Predictors of father-son communication about sexuality.

Teen sexual activity in the United States is a major societal concern. Much of this concern is related to the young age of initiating sexual intercourse and the unsafe nature of many sexual behaviors, particularly those that put teens at risk for unintended pregnancy and sexually transmitted diseases (STDs), including HIV/AIDS. According to The Alan Guttmacher Institute (2003, Beginning section, para. 3), U.S. teens are more likely than teens in other developed countries to have sexual intercourse before age 15 and to have shorter and more sporadic sexual relationships. The Youth Risk Behavior Surveillance System (YRBSS) notes a continuing high incidence of unsafe sexual practices in this age group (Centers for Disease Control and Prevention [CDC], 2002, June 28). As a result of the frequent finding that knowledge alone has not led to a significant decrease in risky sexual behaviors in adolescents, researchers have begun to examine other factors that impact young people's sexual activities. One of these factors is family communication and its relationship to young people's sexual behaviors.

Lefkowitz, Kahlbaugh, Au, and Sigman (1998) acknowledged the importance of parent-adolescent communication in impacting adolescent AIDS-related knowledge, attitudes, and behaviors, because parents are generally available and willing to give their offspring information. Tinsley, Lees, and Sumartojo (2004) recognized the potential of family communication in decreasing HIV risk in youth and in encouraging the development of positive sexuality. Wilson (1994), a Board Member of the Sex Information and Education Council of the United States (SIECUS), noted that parents want to educate their children about sexuality but in many cases do not feel adequately informed to do this. At the same time, adolescents are asking their parents, teachers, and other adults to give them information and guidance in how to protect themselves from negative consequences of sexual behaviors (Billings, 1996).

The complex nature of communication and methods for examining the communication process have presented ongoing challenges to researchers. Fisher (1993) noted the inconsistency of research findings regarding parent-child communication about sexuality, and she stressed the need for instrument development in this area. Raffaelli, Bogenschneider, and Flood (1998) examined previously neglected predictors of parent-teen sexuality communication and suggested that there may be no single set of predictors of this communication. Moreover, Rosenthal and Feldman (1999) pointed out that in addition to the complex nature of communication, the area of sexuality may also consist of a variety of different domains, including physical development, dangers of sexuality, and perhaps psychological or interpersonal issues. Despite the complexity of these issues, there has been continued interest in examining family communication about sexuality as clinicians, educators, researchers and parents have searched for methods to promote and support adolescent health.

Centers for Disease Control and Prevention (CDC) research (2002, March 11) noted the importance of early, clear communication between parents and young people about sex in helping adolescents adopt and maintain protective sexual behaviors. DiIorio, Kelly, and Hockenberry-Eaton (1999) found that adolescents who communicated with mothers about a variety of sexual topics were less likely to have begun having sexual intercourse. Leland and Barth (1993) found that adolescents who reported talking with their parents about sexuality-related topics, including pregnancy, birth control, abstinence, and STD/AIDS prevention, also reported greater use of safer sex behaviors. Dittus, Jaccard, and Gordon (1999) suggested that mothers' direct discussions of sex-related topics with their teens were important in influencing teens' future protective sexual behaviors. In addition, Miller, Levin, Whitaker, and Xu (1998) reported that mother-adolescent discussions about condoms prior to the adolescent's initiation of sexual intercourse were strongly associated with greater condom use over time.

Conversely, other researchers (Newcomer & Udry, 1985; O'Sullivan, Jaramillo, Moreau, & Meyer-Bahlburg, 1999) found that mother-child communication about sex did not affect teenagers' subsequent sexual behavior. Further, Somers and Paulson (2000) found that although greater parental communication was associated with adolescents' greater sexual knowledge, the communication was also associated with more sexual behavior. It was unclear whether the communication preceded or followed the behavior.

Feldman and Rosenthal (2000) noted that although research has produced mixed results about the impact of parent-teen sexuality communication in reducing risky adolescent sexual behaviors, there is a greater likelihood that effective communication will have a more positive influence than will ineffective communication. Several studies regarding parent-child communication about sexuality have shown that there is frequently a lack of communication about many, if not most, aspects of sexuality (Downie & Coates, 1999; Rosenthal & Feldman, 1999). When communication does occur, mothers appear to be the key communicators with their offspring of both genders, but primarily with their daughters (DiIorio, Kelley, et al., 1999; Miller, Kotchick, Dorsey, Forehand, & Ham, 1998; Raffaelli et al., 1998; Rosenthal & Feldman, 1999).

DiIorio, Resnicow, Dudley, et al. (2000) found that mothers who had greater self-efficacy and favorable outcome expectancies associated with talking with their children were more likely to do so. DiIorio et al. (1999) found that adolescents who reported a greater number of topics discussed with mothers were less likely to have initiated sexual intercourse and more likely to have conservative values. Miller, Forehand, and Kotchick (1999) reported that family-structural variables (family income, parental education, and maternal marital status) did not significantly predict adolescent sexual behavior, but family-process variables (maternal monitoring, mother-adolescent general and sexual communication, and maternal attitudes about adolescent sexual behavior) predicted multiple components of adolescent sexual behavior, including decreased risk-taking.

As noted previously, most of the family communication about sexuality seems to occur between mothers and daughters. It appears that sons receive little information from either parent about sexuality and that the information they do receive comes primarily from mothers. Nolin and Peterson's (1992) focus group data suggested that because there is little parental communication with sons about sex, sons may not be aware of family norms and parental expectations regarding sexual behaviors. Consequently, they may be more likely to engage in casual sexual encounters. These researchers concluded that gender differences in family communication about sexuality may occur as a result of a sexual double standard, and they may also perpetuate this standard.

In several instances, recruitment of fathers to participate in studies of father-son communication has been a challenge. For example, Brock and Beazley's study (1995) on parent involvement in an at-home sexuality education program yielded a sample of 176 mothers and only 34 fathers. Rosenthal and Feldman (1999) noted difficulty in recruiting father participants for their study, citing the fathers' frequent comment that their wives were responsible for sex education. Many of the fathers volunteered their wives for the study. Despite recruitment difficulties, Miller, Kotchick, et al. (1998) stressed that fathers should not be ignored in research or in educational programs since they appear to communicate with their children, particularly their sons, about sexual topics (e.g., condom use).

