Journal of Pediatric Psychology Advance Access originally published online on February 1, 2006
Journal of Pediatric Psychology 2006 31(9):917-927; doi:10.1093/jpepsy/jsj101
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
An Exploration of Social Cognitive Theory Mediators of Father–Son Communication About Sex
Department of Behavioral Sciences and Health Education, Rollins School of Public Health, Emory University
All correspondence concerning this article should be addressed to Colleen DiIorio, PhD, RN, FAAN, Department of Behavioral Sciences and Health Education, Rollins School of Public Health, Emory University, Room 262, 1520 Clifton Road, N.E., Atlanta, Georgia 30322. E-mail: cdiiori{at}sph.emory.edu.
Received October 3, 2005; revisions received December 28, 2005 and January 4, 2006; accepted January 4, 2006
 |
Abstract |
|---|
 Top Abstract Theoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
Objectives R.E.A.L. MEN (responsible, empowered, aware, living) is an HIV prevention project developed for fathers and their sons. The primary aim of R.E.A.L. MEN was to test the efficacy of an intervention to promote delay of sexual intercourse among 11- to 14-year-old adolescents and to enhance the father’s role as a sexuality educator. In this article, we report the outcomes related to father–son communication about sex and the results of an exploratory analysis to test the hypothesis that the intervention effect on father–son communication about sex-related topics is mediated by the social cognitive variables of self-efficacy and outcome expectations. Methods Two hundred seventy-seven fathers and their sons enrolled in the study. Fathers randomized to the intervention group attended seven group sessions, and their sons attended the final session with them. Fathers and sons completed baseline and 3-, 6-, and 12-month follow-up assessments. The fathers who completed the 3-month follow-up assessment were included in the mediation analyses described in this article. Results All the participants were male, most were African American (97%), and most fathers and sons reported residing in the same household. The mean age of father participants was 40.5 years; most had graduated from high school, and about 60% reported a yearly income over $30,000. Based on the results of mediation analyses, the data supported a mediation model that suggests that the effect of the intervention on father–son communication was mediated by differences in self-efficacy. A mediation effect for outcome expectations was weak and could not be validated using additional tests. Conclusions The mediation analysis provides some understanding of the role of self-efficacy and outcome expectations in promoting behavioral change. Understanding the active components of interventions can help refine the interventions to eliminate nonfunctioning components and enhance the active components leading to more streamlined programs.
Key words: fatherhood; HIV prevention; parent–adolescent sexual communication.
Parents play an important role in the sexuality education of their children. Through their words and actions, they communicate both information and values to their children. This communication is particularly important for early adolescents who are becoming more aware of themselves as sexual beings. Recognizing the importance of parents as the primary sexuality educators for their children, researchers and health educators have developed and tested a variety of programs to improve parent–adolescent discussions about sexuality and sex-related topics including HIV and sexually transmitted diseases (STDs). In general, the programs have demonstrated an increase in frequency and quality of sex-based discussions between parents and their children (Anderson et al., 1999
; Blake, Simkin, Ledsky, Perkins, & Calabrese, 2001; Crawford et al., 1990; Lefkowitz, Sigman, & Au, 2000; Miller et al., 1993). The evaluation of these programs has focused primarily on communication outcomes, such as frequency, quality, and type of discussions about sexuality. Some investigators have also examined adolescent behaviors such as initiation of sexual intercourse and preventive practices such as contraception (Anderson et al., 1999; Miller et al., 1993). Although most programs were based on theoretical frameworks and included a variety of intervention components, the program developers have not paid much attention to the manner in which the outcomes were achieved. That is, they have not determined whether the intervention components designed to promote change were actually responsible for the changes noted in behavior. As we refine our programs to promote parent–adolescent discussions about sexuality, it is important to understand how intervention components work and which components are most influential in achieving behavior change. By identifying and using active intervention components, we will be able to present more powerful interventions and perhaps reduce delivery time.
A recent study to evaluate an HIV prevention program designed for fathers and their sons afforded us the opportunity to identify intervention components associated with an increase in father’s discussions with their adolescent sons about HIV and other sexuality issues. The objectives of the R.E.A.L. MEN (responsible, empowered, aware, living) program were to increase the fathers’ level of communication with their sons about sexual topics to promote the delay of sexual intercourse among 11- to 14-year-old adolescent males and to promote condom use among sexually active adolescents. Two hundred seventy-seven fathers and their sons enrolled in the study. Fathers attended six program sessions, and sons accompanied their fathers to the final (seventh) session. Adolescent participants enrolled in the intervention condition demonstrated significantly higher rates of abstinence and condom use than those in the control condition, and father participants in the intervention condition reported significantly more discussions about sexuality and greater intentions to discuss than fathers in the control condition (DiIorio, McCarty, Resnicow, Lehr, & Denzmore, 2005).
