The Ruth L. Kirschstein National Research Service Award (NRSA) program
Joslyn Armstrong
Florida State University
Specific Aims
African American adolescent girls are more likely to have increased risk for negative sexual behaviors than any other racial group. They reach puberty at earlier ages than other racial groups, putting them at greater risk for activity; 19% of African American adolescent girls reported engaging in sexual intercourse before the age of 13 (CDC, 2008; Vidourek, Bernard, & King, 2009). Surprisingly, 50% of African American girls aged 14 through 19 reported having an STI compared to 20% of White females within that same age range (CDC, 2008;Kapungu et al., 2010;Vidourek et al., 2009). Finally, African American girls are disproportionally at risk for unplanned pregnancy, STIs infection, and HIV/AIDS. Scholars have identified the father as a protective factor for adolescent girls to eliminate the prevalence of sexual risk-taking behaviors.
The proposed research study will investigate father-daughter relationship quality and daughter’s self-esteem as potential mechanisms that might account for the association between the father-daughter sexual communication and reduced sexual risk-taking behaviors. Currently, no empirical studies exist examining these mechanisms; however, relationship quality is associated with the open communication that the father has with his daughter; thereby, improving their relationship quality (Nielson, Latty, Angera, 2013). Without their close relationship, a father may not engage his daughter in an open dialogue about sex. This significant association suggests that without the relationship quality between father and daughter, there would be no communication. In sum, simply having a close and significant relationship with her father influenced daughter’s sexual decision-making, but what other factors account for this association. Daughter’s self-esteem has been significantly correlated with reduced sexual risk-taking in the daughter (Peterson, 2006). Daughters with higher self-esteem were more likely to use protective practices against sexual risk than those with lower self-esteem (Peterson, 2006). Additionally, daughters that engaged in sexual communication with their father reported increased self-esteem. These dialogues increased the daughter’s perception of themselves and their role in sexual relationships. Also, those with increased self-esteem may initiate conversations with father about sexual relationships, even those engaging in safer sex practices. Yet in research, the association between self-esteem and sexual communication and sexual risk has not been addressed in the literature.
The purpose of this research study is to examine how father-daughter relationship quality and daughter’s self-esteem explains the relationship between father-daughter sexual communication and the daughter’s sexual risk.
Aim 1. To test the direct relationship between the father-daughter’s sexual communication to daughter’s sexual risk variable to assess the significance with daughter’s sexual risk-taking behaviors.
Aim 2. Use relationship quality of the father-daughter and the daughter’s self-esteem as mediators in order to analyze its impact on father-daughter sexual communication for the sexual risk-taking of the daughter.
Significance
African American girls are at an increased risk for negative sexual behaviors in comparison to other racial groups. African American adolescent girls reach puberty at earlier ages than other racial groups. These increased hormonal changes lead to engagement in unsafe sex behaviors and sexual risk. Researchers suspect fathers have a unique influence over their daughters’ sexual choices. Researchers have found that when fathers knew of their daughter’s sexual behaviors, this led to a decrease in risky sexual behavior of the daughter (Coley, Drazal-Votruba & Schindler, 2009). Depending on the strength of the father-daughter relationship, the daughter may be less likely to engage in unsafe sex practices. The father has a significant role in her sexual decision-making due to her close relationship to him. Yet, this is related to having a relationship with their father. In fact, a close father-daughter relationship could decrease the likelihood of the daughter’s engagement in sexual risky behaviors (Nielson, Latty, Angera, 2013). Further, African American fathers’ sexual risk communication was distinct from mothers’ depending upon his relationship quality with his daughter. Wright (2009) reported that African American fathers engaged in father-child sexual communication more than Latino, White, and Asian fathers. Interestingly, adolescent girls reported higher levels of comfortability in discussing sex with their father, even if those conversations were low in frequency (Hutchinson & Cooney, 1998; Hutchinson & Cederbaum, 2011; Somers & Vollmar, 2006). However, the father-daughter sexual communication could be determined by the relationship quality between father and daughter. If no relationship exists, then sexual risk communication is less likely to occur. Despite this evidence, relatively few studies have shown associations between the father-daughter relationship, father-daughter communication, and daughter’s sexual risk behaviors using African American samples. Few empirical studies have examined the African American father-daughter sexual risk communication and daughters’ sexual risk behaviors. Essentially the father could have a distinct function in reducing the sexual risk-taking of the daughter.
Another mechanism to explain a reduction in sexual risk-taking and the amount of sexual risk communication between father and daughter is the daughter’s self-esteem. African American adolescent girls with low self-esteem have been identified as having early debut of sexual behaviors (Peterson, 2007). Thus, researchers found that daughters with higher self-esteem were less likely to engage in risky sexual behaviors (Peterson, 2007). Also, fathers play key roles in increasing their daughter’s self-esteem in sexual relationships. In the sexual risk communication, fathers give their daughters confidence and support in sexual relationships. These daughters were more likely to receive messages on self-agency and resisting sexual pressure in sexual relationships resulting in more vocalization of their sexual needs with partners. Brown and colleagues (2014) found fathers use sexual risk communication that is supportive. The dialogue contained messages of strategies on how to sexually protect and safely enjoy sexual relationships; thereby increasing the daughter’s self-esteem (Tielman, Bohinski, Boente, 2009). Little research has explored the impact of self-esteem on the father-daughter sexual communication and daughter’s sexual risk-taking.
