The current dissertation focused on the risks and rewards of the digital context in adolescent romantic relationships. Adolescent romantic relationships are a pivotal developmental milestone and a foundation for future relationship functioning. Thus, it is vital to understand how adolescents function within their romantic relationships to identify potential intervention points that can improve adolescents’ relationship skills. Adolescents frequently utilize technology within their relationships, with positive and negative implications. Thus, the digital context is an important area of research for adolescent romantic relationship functioning. The neo-ecological theory and the transformation framework help contextualize the digital context's impact on adolescent romantic relationships. The first study utilized two experiments to test the effects of an adolescent’s romantic partner hypothetically ”liking” a digital relationship threat’s Instagram post on their feelings of jealousy and digital dating abuse behaviors. Adolescents reported greater feelings of jealousy and engagement in digital dating abuse behaviors when their romantic partner “liked” a post from a different-gendered individual, and effects were exacerbated when that individual was high on attractiveness. While the digital context may serve as a risk context for adolescent relationships, the risk conferred may depend on the couple's functioning. Thus, the second study examined how sexting among adolescent couples was associated with their daily affect. Results demonstrate that while sexting may boost an adolescent’s affect on the same day, it is related to worse affect as the days pass. Lastly, the digital context can also be an external stressor that impacts the relationship. Thus, the third study examined how daily digital stress exposure during the COVID-19 pandemic is associated with late adolescent romantic couples’ substance use and mental health. This study examined actor and partner effects to assess the dyadic nature of stress contagion between romantic partners. This dissertation advances the current literature on associations between the digital context, adolescent development, and adolescent romantic relationship functioning.

Attention-Deficit/Hyperactivity Disorder (ADHD) impacts 7% of children and is associated with serious impairment throughout the lifespan. Though considered a chronic, stable condition, symptoms fluctuate substantially during adolescence, and symptom trajectory is linked to adult outcomes. A small number of studies have examined symptom trajectory during adolescence, but these have predominately examined demographic predictors. As a neurodevelopmental disorder, ADHD is theorized to arise from deficits in executive functions (EFs). Extant literature identifies three major components of EF- working memory, behavioral inhibition, and set shifting- as interrelated constructs underlying ADHD symptom expression. This study aimed to 1.) identify trajectories of ADHD symptoms, 2.) examine demographic predictors of trajectories, and 3.) examine whether EF predicts symptom trajectory using five timepoints from the Adolescent Brain Cognitive Development study, a large-scale, representative, population-based sample from the United States. 1,605 participants meeting criteria for ADHD included in analyses. ADHD symptoms were measured by parent report on the widely used Achenbach Child Behavior Checklist. Growth Mixture Modeling was used to model trajectories of ADHD symptoms. However, poor entropy indicated trajectories were not clearly differentiated and predictors could not be examined. Therefore, exploratory regression was conducted to examine predictors of ADHD symptom change from baseline to 3-year follow-up. Male sex, medication use, greater than high school parental education, and better EF all predicted increasing ADHD symptoms. Findings must be interpreted with caution due to their exploratory nature and poor validity of the measure used for ADHD symptoms, which was found to have sensitivity of only 22.58%. Given the strong theoretical and empirical link between ADHD symptoms and EF, additional research on the connection between EF and disorder trajectory with more robust measures of EF is warranted.

Psychologists report effect sizes in randomized controlled trials to facilitate interpretation and inform clinical or policy guidance. Since commonly used effect size measures (e.g., standardized mean difference) are not sensitive to heterogeneous treatment effects, methodologists have suggested the use of an alternative effect size δ, a between-subjects causal parameter describing the probability that the outcome of a random participant in the treatment group is better than the outcome of another random participant in the control group. Although this effect size is useful, researchers could mistakenly use δ to describe its within-subject analogue, ψ, the probability that an individual will do better under the treatment than the control. Hand’s paradox describes the situation where ψ and δ are on opposing sides of 0.5: δ may imply most are helped whereas the (unknown) underlying ψ indicates that most are harmed by the treatment. The current study used Monte Carlo simulations to investigate plausible situations under which Hand’s paradox does and does not occur, tracked the magnitude of the discrepancy between ψ and δ, and explored whether the size of the discrepancy could be reduced with a relevant covariate. The findings suggested that although the paradox should not occur under bivariate normal data conditions in the population, there could be sample cases with the paradox. The magnitude of the discrepancy between ψ and δ depended on both the size of the average treatment effect and the underlying correlation between the potential outcomes, ρ. Smaller effects led to larger discrepancies when ρ < 0 and ρ = 1, whereas larger effects led to larger discrepancies when 0 < ρ < 1. It was useful to consider a relevant covariate when calculating ψ and δ. Although ψ and δ were still discrepant within covariate levels, results indicated that conditioning upon relevant covariates is still useful in describing heterogeneous treatment effects.

