A Perception-Augmented Hidden Markov Model for Parent–Child Relations in Families of Youth with Type 1 Diabetes

Abstract In youth with Type 1 diabetes, adherence to medical treatment regimens requires the involvement of both parent and child. A clinic-integrated behavioral intervention in Family Management Diabetes (FMOD) trial was shown be effective controlling deterioration glycemic level; yet mechanism remains unknown. It is possible that effectiveness through improved parent–child relation. To investigate whether improves relations, we proposed a novel approach allows differential perceptions child toward unobserved relationship. Leveraging manifesto data collected from FMOD trial, extended standard hidden Markov model by inserting layer parent- child-specific states. We took Bayesian perspective estimation developed an efficient computational algorithm sample joint posterior distribution. Extensive simulations were conducted demonstrate performance modeling framework. Application reveals families arm are more likely stay Harmonious relation state less transition Indifferent state. Compared parent, tends have heterogeneous perception