Study design
The current research study will examine the association between father-daughter sexual communication in reducing sexual risk-taking behaviors. The researcher will investigate what other mechanisms exist to further explain how the father-daughter sexual communication can assist in decreasing the daughter’s risky sexual behaviors. The researcher hypothesizes that father-daughter relationship quality and daughter’s self-esteem are those mechanisms that can further clarify the hypothesized relationship. The researcher hypothesized that there is 1) a direct relationship between father-daughter’s sexual communication to reduce sexual risk-taking and 2) that father-daughter relationship quality and daughter’s self-esteem would mediate the relationship between father-daughter sexual communication and daughter’s sexual risk-taking. The hypothesized model for the study is:
Methods
Sample
The sample was a nationally representative sample surveyed from cities with high concentrations of African American families in the United States. Father-daughter pairs were recruited from neighborhoods that were predominately African American. The survey information was posted and advertised through mechanisms such as local ethnic-centered grocery stores, barber shops, beauty salons, churches, schools, community centers, and neighborhoods. The local radio stations were contacted to advertise the study and family-centered programs in the local community were points of contact as well to collect the sample. The sample consisted of 200 African American father-daughter pairs from low to middle socioeconomic statuses (SES). Father participants reported a mean age of 45 years old, and a majority had a high school diploma and a bachelor’s degree. Daughter participants reported a mean age of 15 years old. The pairs were instructed to complete the paper surveys separately and then instructed to either mail back the completed surveys or drop them off in various sealed location boxes in each of the cities for collection. The response rate was 70%. The study was completed in 9 months.
Measurements
Several measures were used to assess for the variables. The Sexual Risk Survey (SRS) was used to assess for sexual risk-taking behaviors in the past six months. The 13-item questionnaire was on a 4-point Likert scale where higher scores indicated higher sexual risk. The questionnaire is divided into two subscales assessing for sexual risk-taking with uncommitted partners and risky sex acts. The reliability coefficient is .88 (TurchikGarske, 2009).
The Parent-Teen Sexual Risk Communication Scale (PTSRC) was used to assess father-daughter sexual communication. The 13-item questionnaire was on a 5-point Likert scale that asked questions on sexual communication with the father. Higher scores indicated higher levels of communication and greater comfortability of the father in discussing sex with the daughter. The reliability coefficient is .94 (Hutchinson & Cooney, 1998).
Taylor’s Measure of Self-Esteem was used to measure the daughter’s self-esteem. The 16-item scale was on an 8-point Likert scale. The questionnaire measured positive and negative aspects of acceptance, pleasure, resourcefulness, regard, pridefulness, support, global support, and cooperativeness (Peterson, 2006). The scale has been tested with an African American women sample. The test-retest reliability is .93 (Peterson, 2006).
The Inventory of Father Involvement (IFI) was used to assess for father involvement in a variety of domains for the father-daughter relationship quality. The questionnaire has nine subscales on a 5-point Likert scale that assess from the behavioral, cognitive, and affective aspects of fathering (Karre, 2015). The scale was designed from the father’s perspective. The reliability coefficient is .98 (Karre, 2015).
Preliminary Results
The researcher investigated the association between father-daughter sexual communication in reducing sexual risk-taking. Additionally the researcher investigated if the father-daughter relationship quality and daughter’s self-esteem explained how the association between the predictor and outcome variables existed. To begin the analysis the researcher used structural equation modeling (SEM) for the hypothesized model (shown on page 1). Structural equation modeling is a useful method of analysis to use to assess for multiple mediators and a relationship among the variables.
Analyses were conducted using Mplus (version 7.0; MuthenMuthen, 2004) the software program and with the robust maximum likelihood (MLE) estimation method. The models were evaluated using a selection of four fit indexes that focus on different aspects of model fit. The fit indexes chosen to evaluate each model was: (a) chi-square statistic, (b) root mean squared error of approximation (RMSEA), (c) standardized root mean residual (SRMR), and (d) comparative fit index (CFI) and tucker-lewis index (TLI). A non-significant chi-square statistic is evidence of good model-data fit. Values of lower than .05 and .08 for RMSEA and SRMR, respectively, are considered evidence of good model-data fit. A value closer to 1 for both CFI and TLI are considered evidence for good model-data fit. In addition, the researcher examined the factor loadings to assist with model-data fit and interpretability of the model as well as modification indexes and standardized residuals.
The model consisted of father-daughter sexual communication on sexual risk-taking. Model 1 was evaluated for model-data fit. The chi-square statistic was non-significant, all the fit indexes were within the acceptable range for good model-fit, and all the parameter estimates were significant. The researcher examined the modification indices and found no additional suggestions to improve the model. Thus, there is an association between father-daughter sexual communication in reducing daughter’s sexual risk-taking.