In this article, we present the findings related to father–son communication about sex and the results of a mediation analysis conducted to determine the extent to which the social cognitive factors of self-efficacy and outcome expectations served to mediate intervention effectiveness in promoting father–son sex-based discussions. We conclude the article with lessons learned.
|
Theoretical Framework |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
The R.E.A.L. MEN program was based primarily on social cognitive theory (SCT). Within SCT, human behavior is conceptualized as the result of interactions among personal factors, environmental factors, and the behavior itself (Bandura, 1997
). Although there are many personal factors that influence behavior, Bandura identified self-efficacy (confidence in performing a behavior), outcome expectations (expectations about the outcomes associated with performance of a behavior), and personal goals related to the behavior as the most salient. Environmental factors that contribute to behavioral enactment include support from others in the form of encouragement and resources. In applying SCT to HIV prevention, Bandura (1992) noted that prevention requires people to control their own sexual behaviors. To do so, they must develop confidence to choose safe rather than nonsafe sexual practices and to expect positive outcomes associated with safer behaviors. HIV and sexually transmitted infection (STI) prevention programs based on SCT should be behavior-specific and provide people with information on safe and risky sexual practices and the consequences of each. The programs should also include skill development in areas, such as setting goals, recognizing stimuli that trigger unsafe behaviors, reinforcing positive behaviors, and negotiating for safer sex (Bandura, 1992).
Although most HIV prevention programs based on SCT are designed for the individual, some programs include support partners such as spouses (El-Bassel et al., 2003). The inclusion of significant others is consistent with the SCT framework as Bandura (1992) argued that knowledge alone is insufficient for people to change behavior. People need opportunities, resources, and guidance from others within one’s social network. Thus, our approach to HIV education for adolescent males was to teach fathers who are important members of an adolescent’s network. Consistent with SCT, we expected that by fostering understanding of HIV and risk-reduction practices among fathers, along with enhancing communication skills, fathers in the intervention group would express more confidence in talking with their sons about sexual issues and more positive outcomes associated with such discussions. These beliefs, in turn, would foster more sex-based discussions with their sons.
The role of self-efficacy and outcome expectations as mediators of behavior change has been examined in many studies. For example, both self-efficacy and beliefs were found to be significant mediators of behavior change in a school-based drug prevention program (Orlando, Ellickson, McCaffrey, & Longshore, 2005), and self-efficacy was found to partially mediate an intervention to promote physical activity among adolescent girls (Dishman et al., 2004). Among adults, elderly participants reporting fewer falls following a Tai Chi intervention also reported higher levels of self-efficacy (Li, Fisher, Harmer, & McAuley, 2005), and both self-efficacy and outcome expectancies were noted to mediate behavior change in a study designed to reduce HIV risk behaviors among adult men and women (The National Institute of Mental Health [NIMH] Multisite HIV Prevention Trial Group, 2001).
|
Methods |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
Intervention Condition
The R.E.A.L. MEN intervention program contents and activities were designed to promote the development of self-efficacy, positive expectations, and intentions to discuss sexual topics with one’s son. To foster self-efficacy, the four sources of self-efficacy—behavioral enactment, vicarious learning, verbal persuasion, and physiological and affective feelings related to behavioral performance—described by Bandura (1997)
were used to guide the selection of program contents and activities. According to Bandura (1997), knowledge about the health behavior is a prerequisite for the performance of behaviors. Thus, the R.E.A.L. MEN program included accurate and current information about topics, such as effective listening and communication skills, adolescent development, puberty, and HIV and STI transmission and prevention practices. Self-efficacy is enhanced when people have the opportunity to perform a behavior. Thus, to promote self-efficacy for discussing sexual topics with their sons, fathers were given take-home activities. The activities began with simple listening and communication exercises related to everyday events and increased in difficulty to include the discussion of sexual topics. Fathers who see others like themselves being able to discuss topics with their sons are likely to believe that they can do so as well. Thus, within the group sessions, there were opportunities to discuss the take-home activities, to share stories about talking with their sons about sexual topics or missed opportunities to do so, and to watch video clips of fathers talking with their sons. Within the group sessions, fathers were able to provide each other with encouragement and support. The group sessions also provided opportunities to practice skills such as communication strategies and discussion of difficult topics thereby increasing the father’s comfort with discussing sexual topics with their sons.
