Matching Items (29)
Filtering by
- Member of: ASU Electronic Theses and Dissertations
Description
This dissertation examines a planned missing data design in the context of mediational analysis. The study considered a scenario in which the high cost of an expensive mediator limited sample size, but in which less expensive mediators could be gathered on a larger sample size. Simulated multivariate normal data were generated from a latent variable mediation model with three observed indicator variables, M1, M2, and M3. Planned missingness was implemented on M1 under the missing completely at random mechanism. Five analysis methods were employed: latent variable mediation model with all three mediators as indicators of a latent construct (Method 1), auxiliary variable model with M1 as the mediator and M2 and M3 as auxiliary variables (Method 2), auxiliary variable model with M1 as the mediator and M2 as a single auxiliary variable (Method 3), maximum likelihood estimation including all available data but incorporating only mediator M1 (Method 4), and listwise deletion (Method 5).
The main outcome of interest was empirical power to detect the mediated effect. The main effects of mediation effect size, sample size, and missing data rate performed as expected with power increasing for increasing mediation effect sizes, increasing sample sizes, and decreasing missing data rates. Consistent with expectations, power was the greatest for analysis methods that included all three mediators, and power decreased with analysis methods that included less information. Across all design cells relative to the complete data condition, Method 1 with 20% missingness on M1 produced only 2.06% loss in power for the mediated effect; with 50% missingness, 6.02% loss; and 80% missingess, only 11.86% loss. Method 2 exhibited 20.72% power loss at 80% missingness, even though the total amount of data utilized was the same as Method 1. Methods 3 – 5 exhibited greater power loss. Compared to an average power loss of 11.55% across all levels of missingness for Method 1, average power losses for Methods 3, 4, and 5 were 23.87%, 29.35%, and 32.40%, respectively. In conclusion, planned missingness in a multiple mediator design may permit higher quality characterization of the mediator construct at feasible cost.
The main outcome of interest was empirical power to detect the mediated effect. The main effects of mediation effect size, sample size, and missing data rate performed as expected with power increasing for increasing mediation effect sizes, increasing sample sizes, and decreasing missing data rates. Consistent with expectations, power was the greatest for analysis methods that included all three mediators, and power decreased with analysis methods that included less information. Across all design cells relative to the complete data condition, Method 1 with 20% missingness on M1 produced only 2.06% loss in power for the mediated effect; with 50% missingness, 6.02% loss; and 80% missingess, only 11.86% loss. Method 2 exhibited 20.72% power loss at 80% missingness, even though the total amount of data utilized was the same as Method 1. Methods 3 – 5 exhibited greater power loss. Compared to an average power loss of 11.55% across all levels of missingness for Method 1, average power losses for Methods 3, 4, and 5 were 23.87%, 29.35%, and 32.40%, respectively. In conclusion, planned missingness in a multiple mediator design may permit higher quality characterization of the mediator construct at feasible cost.
ContributorsBaraldi, Amanda N (Author) / Enders, Craig K. (Thesis advisor) / Mackinnon, David P (Thesis advisor) / Aiken, Leona S. (Committee member) / Tein, Jenn-Yun (Committee member) / Arizona State University (Publisher)
Created2015
Description
Internalizing symptoms are prevalent among adolescents, especially among Latinos, and can have negative consequences on health and development. Understanding the risk and protective factors leading to internalizing difficulties among Latino youth is critical. The current study sought to assess the effects of family risk and peer social rejection in the seventh grade on internalizing symptoms in the tenth grade, and the potential buffering effects of social support from family and from friends, among a sample of 749 Mexican American youth. Structural equation modeling was used to examine pathways from seventh grade family risk and peer social rejection to internalizing symptoms in the tenth grade. Perceived social support from family and perceived social support from friends were tested as moderators of these relations. Gender differences in these pathways were also assessed. Results showed that family risk did not predict tenth grade internalizing symptoms, but that peer social rejection predicted increased internalizing symptoms for girls. Furthermore, buffering effects were not confirmed; rather social support from both friends and family had no effect on the relation between family risk and internalizing symptoms, and high levels of social support from both sources amplified the effect of peer social rejection on internalizing symptoms. Secondary analyses suggested that at low levels of social support from both sources, peer social rejection predicted decreased internalizing symptoms for males. Limitations and implications for prevention and future research are discussed.
