Bayesian Family Factor Models for Multiple Outcomes

• Activity Number: 314
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Qiaolin Chen*+ and Robert E Weiss and Catherine Ann Sugar and Keith Nuechterlein and Asarnow Robert
• Companies: University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles
• Keywords: Bayesian Inference ; Familial Data ; Multitrait-Multimethod ; Confirmatory Factor Analysis ; Multiple Outcomes

Abstract

The UCLA Neurocognitive Family Study collected more than 100 neurocognitive measurements on relatives of schizophrenia patients and relatives of matched control subjects, to study the transmission of vulnerability factors for schizophrenia. There are two types of correlations: measurements on individuals from the same family are correlated, and outcome measurements within subjects are also correlated. Standard analysis techniques for multiple outcomes do not take into account associations among members in a family, while standard analyses of familial data usually model outcomes separately and does not provide information about the relationship among outcomes. Therefore, I constructed new Bayesian Family Factor Models (BFFMs), which apply Bayesian inferences to confirmatory factor analysis (CFA) models with inter-correlated family-member factors and inter-correlated outcome factors. Results of model fitting on synthetic data show that the Bayesian factor analysis model can reasonably estimate parameters and that it works as well as the classic confirmatory factor analysis model using SPSS AMOS.