The intervention consisted of seven group sessions each 2 h in length and held once each week between 7 and 9 p.m. Fathers in groups of 10–12 attended the first six sessions alone, and fathers and sons attended the final session together. Each session began with an introductory component that consisted of an overview of the session and an icebreaker. In all but the first session, the introduction was followed by a review of the previous session, a discussion of the take-home activities, and a review of personal goals set by the fathers. The content was delivered using a combination of lectures, discussions, role-plays, games, and videotapes.
Session 1 was begun by introducing the facilitators and participants to one another and providing information about the program, including a preview of the seven sessions and establishing group rules. The remainder of the session was devoted to discussion and skill-building exercises related to effective listening. Session 2 was designed to increase the participants understanding of basic types of communication and develop effective communication skills, and Session 3 was designed to increase the understanding of both physical and emotional issues associated with adolescent development and ways in which fathers could be supportive of their sons. Session 4 began with an activity designed to increase the father’s appreciation of HIV risk and was followed by a discussion of HIV transmission and prevention. The purpose of Session 5 was to increase the father’s awareness of their own sexual values, to foster positive attitudes about talking to their sons about sex and HIV, and to promote the development of skills and confidence to discuss sexual topics with their sons. Session 6 was designed to increase the father’s comfort in talking about difficult topics with his son. Sons were invited to attend the last (seventh) session, which was designed to increase the awareness of both the father and the son of issues that were important to each other, particularly as they relate to peer pressure and parental monitoring. The session ended with a celebration of the completion of the program.
Control Condition
Participants in the control condition received a 7-session nutrition and exercise program (N&E) held once each week for 2 h. Fathers attended the first six sessions alone, and their sons attended the final session with them. Briefly, content included basic information about nutrition and exercise, serving sizes, food labels, screening for diseases, blood pressure and heart rate assessment, simple exercises, food components (fats, carbohydrates, and protein), and supplements (steroids). The content was presented using interactive techniques including preparing foods, reading labels, calculation of food components, and doing simple exercises. Videos and games were also included to enhance the interest of the sessions. The fathers were encouraged to share the information with their sons through the use of take-home activities.
Two facilitators were trained to deliver each curriculum—HIV and N&E. Facilitator manuals were developed and used to guide the training. Practice sessions were videotaped and reviewed by the facilitators to evaluate facilitation skills before the start of the intervention. During the conduction of the study, facilitators met two to four times per month to discuss issues with a co-investigator, to review fidelity to the curriculum, and to watch taped segments to evaluate and improve facilitation skills.
Procedures
Approval for this study was obtained from the researchers’ institutional review board and the Boys & Girls Clubs of Metro Atlanta (BGCMA), our collaborating partner. Seven BGCMA sites were selected for participation in the R.E.A.L. MEN research project. These sites were randomized to the intervention and control group creating a randomized cluster design.
To participate in the study, children and adolescents were required to be a member of BGCMA and between 11 and 14 years of age. They were also required to participate with their father. Because not all interested boys lived with their fathers, stepfathers, and other father figures such as uncles, grandfathers, and family friends were allowed to participate with the mother’s signed consent. In cases where the participating father was a legal guardian, the father provided a written informed consent with written assent being obtained from the adolescent. When the participating father figure was not a legal guardian, the father figure provided written informed consent for himself, and the adolescent’s mother or legal guardian provided written informed consent for the adolescent. Again, written assent was obtained from the adolescent.
Detailed information about recruitment, retention, and data collection activities are described elsewhere (Denzmore, McCarty, & DiIorio, 2005; DiIorio et al., 2005). Fathers and their sons began the study by completing a baseline questionnaire administered by computer using the Questionnaire Development Software (Questionnaire Design Systems [QDS] Software, Version 2.1). Following the programs offered for both the intervention and the control groups, fathers and sons completed three interviews during the following 12 months. The father–son communication outcomes are presented along with the mediation analyses using data from the father’s first follow-up interview, 3 months after baseline.