Although most research has focused on mother-child, particularly mother-daughter, communication about sexuality, a growing number of studies has examined the communication contribution of fathers (Brock & Beazley, 1995; Feldman & Rosenthal, 2000; Fisher, 1990; Miller, Kotchick, et al., 1998; Nolin & Peterson, 1992; Somers & Paulson, 2000; Whalen, Henker, Hollingshead, & Burgess, 1996). Rosenthal and Feldman (1999) found that a majority of boys reported that their fathers never discussed 19 of 20 sex-related topics with them. Feldman and Rosenthal (2000) found that teens' evaluations of their fathers as sex educators were influenced by the fathers' good general communication. These researchers suggested helping parents to improve their communication skills in order to enhance their effectiveness as sex educators.

Fisher (1990) found three major predictors of fathers talking with their adolescents about sexuality: (a) openness in general family communication, (b) the extent to which their own fathers discussed sexuality with them, and (c) less education. In addition, Raffaelli et al. (1998) stated that open communication appeared to be an important variable in promoting parent-adolescent sexuality discussions. Whalen et al. (1996), in their direct observations of parent-adolescent communication about sexuality, found that even though both parents had give-and-take conversations with daughters and were more directive toward sons, fathers appeared to be as competent as mothers when they were given a prescribed vehicle and directions for discussion.

Other factors that generate parent-child communication about sexuality are the age and pubertal development of the child and parental concerns about budding sexuality (Raffaelli et al., 1998). According to Somers and Paulson (2000), "perhaps the single most important reason that teens have sex is because it is a natural drive that manifests during and after biological and cognitive maturation" (p. 640). Miller, Norton, Fan, and Christopherson (1998) found that pubertal development relative to peers was directly related to sexual behaviors in both genders. For early maturing children, the physical manifestations of puberty, rather than chronological age, may alert parents to the prudence of talking with their children about sex. Fathers have an important role at this developmental junction, and Miller, Kotchick, et al. (1998) stressed the importance of including fathers in research and programming.

In summary, parents recognize the need to communicate with their children about sexual topics. However, many parents feel inadequate as sexual educators and thus avoid these conversations. Although there are conflicting findings regarding the value of parent-teen communication in reducing sexual activity in adolescents, there is strong support that communication increases safer sex behaviors. Current studies reveal that mothers participate more in communicating about sexual issues than do fathers. The role of fathers and their contribution to their sons' development of sexual knowledge, attitudes, and behaviors warrants further research. The literature supports the relationship of fathers' general communication with their sons, fathers' communication with their fathers about sexuality, fathers' educational levels, and sons' physical maturity to father-son communication about sexuality. Thus, the current study will investigate whether these variables and others are predictors of amount of father-son communication about sexuality.

Theoretical Framework

Bandura's (1986) social cognitive theory guided the study. Researchers in a number of social sciences and health related disciplines have extensively studied two concepts of this theory: self-efficacy and outcome expectations. Self-efficacy is defined as an individual's judgment of his or her ability to accomplish a certain level of performance (Bandura). For example, in communicating about sexuality, the father may believe that he can explain to his son (a) the reasons he should wait until he is older to have sexual intercourse and (b) the importance of using condoms if he decides to have sexual intercourse.

In contrast, an outcome expectation is the perception of the consequence of an act, not the act itself (Bandura, 1986). With regard to sexuality communication, a father's outcome expectations could be that (a) he would feel like a responsible parent and (b) his son would be less likely to have sexual intercourse as a young teen. Self-efficacy and outcome expectations work together to determine behavior (Bandura). Individuals may believe that certain actions will produce positive outcomes, but they may not act on the outcome beliefs because they question their ability to perform the necessary actions.

Combining the variables identified in the literature review with the self-efficacy and outcome expectations variables from Bandura's theory provides a better understanding of predictors of fathers' communication with their sons. Knowledge of these predictors may be useful in developing programs to promote father-son communication.

Purpose and Rationale

The purpose of this study was to identify variables that influence father-son communication about sexuality by testing predictors of this communication. Based on the review of the literature and Bandura's (1986) theory, we proposed three hypotheses:

1) Son's pubertal development would be a positive predictor of amount of father-son sex-based communication.

2) Father's sex-based communication with his father, openness of father's sexual communication with his father, self-efficacy for communicating about sex, outcome expectations for communicating about sex, frequency of father-son contact, and general father-son communication would be positive predictors of amount of father-son sex-based communication.

3) Father's knowledge of AIDS and safer sex and father's permissive sex-based values would be positive predictors, and father's education and age at first intercourse would be negative predictors of amount of father-son sex-based communication.

Figure 1 presents a diagram of relationships explored.

[FIGURE 1 OMITTED]

METHOD

Description of Participants

The target population was fathers who were involved in various community and workplace groups and activities, as well as fathers whose addresses were on community and professional association membership lists. Participants had to have one or more sons between the ages of 11 and 17. We asked fathers to respond based on experience with their oldest son who was currently in the target age group.

A majority of fathers were biological fathers (94%), and 90% of fathers lived with their sons. Also, most (85%) fathers were currently living with their sons' mothers. A majority (80%) of fathers were White, 29% had household incomes ranging from $50,000 to $99,999, and 53% had household incomes over $100,000. In addition, 77% were college graduates or held postgraduate degrees. See Table 1 for participant demographics.

Procedure

We recruited participants through face-to-face contact and mailings. In the face-to-face contact method, we went to selected university classes and community groups, presented verbal information about the study, and distributed survey packets to potential participants or their acquaintances. University classes included those in which the class faculty had given us permission to announce the study. Community groups included little league and high school football teams and residents of a local county who were assembled for jury duty. Also, in the face-to-face method, the researchers' friends and professional colleagues distributed survey packets to their relatives, colleagues, and neighborhood contacts.

In the mailing procedure, we sent survey packets to fathers whose addresses were on selected community and professional association membership lists. These membership lists consisted of a local Boy Scout troop, a church youth group, the neighborhood association of a professional colleague, and the advanced practice psychiatric nurses' association.

Survey packets contained a letter introducing the researchers and the study, the survey instrument (a professionally prepared booklet), a small No. 2 pencil, and a postage paid return envelope. The introductory letter invited fathers who had sons between the ages of 11 and 17 to be a part of a study of father-son communication about sexuality. The letter stressed anonymity, confidentiality, and the voluntary nature of participation and provided contact information for the primary researcher. The cover of the survey booklet noted the inclusion of sensitive topics, asked for frank responses, and thanked respondents for their thoughtful answers. Instructions stated, "Please respond based on your experience with your son who is currently in the age group 11-17 that you see most frequently. If you see two or more sons (of this age) the same amount, select the oldest son."

