Advancing Knowledge and Building the Research Infrastructure in Education and STEM Learning

Principal Investigator: 
Co-Investigator: 
Project Overview
Background & Purpose: 

The American Educational Research Association (AERA) Grants Program is a research and training program designed to advance knowledge and build the research infrastructure in education and STEM learning. For 20 years, the Program has supported the advanced statistical use of major federal data sets, especially those of the U.S. Department of Education’s National Center for Education Statistics (NCES) and the National Science Foundation (NSF). The Program continues to offer competitive small grants (Research Grants for doctoral-level researchers and Dissertation Grants for advanced doctoral students), training, and think tanks as in previous grants. Two new initiatives include (1) a programmatic component to enable access to and use of state longitudinal data bases for scientific research on student learning, performance, and achievement; and (2) a programmatic component to foster secondary analysis of data emanating from NSF-funded investigator-initiated STEM projects that remain rich with opportunities for further discovery.

Setting: 

The AERA Grants Program supports a range of research projects and related capacity building and infrastructural activities related to education and learning and to U.S. education policy at primarily the national level, but also at the state and international level.

Research Design: 

The research design for this project is longitudinal, cross-sectional, and comparative. The project is designed to generate evidence that is causal [statistical modeling].

This project also involves secondary data analysis, which includes the analysis of at least one of the large-scale, national or international data sets supported. Each small grant project uses sophisticated quantitative methods and techniques (i.e., structural equation modeling, hierarchical linear modeling, propensity score matching) with data from the large-scale federal data sets. The PIs for each project determine the methods that are appropriate for their specific research question(s) and available data.

Findings: 

The AERA Grants Program supports a wide variety of research projects, covering an array of topics and areas of interest within the field of education. The following recent examples are representative of the types of research the AERA Grants Program supports through the small grants component.

Chueh-An Hsieh from Michigan State University received $20,000 for the project “A unified model for the analysis of latent growth trajectories.” The application of item response theory models to repeated observations has demonstrated great promise in developmental research. It allows researchers to take into consideration the characteristics of item response and measurement error in longitudinal trajectory analysis, both of which improves the reliability and validity of the latent growth curve (LGC) model. This study will demonstrate the potential of Bayesian methods and propose a comprehensive modeling framework, combining a measurement model with a structural model. That is, through the incorporation of a commonly used link function and Bayesian estimation, an item response theory (IRT) model can be naturally introduced into a latent variable model (LVM). As preliminary results indicate, the IRT-LVM utilizes information from individual items of the scales at each point in time, allowing the utilization of item response characteristics from distinct psychometric models, permitting the separation of time-specific error and measurement error, and giving researchers a way to evaluate the factorial invariance of latent constructs across different assessment occasions.

Odis Johnson from the University of Maryland, College Park, received $34,999 to study the ecological determinants of the achievement gap in mathematics and reading in elementary school. Using a sample of 3075 kindergarteners from the Early Childhood Longitudinal Study, Kindergarten Cohort of 1998-1999, this study examines growth in the racial, gender and social class test-score gaps according to the qualities of the neighborhoods children occupy in the summer and the academic year. The central premise of this study is that seasonal investigations of the achievement gap have not considered how the social organization of neighborhoods also changes as the seasons do, and is likely to have a greater impact on the cognitive development of children during the summer. Research that generates knowledge about seasonal variation in neighborhood effects may help policy-makers better direct resources to the appropriate context of inequality at a time when cognitive disparities are produced

Michael Kieffer from Teachers College, Columbia University received $35,000 for the project “English language learners' growth in mathematics and reading during early adolescence: Do K-8 schools make a difference?” This study evaluates the impact of the reform strategy of replacing traditional middle schools with K-8 schools. Prior research conducted with native English speakers suggests that K-8 schools allow for more student-centered, cooperative, and developmentally appropriate instruction than do traditional middle schools. However, very few studies have investigated the effects of K-8 schools for ELLs, who are particularly likely to benefit from individualized instruction and cooperative learning opportunities. Using seven waves of data from the Early Childhood Longitudinal Study Kindergarten (ECLS-K) cohort, this study will investigate whether continuously attending a K-8 school, as opposed to transitioning from an elementary school to a middle school, produces greater rates of growth in mathematics and/or reading between fifth and eighth grade. Individual growth modeling with a piecewise specification of time will be used to determine whether attending a K-8 school produces a greater inflection in students' growth trajectories between fifth and eighth grade than does attending a middle school. Propensity score matching using ECLS-K's extensive data on family and child characteristics will be used to control for observable background characteristics that predict students’ selection into K-8 schools. Potential mechanisms for this effect will also be explored, including differences in school size, reported instructional practices, and ELL support services. Findings will have implications for reforming schools to better serve ELLs.

Publications & Presentations: 

Hsieh, C. von Eye, A.A., and Maier, K.S. (In press). Using a multivariate multilevel polytomous item response theory model to study parallel processes of change: The dynamic association between adolescents’ social isolation and engagement with delinquent peers in the NYS. Multivariate Behavioral Research.

Kieffer, M.J. (in press). Socioeconomic status, English proficiency, and late-emerging reading difficulties. Educational Researcher.

Other Products: 

Abstracts of all funded projects are available on the website below, as are the calls for proposals, training announcements, lists of grant-related publications, and reports.