Measures
The 3-month assessment included items to measure father–son sex-based communication and the mediator variables of self-efficacy and outcome expectations. Based on previous experience with a mother–adolescent intervention study conducted by our group, we decided not to include questions about sex-based communication, self-efficacy, or outcome expectations in the baseline questionnaire. This decision was made to reduce the chance that the questions themselves would provide an intervention effect (DiIorio et al., in press).
The sex-based communication measure consisted of 16 sex-specific topics to which participants responded on a 3-point rating scale with 0 indicating that they had not discussed the topic at all and 3 indicating that it had been discussed a lot (DiIorio et al., 2001). A sample item is "Have you ever talked to your son about how you know if you are ready to have sexual intercourse?" The adolescent’s first name was substituted for the term your son so that the responses were specific to the adolescent enrolled in the study. Positive responses are summed to yield a total score ranging from 0 to 48 with higher scores corresponding to a greater amount of sex-related discussion. Cronbach’s alpha coefficient for the fathers’ responses was .97.
The fathers’ degree of confidence in talking with their sons about sex was measured by the parenting self-efficacy scale (DiIorio et al., 2001). Each of the 17 items is rated using a 7-point response scale anchored with the terms 1 (not sure at all) and 7 (completely sure). A sample item is "You can always explain to [son’s name] how to tell a girl no if he does not want to have sex." Total scores are computed by adding responses to each item. Total scores can range from 17 to 119 with higher scores corresponding to more confidence to discuss sex-related issues with their sons. Cronbach’s alpha coefficient for responses from mothers using a 16-item version in an earlier study was .85 and from father participants in this study was .85.
The fathers’ perception of outcomes that he expects to occur after talking with his son about sexual topics was assessed using a scale developed by DiIorio et al. (2001). Each of the 23 items item is rated on a 5-point scale anchored with the terms 1 (strongly disagree) and 5 (strongly agree). A sample item is "If you talk with [son’s name] about sexual topics, you will feel comfortable." Total possible scores are found by summing responses to individual items and range from 23 to 115. Higher scores correspond to more positive outcomes associated with talking with one’s son about sexual topics. Cronbach’s alpha coefficient for responses from mothers using a 17-item version in an earlier study was .83 and for responses of father participants in this study was .83.
Data Analysis
The goals of the data analysis for this study were to identify treatment group differences on sex-based communication outcomes at each follow-up and, through an exploratory analysis, to test the hypothesis that the intervention effect on communication was mediated by self-efficacy and outcome expectations. To achieve the first goal, independent sample t tests were used to test for group differences for the communication outcomes at each follow-up using site-aggregated data. Because the conceptual design of the study was a nested cohort design (Murray & Hannan, 1990) in which the BGCMA sites were randomly allocated to either the intervention or the control condition, this analytic strategy was one that would account for randomization at the site level as opposed to the individual level. Although many strategies are available for the analysis of data with a nested structure, the small number of sites randomized to the conditions was an important consideration in choosing this method of analysis. When the number of clusters is small (<10), estimation of the intraclass correlation coefficient and typical modeling methods (e.g., mixed-model regression using SAS Proc MIXED) may be unreliable (Murray & Wolfinger, 1994; Wears, 2002). Therefore, analyses for the main treatment effects were conducted using data aggregated at the site level that served as the allocation unit (DiIorio et al., 2005). Before conducting the main analyses, comparison tests using t tests and chi-square analyses were conducted to determine the equivalence of the intervention and control groups at baseline testing.
A mediation analysis using the 3-month follow-up data was used to address the second goal of this study. Given the exploratory nature of the mediation analysis, the mediation analyses employed the individual as the unit of analysis as opposed to the site as used in the main outcome analysis. The mediation analyses presented here were based on the mediation framework proposed by Baron and Kenny (1986). This framework is usually specified within a multiple regression model focusing on three variables: A (the independent variable—group membership), B (the mediator—self-efficacy or outcome expectations), and C (the outcome—sex-based communication). Briefly, the framework requires that three associations be established. First, for construct B to mediate the relationship between A and C, A must be associated with C. Second, A must also be associated with the mediator (B). Finally, when both A and B are included in the regression model predicting C, the magnitude of A must be significantly smaller than when B is not in the model. Applied to our analysis, in three separate models, the treatment groups (A) should differ on sex-based discussion (outcome C) and on self-efficacy and outcome expectations (mediators B). When the treatment groups and the mediators are included in the regression model predicting sex-based discussion, the effect of the treatment on sex-based discussion will be diminished if self-efficacy or outcome expectations are functioning as a mediator. Besides the standard regression analyses, the Sobel test as specified by MacKinnon and Dwyer (1993) and as implemented in the SPSS program developed by Dudley, Benuzillo, and Carrico (2004) was used to provide further information regarding the mediation effect. In addition to the mediation analysis, preliminary descriptive analyses were conducted for the variables included in the mediation model. In interpreting the results of the mediation analysis, it is extremely important to note that the analyses are based on the 3-month follow-up cross-sectional data.