The amount of time necessary to complete the survey was approximately 20 minutes. We distributed 426 surveys through face-to-face contact or mailing, and 177 completed surveys were returned by mail, yielding a response rate of 42%. Twenty-two questionnaires were omitted from the data analysis because of missing data, resulting in a sample size of 155.

Instrument

The study instrument, Father-Son Communication about Sexuality, was a questionnaire composed of 114 items that yielded 12 variables. Two measured self-efficacy and outcome expectations, principal concepts of social cognitive theory.

The Two Factor Sex-Based Communication Scale (SBCS). Sixteen items of the questionnaire measure how much fathers talk about specific topics with their sons. Each item begins with the stem, "While your son has been growing up, how much have you talked to him about ...?" Items are rated from 1 (not at all) to 5 (a lot). Items are summed, and this sum is divided by 16, yielding a mean score range from 1 to 5. Higher mean scores indicate a greater amount of sex-based communication. This scale, developed by DiIorio et al. (2000) for a study with mothers and adolescents, demonstrated construct validity and an alpha value of .90 for mother-adolescent discussions. Using the data collected from this sample, we conducted a factor analysis of these 16 items to determine if more than one construct was measured by the original 16 items of the SBCS scale. The factor analysis yielded two communication scales: Information Sharing (6 items) and Values Sharing (4 items). We used these 10 items as two separate scales in place of the original 16-item SBCS scale. The Information Sharing scale was composed of items that measured factual information related to sexuality. An example is, "While your son has been growing up, how much have you talked to him about how to prevent pregnancy?" The Values Sharing scale was made up of items that measured parental values related to teenage relationships and sexual behaviors. An example is, "While your son has been growing up, how much have you talked to him about what you think about teenagers having sexual intercourse?" These two scales became the dependent variables in the analysis described below. For the current study sample, the Cronbach's alpha coefficients for responses to the Information Sharing scale and the Values Sharing scale were .91 and .80, respectively.

Self-Efficacy for Communicating about Sex (SECS). SECS is a 14-item scale that measures fathers' perception of their ability to explain sex-related topics to their sons. The original scale of 16 items was developed by DiIorio, Dudley, Wang, Wasserman, Eichler, Belcher, et al. (2001) for a mother-adolescent study. We deleted two items pertaining to mother-daughter topics from the present scale. Each of the items begins with the stem, "I can always explain to my son.... "Two item examples are, "... why he should wait until he is older to have sexual intercourse" and "... that he should use condoms if he decides to have sexual intercourse." Responses to the items range from 1 (not sure at all) to 5 (completely sure). Items are summed, and this sum is divided by 14, yielding a mean score range from 1 to 5. Higher mean scores indicate greater self-efficacy for talking with sons about sex. For responses from mothers, the alpha coefficient for the scale was .85 (DiIorio et al., 2001), and in the current study, the alpha coefficient was .92.

Outcome Expectations for Communicating about Sex (OECS). OECS is a 14-item scale that measures fathers' perceived consequences of communicating with their sons about sex. The original scale of 15 items was developed by DiIorio et al. (2001) for a mother-adolescent study. We deleted one item pertaining to a mother-daughter topic from the present scale. The scale begins with the question, "What are your expectations about talking with your son? If you talk with your son about sexual topics (such as waiting to have sexual intercourse until he is older, birth control, physical changes of puberty, wet dreams, HIV/AIDS, etc.)." Example items are "... I will feel like a responsible parent," and "... my son will be less likely to have sexual intercourse as a young teen." Items are rated on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree). Items are summed, and this sum is divided by 14, yielding a mean score range from 1 to 5. Higher mean scores indicate greater outcome expectations for talking with sons about sex. The scale demonstrated an alpha coefficient of .83 for responses from mothers (DiIorio et al., 2001) and an alpha coefficient of .83 in the current study.

General Father-Son Communication (GFSC). GFSC, a 17-item scale, measures fathers' perception of their general communication with their sons. The scale was developed by DiIorio, Resnicow, Thomas, et al. (2002) for a mother-adolescent study. The DiIorio et al. (2002) scale was based on the work of Armsden and Greenberg (1987) and Barnes and Olson (1985). The scale begins with the question, "How do you think you and your son communicate with each other?" Two examples are, "When your son asks questions, he gets honest answers from you" and "You help your son to understand himself better." Items are rated from 1 (never true) to 5 (always true). Items are summed, and this sum is divided by 17, yielding a mean score range from 1 to 5. Higher mean scores reflect a greater amount of general communication. The alpha coefficient for mothers in DiIorio et al.'s study (2002) was .90, and in the current study the alpha coefficient was .83.

Father's Sex-Based Communication With His Father (FSBCWF). FSBCWF is a 16-item scale that measures how much fathers talked about specific topics with their fathers. Each item begins with the stem, "When you were growing up, how much did your father talk to you about ..." Two examples are "... what the act of sexual intercourse is?" and "... dangers of having many sex partners?" Items are rated from 1 (not at all) to 5 (a lot). Items are summed, and this sum is divided by 16, yielding a mean score range from 1 to 5. Higher mean scores indicate a greater amount of sex-based communication. This scale, developed by DiIorio et al. (2000), demonstrated construct validity and an alpha value of .90 for mother-adolescent discussions. The alpha coefficient for the current study was .94.

Openness of Sexual Communication Scale (OSCS). OSCS is a 4-item scale developed by Lehr, DiIorio, Dudley, and Lipana (2000) that measures father's perceived openness with his father about sexual topics. Each item begins with the stem, "When I was growing up, my father ..." Two examples are, "... openly discussed sex with me and my siblings" and "... would answer any question I asked about sex." Items are rated from 1 (strongly disagree) to 5 (strongly agree). Items are summed, and this sum is divided by 4, yielding a mean score range from 1 to 5. A higher score corresponds to more perceived openness of communication. The scale demonstrated construct validity and an alpha value of .85 for discussions with fathers in a retrospective study with college students (Lehr et al., 2000) and an alpha value of .79 in the current study.

Knowledge of AIDS and Safer Sex (KASS). KASS was measured by a 9-item true/false scale developed for a father-son HIV prevention intervention study. Items are scored as 0 (incorrect) and 1 (correct). Correct answers are summed, and the scale score can range from 0 to 9.

