Optimal Design for Two-Level Random Assignment and Regression Discontinuity Studies

An important concern when planning research studies is to obtain maximum precision of an estimate of a treatment effect given a budget constraint. When research designs have a multilevel or hierarchical structure changes in sample size at different levels of the design will impact precision differently. Furthermore, there will typically be differential costs of enrolling additional units at different levels of the hierarchy. The optimal design problem in multilevel research studies involves determining the optimal sample size at each level of the design given specified design parameters and a specified marginal cost of recruitment at each level. The current work extends existing results by considering optimal design for (a) unbalanced random assignment designs and (b) regression discontinuity designs.     

Predicting First Grade Reading Performance from Kindergarten Response to Tier 1 Instruction

Many schools are beginning to implement multi-tier response to intervention (RTI) models for the prevention of reading difficulties and to assist in the identification of students with learning disabilities (LD). The present study was part of our larger ongoing longitudinal RTI investigation within the Florida Learning Disabilities Center grant. This study used a longitudinal correlational design, conducted in 7 ethnically and socio-economically diverse schools. We observed reading instruction in 20 classrooms, examined response rates to kindergarten Tier 1 instruction, and predicted students' first grade reading performance based upon kindergarten growth and end of year reading performance (n = 203). Teachers followed an explicit core reading program and overall, classroom instruction was rated as effective. Results indicate that controlling for students' end of kindergarten reading, their growth across kindergarten on a variety of language and literacy measures suppressed predictions of first grade performance. Specifically, the steeper the students' trajectory to a satisfactory outcome, the less likely they were to demonstrate good performance in first grade. Implications for future research and RTI implementation are discussed.     

Supporting Programming and Learning-to-Program with an Integrated CAD and Scaffolding Workbench*

Programming is a complex cognitive task for students, because of the difficulty of finding the appropriate elements (the “decomposition” problem) and integrating them correctly into a whole (the “composition” problem). Programming is also hard to learn, because so much of the thinking behind a program is implicit and the process of programming is long and complicated. Our approach is to integrate a computer-aided design tool (CAD) with scaffolding to create a single, cohesive, and coherent workbench for the entire process. Our workbench is called the GPCeditor. We have evaluated its use by high-school students. Our results suggest that students program well in the GPCeditor, they learn good programming practices, and that the learning is occurring through use of the GPCeditor.     

The assessment of the readiness of five countries to implement child maltreatment prevention programs on a large scale

This study aimed to systematically assess the readiness of five countries – Brazil, the Former Yugoslav Republic of Macedonia, Malaysia, Saudi Arabia, and South Africa – to implement evidence-based child maltreatment prevention programs on a large scale. To this end, it applied a recently developed method called Readiness Assessment for the Prevention of Child Maltreatment based on two parallel 100-item instruments. The first measures the knowledge, attitudes, and beliefs concerning child maltreatment prevention of key informants; the second, completed by child maltreatment prevention experts using all available data in the country, produces a more objective assessment readiness. The instruments cover all of the main aspects of readiness including, for instance, availability of scientific data on the problem, legislation and policies, will to address the problem, and material resources. Key informant scores ranged from 31.2 (Brazil) to 45.8/100 (the Former Yugoslav Republic of Macedonia) and expert scores, from 35.2 (Brazil) to 56/100 (Malaysia). Major gaps identified in almost all countries included a lack of professionals with the skills, knowledge, and expertise to implement evidence-based child maltreatment programs and of institutions to train them; inadequate funding, infrastructure, and equipment; extreme rarity of outcome evaluations of prevention programs; and lack of national prevalence surveys of child maltreatment. In sum, the five countries are in a low to moderate state of readiness to implement evidence-based child maltreatment prevention programs on a large scale. Such an assessment of readiness – the first of its kind – allows gaps to be identified and then addressed to increase the likelihood of program success.     

Estimation of MIMIC Model Parameters with Multilevel Data

The purpose of this simulation study was to assess the performance of latent variable models that take into account the complex sampling mechanism that often underlies data used in educational, psychological, and other social science research. Analyses were conducted using the multiple indicator multiple cause (MIMIC) model, which is a flexible and effective tool for relating observed and latent variables. The data were simulated in a hierarchical framework (e.g., individuals nested in schools) so that a multilevel modeling approach would be appropriate. Analyses were conducted accounting for and not accounting for the nested data to determine the impact of ignoring such multilevel data structures in full structural equation models. Results highlight the differences in modeling results when the analytic strategy is congruent with the data structure and what occurs when this congruency is absent. Type I error rates and power for the standard and multilevel methods were similar for within-cluster variables and for the multilevel model with between-cluster variables. However, Type I error rates were inflated for the standard approach when modeling between-cluster variables.     

Children's heart rate reactivity responses to three school tasks

The effects of three routine classroom tasks upon heart rate reactivity were investigated with a sample of 30 typical fifth-grade children. Instructions for a mental arithmetic task, the mental arithmetic task itself, and silent reading were administered under standardized conditions, with a reward being offered for performance on the mental arithmetic task to enhance competition. Heart rate was individually monitored each second during these tasks, and data were collected on mental arithmetic performance and reading ability. Results indicated that some children showed large increases in their heart rates during the three tasks, and that these children should be considered as at-risk in terms of their cardiac health.