|
Results |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
Sample Description
At baseline, the father participants who also completed the 3-month follow-up interview ranged in age from 18 to 80 with a mean of 40.5 years (SD = 11.8). The majority was African American (97%), and 69.9% lived with their sons (Table I). About 43% of the fathers were the biological father, 16% were stepfathers, 22% were other male relatives (brother, uncle, grandfather), and 19% were some other male role model. Almost one half of the fathers (51%) reported additional schooling beyond high school, whereas an additional 32% completed high school. Almost 49% of fathers reported yearly income between $30,000 and $69,999, whereas about 39% reported income less than $30,000, and about 12% reported income greater than or equal to $70,000. For the variables reported in Table I, chi-square tests and independent sample t tests were conducted to identify any statistically significant (p < .05) baseline differences between the two intervention groups. Although not statistically significant (p > .05), the fathers in the control group reported a somewhat higher yearly income compared with those in the intervention group. For the remaining demographic variables listed in Table I, there were no statistically significant differences noted.
|
Intervention Effects on Communication
Analyses of the father–son communication variables revealed statistically significant differences between the intervention and control groups (DiIorio et al., 2005). The results of the analyses are reported in Table II. The mean level of discussion reported by the father was higher for the intervention group compared with the control group at each time period with the difference between the groups being statistically significant (p < .05, one-tailed) at the 3- and 12-month time points. Interestingly, fathers in the intervention and control groups were similar at the 3- and 6-month follow-up assessments with respect to intent to discuss sex topics but were found to differ at the 12-month follow-up assessment. Fathers in the intervention group reported a higher level of intent to discuss sex-related topics at the 12-month follow-up. Although not statistically significant at any time point, a similar pattern was noted for the adolescent report of communication about sex with adolescents in the intervention group reporting more sex-related discussion.
|
Mediating Factors
Preliminary descriptive analyses were conducted to examine the bivariate associations among the variables included in the mediation model. To examine the relationship between intervention group membership and the outcome and mediator variables, independent sample t tests were conducted to identify group differences on these variables with the individual as the unit of analysis. The results of these analyses are reported in Table III. Based on one-tailed hypotheses tests, the fathers in the intervention group scored significantly higher on sex-based communication, self-efficacy for communication, and outcome expectations for communication. Also summarized in Table III are the bivariate correlations between the outcome variable (sex-based discussion) and the two proposed mediator variables (self-efficacy and outcome expectations). Both proposed mediators were positively related (p < .01) to sex-based communication with self-efficacy showing a slightly stronger relationship than outcome expectations although the magnitude for both is relatively small. There was a moderate relationship between the two proposed mediators with a correlation between self-efficacy and outcome expectations of .51 (p < .001).
|
Given the exploratory nature of the proposed mediation analyses and the results of the preliminary analyses examining the associations among the study variables, the necessary regression analyses were conducted to test the proposed mediation models. Separate analyses were conducted with self-efficacy and outcome expectations as the respective mediator. In the first mediation model with self-efficacy as the mediator, the results for each of the three equations can be found in Table IV. In the first equation, group membership was used to predict self-efficacy. The F statistic for this model was 10.6 with a p value of .001 and an R2 of .04. In the next equation, group membership was used to predict the sex-based communication score. The F statistic for this model was 3.9 with a p value of .05 and an R2 of .02. In the final equation, both group membership and self-efficacy were used to predict sex-based communication. The F statistic for this model was 14.27 with a p value of <.001 and an R2 of .11. In terms of testing the mediation model, the regression coefficient for group membership was not statistically different from zero (p = .34), and the regression coefficient for self-efficacy was statistically significantly different from zero (p < .001). In addition to the regression analyses, the Sobel and Goodman tests indicated that the mediation pathway was statistically different from zero (p < .01). The percent of the total effect of the intervention that was mediated was 51. In this case, a little over half (a little over 1% of the 2% attributed to the intervention) of the intervention effect was accounted for by mediation. Based on the results of these analyses, the data supported a mediation model that suggests that the effect of the intervention may be mediated by differences in self-efficacy.
