Modeling integrated clothing shopping destination choice using structure equation models

The destination choice problem is an essential element in transportation planning processes. The problem is to find the probability that a person traveling from a given origin will choose a destination among many available alternatives. The focus of this paper is the destination choice of non-work (shopping purpose) trips, as part of the transportation planning process, in particular in trip- based and activity- based models. In general, destination choice models are estimated and applied at the traffic zone level, although the actual destination is an elemental alternative inside a traffic zone. Therefore, the number of explicitly modeled choice alternatives is usually the number of traffic zones. Most destination choice models assume a Multinomial Logit (MNL) form for the problem. The Multinomial Logit is not capable of accounting for unobserved similarities among alternatives, since the covariance matrix of the MNL model has only elements in the diagonal. The purpose of this paper is to investigate alternative destination choice model structures, focusing on structure equation models. The non-work destination choice problem is studied in spatial choice modeling. The literature concerned with spatial choice models covers several disciplines and important insights can be found in spatial behavior and spatial interaction models. Trip distribution models are expressed indirectly in terms of behavior models and this issue in trip generation and trip distribution is bolder. Considering new approach to transport planning (activity- based) and modeling behavior of passengers, the activity location choice is more attended and usually discrete choice models are used. Many studies describe zonal utility (simple decision structure) by using land- use and socioeconomic variables and thus cannot describe individual behavior in disaggregate level. For describe more accurate of individual utility, recent studies, have used simultaneous choice process concept and in other hand few studies used structural equation models and latent variables in describe choice of activity location. Investigating of individual features in activity location choice by using of structure equation models considered in recent studies. Considering the importance of determining activity location in activity- based approach, use of exogenous and explainer variables are bolded. Variable in classic destination choice models firstly are supposed independently and secondly have less attention to psychological and personal feature of passengers. Considering these two points, the power and efficiency of representation of behavior are reduced. Studies on the consumer behavior in shopping centers, showed that in addition to observable demographic and socio-economic variables, latent individual variables like to psychological variable, Attitude lifestyle and shopping orientation are important and must be attendant (complex decision structure).the idea of applying these variables in modeling the individual clothes shopping destination choice by using structure equation models was sourced from ethology studies on customers of shopping centers (novelty of paper). In this paper 213 sample are collected by internet- based questionnaire and individuals socio- economic, attitude, lifestyle and shopping orientation were asked. This integrated model is able to correctly predict the 42 percent of observation in which destination number 1 (Bazar of Tehran and Plasko shopping center) has the highest percent correct.