a b s t r a c t
Whilst Geographical Information Systems (GIS) are now used more commonly in transport research and
modelling, GIS techniques were used in this study to select similar sample areas (in terms of geography
and census attributes) for data collection. For this purpose, a GIS mapping system for Tyne and Wear, UK,
was built. The system included topographic maps of the area, boundary maps of Lower Super Output
Areas (LSOA), and aggregated census statistics datasets for LSOAs. Criteria relating to census attributes
and the nature of transport were employed to identify ‘hotspots’ by GIS enquiry to provide suitably
matching areas, which then formed the basis of the sampling frame.
The research project was concerned with commuters’ travel choices and so the study needed to identify
commuters. In this case-study context, it is not possible to select fully homogeneous areas, so the GIS
‘hotspots’ approach allowed the identification of areas where there were a high concentration of commuters
with multiple alternatives for travel to work. A pilot study showed that the GIS origin-based
approach was good in collecting a balanced sample, as compared to an employment-based destination
survey. This paper explores the benefits and costs of these origin- and destination-based approaches.
In the origin-based home sample, households with paper-based surveys were targeted after identification
by GIS. This origin approach requires more data preparation compared to the alternative of an employerbased,
destination-based sample that could use online survey methodologies.
The paper concludes by identifying GIS as an important tool in selecting a sample area for data collection
using multiple criteria, but argues that plans for data collection need to be flexibly constructed to
overcome unexpected challenges. Although this paper focuses on a transport research case study, the
methodology presented can be applied to survey design and selection of sample areas in other disciplines.
a b s t r a c t
Whilst Geographical Information Systems (GIS) are now used more commonly in transport research and
modelling, GIS techniques were used in this study to select similar sample areas (in terms of geography
and census attributes) for data collection. For this purpose, a GIS mapping system for Tyne and Wear, UK,
was built. The system included topographic maps of the area, boundary maps of Lower Super Output
Areas (LSOA), and aggregated census statistics datasets for LSOAs. Criteria relating to census attributes
and the nature of transport were employed to identify ‘hotspots’ by GIS enquiry to provide suitably
matching areas, which then formed the basis of the sampling frame.
The research project was concerned with commuters’ travel choices and so the study needed to identify
commuters. In this case-study context, it is not possible to select fully homogeneous areas, so the GIS
‘hotspots’ approach allowed the identification of areas where there were a high concentration of commuters
with multiple alternatives for travel to work. A pilot study showed that the GIS origin-based
approach was good in collecting a balanced sample, as compared to an employment-based destination
survey. This paper explores the benefits and costs of these origin- and destination-based approaches.
In the origin-based home sample, households with paper-based surveys were targeted after identification
by GIS. This origin approach requires more data preparation compared to the alternative of an employerbased,
destination-based sample that could use online survey methodologies.
The paper concludes by identifying GIS as an important tool in selecting a sample area for data collection
using multiple criteria, but argues that plans for data collection need to be flexibly constructed to
overcome unexpected challenges. Although this paper focuses on a transport research case study, the
methodology presented can be applied to survey design and selection of sample areas in other disciplines.
การแปล กรุณารอสักครู่..
