Alexander Savi

My research

My research aims to describe, explain, and improve learning for many individuals. To accomplish this, I analyze longitudinal practice data from thousands of learners, develop formal idiographic theories of cognitive development, and create and implement adaptive algorithms for large-scale online learning.

IAS Fellowship

In a joint fellowship with Marie Deserno, I will explore how developmental psychological theories can be formalized as evolutionary networks.

The project is structured in two tightly connected sub-projects, one theoretical and one empirical. On the theoretical side, we aim to identify how evolutionary network principles—including change in node and link states, birth-death processes, and preferential attachment mechanisms—can inform psychological theory formation. These principles offer a dynamic view on how cognitive systems grow and reconfigure over time, moving beyond static representations to capture the developmental pathways that shape psychological constructs.

On the empirical side, we focus on characterizing and modeling developmental phenomena in large-scale longitudinal developmental data. A central mechanism we aim to investigate is the differentiation-dedifferentiation hypothesis. This hypothesis posits that cognitive abilities are initially undifferentiated in early development, gradually differentiate into distinct domains during childhood and adolescence, and may later dedifferentiate with aging. Despite its theoretical importance, the diff-dediff processes remain underexamined at the within-person level, especially during early development.