Spatial Dependencies in Item Response Theory: Gaussian Process Priors for Geographic and Cognitive Measurement

2025-07-13

AI Summary

Existing spatial item response theory (SIRT) methods suffer from two key limitations: (i) they accommodate only binary responses, hindering application to polytomous item formats; and (ii) they rely on conditional autoregressive (CAR) priors, which enforce local spatial dependence and fail to capture complex, non-local dependencies in geographic or abstract cognitive spaces (e.g., theory-driven construct structures). To address these, we propose Spatial Gaussian Process IRT (SGP-IRT), which replaces the CAR prior with a Gaussian process prior—enabling flexible modeling of global and non-neighboring spatial or cognitive dependencies—and naturally extends to polytomous responses. Simulation studies demonstrate substantially improved parameter recovery accuracy. Empirical analysis shows that SGP-IRT yields more precise estimates of item difficulty surfaces, thereby enhancing measurement reliability and validity. The model is broadly applicable across diverse domains, including geospatial assessment, psychological construct mapping, and ecological evaluation.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Probabilistic Circuits and Graphical Models

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Measurement validity in Item Response Theory depends on appropriately modeling dependencies between items when these reflect meaningful theoretical structures rather than random measurement error. In ecological assessment, citizen scientists identifying species across geographic regions exhibit systematic spatial patterns in task difficulty due to environmental factors. Similarly, in Author Recognition Tests, literary knowledge organizes by genre, where familiarity with science fiction authors systematically predicts recognition of other science fiction authors. Current spatial Item Response Theory methods, represented by the 1PLUS, 2PLUS, and 3PLUS model family, address these dependencies but remain limited by (1) binary response restrictions, and (2) conditional autoregressive priors that impose rigid local correlation assumptions, preventing effective modeling of complex spatial relationships. Our proposed method, Spatial Gaussian Process Item Response Theory (SGP-IRT), addresses these limitations by replacing conditional autoregressive priors with flexible Gaussian process priors that adapt to complex dependency structures while maintaining principled uncertainty quantification. SGP-IRT accommodates polytomous responses and models spatial dependencies in both geographic and abstract cognitive spaces, where items cluster by theoretical constructs rather than physical proximity. Simulation studies demonstrate improved parameter recovery, particularly for item difficulty estimation. Empirical applications show enhanced recovery of meaningful difficulty surfaces and improved measurement precision across psychological, educational, and ecological research applications.

Problem

• Modeling spatial dependencies in Item Response Theory
• Overcoming binary response and rigid correlation limitations
• Accommodating polytomous responses in cognitive and geographic spaces

Innovation

• Uses Gaussian process priors for dependencies
• Accommodates polytomous response data
• Models geographic and cognitive spatial dependencies