The proliferation of intensive longitudinal datasets has necessitated the development of analytical techniques that are flexible and accessible to researchers collecting dyadic or individual data. Dynamic structural equation models (DSEMs), as implemented in Mplus, provides the flexibility researchers require by combining components from multilevel modeling, structural equation modeling, and time series analyses. This dissertation project presents a simulation study that evaluates the performance of categorical DSEM using a probit link function across different numbers of clusters (N = 50 or 200), timepoints (T = 14, 28, or 56), categories on the outcome (2, 3, or 5), and distribution of responses on the outcome (symmetric/approximate normal, skewed, or uniform) for both univariate and multivariate models (representing individual data and dyadic longitudinal Actor-Partner Interdependence Model data, respectively). The 3- and 5-category model conditions were also evaluated as continuous DSEMs across the same cluster, timepoint, and distribution conditions to evaluate to what extent ignoring the categorical nature of the outcome impacted model performance. Results indicated that previously-suggested minimums for number of clusters and timepoints from studies evaluating continuous DSEM performance with continuous outcomes are not large enough to produce unbiased and adequately powered models in categorical DSEM. The distribution of responses on the outcome did not have a noticeable impact in model performance for categorical DSEM, but did affect model performance when fitting a continuous DSEM to the same datasets. Ignoring the categorical nature of the outcome lead to underestimated effects across parameters and conditions, and showed large Type-I error rates in the N = 200 cluster conditions.

Decision trees is a machine learning technique that searches the predictor space for the variable and observed value that leads to the best prediction when the data are split into two nodes based on the variable and splitting value. Conditional Inference Trees (CTREEs) is a non-parametric class of decision trees that uses statistical theory in order to select variables for splitting. Missing data can be problematic in decision trees because of an inability to place an observation with a missing value into a node based on the chosen splitting variable. Moreover, missing data can alter the selection process because of its inability to place observations with missing values. Simple missing data approaches (e.g., deletion, majority rule, and surrogate split) have been implemented in decision tree algorithms; however, more sophisticated missing data techniques have not been thoroughly examined. In addition to these approaches, this dissertation proposed a modified multiple imputation approach to handling missing data in CTREEs. A simulation was conducted to compare this approach with simple missing data approaches as well as single imputation and a multiple imputation with prediction averaging. Results revealed that simple approaches (i.e., majority rule, treat missing as its own category, and listwise deletion) were effective in handling missing data in CTREEs. The modified multiple imputation approach did not perform very well against simple approaches in most conditions, but this approach did seem best suited for small sample sizes and extreme missingness situations.

Introduction: Edibles, THC-infused food products, are a popular type of cannabis. However, there is limited research on how acute effects of edibles differ from more traditional cannabis types, such as smoked flower (e.g., dried bud). The current study examined the subjective response of cannabis between smoked flower and edibles using a two-week long ecological momentary assessment (EMA). Sex differences were also examined.Method: Individuals (n=101) using both edibles and flower at least once weekly completed a cannabis report within 30 minutes (T1) of first cannabis use each day as well as two follow-up reports sent 1.5 (T2) and 3 hours (T3) after initial use. Participants additionally completed assessments throughout the day for fourteen consecutive days to examine daily affect. Multi-level models examined whether overall high, low-arousal negative effects, high-arousal negative effects, and general positive effects differed by edibles and flower. Given time differences in effects between cannabis types, subjective effects were examined at T1, T2, and T3, as well as for the peak effects across the three-hour time window. Covariates included demographics, variant- and invariant- cannabis use characteristics, and daily affect. Results: At T1, edibles produced lesser positive effects (b=-0.60, S.E.=0.16, p=0.001) and overall high (b=-2.00, S.E.=0.27, p<0.001) relative to flower. At T2, edibles produced greater positive effects (b=0.52, S.E.=0.21, p=0.01) relative to flower. At T3, edibles produced greater low-arousal negative effects (b=0.63, S.E.=0.23, p=0.01) relative to flower. Edibles produced greater peak low-arousal effects relative to flower (b=0.59, S.E.=0.21, p=0.01), With respect to sex differences, there was an interaction between sex and cannabis type at T1 for positive effects (b=-0.99, S.E.=0.31, p=0.001), such that males reported greater positive effects for flower. Males additionally reported lesser low-arousal effects at T1 (b=-0.60, S.E.=0.30, p=0.05) and greater overall high at T3 relative to females (b=1.24, S.E.=0.56, p=0.03). Discussion: Smoked flower produced greater effects immediately and edibles produced greater delayed effects. Edibles appear to have greater peak levels of low-arousal effects (e.g., sluggish, drowsy, slow) relative to smoked flower. Males may be more sensitive to the rewarding effects of cannabis, particularly when smoking flower.