ContributorsJenchura, Emily C (Author) / Gonzales, Nancy (Thesis advisor) / Tein, Jenn-Yun (Committee member) / Luecken, Linda (Committee member) / Arizona State University (Publisher)
Created2015
Description
Longitudinal data from European-American (EA) and Mexican-American (MA) families (n = 179 mothers, fathers, and youth; 41% MA) was used to test a bio-psycho-cultural model of the effect of non-responsive parenting on externalizing problems in young adult offspring through the effect on the stress response system. Parenting behavior (acceptance, rejection, harsh discipline) was assessed when children were in late childhood (12-13 years), cortisol samples were collected during late adolescence (18-19 years), and externalizing problems were measured in young adulthood (21-22 years). Latent profile analyses were used to examine patterns of parenting behavior in EA and MA families. A path analysis framework was used to examine how non-responsive parenting interacted with acceptance to predict adolescent stress response and subsequent externalizing problems in EA and MA young adults. Results showed different patterns of parenting behavior in EA versus MA families, with MA families demonstrating a profile of high acceptance and high non-responsiveness at higher rates than EA families. In MA families, youth adherence to the traditional cultural value of familismo related to more positive perceptions of parenting behavior. Across ethnic groups, parent rejection only predicted higher externalizing problems in young adults when acceptance was high. The effect of parent harsh discipline on offspring stress response differed by ethnicity. In MA families, harsh discipline predicted dysregulated stress response in youth when acceptance was low. In EA families, harsh discipline did not relate to youth stress response. Overall, results increase the understanding of normative and adaptive parenting behaviors in MA families. Findings inform the development of culturally-competent parenting-focused interventions that can better prevent dysregulated stress response and externalizing behavior problems in ethnically diverse youth.
ContributorsMahrer, Nicole Eva (Author) / Luecken, Linda (Thesis advisor) / Wolchik, Sharlene (Thesis advisor) / Tein, Jenn-Yun (Committee member) / Pina, Armando (Committee member) / Arizona State University (Publisher)
Created2015
Description
Asian American (AA) adolescents and young adults are at risk for poor psychological adjustment and diminished health. Parental involvement and intergenerational gap in acculturation (IGA) have been independently associated with intergenerational acculturative conflict, a common stressor in the AA population. However, few studies have tested how the influence of parental involvement on intergenerational acculturative conflict/family cohesion and subsequent psychological adjustment may vary depending on IGA; and even fewer studies have investigated how such models apply to AA general health. The goals of the present study were, therefore, to identify pathways linking these acculturative family processes to AA young adult general health in a large sample of Filipino and Southeast Asian (SEA) families. Analyses utilized data from the Children of Immigrants Longitudinal Study (CILS; Portes & Rumbaut, 2001), a national longitudinal study of children from immigrant families. Results suggested that although Filipino and SEA families may differ in the acculturative processes that contribute to intergenerational acculturative conflict and family cohesion, depressive symptoms are an important mechanism through which these family outcomes in adolescence influence young adult general health outcomes in both Filipino and SEA families. This investigation serves to inform future programs aimed at providing targeted interventions for AAs at risk for long-term psychological disorders and physical health problems.
ContributorsTanaka, Rika (Author) / Luecken, Linda J. (Thesis advisor) / Gonzales, Nancy (Committee member) / Doane Sampey, Leah (Committee member) / Tein, Jenn-Yun (Committee member) / Arizona State University (Publisher)
Created2015
Description
The present study applied latent class analysis to a family-centered prevention
trial in early childhood to identify subgroups of families with differential responsiveness to the Family Check-up (FCU) intervention. The sample included 731 families of 2-year- olds randomized to the FCU or control and followed through age five with yearly follow up assessments (Dishion et al., 2014; Shaw et al., 2015). A two-step mixture model was used to examine whether specific constellations of family characteristics at age 2 (baseline) were related to intervention response at age 3, 4, and 5. The first step empirically identified latent classes of families based on a variety of demographic and adjustment variables selected on the basis of previous research on predictors of response to the FCU and parent training in general, as well as on the clinical observations of FCU implementers. The second step modeled the effect of the FCU on longitudinal change in children's problem behavior in each of the empirically derived latent classes. Results suggested a five-class solution, where a significant intervention effect of moderate-to- large size was observed in one of the five classes. The families within the responsive class were characterized by child neglect, legal problems, and mental health issues. Pairwise comparisons revealed that the intervention effect was significantly greater in this class of families than in two other classes that were generally less at risk for the development of disruptive behavior problems, and post hoc analyses partially supported these results. Thus, results indicated that the FCU was most successful in reducing child problem behavior in the highly distressed group of families. We conclude by discussing the potential practical utility of these results and emphasizing the need for future research to evaluate this approach's predictive accuracy.
