Four directions.
One methodological purpose.

We examine factors, components, and their integration as representations of psychological and learning constructs, recognizing that different scientific questions may call for different representations. Our work considers the assumptions that connect constructs with observed data, the interpretation of resulting scores, and uncertainty in individual-level measurement.

These methodological tools enable empirical investigations of beliefs and faith, knowledge, competencies, cognitive strategies, and other psychological attributes. Mathematical analysis, simulation, and collaborative empirical validation help clarify when the tools are informative and where their limitations lie.

We develop and evaluate estimation methods, hypothesis tests, model-fit criteria, prediction measures, and local diagnostics for structural models with factors, components, or both. We consider parameter recovery, score interpretation, uncertainty, and the distinction between reproducing observed data and predicting new observations.

Longitudinal, multilevel, and heterogeneous data motivate extensions for studying dynamics, individual differences, and change across time and levels. Methodological analysis, computational simulation, and empirical collaboration help assess the assumptions, performance, and practical use of these models.

Games, simulated tasks, and interactive environments can generate behavioral evidence about how people respond to changing situations, make decisions, and learn over time. We seek to combine statistical and psychometric models with neural-network and reinforcement-learning approaches to measure these processes and evaluate the resulting inferences.

The intended systems would use this evidence to provide timely, personalized feedback supporting self-understanding, effective learning, and independent judgment. Empirical validation, interpretability, uncertainty, privacy, and autonomy guide their development.

Our aim is to explain the ideas behind methods, clarify their assumptions and limitations, and help researchers apply and interpret them appropriately. Software and published tutorials translate methodological advances into tools and guidance that researchers can use.

Workshops and collaborative projects provide opportunities to exchange questions and experience in both directions: methodological knowledge informs applied studies, while substantive research problems motivate new methods and improvements.