The Application of Multilevel Modeling in QuantCrit and CritQuant Frameworks: A Critical-MLM Approach (CritMLM)

• In Event: Critical Methodological Innovations in Quantitative Research

Abstract

This study introduces Critical-MLM (CritMLM), an application of multilevel modeling (MLM) grounded in QuantCrit and CritQuant frameworks to challenge traditional, causally framed approaches that often reify hierarchical and deficit-oriented narratives about social groups. In CritMLM, race and ethnicity are included as nesting variables rather than fixed predictors to emphasize the structural and contextual forces shaping outcomes. Reading comprehension data on 790,208 students in grades 3-8 were used to demonstrate CritMLM, nesting students within race rather than treating race as a traditional covariate. Findings reveal how individual-level predictors interact with group-level and contextual variables, exposing nuanced patterns of inequity frequently overlooked in conventional methods. These results demonstrate the value of CritMLM, advancing equity-focused research while maintaining rigorous statistical methods.

Authors

• Jeffrey Shero, Vanderbilt University

• Mason Shero, Denison University

• Andres Pinedo, Vanderbilt University

• Jessica Logan, Vanderbilt University