This study investigated the relationship between mother-adolescent sexual communication and adolescents' engagement of sexual behavior among a sample of 2,669 European American female adolescents, ages 13 to 18 years, and their mothers, from the first Wave of the National Longitudinal Study of Adolescent Health. Communication was assessed by the frequency that mothers discussed the negative consequences of intercourse with their daughters, one year prior to engagement of sexual risk. Additional family factors including mothers' knowledge of adolescent dating, family structure, and socioeconomic status were examined on the effectiveness of sexual communication and pregnancy risk. In addition to maternal influence, this study acknowledged the importance of adolescent predictors on sexual risk, including adolescents' dating experience, decision making, and awareness of the negative consequences of sexual intercourse. This study also determined if decision making/awareness of sexual consequences mediated the relationship between communication and pregnancy risk. Finally, the moderating effects of adolescents' age and mother-daughter closeness were examined in the associations between communication and sexual risk, and between decision making/awareness and sexual risk. Sexual communication influenced engagement of sexual risk. However, contrary to the hypothesis, communication around sexual risk contributed to adolescents' increased engagement of intercourse, and was ineffective in increasing condom use. Regarding adolescent predictors, decision making and awareness of sexual consequences decreased the likelihood of sexual risk, and decision-making partially mediated the relationship between communication and intercourse. There was no moderation of age or mother-daughter relationship quality in the above associations, indicating that mother and adolescent predictors had similar effects for younger and older adolescents, and across groups of mothers/adolescents who reported low and high levels of closeness. These findings suggest that discussing the negative consequences of intercourse does not prevent adolescents' engagement in sexual behavior, and adolescents may interpret this style of communication as controlling and dramatic, and rebel against parents' advice by becoming sexually active. Furthermore, adolescents' cognitive skills and ability to recognize consequences of sex played an important role in deterring sexual engagement. Thus, adolescent predictors may be more of a protective factor against engagement in sexual risk above and beyond that of maternal influence.

This study investigated the trajectories of school GPA among immigrant youth during high school and the associations between these trajectories and financial, human, and social capital factors, after accounting for demographic differences. The sample consisted of 3,454 high school students (mean age = 14.51 year-old in the fall of 9th grade), who identified themselves as first- or second-generation Chinese, Korean, Filipino, Mexican, or White immigrants, as well as third-plus-generation White youth.

Ethnic-generational variations in the latent factors (i.e., initial level and growth) of academic trajectories were first assessed between the third-plus-generation White youth, the reference group, and the ten immigrant groups, using Latent Growth Modeling (LGM). Generational variations were then examined within ethnic/racial groups. Finally, variations in academic trajectories were explored within ethnic/ethnic-generational groups using Latent Class Growth Analysis (LCGA). Once multiple trajectory classes were identified, multinomial logistic regression was used to test the associations between the trajectory classes and financial, human, and social capital factors, over and above demographic differences.

The results of LGM models indicated higher initial GPAs among both generations of Chinese, Korean, and Filipino youth, lower initial GPAs among both generations of Mexican youth, and similar initial GPAs among both generations of White immigrant youth, compared to their third-plus-generation White counterparts. For the growth of GPAs, Asian youth had more negative, Mexican youth had somewhat more positive, and White immigrant youth had similar growth, compared to the reference group, which had slightly negative change over time. The capital factors did not explain the ethnic-generational variations. Generational variations within ethnic groups were only found among Chinese, Koreans, and Filipinos. When such variations were found, first-generation youth had higher initial GPAs and more negative growth than their second-generation counterparts. Unlike the ethnic-generational variations, however, the generational variations in growth of GPAs were explained by the capital factors. The results of LCGA indicated 3 to 4 classes of academic trajectories in each ethnic/ethnic-generational group. They represented high, middle, and low achieving classes. Finally, the results of multinomial logistic regression indicated that some of the social capital factors were associated with the trajectory classes of Chinese and Korean youth.