Permissive Sex-Based Values (PSBV). PSBV is an 8-item scale expanded from a scale developed by DiIorio et al. (1999) assessing the extent to which fathers believe it is all right or not all right for teenagers to have sexual intercourse. Two examples are, "It is ok for teenagers to have sexual intercourse if they protect themselves by using condoms" and "Having sexual intercourse is a normal part of growing up for teenagers." Items are rated from 1 (strongly disagree) to 5 (strongly agree). Items are summed, and this sum is divided by 8, yielding a mean score range from 1 to 5. Higher mean scores indicate more permissive attitudes toward teen sex. The alpha coefficient for the scale in DiIorio et al.'s study (1999) was .80, and in the current study the alpha value was .87.

Father's Age at First Intercourse (FAFI). FAFI was assessed by the father's response to the item, "At what age did you first have sex (sexual intercourse)?"

Pubertal Development Scale (PDS). The father's perception of his son's physical development was measured by a 5-item scale, the Pubertal Development Scale (Petersen, Crockett, Richards, & Boxer, 1988). Two examples are, "How would you describe your son's 'growth spurt' (sudden increase in height)?" and "Have you noticed a change in your son's voice?" Responses to the items range from 1 (has not started) to 4 (is completed). Items are summed, and this sum is divided by 5, yielding a mean score range from 1 to 4. Higher scores indicate more physical development. Originally developed by Petersen et al. (1988) as a self-report measure of pubertal status for adolescents, the scale showed alpha coefficients ranging from .68 for a group of seventh grade boys to .83 for a group of eighth grade girls (Petersen). In the current study, the alpha value was .90.

Frequency of Father-Son Contact (FFSC). We created this variable from a question that asked the fathers how many days in the previous month they had talked with or seen their sons. A large percentage (70%) reported 30 days, and another 17% reported 25 days. This variable was recoded as a dichotomous variable where those who reported seeing their sons less than daily were coded 0. Those who reported 30 days were coded as 1.

DATA ANALYSIS

Prior to data analysis, we tested all variables for normality. Three variables--father's sex-based communication with his father, openness of father's communication with his father, and self-efficacy--were transformed because their distributions were skewed (Tabachnick & Fidell, 1996). The two father's communication variables were transformed using the log transformation. The variable self-efficacy was transformed through reflection and the log transformation. The interpretation of self-efficacy is therefore reversed. A lower transformed self-efficacy value represents higher self-efficacy for talking with sons about sex. These transformed variables are used in the correlation matrix and the regression analyses. Most of the means were in the mid range, with knowledge very high, self-efficacy somewhat high, and father's communication openness with his father somewhat low.

We conducted two hierarchical regression analyses to examine the proposed predictors of sex-based communication. The two Sex-Based Communication Scales, Information Sharing and Values Sharing, were dependent variables in the analyses. We entered the predictor variables as Blocks determined by the model in Figure 1. We entered the son variable, pubertal development, first because of the recognized importance of this factor in the initiation of sexual activity (Miller, Norton, et al., 1998) and our desire to control for this factor. We entered father's characteristics second in order to control for father's background variables. Because the primary research question related to communication, we entered aspects of communication as the third and last block in order to see if these variables explained a significant amount of variance over and above that explained by the background characteristics of sons and fathers. Both regressions were statistically significant (F [11,143] = 8.85,p < .01 and F [11,143] = 4.59,p < .01 for Information Sharing and Values Sharing, respectively).

RESULTS

Means, standard deviations, and reliability coefficients for the variables are presented in Table 2.

Table 3 shows the intercorrelations of the 13 variables. Five of the independent variables were significantly (ranging from low to moderate) correlated with the Information Sharing scale of Sex-Based Communication, and six were significantly (low to moderate) correlated with the Values Sharing scale of Sex-Based Communication.

Tables 4 and 5 present the results of the regression analyses. Values for all variables are included. Son's pubertal development; father's permissiveness and level of education; and father's communication with his father, outcome expectations, and general communication were all statistically significant variables in the equation predicting information sharing communication. The variables in the model together explained 36% (adjusted) of the variance in information sharing. The addition of father's characteristics in Block 2 and communication variables in Block 3 each accounted for a significant increase in the percent of variance explained. A trimmed model was run with only the significant variables entered, and [R.sup.2] changes remained largely the same, that is, .13 for Block 1, .09 for Block 2, and .16 for Block 3.

Values sharing was predicted by son's pubertal development. The addition of father's characteristics did not explain a significant percent of the variance. Finally, outcome expectations, general communication, and the amount of father-son contact were all significant variables in the equation predicting values sharing communication. The variables in the model together explained 20% (adjusted) of the variance in values sharing. The addition of communication variables to the son's pubertal development accounted for a significant increase in the percent of variance explained. Again, when a trimmed model was run, [R.sup.2] changes remained essentially the same, that is, .05 for Block 1 and .17 for Block 3.

In summary, we found seven of the hypothesized predictor variables to be predictors of either information sharing, values sharing, or both factors of amount of father-son sex-based communication. Study results did not support other hypothesized predictors of this communication, such as father's knowledge of AIDS and safer sex, father's age at first intercourse, openness of sexual communication between fathers and their fathers, or self-efficacy for communicating about sex.

Examination of responses to specific items in the Information Sharing and Values Sharing scales revealed interesting aspects of amount of father-son sex-based communication (see Table 6). With regard to the six information topics, fathers talked most with sons about AIDS and other sexually transmitted diseases and somewhat less about pregnancy prevention and condom use. Few fathers talked much about wet dreams and spontaneous erections, and over half of respondents said they had not discussed these topics with their sons at all. Of the four values items, fathers were most likely to report giving their sons advice about personal problems. What fathers think about teenagers having sex was also discussed a moderate amount. Topics that were discussed less often were when to have sexual intercourse and what mothers think about teenagers having sex. It appears that topics that are less personal and perhaps less emotionally charged are topics that fathers are more likely to address. Of the 10 items comprising the two scales, the item that was reported the most was giving the son advice on his personal problems.

DISCUSSION

Discussion of Hypotheses

The purpose of this study was to identify variables that influence father-son communication about sexuality. The current study differentiated between information sharing and values sharing aspects of this communication. As hypothesized, son's pubertal development was a significant predictor of both information sharing and, to a lesser extent, values sharing, with fathers more likely to talk with sons who had attained more physical development. The father's recognition of his son's physical development appears to be an important factor in talking about sexuality. When fathers see their sons maturing physically, they may become aware of the increased possibility of sexual initiation, and this possibility spurs them to discussion of sexual topics.