trial in early childhood to identify subgroups of families with differential responsiveness to the Family Check-up (FCU) intervention. The sample included 731 families of 2-year- olds randomized to the FCU or control and followed through age five with yearly follow up assessments (Dishion et al., 2014; Shaw et al., 2015). A two-step mixture model was used to examine whether specific constellations of family characteristics at age 2 (baseline) were related to intervention response at age 3, 4, and 5. The first step empirically identified latent classes of families based on a variety of demographic and adjustment variables selected on the basis of previous research on predictors of response to the FCU and parent training in general, as well as on the clinical observations of FCU implementers. The second step modeled the effect of the FCU on longitudinal change in children's problem behavior in each of the empirically derived latent classes. Results suggested a five-class solution, where a significant intervention effect of moderate-to- large size was observed in one of the five classes. The families within the responsive class were characterized by child neglect, legal problems, and mental health issues. Pairwise comparisons revealed that the intervention effect was significantly greater in this class of families than in two other classes that were generally less at risk for the development of disruptive behavior problems, and post hoc analyses partially supported these results. Thus, results indicated that the FCU was most successful in reducing child problem behavior in the highly distressed group of families. We conclude by discussing the potential practical utility of these results and emphasizing the need for future research to evaluate this approach's predictive accuracy.
ContributorsPelham, William (Author) / Dishion, Thomas J (Thesis advisor) / Tein, Jenn-Yun (Committee member) / Crnic, Keith A (Committee member) / Arizona State University (Publisher)
Created2016
Description
Comorbidity is a significant problem for current classification systems of psychopathology (i.e., DSM-V). One issue is that the underlying mechanisms shared among comorbid disorders are poorly understood, especially potential psychosocial mechanisms such as family dynamics. The current study used latent class analysis to empirically classify patterns of psychopathology within a large community sample of late adolescents (age 18-19) based on their lifetime psychological adjustment measured using the World Health Organization Composite International Diagnostic Interview. Videotaped family interactions of adolescents (age 16-17) and their parents were micro and macro coded and the resulting family dynamics were compared across the three empirically defined groups of psychological adjustment which emerged from the latent class analysis: 1) an early onset, persistent antisocial behavior class; 2) an emotionally distressed and substance using class; and 3) a typically developing class. It was found that some directly observed family dynamics, including parental monitoring, dyadic positive engagement and coercive engagement discriminated among empirically derived classes. It was also found that particular tasks better discriminated among classes with regard to specific family dynamics (e.g., family activity task best discriminated among classes on dyadic positive engagement). Overall, findings suggest that novel methodologies like latent class analysis can be useful in attempting to map underlying transdiagnostic mechanisms onto the current diagnostic framework. The findings also highlight the importance of taking many variables into consideration when attempting to understand how family dynamics are associated with psychological adjustment.
ContributorsPanza, Kaitlyn E (Author) / Dishion, Thomas J (Thesis advisor) / Crnic, Keith A (Committee member) / Tein, Jenn-Yun (Committee member) / Arizona State University (Publisher)
Created2015
Description
Random Forests is a statistical learning method which has been proposed for propensity score estimation models that involve complex interactions, nonlinear relationships, or both of the covariates. In this dissertation I conducted a simulation study to examine the effects of three Random Forests model specifications in propensity score analysis. The results suggested that, depending on the nature of data, optimal specification of (1) decision rules to select the covariate and its split value in a Classification Tree, (2) the number of covariates randomly sampled for selection, and (3) methods of estimating Random Forests propensity scores could potentially produce an unbiased average treatment effect estimate after propensity scores weighting by the odds adjustment. Compared to the logistic regression estimation model using the true propensity score model, Random Forests had an additional advantage in producing unbiased estimated standard error and correct statistical inference of the average treatment effect. The relationship between the balance on the covariates' means and the bias of average treatment effect estimate was examined both within and between conditions of the simulation. Within conditions, across repeated samples there was no noticeable correlation between the covariates' mean differences and the magnitude of bias of average treatment effect estimate for the covariates that were imbalanced before adjustment. Between conditions, small mean differences of covariates after propensity score adjustment were not sensitive enough to identify the optimal Random Forests model specification for propensity score analysis.
ContributorsCham, Hei Ning (Author) / Tein, Jenn-Yun (Thesis advisor) / Enders, Stephen G (Thesis advisor) / Enders, Craig K. (Committee member) / Mackinnon, David P (Committee member) / Arizona State University (Publisher)
Created2013
Description
An ever expanding body of research has shown that children of divorce are at increased risk for a range of maladaptive outcomes including academic failure, behavior problems, poor psychological adjustment, reduced self-concept, and reduced social competence (Amato, 2001). Furthermore, the widespread prevalence of divorce makes preventing these poor outcomes a pressing public health concern. The Children of Divorce-Coping with Divorce (CoD-CoD) program is an internet-based selective prevention that was derived from recent research identifying modifiable protective factors in children of divorce including active and avoidant coping, divorce appraisals, and coping efficacy. CoD-CoD addresses these putative mediators through careful adaptation of intervention components previously demonstrated to be effective for children from disrupted families (Pedro-Carroll & Alpert-Gillis, 1997; Stolberg & Mahler, 1994; Sandler, et al., 2003). In the CoD-CoD efficacy trial, 147 children ages 11-16 whose family had received a divorce decree within 48 months of the intervention start date served as participants. Participants were assessed in two waves in order to test the small theory of the intervention as well as the interventions effects on internalizing and externalizing behaviors. Analyses indicated that the program effectively reduced the participants total mental health problems and emotional problems as reported on the Strengths and Difficulties Questionnaire (SDQ) (d = .37) and for total mental health problems this effect was stronger for children with greater baseline mental health problems (d = .46). The program also had mediated effects on both child and parent-reported total mental health problems whereby the program improved coping efficacy for children with low baseline coping efficacy which led to reduced parent-reported mental health problems. To the author's knowledge this is the first randomized controlled trail of internet-based mental health program for children or adolescents which utilizes an active control condition.