ABSTRACT OF THE DISSERTATION

Mindfulness Predicting Physical and Psychological Health in Emerging Adulthood: Understanding Mechanisms of Mindfulness using a Mediation Model

by Christie Lynn Lundwall

Doctor of Philosophy, Graduate Program in Psychology

University of California, Riverside, December 2011

Dr. Ruth K. Chao, Chairperson

This study tested a theory derived multiple step structural equation model of four potential mediators of the relation from mindfulness to emotional distress, physical health, and well-being during emerging adulthood in 440 college students. Specifically, two models (a basic mediation model and the hypothesized theory based model) were tested and compared to determine whether and in what ways goal-directed self-regulation, cognitive emotion regulation, stress reduction, and sleep quality mediated the relation of mindfulness to health and well-being variables. Separate models were run for the three outcome variables, and in each case the hypothesized theoretical mediation model was a better fit of the data than the basic mediation models was.

As hypothesized by the model, the two types of self-regulation, goal-directed and cognitive emotional, fully mediated the relation from mindfulness to stress-reduction, as the direct relation became non-significant once these variables were included in the model. Next, stress reduction was found to either fully or partially mediate the relations from both types of self-regulation to sleep quality, emotional distress, physical health, and well-being. Stress reduction then had a strong direct relation to all three latent health outcomes. While the patterns of results for the hypothesized models were somewhat similar across the three outcomes, there were also some interesting differences in the degree to which each mediator was influential. For example, while cognitive emotion regulation had substantial indirect relations through stress reduction to all outcomes, it was only directly related to emotional distress. In contrast, goal-directed self-regulation had smaller indirect relation to the outcomes, and was directly related to physical health and well-being, but not emotional distress. Moreover, sleep quality had small significant direct relations to emotional and physical health, but not well-being. Altogether, these data supported a meaningful role of mindfulness in emotional distress, physical health, and well-being during emerging adulthood, while illuminating several mechanisms by which it likely influenced these outcomes.

This study investigated cross-lagged associations between child effects (i.e., adolescents' internalizing and externalizing behaviors) and parenting behaviors (i.e., parental support and monitoring) during high school (i.e., 9th, 10th, and 11th grades). The focus of this study was to determine whether child effects paths were greater than parent effects path, whether the cross-lagged relationship suggests erosion of parental support and monitoring, and whether ethnic differences are found for these effects. The sample consisted of 2,939 adolescents from a larger longitudinal study who have identified themselves as Chinese (n = 611), Korean (n = 572), Mexican (n = 612), Filipino (n = 355), and European American (n = 789).

A series of cross-lagged simplex models with three time points (i.e., 9th, 10th, and 11th grades) and two constructs (i.e., parental behaviors and adolescents' outcome) using maximum likelihood estimation was used to examine the parent-adolescent relationship. First, a parsimonious model in which the cross-lagged paths for child and parent effects were constrained to be equal was examined for each pairing of parent behaviors and adolescents' outcomes. Second, models in which child effects were unconstrained were used to assess ethnic group differences. A chi-square difference test (Δχ2) between the parsimonious model and the series of unconstrained model determined which model best fit the data. In the base model, parent and child effects were constrained to be equal and then these effects were allowed to vary over time. Ethnic differences in these cross-lagged paths were then examined and reported.

The results of the cross-lagged simplex models provide evidence that child effects paths were greater than parent effects path for three of the four models (i.e., parental support and adolescents' externalizing behaviors, parental monitoring and adolescents' internalizing behaviors, and parental support and adolescents' internalizing behaviors). Two of the four models (i.e., parental support and adolescents' externalizing behaviors, and parental support and internalizing behaviors) provide evidence of support erosion (i.e., child to parent cross-lagged paths are negative and significant), where initial and continued problem behaviors eroded parenting behaviors. This pattern of erosion suggests that only erosion of parental support occurred and not erosion of parental monitoring. Additionally, there was evidence of increased parental monitoring as adolescents displayed initial and continued internalizing behaviors. Ethnic differences for East Asians, Mexican, Filipino, and European Americans are discussed in detail.