Both Fisher (1990) and DiIorio et al. (1999) found that parents' sex-based values were not a predictor of sex-based communication with adolescents. In the current study, however, permissive values were a predictor of one factor of sex-based communication, information sharing, but not the other factor, values sharing. Perhaps fathers who are more permissive are comfortable discussing the informational aspects of sexuality and do communicate information to their sons. At the same time, fathers may consider their values to be self-evident and not necessary to discuss. Another possible reason that values sharing was not predicted by permissiveness is that talking about one's values might be difficult and emotionally laden. To share values might lead to a son's questions about the father's own sexual behaviors, behaviors that the father may be reluctant to discuss.

With regard to education, Fox (1981) reported that previous research showed mothers with less education were more likely to talk with children about sex, and Fisher (1990) found that fathers with less education reported more communication with their sons about sex. The current study supports these findings. Also in this study, less education in the father was significantly correlated with younger age of father's first intercourse. This being so, perhaps fathers with less education may be more aware that young teens do, in fact, have sex, and may see more of a need to share information. At the same time, more educated parents might assume that their teens are engaged in intellectual activities and preoccupied with educational pursuits.

The majority of variables that emerged as significant predictors of both information sharing and values sharing factors of father-son communication about sexuality were related to variables included in the Aspects of Communication Block. Significant predictors for information sharing were father's sex-based communication with his father, outcome expectations, and general communication. Communication predictors for values sharing were frequency of father-son contact, outcome expectations, and general communication. Outcome expectations and general communication were predictors of both factors.

The importance of general communication and father's communication with his father in predicting father-son sex-based communication supports the findings of Fisher (1990), who reported that two of the three major predictors of father-son communication about sexuality were openness in general family communication and the amount that father's own father discussed sexuality with him. Also, general communication was moderately correlated with both self-efficacy for communicating about sex (r = .45) and outcome expectations for communicating about sex (r = .52), lending further support to the importance of general communication in promoting communication about sexuality in families.

The frequency of father-son contact in the study sample was high (average days of contact in past month = 27.7). This variable did not predict information sharing, but it did predict values sharing. Fathers who had less contact with their sons were more apt to share their values. Perhaps fathers who spend less time with their sons feel more of a need to impart values in a direct way than do fathers who spend greater amounts of time with their sons, while fathers who have frequent contact may consider their values to be self-evident.

Bandura (1986) stated that knowledge and cognitive skills are essential for skilled performance. The finding that knowledge of AIDS and safer sex was not a predictor of amount of father-son sex-based communication supports previous research findings and statements by noted authorities that although knowledge is necessary, it is not sufficient to influence behavior (Bandura, 1986; Hedgepeth & Helmich, 1997; Wulfert & Wan, 1993). Interestingly, in the current study, the mean score for fathers' knowledge on the Knowledge of AIDS and Safer Sex scale was 8.05 of 9, and the standard deviation was .99, indicating a high level of knowledge about the items on this particular scale. The lack of variability on this scale detracts from its value as a predictor and may account for its failure in the current study.

We had expected to find that father's age at first intercourse would predict father-son sex-based communication. The rationale was that fathers who were sexually active at an earlier age would remember their experiences and would see their sons as needing information. Data did not support this hypothesis. A possible explanation is that fathers who had sexual intercourse at a later age may believe it is best to wait, and they may talk with sons to instill this same value, while fathers who had sex at an early age might believe it best to inform their sons about sex in order to prepare them for it. These opposite trends would cancel each other or render the relationship curvilinear. Either scenario would attenuate any observed correlation between fathers' age at first intercourse and communication with their sons about sex. One father in the study commented that he had had sex early (age 16) and had found the experience both positive and negative; he now finds himself ambivalent about the prohibition of sex before marriage, but thinks that teenagers are too young emotionally to engage in sex. This statement illustrates the challenges posed by attempts to examine the relationship between fathers' age of sexual initiation and subsequent sex-based communication with their sons.

Theoretical Implications

Outcome expectations, one of the two most studied concepts of social cognitive theory, was also a significant predictor of sex-based communication. Social cognitive theory posits that self-efficacy and outcome expectations are both important determinants of health-promoting behaviors, and studies related to safer sex behaviors in adolescents have supported this position (DiIorio et al., 2000; Jemmott & Jemmott, 1992). Despite the recognized importance of both concepts, Bandura (1995), the developer of the theory, believes that self-efficacy predicts performance better than outcome expectations. In the current study, however, self-efficacy did not emerge as a predictor of father-son sex-based communication.

A close examination of self-efficacy found that the group of respondents was very homogeneous in replies given on this scale, with 70% choosing 4 or 5 (on a 5 point scale), indicating fathers' high levels of confidence about their ability to explain the items to their sons. Only 11% of respondents indicated they were not sure of their ability to explain the items. It was noted that self-efficacy and outcome expectations were correlated (-.50). Because the self-efficacy variable was transformed prior to data analysis, the -.50 correlation means that increases in self-efficacy were associated with increases in outcome expectations. Further investigation revealed that when outcome expectations were omitted from the regression analyses, self-efficacy still did not emerge as a predictor of amount of information sharing or values sharing father-son sex-based communication. It appears that the range of difficulty for the self-efficacy scale used in this study did not capture enough variability in the sample to be a useful tool with this particular group of respondents. Since the current study sample differs from others in the literature, further research is needed to understand the relationship between self-efficacy and outcome expectations and to determine whether or not the relationship varies for this particular group.

In the present study, perhaps outcome expectations emerged as a significant predictor because the father expects that his communicating with his son will enhance his own feelings of being a responsible parent and, more importantly, that this communication will have a positive impact on his son's life. In the current era of life-threatening STDs, the consequences of not acting could be serious. Fathers' beliefs about the importance of acting to ensure sons' safety, health, and well-being appear to be a key element in fostering information sharing and values sharing sex-based communication.

Limitations

Limitations of the study include the use of a convenience sample of mostly White (80.0%), high household income (53% > $100,000) fathers with college or postgraduate degrees (77%), limiting generalizability and interpretation of findings to other groups. In addition, 89% of boys were living in two-parent households with their mothers and fathers or stepfathers. This number contrasts with 2000 census figures that report only 69% of children living in two-parent households (Fields & Casper, 2001). This figure makes the sample non-representative and suggests that the study subjects may have more frequent contact than fathers and sons in the general population. Also, three of the measures (two of the scales regarding father's communication with his father and the variable of father's age at first intercourse) called for retrospective reporting, which is subject to error. In addition, we did not obtain information about other elements of communication (for example, depth, frequency, and length of conversations) and communication style (for example, father-initiated vs. son-initiated and didactic vs. interactive). Finally, several of the variables appeared to have ceiling effects, limiting their ability to predict the outcome variable. For example, the variables of self-efficacy and father's knowledge had little variance, with many fathers answering at very high levels. Additional work needs to be done with a more heterogeneous sample of respondents.

