Repeated measurement problem in stated choice surveys used for the estimation of discrete choice models – with applications to transport demand


In the preliminary activities, the research reviews the methods that are used to account for correlation across observations by the same individual in stated choice experiments. A new approach based on the model proposed by Bhat and Sener is investigated. The joint distribution of the random terms is based on a multivariate version of the Farlie-Gumbel-Morgenstern copula with logit marginals. The sequence probability is in closed form and differentiable. The Price algorithm is used for estimation. The expected results consist in the validation  of the consistency conjecture relative to the logit estimated on the basis of the standard independence across observations assumption. The approach is applied in two contexts, one related to urban passenger transport (users’ attitudes towards automated vs conventional minibuses), and one related to urban freight transport (stakeholders’ attitudes towards pricing and parking policies).

Dati sulla ricerca

  • Prima Tipologia di ricerca: Ricerca applicata o industriale
  • Seconda Tipologia di ricerca (Metodologia):Ricerca quantitativa
  • Terza Tipologia di ricerca:Ricerca sperimentale
  • Coordinamento interno: paolo delle site
  • Settore disciplinare:
  • Area disciplinare:
  • Corso di studi: Corso di Laurea Magistrale in Ingegneria Civile (biennale - classe LM-23)
  • Collaborazioni interne:
  • Collaborazioni esterne: Prof. Edoardo Marcucci e Dott. Valerio Gatta, Università Roma Tre; Ing. Adriano Alessandrini e Dott. Qing Zhang, Università La Sapienza
  • Coordinamento esterno:
  • Enti partecipanti: Università Niccolò Cusano, Università Roma Tre, Università La Sapienza
  • Stato della ricerca: 75%
  • Finanziamento esterno: no
  • Data inizio ricerca:
  • Data fine ricerca: