The project aims is to identify a set of psychometric and growth models that will permit researchers and curriculum developers to describe and test the interrelationships between students' progress along two or more learning progressions. Unlike the current thinking in the field, these psychometric and growth models will not assume linearity, continuity, or unidimensionality of student learning. The project proposes three goals: gathering evidence for the existence of non-linear patterns of learning that demonstrate branching of the learning progressions; gathering evidence for the existence of non-linear patterns of learning that demonstrate parallel learning progressions, and proposing the psychometric and growth models for these types of learning progressions.
The psychometric models will be developed to estimate the cognitive models underlying complex learning patterns. The work will progress in four phases: (1) exploring, statistically, existing datasets from NSF-funded projects containing simultaneous measures of learning on more than one learning progression; (2) testing the psychometric and growth models using the existing data; (3) collecting and analyzing new data; and (4) testing the psychometric and growth models using new data.