“…Early studies mainly focused on either "preference" or actual "choice" (Sullivan et al, 1993;Bernardino et al, 1993;Mokhtarian & Salomon, 1994, 1996a, 1996bMokhtarian et al, 1998;Belanger, 1999;Wells et al, 2001;Grippaldi, 2002), while aspects such as "frequency" or "option" were gradually added to the literature (Mannering & Mokhtarian, 1995;Yen, 2000;Drucker & Khattak, 2000;Peters et al, 2001;Popuri & Bhat, 2003;Wernick, 2004;Walls et al, 2006;Mamdoohi et al, 2006;Zhou, 2008;Vana et al, 2008;Haddad et al, 2009;Tang et al, 2011;Sener & Bhat, 2011;Singh et al, 2012 should be prepared and sent out separately to each company. As it is not plausible to do this at a national level and includes high expenses in terms of both time and money, such surveys usually lead to relatively small sample sizes, which counteract the models' reliability and transferability.…”