Eje Eric Molin

Group-based versus individual-based conjoint preference models of residential preferences: a comparative test

The conjoint preference approach, measuring individual preferences, has a long history in the study of residential decisionmaking processes. Residential choice behaviour, however, is often the result of a group decisionmaking process. In this paper we investigate whether conjoint preference models derived from group responses are different from and predict better than conventional conjoint models derived from the responses of individuals who do not interact during the data-collection process. In particular, we propose a new approach to modelling group preferences for residential choice alternatives that extends previous work of Timmermans et al. The new approach is illustrated in an application among 193 families with children. The results confirm that preference structures of individual family members differ from group preference structures and that the proposed group-based model predicts family preferences for new residential environments better than do conventional models.

Predicting consumer response to new housing : a stated choice experiment

This paper discusses problems typical of eliciting housing preference. It will be argued that stated preference and choice models are potentially powerful in eliciting consumer housing preferences. This approach is illustrated in an example of new housing construction in Meerhoven. The design of the stated choice experiment is outlined and the estimated part-worth utilities of the attributes are presented. Furthermore, choices for houses in low- and high-density environments are predicted and its is examined how much more households are willing to pay for low-density housing.

Analyzing heterogeneity in conjoint estimates of residential preferences

Conjoint estimates of residential preferencesare typically based on aggregate responses. Forreasons of segmentation or to differentiatebetween non-significant housing attributes andattributes that are characterized byconflicting preferences, further analysis isrequired. In this paper, the heterogeneity inthe conjoint estimates of residentialpreferences of families is analyzed. It isexamined how the estimated part-worth utilitiesof the housing attributes are related to thesocio-economic variables and current housingattributes by applying regression analysis.Although the explained variance was rather low,suggesting that residential preferences offamilies are highly idiosyncratic, therelationships that were significant could beinterpreted well.

A Comparison of Full Profile and Hierarchical Information Integration Conjoint Methods to Modeling Group Preferences

This paper presents a comparative test of full profile (FP), original hierarchical information integration (HII-O), and integrated hierarchical information integration (HII-I) conjoint methods for modeling group preferences. It is hypothesized that in settings where groups need to decide about complex multi-attribute alternatives, HII-O will predict holdout profiles better than FP, and that HII-I will perform better than HII-O. The predictive ability of the three methods is tested for the case of housing preferences of housing co-ops, which are groups consisting of three to five people who jointly own a house. The results confirm that HII-I outperforms the other two methods and further suggest that FP and HII-O perform equally well. In addition, two variations of HII-I are developed. One of these provides independent estimates of the relative influence of group members on the decision outcomes, however at the expense of requiring larger designs. The two HII-I variations are also tested and found to be equivalent in predictive ability.

Modeling group preferences using a decompositional preference approach

This article examines an extension of the decompositional, conjoint, or stated-preference approach to model group decisions. In the conventional approach, only one member is chosen to be the groups representative and provide answers for the group as a whole. In this study, all group members are brought together and asked to jointly complete a conjoint preference experiment. The hypothesis is tested that this joint group approach predicts group behavior better than the conventional approach with representatives. The paper presents the estimated part-worth utilities of the group model and compares preference structures of individual group members and groups. Finally, group preference models are tested to determine whether they outperform representative-based preference models in terms of the ability to correctly predict the group preferences for new alternatives. These analyses are performed in the context of residential preferences of co-ops, which are groups of young people, usually not partners, who live together in owner-occupied houses.

Internet-based travel surveys: selected evidence on response rates, sampling bias and reliability

Due to decreasing response rate to conventional mail questionnaires, transportation researchers should explore the use of alternative means of administration. Internet-surveys provide one of these alternatives. This paper reports some experiences with the use of Internet-surveys in a few studies of different complexity. Experiences suggest that sampling bias is a topic of concern. It also shows however that even quite complex stated adaptation experiments can be conducted reliably using Internet-surveys.

Predicting P&R use: an application of HII for mode choice modelling

This paper looks at how Park and Ride (P&R) facilities can improve the accessibility of cities. A Hierarchical Information Integration (HII) approach is utilized to gain insight into the questions of how drivers weigh the different P&R characteristics and how they affect travel mode decisions. Three experiments were conducted to apply the HII method. The paper presents and discusses the models estimated from these 3 experiments and illustrates how these models can be applied to predict P&R use.