Special Education Research Accelerator Phase 2: Identifying Generalization Boundaries

Co-Principal Investigators: Therrien, William; Wong, Vivian

In this project, researchers will expand an existing platform for conducting crowdsourced research focused on learners with disabilities, the Special Education Research Accelerator (SERA), to generate evidence about the generalizability of intervention effects. Crowdsourced research has emerged to address limitations in the traditional research model as it involves multiple research teams combining resources, expertise, and skills to answer research questions that are not easily addressed by individual teams. During the first phase of SERA, the team developed the platform and infrastructure supports, pilot tested SERA with multiple research partners, and demonstrated that SERA could feasibly conduct crowdsourced replication studies. However, crowdsourcing on its own is limited for creating generalizable knowledge. Making valid and reliable inferences about the generalizability of effects requires carefully planned research designs to test generalization boundaries, or the conditions (such as learner characteristics, instructional settings, intervention configurations) under which effects are expected to replicate.

The research team will conduct a 2-year project to extend the SERA platform to incorporate processes, methods, and resources to support crowdsourced research that will yield evidence about the replicability and the generalizability of results. They will demonstrate the approach by developing materials and a research plan to estimate the generalizability boundaries of one widely used instructional practice, repeated reading, on reading fluency for students with learning disabilities (LD). Although repeated reading has demonstrated positive impacts for students with LD, the generalizability of its effects across moderating variables has not been clearly established. First, the research team will convene a consensus panel of approximately six researchers with expertise in repeated reading for students with LD to identify critical student populations, instructional settings, intervention configurations, and other conditions hypothesized to moderate the effects of repeated reading. A secondary panel of approximately 20 practitioners and researchers with experience in repeated reading and/or reading fluency intervention for culturally and linguistically diverse learners will be convened to validate and provide feedback on the conclusions of the consensus panel. Based on these results, the research team will design a series of integrated conceptual replication studies to identify generalization boundaries of repeated reading effects for students with LD. Lastly, the research team will develop the materials, procedures, and data collection infrastructure for conducting these conceptual replication studies. Approximately four research partners will then conduct the conceptual replication studies. Surveys and interviews will be conducted at each step in the process (consensus panel, integrated replication design, and pilot studies) to determine the feasibility and usability of the procedures and supports. The research team will analyze data to determine the replicability and generalizability of the effects of repeated reading as well as the feasibility of procedures and supports. The research plan developed through this project will be applied in future large-scale crowdsourced systematic replications using the SERA platform.