|
In similar analyses also reported in Table IV, outcome expectations was included as a mediator. In the first equation, group membership was used to predict the outcome expectations score. The F statistic for this model was 3.5 with a p value of .062 and an R2 of .02. In the next equation, group membership was used to predict the sex-based communication score with the same result as described above in the analysis with self-efficacy as the mediator. The F statistic for this model was 3.9 with a p value of .05 and an R2 of .02. In the final equation, both group membership and outcome expectations were used to predict sex-based communication. The F statistic for this model was 7.25 with a p value of .001 and an R2 of .06. In terms of testing the mediation model, the regression coefficient for group membership was not statistically different from zero (p = .12), and the regression coefficient for outcome expectations was statistically significantly different from zero (p = .001). In addition to the regression analyses, the Sobel and Goodman tests indicated that the mediation pathway was not statistically different from zero (p > .10). The results of these analyses, although suggestive of a mediation effect, were considerably weaker than the results obtained for self-efficacy.
|
Discussion |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
R.E.A.L. MEN was an HIV prevention program designed for adolescent males and conducted in collaboration with BGCMA. The primary objectives of the program were to promote the delay of initiation of sexual intercourse among 11- to 14-year-old adolescents and to promote the use of prevention practices for adolescent males who were already sexually active. The unique aspect of this program was the delivery of the intervention to the fathers of adolescent males, with the intent of bolstering the father’s role as a sexuality educator.
The program was based on SCT, which dictates that self-efficacy and expectations about potential outcomes are related to behavioral change (Bandura, 1997). The program appeared to have its intended effect as our findings showed that both self-efficacy and outcome expectations related to talking to sons about sexual topics were higher among fathers in intervention group than among those in the control group. The results of the mediation analysis provided evidence that the R.E.A.L. MEN intervention enhanced self-efficacy and promoted positive outcome expectations as expected according to SCT. Self-efficacy functioned as a mediator, that is, self-efficacy of fathers increased because of the intervention and, in part, accounted for the increased sex-based communication noted between fathers and sons. Although the analysis was generally supportive of a mediation effect for self-efficacy, a mediation effect for outcome expectations was weak and could not be validated using the Sobel and Goodman tests. A possible reason for the weak results is the relatively positive attitudes that fathers in general possessed about talking to their sons about sex, creating somewhat of a ceiling effect. Nonetheless, the mediation analysis does provide some understanding of the role of self-efficacy and outcome expectations in promoting behavioral change.
The results of this study add to those of previous studies in which the relationship between self-efficacy and sex-based communication was examined. DiIorio, Dudley, Lehr, and Soet (2000) found that college students who reported higher levels of self-efficacy and more positive outcome expectations were also more likely to report higher levels of safer-sex communication with their partners. Their findings also demonstrated the role of the parents. Participants who discussed sexual topics with their parents reported higher levels of self-efficacy than those who reported little or no conversations with parents. In another study, DiIorio, Resnicow et al. (2000) found mothers’ self-efficacy beliefs and outcome expectations about talking with their children about sex were important factors associated with actual discussions. The role of both self-efficacy and outcome expectations (or beliefs) as mediators of behavior change has also been examined in several studies (Dishman et al., 2004; Li et al., 2005; Orlando et al., 2005). For example, both were noted to mediate behavior change in a study designed to reduce HIV risk behaviors among adult men and women (The NIMH Multisite HIV Prevention Trial Group, 2001).
Although the primary focus of the analysis was the mediation effect of self-efficacy and outcome expectations, the outcomes related to father–son discussions are also interesting. The results showed that fathers in the intervention were more likely to talk with their sons about sexual topics and had greater intentions to do so. The differences were not statistically significant for all assessment periods. However, the consistency of the findings for these two outcomes along with the pattern of results for the adolescents’ reports of sex-based discussions with their fathers provide evidence to support the efficacy of the intervention.