Recommendations for Future Research

Future studies should examine father-son communication factors such as types of communication (open vs. closed, verbal vs. non-verbal, direct vs. indirect, negative vs. positive), initiation of sexuality-related conversations when sons are young (prior to puberty), and methods of increasing general father-son communication as well as sex-based communication. The current study focused on sex-based communication as the outcome variable, but future studies should address actual safer sex behaviors and the impact of father-son communication on these behaviors.

In addition, future research should examine other father variables that might impact father-son communication, such as father's awareness of teachable moments for broaching the subject of sex and father's willingness and motivation to communicate about sexual topics. One father in the study stated that he had difficulty knowing when to initiate a conversation with his son, and another father commented, "I think it would be a good idea to address 'willingness' as well as 'ability.' I have the ability to talk to my son about condoms or AIDS and sex, but have low willingness or motivation to do so." Social cognitive theory would be useful in researching this area since motivation is another of the concepts included in the theory.

There may be great diversity in father-son communication about sexuality across socioeconomic and cultural groups. Future research should be conducted with populations of diverse racial, economic, and educational backgrounds in order to determine the factors that are involved in father-son communication in these groups. Much valuable information was elicited from comments on the father surveys. Qualitative research methods, including the use of father focus groups, could provide additional data for researchers, educators, and health care providers to use in developing programs to enhance the fathers' role in the sexual knowledge, attitudes, and behavior of their sons. Lastly, the continued development of research instruments is essential in order to address more fully the exceedingly broad construct of father-son communication about sexuality.

Conclusion

Since several communication factors were predictors of amount of father-son sex-based communication, the importance of increasing communication now has implications for today's young people as well as for future generations. In the struggle to ensure the health and well-being of our youth, it is crucial to approach health promotion efforts from many fronts, and parents are generally a ready and willing resource for their children. Study findings support the involvement of fathers in the sexual education of their sons and point to the need for development of programs to facilitate both general and sex-based communication between fathers and sons.

Table 1. Demographic Characteristics of Participants (n = 155)

Characteristic                              %      M     Range   SD

Son's age                                         14.2   11-17   2.0
Frequency of contact (days)                       27.7    3-30   5.0
Age father had sex                                18.1    8-33   3.1
Father's age                                      45.7   28-64   5.8
Father's relationship to son
  Biological father                        94.2
  Stepfather                                5.8
Where son lives
  With father                              90.3
  With family member                        5.2
  Other                                     3.2
  Missing responses                         1.3
Father's living situation
  With son's mother                        85.2
  With stepson's mother                     3.8
  Single                                   11.0
Father's race
  White                                    80.0
  Black                                    15.5
  Other                                     3.8
  Missing responses                         0.6
Father's annual household income
  Under $30,000                             7.1
  $30,000-$49,999                          10.3
  $50,000-$99,999                          29.0
  Over $100,000                            52.9
  Missing responses                         0.6
Father's level of education
  High school graduate                      5.8
  Some college/technical school graduate   16.8
  College graduate                         41.3
  Postgraduate degree                      36.1

Table 2. Means, Standard Deviations, and Alpha
Coefficients of Variables

                                        Poten-
                                         tial     Actual     Cronbach's
Variable                   M      SD    Range      Range       Alpha

Sex-Based Communication
  Values Sharing          2.92    .96    1-5     1.00-5.00      .80
  Information Sharing     2.48   1.07    1-5     1.00-5.00      .91
Self Efficacy             4.20    .69    1-5     2.14-5.00      .92
Outcome Expectations      3.75    .54    1-5     1.93-5.00      .83
General Communication     3.78    .40    1-5     2.82-4.76      .83
Father's Communication
  with His Father         1.64    .68    1-5     1.00-4.38      .93
Openness of Sexual
  Communication
  Between Fathers
  & Their Fathers         1.96    .81    1-5     1.00-5.00      .79
Father's Permissive
  Values                  2.20    .81    1-5     1.00-4.75      .87
Son's Physical
  Development             2.57    .78    1-4     1.00-4.00      .90
Father's Knowledge        8.05    .99    0-9     4.00-9.00

Table 3. Correlations Among Variables of Interest

    Scale                             1     2     3     4        5

 1. Information Sharing               --   .63   .01    .20     -.11
 2. Values Sharing                         --    .02   -.09      .02
 3. Knowledge of AIDS & Safer Sex                --    -.18 *    .15
 4. Permissive Sex-Based Values                         --      -.25 **
 5. Father's Age at First
    Intercourse                                                  --
 6. Father's Education
 7. Father's Sex-Based
    Communication
    with His Father
 8. Openness of Father's Sexual
    Communication with His Father
 9. Self Efficacy for Communicating
    about Sex
10. Outcome Expectations for
    Communicating about Sex
11. Frequency of Father-Son Contact
12. General Father-Son
    Communication
13. Pubertal Development

    Scale                              6        7         8

 1. Information Sharing               -.21 *    .27 **    .15
 2. Values Sharing                    -.10      .23 **    .16 *
 3. Knowledge of AIDS & Safer Sex      .18 *   -.08      -.03
 4. Permissive Sex-Based Values       -.11     -.09      -.02
 5. Father's Age at First
    Intercourse                        .35 *   -.05      -.02
 6. Father's Education                         -.20 *    -.10
 7. Father's Sex-Based
    Communication
    with His Father                    --                 .66 **
 8. Openness of Father's Sexual
    Communication with His Father               --
 9. Self Efficacy for Communicating
    about Sex                                             --
10. Outcome Expectations for
    Communicating about Sex
11. Frequency of Father-Son Contact
12. General Father-Son
    Communication
13. Pubertal Development

