Modelling partially cross-classified multilevel data

Effects of Varying Levels of Data Use to Intensify a Multisyllabic Word Reading Intervention for Upper Elementary Students With or at Risk for Reading Disabilities

For students who show inadequate response to research-based intervention, intensification of intervention using data-based decision making (DBDM) is recommended. There is a paucity of research on upper elementary students related to the efficacy of (a) word reading interventions and (b) DBDM procedures. This randomized controlled trial examined the differential effects of data use at two timepoints to intensify a multisyllabic word reading intervention for Grades 4 and 5 students with or at risk for reading disabilities (RD). Eighty-eight students were randomly assigned to one of three conditions: initial customization of the intervention (IC-only), initial customization with DBDM (IC + DBDM), or a business-as-usual comparison condition. Results indicated that (a) students in both treatment conditions outperformed the comparison condition on multisyllabic word reading and (b) students in the IC + DBDM condition also outperformed comparison students on decoding. Implications, including the use of DBDM in addition to IC-only as well as methods for decision-making within a small-group context, are discussed.

Examining the Effects of Integrating Anxiety Management Instruction Within a Reading Intervention for Upper Elementary Students With Reading Difficulties

We present findings from the first cohort of third- and fourth-grade students with reading difficulties (128 students from 31 classrooms) who participated in a 2-year intervention examining the effects of a reading intervention with and without anxiety management. Using a randomized controlled trial, students were assigned to one of three conditions: (a) small-group reading intervention with anxiety management instruction (RANX), (b) small-group reading intervention with math fact practice (RMATH), and (c) business-as-usual (BAU) comparison condition (no researcher provided treatment). Personnel from the research team provided participants in the RANX and RMATH the same reading intervention with the variation in the two treatments being whether the same amount of time per lesson was allocated to anxiety management (RANX) or practicing math facts (RMATH). Students in the RANX significantly outperformed students in the BAU on reading comprehension (effect size [ES] = 1.22) and students in the RMATH outperformed BAU on reading comprehension (ES = 0.77). Groups did not differ significantly on other reading outcomes. Reading anxiety moderated the main effect of the RANX intervention on the Test of Word Reading Efficiency (TOWRE) word reading when contrasted against the BAU group, indicating a significant difference favoring RANX where treatment's effect decreased by 0.94 units (about 1 point on the outcome) on word reading for each additional point increase in reading anxiety.

Parent-focused prevention of adolescent health risk behavior: Study protocol for a multisite cluster-randomized trial implemented in pediatric primary care

Evidence-based parenting interventions play a crucial role in the sustained reduction of adolescent behavioral health concerns. Guiding Good Choices (GGC) is a 5-session universal anticipatory guidance curriculum for parents of early adolescents that has been shown to reduce substance use, depression symptoms, and delinquent behavior. Although prior research has demonstrated the effectiveness of evidence-based parenting interventions at achieving sustained reductions in adolescent behavioral health concerns, public health impact has been limited by low rates of uptake in community and agency settings. Pediatric primary care is an ideal setting for implementing and scaling parent-focused prevention programs as these settings have a broad reach, and prevention programs implemented within them have the potential to achieve population-level impact. The current investigation, Guiding Good Choices for Health (GGC4H), tests the feasibility and effectiveness of implementing GGC in 3 geographically and socioeconomically diverse large integrated healthcare systems. This pragmatic, cluster randomized clinical trial will compare GGC parenting intervention to usual pediatric primary care practice, and will include approximately 3750 adolescents; n = 1875 GGC intervention and n = 1875 usual care. The study team hypothesizes that adolescents whose parents are randomized into the GGC intervention arm will show reductions in substance use initiation, the study's primary outcomes, and other secondary (e.g., depression symptoms, substance use prevalence) and exploratory outcomes (e.g., health services utilization, anxiety symptoms). The investigative team anticipates that the implementation of GGC within pediatric primary care clinics will successfully fill an unmet need for effective preventive parenting interventions. Trial registration: Clinicaltrials.govNCT04040153.

Mandatory Physical Education Classes of Two Hours per Week Can Be Comparable to Losing More than Five Kilograms for Chinese College Students

Despite releases of governmental guidelines for promoting physical fitness among the youth in China, the performance of college students in fitness tests has been declining over the past three decades. Obesity and physical inactivity have been proposed as two main causes. However, their relative importance for improving physical fitness remains unclear. To address this knowledge gap, we collected longitudinal data spanning four consecutive years on the physical fitness test for students from Nanjing University, China. Physical education classes of two hours per week were mandatory for the first two years. Using mixed effects models, we quantify the within-subject effects of weight, muscular endurance, sex, and mandatory physical education courses, among other variables, on physical fitness total score. We found that, in spite of the dominance of normal weight among the students, losing weight was positively associated with the total score, with significant sex differences in the associations. Compulsory exercise provided by physical education classes per week had strong positive impacts on the total score, comparable to losing weight of roughly 15-17 kg for males and 5-10 kg for females. Half sex difference in the total score was explained by male students' poor performance in the muscular endurance represented by pull-ups. Our results suggest that college students in China should engage in physical activity of higher levels to improve their physical fitness, with a heightened awareness of extra fat under normal weight and insufficient muscular endurance.

