Project Activities
The researchers will conduct a systematic review and component network meta-analysis to synthesize the evidence base on multi-component mathematics interventions for grade K–6 students who struggle with mathematics. The researchers will first update and extend an extant dataset of studies of mathematics interventions. They will then extract data on the study's instructional components and intervention features through descriptive coding. The team will also appraise study quality using the What Works Clearinghouse group design standards. The researchers will then conduct the component network meta-analysis, interpret, and disseminate the findings.
Structured Abstract
Setting
The synthesis will include studies that use group designs (randomized control trials, quasi-experimental studies) from 2022 to the present on grades K–6 mathematics interventions that were conducted in the United States and reported in the English language.
Sample
The studies will have samples of students in grades K-6 who struggle with mathematics (e.g., performing below 40th percentile, performing below proficient on a state mathematics assessment). It will include students in both public and private schools and in urban and rural areas across many geographic locations. Students in the samples will come from widely varying socio-economic, racial, and cultural backgrounds.
The researchers will examine multi-component mathematics interventions to identify which instructional components (for example, teaching mathematics vocabulary, representing numbers on a number line, using concrete and semi-concrete representations) or combinations of instructional components (for example, number line paired with mathematics vocabulary) are associated with the strongest positive impacts on mathematics outcomes for low-performing learners. The research team will also examine whether these impacts vary by outcome domain (such as whole number computation, rational number magnitude) or by content area (for example, fractions, whole numbers).
Research design and methods
The researchers will leverage the dataset of studies used in a prior network meta-analysis and narrative synthesis (R305U210004), which included 74 studies published between January 2004 to January 2022. Using the same procedures as in the previous meta-analysis, the project will progress through three phases. In phase 1, the researchers will establish the full set of eligible studies. During this phase, they will update and extend an extant dataset by conducting a systematic literature search to identify published and unpublished studies since January 2022 conducted in the U.S. that use group design (randomized control trials, quasi-experimental studies) to examine interventions that target students struggling with mathematics in grades K–6. They will extract study and intervention data through descriptive coding and appraise study quality using the What Works Clearinghouse group design standards. During Phase 2, they will customize the metafor statistical package in R to conduct component network meta-analysis along with robust variance estimation to handle two types of statistical dependencies (multi-arm and multi-outcome). In phase 3, they will conduct the component network meta-analysis and interpreting and disseminating the results.
Control condition
Studies in the synthesis will have control conditions that are business-as-usual or active controls that include instructional components within mathematics interventions (for example, a treatment condition contrasted with a differing treatment condition).
Key measures
Studies in the synthesis will include assessments of student mathematics performance from which impact estimates can be calculated. The researchers will categorize the measure based on mathematics domains (for example, whole numbers computation, rational numbers knowledge, algebra and algebraic reasoning) delineated in the search and review protocol guiding the synthesis work.
Data analytic strategy
First, the researchers will use standard pairwise meta-analysis techniques to examine the statistical heterogeneity of all directly compared interventions. Second, they will execute an additive model component network meta-analysis to identify and rank the individual instructional components. Third, they will execute an interaction model component network meta-analysis to include pairs and trios of instructional components that might interact. Fourth, they will conduct sensitivity analyses to examine the influence of outliers and also the influence of study quality rating on analysis outcomes. Fifth, they will identify and address sources of bias in the dataset. Finally, the researchers will develop visual elements to aid in understanding of the data structure in the network (e.g., network graph).
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
This project will result in a synthesis that identifies the instructional components within multi-component interventions that are the most impactful for improving student learning outcomes in mathematics. The researchers will disseminate the findings in scholarly researcher-oriented and practitioner-oriented publications. In addition, the researchers will produce a software tutorial illustrating how and why researchers might implement component network meta-analysis models with robust variance estimation, using the customized code developed from this project.
Publications:
ERIC Citations: Find available citations in ERIC for this award here.
Related projects
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
