A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Inferring proxy response in household travel surveys with unknown completer using a group-based choice model
https://orcid.org/0000-0001-8660-1728
Abstract A proxy response is often accepted for household travel surveys to reduce the survey cost and increase the sample size, but proxy-response biases may be introduced into the sample data. To investigate and correct the bias, completer information for the survey is important, but such information is not always available in practice. This study proposes a novel model that can be applicable in situations where completer information is unavailable. The method introduces group-decision modeling in analyzing the response choices of the household travel survey, where the survey response is considered to be a task allocation among household members. The proposed model can infer the probability of proxy response and the proxy-response bias of trip-related records without completer information. The potential of the proposed model was confirmed by application to a household travel survey in Japan. The inferred probability of the proxy response and the inferred bias without completer data demonstrated surprisingly similar results to the existing study with actual proxy-response data. Specifically, the model inferred a high probability of proxy response in young adults and a low proxy probability in middle-aged females, and the model inferred the proxy-response bias that female proxy respondents in the middle-aged group report lower trip rates than self-respondents. This method will be valuable not only in travel surveys, but also in the general research and practice of social surveys.
Inferring proxy response in household travel surveys with unknown completer using a group-based choice model
https://orcid.org/0000-0001-8660-1728
Abstract A proxy response is often accepted for household travel surveys to reduce the survey cost and increase the sample size, but proxy-response biases may be introduced into the sample data. To investigate and correct the bias, completer information for the survey is important, but such information is not always available in practice. This study proposes a novel model that can be applicable in situations where completer information is unavailable. The method introduces group-decision modeling in analyzing the response choices of the household travel survey, where the survey response is considered to be a task allocation among household members. The proposed model can infer the probability of proxy response and the proxy-response bias of trip-related records without completer information. The potential of the proposed model was confirmed by application to a household travel survey in Japan. The inferred probability of the proxy response and the inferred bias without completer data demonstrated surprisingly similar results to the existing study with actual proxy-response data. Specifically, the model inferred a high probability of proxy response in young adults and a low proxy probability in middle-aged females, and the model inferred the proxy-response bias that female proxy respondents in the middle-aged group report lower trip rates than self-respondents. This method will be valuable not only in travel surveys, but also in the general research and practice of social surveys.
Inferring proxy response in household travel surveys with unknown completer using a group-based choice model
https://orcid.org/0000-0001-8660-1728
Abstract A proxy response is often accepted for household travel surveys to reduce the survey cost and increase the sample size, but proxy-response biases may be introduced into the sample data. To investigate and correct the bias, completer information for the survey is important, but such information is not always available in practice. This study proposes a novel model that can be applicable in situations where completer information is unavailable. The method introduces group-decision modeling in analyzing the response choices of the household travel survey, where the survey response is considered to be a task allocation among household members. The proposed model can infer the probability of proxy response and the proxy-response bias of trip-related records without completer information. The potential of the proposed model was confirmed by application to a household travel survey in Japan. The inferred probability of the proxy response and the inferred bias without completer data demonstrated surprisingly similar results to the existing study with actual proxy-response data. Specifically, the model inferred a high probability of proxy response in young adults and a low proxy probability in middle-aged females, and the model inferred the proxy-response bias that female proxy respondents in the middle-aged group report lower trip rates than self-respondents. This method will be valuable not only in travel surveys, but also in the general research and practice of social surveys.
Inferring proxy response in household travel surveys with unknown completer using a group-based choice model
Maruyama, Takuya (author) / Hosotani, Kenta (author) / Kawano, Tomoki (author)
Transportation ; 48
2019
Article (Journal)
English
| Online Contents | 2019
Factoring Household Travel Surveys
| British Library Online Contents | 1993
Household Travel Surveys with GPS: An Experiment
| British Library Online Contents | 2009
Nonresponse Issues in Household Travel Surveys
| British Library Conference Proceedings | 1996
Proxy Responses in Self-Completion Travel Diary Surveys
| British Library Online Contents | 2006