Essential Ingredients and Innovations in the Design and Analysis of Group-Randomized Trials

This article reviews the essential ingredients and innovations in the design and analysis of group-randomized trials. The methods literature for these trials has grown steadily since they were introduced to the biomedical research community in the late 1970s, and we summarize those developments. We review, in addition to the group-randomized trial, methods for two closely related designs, the individually randomized group treatment trial and the stepped-wedge group-randomized trial. After describing the essential ingredients for these designs, we review the most important developments in the evolution of their methods using a new bibliometric tool developed at the National Institutes of Health. We then discuss the questions to be considered when selecting from among these designs or selecting the traditional randomized controlled trial. We close with a review of current methods for the analysis of data from these designs, a case study to illustrate each design, and a brief summary.

Area effects on the level and development of social exclusion in later life

Social exclusion in later life is associated with decreased quality of life and poorer health outcomes. Reducing the number of people at risk of exclusion is a key theme in European social policy, but there is limited understanding of the relationship between neighbourhood characteristics, personal attributes and the level and development of social exclusion in later life. In this paper, cross-classified multilevel growth curve models for predicting exclusion are fitted to seven waves of data from the English Longitudinal Study of Ageing, collected between 2002 and 2015, allowing for the investigation of causal mechanisms linking area characteristics and exclusion in later life, including access to services and amenities, participation in civic, cultural and leisure activities and relationships with friends and family. Results show that living in a deprived area is associated with increased levels of exclusion, and this explained the higher levels of exclusion found for urban compared with rural areas. Population turnover among local residents did not impact on exclusion levels, but length of residence and the degree to which a person feels attached to their neighbourhood did, with ageing in place and stronger attachments predicting lower levels of social exclusion. In terms of individual characteristics, men, those in poor health, people with low levels of wealth or education, and those aged 80 or older, were more likely to experience increased levels of exclusion, while retirement and marriage provided a protective effect. The paper contributes new insights into the pathways through which characteristics of both individuals and neighbourhoods predict social exclusion in later life, and concludes by discussing the policy implications raised by the research.

Using n-Level Structural Equation Models for Causal Modeling in Fully Nested, Partially Nested, and Cross-Classified Randomized Controlled Trials

Complex data structures are ubiquitous in psychological research, especially in educational settings. In the context of randomized controlled trials, students are nested in classrooms but may be cross-classified by other units, such as small groups. Furthermore, in many cases only some students may be nested within a unit while other students may not. Such instances of partial nesting requires a more flexible framework for estimating treatment effects so that the model coefficients are correctly estimated. Although several recommendations have been offered to the field on handling partially nested data, few are comprehensive in their treatment of manifest and latent variables in the context of partial nesting, full nesting, and cross-classification. The present study introduces n-level structural equation modeling (SEM) as a flexible measurement and analytic framework for the estimation of treatment effects for complex data structures that frequently present in randomized controlled trials. In this tutorial, we explore how the notation of n-level SEM allows for parsimonious model specification whether data are observed or latent and in the presence of partial nested or cross-classified designs. By using the xxm package in R, the advantage of using n-level SEM framework is demonstrated through five examples for single outcome manifest variables, as in the traditional multilevel model, as well as latent applications as in multilevel SEM.

A New Multilevel CART Algorithm for Multilevel Data with Binary Outcomes

The multilevel logistic regression model (M-logit) is the standard model for modeling multilevel data with binary outcomes. However, many assumptions and restrictions should be considered when applying this model for unbiased estimation. To overcome these limitations, we proposed a multilevel CART (M-CART) algorithm which combines the M-logit and single level CART (S-CART) within the framework of the expectation-maximization. Simulation results showed that the proposed M-CART provided substantial improvements on classification accuracy, sensitivity, and specific over the M-logit, S-CART, and single level logistic regression model when modeling multilevel data with binary outcomes. This benefit of using M-CART was consistently found across different conditions of sample size, intra-class correlation, and when relationships between predictors and outcomes were nonlinear and nonadditive.

Testing mediation effects in cross-classified multilevel data

In this article, we propose an approach to test mediation effects in cross-classified multilevel data in which the initial cause is associated with one crossed factor, the mediator is associated with the other crossed factor, and the outcome is associated with Level-1 units (i.e., the 2^(A)➔2^(B)➔1 design). Multiple-membership models and cross-classified random effects models are used to estimate the indirect effects. The method is illustrated using real data from the Early Childhood Longitudinal Study-Kindergarten Cohort (1998). The results from the simulation study show that the proposed method can produce a consistent estimate of the indirect effect and reliable statistical inferences, given an adequate sample size.

Partially nested designs in psychotherapy trials: A review of modeling developments

OBJECTIVE

Individually-randomized psychotherapy trials are often partially nested. For instance, individuals assigned to a treatment arm may be clustered into therapy groups for purposes of treatment administration, whereas individuals assigned to a wait-list control are unclustered. The past several years have seen rapid expansion and investigation of methods for analyzing partially nested data. Yet partial nesting often remains ignored in psychotherapy trials.

METHODS

This review integrates and disseminates developments in the analysis of partially nested data that are particularly relevant for psychotherapy researchers.

RESULTS

First, we differentiate among alternative partially nested designs. Then, we present adaptations of multilevel model specifications that accommodate each design. Next, we address how moderation by treatment as well as mediation of the treatment effect can be investigated in partially nested designs. Model fitting results, annotated software syntax, and illustrative data sets are provided and key methodological issues are discussed.

CONCLUSIONS

We emphasize that cluster-level variability in the treatment arm need not be considered a nuisance; it can be modeled to yield insights about the treatment process.