El estadístico educativo que aplica teorías al aprendizaje en el mundo real.
Ethan M. McCormick is using complex data to inform educational interventions
Entrevistado:
Entrevista realizada por:
Ethan M. McCormick builds statistical tools to measure how children change over time, so that support can be targeted effectively. Annie Brookman-Byrne finds out more.
Annie Brookman-Byrne:What are you trying to understand about children’s development over time, and how?
Ethan McCormick: A child’s abilities, behaviors, and experiences evolve over days, months, and years. I want to understand how and why. Many theories of how children develop and learn are nonlinear – we talk about growth spurts, plateaus, and tipping points. Development is also dynamic – successes or setbacks today shape children’s performance and engagement tomorrow.
“Many theories of how children develop and learn are nonlinear – we talk about growth spurts, plateaus, and tipping points.”
My work creates statistical models for these complex processes. I aim to estimate children’s outcomes from complex nonlinear models; distinguish short‑term cycles in behavior from long‑term growth trends; and link individual differences in development and learning to later life outcomes. This matters because poor statistical modeling doesn’t just produce noisy answers, it can also result in misleading conclusions. With good statistical models we can understand when a given child needs a specific intervention to help them acquire an important skill.
ABB: What drew you to work on these questions?
EMC: Early in my career I worked on applied projects related to adolescent learning and decision‑making. Repeatedly, the science I wanted to conduct ran up against the limits of statistical tools. We were able to describe average differences but struggled to test how individuals changed. We wanted to know how feedback today changes learning tomorrow, for example. That tension drew me to the field of quantitative methods.
“We wanted to know how feedback today changes learning tomorrow.”
I wanted statistical models that were sufficiently complex to match our theories of development, but accurate and accessible enough that applied researchers would actually use them. I move back and forth between developing statistical methods and applying them in the real world. Each side informs the other. The translation component is both a central part of my research and an element in the value I add to this area of research.
ABB: ¿Cómo ayudará su investigación a los niños?
EMC: Much of my work focuses on equipping researchers and educators with more precise tools to measure behavior and learning “as they are lived” – rather than in a lab . The wealth of data available to inform decisions and interventions has grown so rapidly that methods for summarizing and forecasting have not had time to catch up.
I aim to advance work in areas that directly impact children’s learning and general wellbeing. For example, I am trying to identify markers of children’s struggles during learning so that teaching resources can be directed effectively. I am also forecasting negative events affecting at-risk youth, so that they might be stopped before they happen. In addition, I’m working to identify early risk factors for children’s long-term failure to thrive, so that early support can be given. All these problems require the highest degree of technical rigor and validation to ensure that decisions are targeted, equitable, and timely.
ABB: Can you give an example of how this work can help educators?
EMC: One of the most important things we find using these dynamic learning models is the importance of variability in behavior, not just in average overall performance. Consider the profiles of two students, one who consistently scores in the 75-85% range and one who scores between 65 and 95. My work and that of others suggests that in order to achieve long-term success, these two profiles of learning and performance require different supports from educators over time. In the case of more variable learners, it can be especially hard to distinguish regular ups and downs from potential long-term challenges, and creating additional tools to help educators differentiate these two patterns is a major goal for future research.
ABB: What ideas are you most excited about pursuing next?
EMC: Children’s day‑to‑day experiences of learning, stress, and sleep interact with their long‑term growth. I’m building models that integrate data from app‑based learning logs with data on longer‑term growth. This will allow us to bridge time scales and ask how short‑term dynamics combine to produce durable change. There are so many fun and challenging problems to tackle to enable these models to capture the complexity of the data.
“Children’s day‑to‑day experiences of learning, stress, and sleep interact with their long‑term growth.”
I’m also excited to work with collaborators to improve equity in educational and mental-health assessments. Cultural, personal, and educational experiences impact how a learner interacts with assessments. We are developing models that reduce bias rooted in an individual’s experiences. The models we are building seek to incorporate that information rather than simply controlling for or discarding it. As a result, decisions will be fairer and better tailored to diverse learners.
Entrevistado:
Entrevista realizada por:
Notas a pie de página
Ethan M. McCormick develops and applies new quantitative methods, especially longitudinal, time‑series, and psychometric models, to understand how learning and development unfold over time. His work focuses on interpretable, reproducible models of change that applied researchers can use to answer theory‑driven questions. Recent papers introduce linear estimation with nonlinear inference, clarify when and how to model “change,” and provide principled tools for predicting distal outcomes from growth. Ethan pairs method development with open software, tutorials, and workshops so these tools reach teachers, clinicians, and researchers working with children and adolescents. Ethan is a 2024-2026 Jacobs Foundation Investigador asociado.
Ethan’s sitio web, and Ethan on Google Scholar.
Esta entrevista fue editada para mayor claridad.