Behavior challenges impact children and educational professionals on a daily basis; however, it is difficult for educators to obtain high quality training in behavior management. The purpose of this study was to compare cognitive apprenticeship and group work, two teaching methods, to determine which provides better knowledge and implementation outcomes for educators taking a course on behavior analysis. Seventeen educational professionals currently working with students who display challenging behavior were randomly assigned to the cognitive apprenticeship or group work conditions. The difference between the conditions is the introduction of a coach in the cognitive apprenticeship condition. The coach guides learners through the process of understanding and using behavior analysis throughout the course by providing feedback, scaffolding, and encouraging reflection and exploration. Participants completed pre-, post-, and post-posttests that measured their knowledge of behavior analysis and how well they implemented the skills taught in the course. Additionally, they completed weekly quizzes and reported how often they used the skills in real-life situations. Overall group differences across time points for knowledge and implementation scores were analyzed using a repeated measures analysis of variance (ANOVA). There were significant differences across time for both scores but not condition or time by condition. A covariance pattern model was used to determine if self-efficacy, self-confidence, previous behavior knowledge, or overall quiz performance predicted the variance in knowledge and implementation scores on the pre-, post-, and post-posttests across conditions. Time was the only significant predictor of knowledge scores, while time, condition and self-efficacy significantly predicted the variance in implementation scores. Additionally, one-way ANOVAs were used to find condition-based differences in quiz scores and practical skill use, neither of which were significant. Finally, a linear regression was used to determine if on quiz performance predicts the use of skills in real-world settings, which it did not. The courses impact on learning, skill use, and student behavior as well as future applications are discussed.

Mediation analysis is integral to psychology, investigating human behavior’s causal mechanisms. The diversity of explanations for human behavior has implications for the estimation and interpretation of statistical mediation models. Individuals can have similar observed outcomes while undergoing different causal processes or different observed outcomes while receiving the same treatment. Researchers can employ diverse strategies when studying individual differences in multiple mediation pathways, including individual fit measures and analysis of residuals. This dissertation investigates the use of individual residuals and fit measures to identify individual differences in multiple mediation pathways. More specifically, this study focuses on mediation model residuals in a heterogeneous population in which some people experience indirect effects through one mediator and others experience indirect effects through a different mediator. A simulation study investigates 162 conditions defined by effect size and sample size for three proposed methods: residual differences, delta z, and generalized Cook’s distance. Results indicate that analogs of Type 1 error rates are generally acceptable for the method of residual differences, but statistical power is limited. Likewise, neither delta z nor gCd could reliably distinguish between contrasts that had true effects and those that did not. The outcomes of this study reveal the potential for statistical measures of individual mediation. However, limitations related to unequal subpopulation variances, multiple dependent variables, the inherent relationship between direct effects and unestimated indirect effects, and minimal contrast effects require more research to develop a simple method that researchers can use on single data sets.

Attendance and engagement in available parenting interventions in both research and community settings is often inconsistent. Recent research suggests that varying the delivery modality of the intervention (i.e., in-person, telehealth, or online) has the potential to increase engagement with evidence-based parenting programs. However, while it is known that both facilitator and parent characteristics also influence engagement, no study has evaluated whether those characteristics moderate the influence that modality has on engagement. Utilizing data from the randomized controlled comparative effectiveness trial of the After Deployment, Adaptive Parenting Tools intervention, this study aimed to assess whether facilitators’ gender, military background, and competence moderated the effect of modality on parents’ engagement. Results suggested that parents were significantly more likely to have attended when they were randomized to the telehealth condition. Additionally, while there were no moderating relationships, female facilitators and facilitators who were more competent had overall higher attendance. Additionally, in the group format, facilitators with military backgrounds had higher engagement than those who did not. Understanding the effects that delivery modality and facilitators have on parental engagement is critical to continue and amplify implementation efforts in community settings.

Latent profile analysis (LPA), a type of finite mixture model, has grown in popularity due to its ability to detect latent classes or unobserved subgroups within a sample. Though numerous methods exist to determine the correct number of classes, past research has repeatedly demonstrated that no one method is consistently the best as each tends to struggle under specific conditions. Recently, the likelihood incremental percentage per parameter (LI3P), a method using a new approach, was proposed and tested which yielded promising initial results. To evaluate this new method more thoroughly, this study simulated 50,000 datasets, manipulating factors such as sample size, class distance, number of items, and number of classes. After evaluating the performance of the LI3P on simulated data, the LI3P is applied to LPA models fit to an empirical dataset to illustrate the method’s application. Results indicate the LI3P performs in line with standard class enumeration techniques, and primarily reflects class separation and the number of classes.