In regard to interventions designed to promote parent–adolescent discussions, the results suggest that it is important to include the intervention components designed to promote self-efficacy. As mentioned earlier, these include opportunities to observe the behavior (i.e., father–son sex-based discussions); opportunities to practice the behavior (e.g., take-home activities of increasing difficulty); and support (e.g., group format allowing for discussion among parents of their concerns and experiences). Although investigators report the main outcomes of their studies, few conduct mediation analysis to determine how the intervention worked. Understanding the active components of interventions can help refine the interventions to eliminate nonfunctioning components while enhancing the active components leading to more streamlined and cost-effective programs.
|
Limitations |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
Our mediation model was limited to social cognitive variables. However, there are other potential mediators that could have contributed to the intervention effect. It is possible that fathers who attended the group sessions supported one another in their efforts to talk with their sons about HIV prevention and other topics of importance to adolescent males. It is also possible that the intervention strengthened the father–son relationship, which in turn promoted more discussion. These and other variables of potential interest were not measured and could not be assessed. They might be considered in future studies. Despite a priori power calculations, the relatively small number of randomized sites may have restricted the power to detect intervention group differences. Given the relatively small effects (small R2) seen with the mediation analysis, it is important to note that these analyses may have been under powered as well. Despite the limitations related to statistical power, the pattern of results suggests that the intervention was effective and provides future researchers with information about expected effect sizes for discussion and social cognitive variables. The specificity of the sample used in this study, to some extent, restricts the generalizability of the findings. The participants in the study were all males, and most were African American. More importantly, they were recruited through a community-based organization that provides programs for personal development. Families who participate in such an organization are not representative of the wider community. Although the results give us some indication of the role of social cognitive mediators in eliciting an intervention effect for African-American fathers, more research is needed to determine how these mediators operate for mothers and members of other cultural groups and among other samples of participants. Other limitations include our inability to control for baseline levels of the mediator variables and the ceiling effect noted particularly for the outcome expectations variable. Without the baseline values, we were unable to examine the differences in change over time, and the skewness of the data may have contributed to the weak results for outcome expectations.
|
Summary |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
The purpose of this study was to develop and evaluate an HIV prevention program for adolescent males. The program was delivered to fathers, who, in turn, disseminated the information to their sons. The program consisted of seven group sessions held once per week. Content of the HIV intervention included normal adolescent development, HIV and STI prevention, and discussions with sons about sexual topics. Outcome analysis showed that fathers in the intervention program talked more with their sons about HIV and other sexuality issues after the intervention. Mediation analysis conducted to determine the active components of the intervention showed that self-efficacy and to a lesser extent outcome expectations served as mediators of behavior change.
|
Acknowledgments |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
We acknowledge the collaboration of the Boys & Girls Clubs of Metro Atlanta in the conduction of this study. We appreciate the assistance of the staff and the contributions of the participants who enrolled in the study. This article was funded by The National Institutes of Mental Health Grant Number 5 R01 MH59010.
|
References |
|---|
|
TopAbstractTheoretical FrameworkMethodsResultsDiscussionLimitationsSummaryAcknowledgmentsReferences |
|---|
Anderson, N. L., Koniak-Griffin, D., Keenan, C. K., Uman, G., Duggal, B. R., & Casey, C. (1999). Evaluating the outcomes of parent-child family life education. Scholarly Inquiry for Nursing Practice, 13, 211–231.[Medline]
Bandura, A. (1992). A social cognitive approach to the exercise of control over AIDS infection. In R. J. DiClimente (Ed.), Adolescents and AIDS (pp. 89–116). Newbury Park, CA: Sage.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.[CrossRef][Web of Science][Medline]
Blake, S. M., Simkin, L., Ledsky, R., Perkins, C., & Calabrese, J. M. (2001). Effects of a parent-child communications intervention on young adolescents’ risk for early onset of sexual behavior. Family Planning Perspectives, 33, 52–61.[CrossRef][Web of Science][Medline]
Crawford, I., Jason, L. A., Riordan, N., Kaufman, J., Salina, D., Sawalski, L. et al. (1990). A multimedia-based approach to increasing communication and the level of AIDS knowledge within families. Journal of Community Psychology, 18, 361–373.
Denzmore, P., McCarty, F., & DiIorio, C. (2005). The P.A.T.I.E.N.C.E. Model: An approach for recruiting African American fathers and sons into behavioral research studies. Challenge – A Journal of Research on African American Men, 11, 38–54.
