.69*

Number of birth control methods (n = 176)
  Gender                                    .00
  Father sexual communication               .11
  Father timing of discussion               .14
  Sexual communication X Timing             .22                .29

Note. The standardized betas reflect the variables in the model at
steps 1, 2, 3, and then 4.

* p < .05. ** p < .01.

Table 3. Hierarchical Binary Logistic Regression of Sexual
Risk-Taking Behaviors Regressed on Sexual
Communication and Timing of Sexual Discussions
With Fathers

                                                     Charge
Step/Variable                      [R.sup.2]           [R.sup.2]

HIV/AIDS testing (2 = 195)
Gender                              .00                 .00
Father sexual communication         .04
Father timing of discussion         .05                 .01
Sexual communication X Timing       .05                 .00
Number of pregnancies (n = 202)
Gender                              .00                 .00
Father sexual communication         .01
Father timing of discussion         .04                 .03*
Sexual communication X Timing       .06                 .02*

Step/Variable                       Beta               Final model
                                                        chi-square
HIV/AIDS testing (n = 195)
Gender                             -.15
Father sexual communication         .04                -.06**
Father timing of discussion        -.72
Sexual communication X Timing       .00                9.38*
Number of pregnancies (n = 202)
Gender                              .51
Father sexual communication         .01                 .06
Father timing of discussion        2.69*
Sexual communication X Timing      -.21*              13.04**

Note. The standardized betas reflect the variables in the model at
steps 1, 2, 3, and then 4.

* p < .05. ** p < .01.

Table 4. Hierarchical Regression of Sexual Risk-Taking Behaviors
Regressed on Sexual Communication and Timing of Sexual Discussions
With Mothers

Step/Variable                        [R.sup.2]      Change [R.sup.2]

First sexual intercourse (n = 184)
  Gender                                .00               .00
  Mother sexual communication           .04               .04**
  Mother timing of discussion           .14               .10**
  Sexual communication X Timing         .15               .01

Condom use in last 30 days (n = 210)
  Gender                                .01               .01
  Mother sexual communication           .01               .00
  Mother timing of discussion           .03               .02
  Sexual communication X Timing         .04               .01

Number of lifetime sexual partners (n = 210)
  Gender                                .01               .01
  Mother sexual communication           .03               .02*
  Mother timing of discussion           .09               .06**
  Sexual communication X Timing         .09               .00

Number of birth control methods (n = 183)
  Gender                                .00               .00
  Mother sexual communication           .08               .08**
  Mother timing of discussion           .11               .03*
  Sexual communication X Timing         .11               .00

Step/Variable                                          Final model
                                        Beta            F value

First sexual intercourse (n = 184)
  Gender                                -.01
  Mother sexual communication           -.20**
  Mother timing of discussion            .40**
  Sexual communication X Timing          .19              7.98**

Condom use in last 30 days (n = 210)
  Gender                                -.10
  Mother sexual communication            .06
  Mother timing of discussion            .15
  Sexual communication X Timing         -.16              1.93

Number of lifetime sexual partners (n = 210)
  Gender                                -.07
  Mother sexual communication            .16*
  Mother timing of discussion           -.31**
  Sexual communication X Timing          .05              5.20**

Number of birth control methods (n = 183)
  Gender                                -.02
  Mother sexual communication            .28**
  Mother timing of discussion            .22**
  Sexual communication X Timing          .05              5.53**

Note. The standardized betas reflect the variable in the model at
steps 1, 2, 3, and then 4.

* p < .05. ** p < .01.

Table 5. Hierarchical Binary Logistic Regression of Sexual Risk-Taking
Behaviors Regressed on Sexual Communication and Timing of Sexual
Discussion With Mothers

Step/Variable                        [R.sup.2]      Change [R.sup.2]

HIV/AIDS testing (n = 204)
  Gender                              .00                 .00
  Mother sexual communication         .06                 .06*
  Mother timing of discussion         .06                 .00
  Sexual communication X Timing       .06                 .00

Number of pregnancies (n = 211)
  Gender                              .00                 .00
  Mother sexual communication         .03                 .03*
  Mother timing of discussion         .05                 .02*
  Sexual communication X Timing       .07                 .02

Step/Variable                                        Final model
                                      Beta            F value

HIV/AIDS testing (n = 204)
  Gender                              .03
  Mother sexual communication        -.07*
  Mother timing of discussion        -.22
  Sexual communication X Timing      -.02               12.79**

Number of pregnancies (n = 211)
  Gender                              .13
  Mother sexual communication         .10*
  Mother timing of discussion        1.94*
  Sexual communication X Timing       .18               14.37**

Note. The standardized betas reflect the variables in the model at
steps 1, 2, 3, and then 4.

* p < .05. ** p < .01.

REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage.

Barnes, H. L., & Olson, D. H. (1985). Parent-adolescent communication and the circumplex model. Child Development, 56, 438--447.

Barnett, J K., Papini, D. R., & Gbur, E. (1991). Familial correlates of sexuality active pregnant and nonpregnant adolescents. Adolescence, 26, 457--472.

Casper, L. B. (1990). Does family interaction prevent adolescent pregnancy? Family Pluming Perspectives, 22. 109-114.

