Vollständiger Abstract
Worum geht es in dieser Arbeit?
Confounding by observation intensity (“coverage”) is a validity risk in smart-home digital biomarkers: coverage can correlate with labels and dominate engineered activity features. We present a confound-aware evaluation protocol that compares coverage-only (Cov3), coverage++ (CovPP), and full-feature (All) models using participant-disjoint out-of-fold prediction, paired cluster-aware inference, and leakage audits. The protocol is demonstrated on CASAS scripted cognitive assessments and Technology Integrated Health Management (TIHM) agitation windows without cross-dataset transfer. TIHM shows strong incremental signal beyond Cov3 (AUC 0.8090 to 0.9138; Δ = + 0 . 1048 , p = 5 × 10 − 5 ), but CovPP matches All, suggesting a compact auditable representation. CASAS gains beyond coverage are modest and uncertain.
Bibliografischer Nachweis
Publikationsdaten
- Autor:innen
- Andrew Choi, Daniel Choi, Junho Park
- Quelle
- Proceedings of the Human Factors and Ergonomics Society Annual Meeting
- Publikation
- 2026-01-01
- Band / Ausgabe
- Nicht angegeben
- Seiten
- Nicht angegeben
- ISSN / ISBN
- 1071-1813, 2169-5067
- Zitationen
- 0 laut Crossref
- Referenzen
- 0 hinterlegt
Zitieren
Zitierfähiger Nachweis
Andrew Choi, Daniel Choi, Junho Park (2026). Confound-Aware Evaluation of Smart-Home Digital Biomarkers: Auditing Observation Intensity Dominance in Cognitive Status and Agitation. Proceedings of the Human Factors and Ergonomics Society Annual Meeting. https://doi.org/10.1177/10711813261484491
Kontext