The researcher then added father-daughter relationship quality and daughter’s self-esteem to the model to test the hypothesized model. The possible mediators were tested together to examine if they were independent of one another on the effect of the mediators. The researcher employed the widely used “the causal steps approach” to establish mediation (Baron & Kenny, 1986; Judd & Kenny, 1981a, 1981b; Kenny et al., 1998). There was a change in the total effect when adding the possible mediators to the model. Thus, the researcher decided to explore them as potential mediators.
Relationship Quality
To test relationship quality, the researcher analyzed the association between father-daughter sexual communication and relationship quality. There were significant path and parameter estimate coefficients. The results indicated a significant association between father-daughter sexual communication and relationship quality. Additionally the association between sexual risk-taking and relationship quality was assessed for significant path and parameter estimate coefficients as well. There were significant parameter estimate coefficients for the association. Lastly, the researcher controlled for relationship quality in re-analyzing the association between sexual communication and sexual risk-taking. The testing produced a zero coefficient. The indirect effect and total effect was computed and the total effect decreased slightly when adding father-daughter relationship quality to the model. Thus, father-daughter relationship quality was a mediator in the association. The results indicated that father-daughter sexual communication was positively significantly correlated with relationship quality. Also, father-daughter relationship quality was negatively significantly correlated with sexual risk-taking. In sum, the researcher found partial mediation with the variable.
Self-Esteem
To examine the mediating relationship of daughter’s self-esteem, the researcher analyzed self-esteem as an outcome for both sexual communication and sexual risk-taking. There were non-significant path and parameter estimate coefficients in the association with sexual communication. The results indicated no association between father-daughter sexual communication and self-esteem. Even though the results suggested that self-esteem would not be a mediator due to a lack of relationship with sexual communication, the researcher continued to explore its relationship to the other variables. Additionally the association between sexual risk-taking and self-esteem was assessed for significant path and parameter estimate coefficients. There were significant coefficients for the association. Lastly, the researcher controlled for self-esteem in re-analyzing the effect between sexual communication and sexual risk-taking. The testing produced a non-zero coefficient. Thus, daughter’s self-esteem was not a mediator in the association. The results indicated that father-daughter sexual communication was positively correlated with self-esteem. Also, daughter’s self-esteem was negatively significantly correlated with sexual risk-taking. Self-esteem may be a possible covariate in that it does not substantially alter the relationship between the predictor and outcome variable but does help describe some of its relationship. The final model is shown below.
PowerAnalysis
Power estimates for SEM are strong, due to the fact that the sample size is fairly large. General guidelines for statistical power for SEM suggest that the sample size must be large enough to reduce error. To compute power of the indirect effects, the joint test of significance was used. Another method to test indirect effect is bootstrapping. Bootstrapping produces a confidence interval, sampling distribution, p value, and standard error. These confidence intervals do not depend on the assumption of normality with significance testing. If zero is not in the confidence interval, the effect is different from zero. To compute the effect size for the indirect effect is the product of the two partial correlations. An adequate effect size is .08.
Limitations
There are several strengths as well as limitations to the study. The study included fathers and daughters in lower SES and education levels. The study results cannot be generalized to all African American girls’ sexual development. Also the results do not apply to African American boy’s experiences. Also the initial sample size was fairly large but the researcher will need to employ other methods to collect a larger sample. The results and analysis are dependent upon the sample size. A larger sample size will produce accurate results and the researcher can make adequate conclusions regarding the sample.
Conclusions
The purpose of the research study is to examine what factors contributed to decreased sexual risk behaviors in African American adolescent girls. The researcher hypothesized that there would be a relationship between father-daughter sexual communication in reducing daughter’s sexual risk-taking. The researcher suspected that the father-daughter relationship quality and daughter’s self-esteem were other mechanisms to explain that relationship as well. The results indicated that father-daughter sexual communication was significantly associated with reducing sexual risk-taking behaviors in the daughter. This means that when fathers communicate with their daughters about sexual choices, this helps in reducing their sexual risk-taking and influencing their sexual decision-making. Also the more likely a father is to discuss safe sex behaviors with their daughter, the daughters were less likely to engage in risky sexual behaviors in adolescence. Then, the father-daughter quality was found to mediate the relationship between the variables. This means that father-daughter relationship quality does explain some of the relation between father-daughter sexual communication and daughter’s sexual risk-taking. Additionally, various aspects of the father-daughter relationship would lead to a reduction in daughter’s sexual risk-taking. The researcher proposed characteristics of the relationship quality such as emotional support, and involvement were factors into the mediating relationship. Unfortunately daughter’s self-esteem was not a mechanism that described the association. Daughter’s self-esteem influences the daughter’s sexual risk-taking; however, self-esteem did not play significant a role in the proposed model. Implications for mental health professionals are to view the father as a protective factor in the daughter’s sexual risk and promote positive father-daughter interactions. Therapists are to include the father in discussions about sex and strategies to optimize the daughter’s sexual health. It is evident that the father plays an essential role in the African American family, specifically in the daughter’s sexual health.