ContributorsBoring, Jesse (Author) / Sandler, Irwin (Thesis advisor) / Crnic, Keith (Committee member) / Tein, Jenn-Yun (Committee member) / Horan, John (Committee member) / Arizona State University (Publisher)
Created2011
Description
The present study utilized longitudinal data from a high-risk community sample (n= 377; 166 trauma-exposed; 54% males; 52% children of alcoholics; 73% non-Hispanic/Latino Caucasian; 22% Hispanic/Latino; 5% other ethnicity) to test a series of hypotheses that may help explain the risk pathways that link traumatic stress, posttraumatic stress disorder (PTSD) symptomatology, and problematic alcohol and drug use. Specifically, this study examined whether pre-trauma substance use problems increase risk for trauma exposure (the high-risk hypothesis) or PTSD symptoms (the susceptibility hypothesis), whether PTSD symptoms increase risk for later alcohol/drug problems (the self-medication hypothesis), and whether the association between PTSD symptoms and alcohol/drug problems is due to shared risk factors (the shared vulnerability hypothesis). This study also examined the roles of gender and ethnicity in these pathways. A series of logistic and negative binomial regressions were performed in a path analysis framework. A composite pre-trauma family adversity variable was formed from measures of family conflict, family life stress, parental alcoholism, and other parent psychopathology. Results provided the strongest support for the self-medication hypothesis, such that PTSD symptoms predicted higher levels of later alcohol and drug problems among non-Hispanic/Latino Caucasian participants, over and above the influences of pre-trauma family adversity, pre-trauma substance use problems, trauma exposure, and demographic variables. Results partially supported the high-risk hypothesis, such that adolescent substance use problems had a marginally significant unique effect on risk for assaultive violence exposure but not on overall risk for trauma exposure. There was no support for the susceptibility hypothesis, as pre-trauma adolescent substance use problems did not significantly influence risk for PTSD diagnosis/symptoms over and above the influence of pre-trauma family adversity. Finally, there was little support for the shared vulnerability hypothesis. Neither trauma exposure nor preexisting family adversity accounted for the link between PTSD symptoms and later substance use problems. These results add to a growing body of literature in support of the self-medication hypothesis. Findings extend previous research by showing that PTSD symptoms may influence the development of alcohol and drug problems over and above the influence of trauma exposure itself, preexisting family risk factors, and baseline levels of substance use.
ContributorsHaller, Moira (Author) / Chassin, Laurie (Thesis advisor) / Davis, Mary (Committee member) / Pina, Armando (Committee member) / Tein, Jenn-Yun (Committee member) / Arizona State University (Publisher)
Created2014
Description
Past literature has indicated that the majority of people with alcohol problems never seek treatment and that this is especially true of women. Relatively few studies have investigated how different types of alcohol-related consequences longitudinally predict men and women's perceived need for treatment and their utilization of treatment services. The current study sought to expand the literature by examining whether gender moderates the links between four frequently endorsed types of consequences and perceived need for or actual utilization of treatment. Two-hundred thirty-seven adults ages 21-36 completed a battery of questionnaires at two time points five years apart. Results indicated that there were four broad types of consequences endorsed by both men and women. Multiple-group models and Wald chi square tests indicated that there were no significant relationships between consequences and treatment outcomes. No gender moderation was found but post-hoc power analyses indicated that the study was underpowered to detect moderation. Researchers need to continue to study factors that predict utilization of alcohol treatment services and the process of recovery so that treatment providers can better address the needs of people with alcohol-related consequences in the areas of referral procedures, clinical assessment, and treatment service provision and planning.
ContributorsBeltran Gonzalez, Iris (Author) / Chassin, Laurie (Thesis advisor) / Tein, Jenn-Yun (Committee member) / Corbin, William (Committee member) / Barrera, Jr., Manuel (Committee member) / Arizona State University (Publisher)
Created2013