Dijk, A. van

model parameter house prices region segment price house variable distribution probability approach class value result number table respondent estimate function level likelihood cluster effect 0.00 sample individual section information journal estimation method response mixture aggregated logit model parameters alternative analysis survey bayesian approach preference heart example aggregated level growth bayesian study utility programme 1.00 series likelihood function difference regression correlation parameter estimates patient density characteristic treatment mixture approach household latent-clas simulation column representative rol model rank-ordered logit model estimator error two-mode panel 0.000 treatment effect lcrol model rank-ordered alcohol consumption application zip-code level 0.02