Centers for Disease Control. (1998). Youth risk behavior surveillance--United States, 1997. Morbidity and Mortality, Weekly Reports, 47(SS--3).

Christopher, F.S. (2000). To dance the dance: A symbolic exploration of premarital sexuality. Mahwah, NJ: Lawrence Erlbaum Associates.

Darling, C. A., & Hicks, M. W. (1982). Parental influence on adolescent sexuality: Implications for parents as educators. Journal of Youth and Adolescence, 11, 231--245.

Davis, S., & Harris, M. (1982). Sexual knowledge, sexual interests, and sources of sexual information of rural and urban adolescents from three cultures Adolescence, 17, 471--492.

Fisher, T. D. (1987). Family Communication and the sexual behavior and attitudes of college students. Journal of Youth and Adolescence, 16, 481--495.

Fox, G. L., & Inazu, J.K. (1980). Patterns and outcomes of mother-daughter communication about sexuality. Journal of Social Issues, 36, 7--29.

Handelsman, C. D., Cabral, R. J., & Weisfeld, G. E .(1987). Sources of information and adolescent sexual knowledge and behavior. Journal of Adolescent Research. 2(4), 455--463.

Jaccard, J., & Dittus, P. (1991). Parent-child communication: Toward the prevention unintended pregnancies. New York: Springer-Verlag.

Jaccard, J., Dinus. P.J, & Gordon, V. V. (1998). Parent-adolescent congruency in reports of adolescent sexual behavior and in communication about sexual behavior. Child Development, 69, 247--261.

Kahn, J., Smith, K., & Roberts, E. (1984). Familial communication and adolescent sexual behavior: Final report to the office of Adolescent Pregnancy Programs Cambridge, MA: American Institutes for Research.

Kallen, D. J., Stephenson. J. J., & Doughty, A. (1983). The need to know: Recalled adolescent sources of sexual and contraceptive information and sexual behavior. The journal of Sex Research, 19, 137--159.

Kotva, H. J., & Schneider, H. G. (1990). Those "talks": General and sexual communication between mothers and daughters. Journal of Social Behavior and Personality, 5,603--613.

Leland, N. L., & Barth, R. P. (1993). Characteristics of adolescents who have attempted to avoid HIV and who have communicated with parents about sex. Journal of Adolescent Research, 8, 58--76.

Miller, B. C., & Fox, G. L. (1987). Theories of adolescent heterosexual Behavior. Journal of Adolescent Research, 2, 269--282.

Miller, B. C., & Moore, K. A. (1990). Adolescent sexual behavior, pregnancy, and parenting: Research through the 1980s. Journal of Marriage and the Family, 52, 1025--1044.

Miller, B.C., Norton, M. C., Curtis, T., Hill, E. J., Schvaneveldt, P., & Young, M. H. (1997). The timing of sexual intercourse among adolescents: Family, peer, and other antecedents. Youth & Society, 29, 54--83.

Miller, K. S., Levin, M. L., Whitaker, D. J., & Xu, X. (1998). Patterns of condom use among adolescents: The impact of mother-adolescent communication. American Journal of Public Health, 88, 1542--1544.

Mueller, K. S., & Powers, W. G. (1990). Parent-child sexual discussion: Perceived communicator style and subsequent behavior Adolescence, 25, 469--482.

Newcomer, S. F., & Baldwin, W. (1992). Demographics of adolescent sexual behavior, contraceptions, pregnancy, and STD's. Journal of school Health, 62, 265--270.

Newcomer, S.F., & Udry, R. (1985). Parent-child communication and adolescent sexual behavior Family Planing Perspectives, 17, 169--174.

Philliber, S. G. (1980). A conceptual framework for population socialization. Population and Environment, 3, 3--9.

Pick, S., & Palos, P. A. (1995). Impact of the family on the sex lives of adolescents. Adolescence, 30, 667--675.

Pistella, C. L., & Bonati, F. A. (1998). Communication about sexual behavior among adolescent wolne0 their family, and peers. Families in Society.: The Journal of Contemporary Human Services, 79, 206--211.

Poppen, P. (1995). Adolescent contraceptive use mid communication: Changes over a decade Adolescence, 29, 503--514.

Rodgers, J.L. (1996). Sexual transitions in adolescence. In J. A. Graber, J. Brooks-Gunn, & A. C. Petersen (Eds.), Transition through adolescence: Interpersonal domains and context (pp. 85--110). Mahwah, NJ: Erlbaum.

Santelli, J. S., & Beilenson, P. (1992). Risk factors for adolescent sexual behavior, fertility, and sexually transmitted diseases. Journal of School Health, 62, 271--279.

Seidman, S. N., Mosher, W. O., & Aral, S. D. (1994). Predictors of high-risk behavior in unmarried American women: Adolescent environment us a risk factor. Journal of Adolescent Health, 15, 126--132.

Somers, C. L., & Paulson, S. E. (2000). Students' perceptions of parent-adolescent closeness and communication about sexuality: Relations with sexual knowledge, attitudes, and behaviors. Journal of Adolescence, 23, 629--644.