DiIorio, C., Dudley, W. N., Lehr, S., & Soet, J. (2000). Correlates of safer sex communication among college students. Journal of Advanced Nursing, 32, 658–665.[CrossRef][Web of Science][Medline]
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 discussion about sex. Journal of Nursing Measurement, 9, 135–149.[Medline]
DiIorio, C., McCarty, F., Resnicow, K., Lehr, S., & Denzmore, P. (2005). R.E.A.L. MEN: A group randomized trial of an HIV prevention program for adolescent males. Manuscript submitted for publication.
DiIorio, C., Resnicow, K., De McCarty, F. A., Dudley, W. N., Wang, D. T., & Denzmore, P. (in press). Keepin? it R.E.A.L. Results of a mother-adolescent HIV prevention program. Nursing Research.
DiIorio, C., Resnicow, K., Dudley, W. N., Thomas, S., Wang, D. T., Van Marter, D. F. et al. (2000). Social cognitive factors associated with mother-adolescent communication about sex. Journal of Health Communication, 5, 41–51.[CrossRef][Web of Science][Medline]
Dishman, R. K., Motl. R. W., Saunders, R., Felton, G., Ward, D. S., Do, M. et al. (2004). Self-efficacy partially mediates the effect of a school-based physical-activity intervention among adolescent girls. Preventive Medicine: An International Journal Devoted to Practice and Theory, 38, 628–636.
Dudley, W., Benuzillo, J., & Carrico, M. (2004). SPSS and SAS programming for the testing of mediation models. Nursing Research, 53, 59–62.[CrossRef][Web of Science][Medline]
El-Bassel, N., Witte, S., Gilbert, L., Wu, E., Chang, M., Hill, J. et al. (2003). The efficacy of a relationship-based HIV/STD prevention program for heterosexual couples. American Journal of Public Health, 93, 963–969.
Lefkowitz, E. S., Sigman, M., & Au, T. K. (2000). Helping mothers discuss sexuality and AIDS with adolescents. Child Development, 71, 1383–1394.[CrossRef][Web of Science][Medline]
Li, F., Fisher, K. J., Harmer, P., & McAuley, E. (2005). Falls self-efficacy as a mediator of fear of falling in an exercise intervention for older adults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 60, 34–40.
MacKinnon, D. P., & Dwyer, J. H. (1993). Estimating mediated effects in prevention studies. Evaluation Review, 17, 144–158.
Miller, B. C., Norton, M. C., Jensen, G. O., Lee, T. R., Christopherson, C., & King, P. K. (1993). Impact evaluation of facts & feelings: A home-based video sex education curriculum. Family Relations, 42, 392–400.[CrossRef][Web of Science]
Murray, D. M., & Hannan, P. J. (1990). Planning for the appropriate analysis in school-based drug-use prevention studies. Journal of Consulting and Clinical Psychology, 58, 458–468.[CrossRef][Web of Science][Medline]
Murray, D. M., & Wolfinger, R. D. (1994). Analysis issues in the evaluation of community health: Progress toward solutions in SAS/STAT MIXED [Special issue]. Journal of Community Psychology, 140–154.
Orlando, M., Ellickson, P. L., McCaffrey, D. F., & Longshore, D. L. (2005). Mediation analysis of a school-based drug prevention program: Effects of Project ALERT. Prevention Science, 6, 35–46.
Questionnaire Design Systems (QDS) Software (Version 2.1). Bethesda, MD: NOVA Research Company. Retrieved XXXX 00, 0000, from http://www.novaresearch.com
The National Institute of Mental Health Multisite HIV Prevention Trial Group. (2001). Social-cognitive theory mediators of behavior change in the National Institute of Mental Health Multisite HIV Prevention Trial. Health Psychology, 20, 369–376.[CrossRef][Web of Science][Medline]
Wears, R. L. (2002). Advanced statistics: Statistic methods for analyzing cluster and cluster-randomized data. Academic Emergency Medicine, 9, 330–341.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. M. Yarcheski Commentary by Yarcheski West J Nurs Res, February 1, 2008; 30(1): 93 - 94. [PDF]
|
|
|
|
 |
|
 L. B. Morgenstern, N. R. Gonzales, K. E. Maddox, D. L. Brown, A. P. Karim, N. Espinosa, L. A. Moye, J. K. Pary, J. C. Grotta, L. D. Lisabeth, et al. A Randomized, Controlled Trial to Teach Middle School Children to Recognize Stroke and Call 911: The Kids Identifying and Defeating Stroke Project Stroke, November 1, 2007; 38(11): 2972 - 2978. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||