An integrated choice and latent variable model for multiple discrete continuous choice kernels: Application exploring the association between day level moods and discretionary activity engagement choices

摘要: Abstract In the recent years, multiple discrete continuous (MDC) models have emerged as a popular framework to simultaneously model choice of goods (that are imperfect substitutes one another) and associated consumption quantities. The paper presents new integrated latent variable (ICLV) implementation called Hybrid Multiple Discrete Continuous (HMDC) that is capable incorporating influence psychological factors (modeled constructs) on MDC behaviors. Estimation ICLV (with single kernels kernels) has been challenge owing high dimensional integrals involved in likelihood function. typically used maximum simulated estimation (MSLE) approach becomes cumbersome when dimensionality integration increases. this research, composite marginal (CML) based proposed for parameter HMDC framework. Unlike implementations with kernel, dimension integral be decomposed varies across observations. This necessitated use weights decomposing function using CML approach. A simulation study was conducted synthetic datasets demonstrate superiority weighted over its unweighted counterpart presence kernel. applicability formulation routine demonstrated an empirical case data from 2013 American Time Use Survey (ATUS). identifies interesting association between day level moods discretionary activity participation decisions.