    Scale                              9        10        11

 1. Information Sharing               -.08      .30 **   -.08
 2. Values Sharing                    -.11      .31 **   -.19 *
 3. Knowledge of AIDS & Safer Sex     -.04     -.02       .04
 4. Permissive Sex-Based Values        .18 *   -.09      -.19 *
 5. Father's Age at First
    Intercourse                        .03     -.14       .00
 6. Father's Education                -.15     -.09       .11
 7. Father's Sex-Based
    Communication
    with His Father                    .04      .09      -.12
 8. Openness of Father's Sexual
    Communication with His Father      .06      .08      -.02
 9. Self Efficacy for Communicating
    about Sex                                  -.50 **   -.13
10. Outcome Expectations for
    Communicating about Sex            --                 .06
11. Frequency of Father-Son Contact             --
12. General Father-Son
    Communication
13. Pubertal Development

    Scale                              12        13

 1. Information Sharing                .22 **    .36 **
 2. Values Sharing                     .26 **    .21 **
 3. Knowledge of AIDS & Safer Sex     -.04       .02
 4. Permissive Sex-Based Values       -.08       .04
 5. Father's Age at First
    Intercourse                       -.15       .16 *
 6. Father's Education                 .00       .11
 7. Father's Sex-Based
    Communication
    with His Father                    .07       .01
 8. Openness of Father's Sexual
    Communication with His Father      .09      -.04
 9. Self Efficacy for Communicating
    about Sex                         -.46 **   -.05
10. Outcome Expectations for
    Communicating about Sex            .52 **   -.04
11. Frequency of Father-Son Contact    .12      -.l0
12. General Father-Son
    Communication                               -.13
13. Pubertal Development

* p < .05

** p < .01.

Table 4. Predictors of Sex-Based Communication
(Information Sharing)

        Variable              B     Beta      F

Block 1
  Son's Pubertal
    Development               .56     .41   37.18
Block 2
  Fathers' knowledge of
    AIDS and HIV              .13     .12    3.04
  Fathers' permissiveness     .32     .24   11.39
  Fathers' highest level
    of education             -.21    -.17    5.24
  Age of fathers' first
    sexual activity           .00     .00     .01
Block 3
  Amount of father-son
    contact                   .05     .02     .12
  Fathers' communication
    with their fathers       1.86    0.27    9.11
  Openness of fathers'
    communication with
    their fathers            -.39    -.06     .54
  Self-efficacy               .62     .09    1.20
  Outcome expectations        .50     .25    9.43
  General communication       .49     .19    5.41
  (Constant)                -4.17   10.35    <.01
Model Statistics

                                   [R.sup.2]   Sig. F
        Variable             p      [DELTA]    [DELTA]

Block 1                               .13        .00
  Son's Pubertal
    Development             <.01
Block 2                               .09        .00
  Fathers' knowledge of
    AIDS and HIV             .08
  Fathers' permissiveness   <.01
  Fathers' highest level
    of education             .02
  Age of fathers' first
    sexual activity          .94
Block 3                               .19        .00
  Amount of father-son
    contact                  .73
  Fathers' communication
    with their fathers      <.01
  Openness of fathers'
    communication with
    their fathers            .47
  Self-efficacy              .28
  Outcome expectations      <.01
  General communication      .02
  (Constant)
Model Statistics                       [R.sup.2] = .41
                              Adjusted [R.sup.2] = .36
                      [F.sub.(11, 143)] = 8.85, p <.01

Note. Tabled values for B, Beta, F, and p are from the final model
in which all predictors were entered in the regression equation.

Table 5. Predictors of Sex-Based Communication (Values Sharing)

                                                    [R.sup.2]   Sig. F
    Variable          B      Beta     F       p      [DELTA]    [DELTA]

Block 1                                                .05        .01
  Son's Pubertal
    Development       .29     .24   10.18    <.01
Block 2                                                .01        .36
  Fathers'
    knowledge of
    AIDS and HIV      .05     .05     .42     .52
  Fathers'
    permissive-
    ness             -.10    -.08    1.07     .30
  Fathers'
    highest level
    of education     -.09    -.08     .94     .34
  Age of fathers'
    first sexual
    activity          .02     .06     .48     .49
Block 3                                                .20        .00
  Amount of
    father-son
    contact          -.40    -.19    6.57     .01
  Fathers'
    communication
    with their
    fathers           .80     .13    1.67     .20
  Openness of
    fathers'
    communication
    with their
    fathers           .13     .02     .06     .82
  Self-efficacy       .59     .10    1.11     .29
  Outcome
    expectations      .43     .24    6.80     .01
  General
    communication     .54     .23    6.58     .01
  (Constant)        -1.718           1.773    .19
Model Statistics                                        [R.sup.2] = .26
                                               Adjusted [R.sup.2] = .20
                                       [F.sub.(11, 143)] = 4.59, p <.01

Note. Tabled values for B, Beta, F, and p are from the final model
in which all predictors were entered in the regression equation.

Table 6. Descriptive Statistics for Information Sharing and Values
Sharing Topics Discussed by Fathers

While your son has been growing up, how
much have you talked to him about ...       Scale    N     M      SD

Advice on his personal problems (problems
  with friends, acne, etc.)                  VS     160   3.60   1.10
What you think about teenagers having sex    VS     160   3.01   1.26
Getting AIDS                                 IS     160   2.98   1.51
Sexually transmitted diseases (STDs)         IS     159   2.91   1.30
How to prevent pregnancy                     IS     160   2.73   1.34
When to have sexual intercourse              VS     160   2.64   1.23
Using a condom during sex                    IS     160   2.62   1.44
What his mother thinks about teenagers
  having sex                                 VS     160   2.49   1.37
Spontaneous erections                        IS     159   1.87   1.11
Wet dreams                                   IS     160   1.83   1.08

Note. VS=Values Sharing; IS=Information Sharing.

REFERENCES

Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer attachment: Individual differences and their relationship to psychological well-being in adolescence. Journal of Youth & Adolescence, 16, 427-454.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1995). Exercise of personal and collective efficacy in changing societies. In A. Bandura (Ed.), Self-efficacy in changing societies. Cambridge: Cambridge University Press.

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

Billings, J. A. (1996). Listen to the plea from students about HIV/AIDS: "Educate us." AIDS Patient Care and STDs, 10, 180-181.

Brock, G. C., & Beazley, R. P. (1995). Using the Health Belief Model to explain parents' participation in adolescents' at-home sexuality education activities. Journal of School Health, 65, 124-128.

Centers for Disease Control and Prevention. (2002). Young people at risk: HIV/AIDS among America's youth. Retrieved March 11, 2002, from http://www.cdc.gov/hiv/pubs/facts/youth.htm.

