Project AIM: The Network Meta-analysis of Active Ingredients of Mathematics Instruction

Many students struggle with mathematics, with a large and increasing percentage of students performing below basic on the National Assessment of Educational Progress. Over the years, researchers and practitioners have created and tested different ways to support students who struggle with learning mathematics. These methods include directly teaching various mathematics skills and strategies that affect mathematics performance, such as alleviating mathematics anxiety and fostering motivation and engagement in mathematics learning. The idea is that teaching mathematics using a mix of these skills or strategies might help students learn better than teaching just one skill or strategy at a time. However, it remains unclear which skills or strategies should be taught together and if mixing different skills or strategies leads to differential effects across different students or contexts. Understanding this is vital because it can help researchers and practitioners determine the best ways to address the need of struggling students in mathematics. Whereas a traditional meta-analysis allows for examination of the effectiveness of a skill/strategy, a network meta-analysis will allow the field to examine different combinations of instructional skills/strategies as well as their interaction effects. This can provide more optimal information about different instructional approaches.

This study is a comprehensive systematic review and network meta-analysis of research on the impact of mathematics interventions on mathematics outcomes for grades PreK-12 students. The goal of this project is to explore 1) the effectiveness of mathematics intervention with different combinations of instructional and affective components on mathematics outcomes among PreK-12 students with and without mathematics difficulties and 2) for whom, on what mathematics outcomes, and under what conditions mathematics interventions and combinations of intervention components work the best. This study will focus on evidence-based mathematics instruction tapping the following components: Worked examples, mathematics language, multiple representation, number lines, word problems, mathematics fluency, mathematics motivation, mathematics anxiety, and engagement. This meta-analysis will include studies that employ rigorous experimental group designs. Methods will include network meta-analysis, moderation analyses, and assessments of bias and missing data. For the network meta-analysis, this project will adopt both standard and component models to estimate the main effect as well as interaction effects among different skills and strategies. Findings from this study will not only help understand how different instructional strategies and skills are connected, but will also guide practitioners in choosing the most effective ways to help students improve mathematics performance.