Centers for Disease Control and Prevention. (2002). Youth Risk Behavior Surveillance Summaries--United States, 2001. Retrieved June 28, 2002, from http://www.cdc.gov/mmwr/preview/mmwrhtml/ss5104al.htm.

DiIorio, C., Dudley, W. N., Wang, D. T., Wasserman, J., Eichler, M., Belcher, L., et al. (2001). Measurement of parenting self-efficacy and outcome expectancy related to discussions about sex. Journal of Nursing Measurement, 9, 135-149.

DiIorio, C., Kelley, M., & Hockenberry-Eaton, M. (1999). Communication about sexual issues: Mothers, fathers, and friends. Journal of Adolescent Health, 23, 181-189.

DiIorio, C., Resnicow, K., Dudley, W., Thomas, S., Wang, T., Van Marter, D., et al. (2000). Social cognitive factors associated with mother-adolescent communication about sex. Journal of Health Communication, 5, 41-51.

DiIorio, C., Resnicow, K., Thomas, S., Wang, D. T., Dudley, W. N., Van Marter, D. F., et al. (2002). Keepin' it R.E.A.L.!: Program description and results of baseline assessment. Health Education & Behavior, 29, 104-123.

Dittus, P. J., Jaccard, J., & Gordon, V. V. (1999). Direct and nondirect communication of maternal beliefs to adolescents: Adolescent motivations for premarital sexual activity. Journal of Applied Social Psychology, 29, 1927-1963.

Downie, J., & Coates, R. (1999). The impact of gender on parent-child sexuality communication: Has anything changed? Sexual & Marital Therapy, 14, 109-121.

Feldman, S. S. & Rosenthal, D. A. (2000). The effect of communication characteristics on family members' perceptions of parents as sex educators. Journal of Research on Adolescence, 10, 119-150.

Fields, J., & Casper, L. (2001). America's families and living arrangements: March 2000. Current population reports, pp. 20-537. U.S. Census Bureau, Washington, D.C.

Fisher, T. (1990). Characteristics of mothers and fathers who talk to their adolescent children about sexuality. Journal of Psychology & Human Sexuality, 3, 53-70.

Fisher, T. (1993). A comparison of various measures of family sexual communication: Psychometric properties, validity, and behavioral correlates. The Journal of Sex Research, 30, 229-238.

Fox, G. L. (1981). The family's role in adolescent sexual behavior. In T. Ooms (Ed.), Teenage pregnancy in a family context (pp. 73-130). Philadelphia: Temple University Press.

Hedgepeth, E., & Helmich, J. (1997). Teaching about sexuality and HIV: Principles and methods for effective education. New York: New York University Press.

Jemmott, L. S., & Jemmott, J. B. (1992). Increasing condom-use intentions among sexually active black adolescent women. Nursing Research, 41, 273-279.

Lefkowitz, E. S., Kahlbaugh, E, Au, T. K., & Sigman, M. (1998). A longitudinal study of AIDS conversations between mothers and adolescents. AIDS Education and Prevention, 10, 351-365.

Lehr, S., DiIorio, C., Dudley, W. N., & LiPana, J. (2000). The relationship between parent-adolescent communication and safer sex behaviors in college students. Journal of Family_Nursing, 6, 180-196.

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., Norton, M. C., Fan, X., & Christopherson, C. R. (1998). Pubertal development, parental communication, and sexual values in relation to adolescent sexual behaviors. Journal of Early Adolescence, 18, 27-52.

Miller, K. S., Forehand, R., & Kotchick, B. A. (1999). Adolescent sexual behavior in two ethnic minority samples: The role of family variables. Journal of Marriage & the Family, 61, 85-98.

Miller, K. S., Kotchick, B. A., Dorsey, S., Forehand, R., & Ham, A. Y. (1998). Family communication about sex: What are parents saying and are their adolescents listening? Family Planning Perspectives, 30, 218-222, 235.

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.

Newcomer, S. F., & Udry, J. R. (1985). Parent-child communication and adolescent sexual behavior. Family Planning Perspectives, 17, 170-174.

Nolin, M., & Peterson, K. (1992). Gender differences in parent-child communication about sexuality: An exploratory study. Journal of Adolescent Research, 7, 59-79.

O'Sullivan, L. F., Jaramillo, B. M., Moreau, D., & Meyer-Bahlburg, H. F. (1999). Mother-daughter communication about sexuality in a clinical sample of Hispanic adolescent girls. Hispanic Journal of Behavioral Sciences, 21, 447-469.

Petersen, A. C., Crockett, L., Richards, M., & Boxer, A. (1988). A self-report measure of pubertal status: Reliability, validity, and initial norms. Journal of Youth and Adolescence, 17, 117-133.

Raffaelli, M., Bogenschneider, K., & Flood, M. F. (1998). Parent-teen communication about sexual topics. Journal of Family Issues, 19, 315-333.

Rosenthal, D. A., & Feldman, S. S. (1999). The importance of importance: adolescents' perceptions of parental communication about sexuality. Journal of Adolescence, 22, 835-851.

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.

Tabachnick, B., & Fidell, L. (1996). Using multivariate statistics (3rd ed.). New York: Harper Collins Publishers, Inc.

The Alan Guttmacher Institute. (2003). Facts in brief: Teenagers' sexual and reproductive health: Developed countries. Retrieved April 3, 2003, from http://www.AGI-USA.org/pubs/fb_teens.html.

Tinsley, B., Lees, N. B., & Sumartojo, E. (2004). Child and adolescent HIV risk: Familial and cultural perspectives. Journal of Family Psychology, 18, 208-224.

Whalen, C. K., Henker, B., Hollingshead, J., & Burgess, S. (1996). Parent-adolescent dialogues about AIDS. Journal of Family Psychology, 10, 343-357.

Wilson, P. (1994). Forming a partnership between parents and sexuality educators. SIECUS Report, 22(3), 1-5.

Wulfert, E., & Wan, C. K. (1993). Condom use: A self-efficacy model. Health Psychology, 12, 346-353.

Manuscript accepted August 11, 2004

Sally T. Lehr

Emory University

Alice S. Demi

Georgia State University

Colleen DiIorio

Emory University

Jeffrey Facteau

Qwiz, Inc.

Address correspondence to Sally T. Lehr, PhD, RN, CS, FAACS, Nell Hodgson Woodruff School of Nursing, Emory University, 1520 Clifton Rd., Atlanta, GA, 30322; e-mail: [e-mail].