The use of exploratory data techniques in the investigation of potential clusters of leukaemia in Wales

Ian Wilson; Paul Jarvis; Gary Higgs; Andrew Ware; David Kidner

The use of exploratory data techniques in the investigation of potential clusters of leukaemia in Wales

Ian Wilson, Paul Jarvis, Gary Higgs, Andrew Ware, David Kidner

Research output: Contribution to conference › Paper › peer-review

16 Downloads (Pure)

Abstract

Spatial data has particular characteristics that allow its distribution to be inspected using analytical techniques. These techniques can be used to determine whether point datasets display a tendency towards spatial clustering. Locations of any clusters can then be ascertained.

The Welsh Leukaemia Registry has provided postcode details of all leukaemia cases in Wales between 1991 and 2001.This paper describes a project that aims to examine the extent to which this dataset exhibits clustering tendencies. The exploratory techniques developed can be used to highlight areas where further investigation is needed in order to account for such clusters.

Building upon previous research, the Cluster Location Analysis Procedure (CLAP) has been developed. CLAP incorporates the best attributes of other cluster detection tools (exhaustiveness, independent analysis), while introducing novel qualities of its own. Requiring no previous knowledge of the dataset, CLAP independently interprets the information in order to produce accurate results. CLAP is efficient and unlikely to disregard potential clusters. While CLAP was inspired by the need to analyse the leukaemia data provided, it is applicable to any type of point dataset containing relatively few cases.

Original language	English
Number of pages	22
Publication status	Published - 2003

Access to Document

GISRUK Paper v8Final published version, 244 KBLicence: Unspecified

Cite this

@conference{e6f893dcd7954030a78707e00354e5cc,

title = "The use of exploratory data techniques in the investigation of potential clusters of leukaemia in Wales",

abstract = "Spatial data has particular characteristics that allow its distribution to be inspected using analytical techniques. These techniques can be used to determine whether point datasets display a tendency towards spatial clustering. Locations of any clusters can then be ascertained. The Welsh Leukaemia Registry has provided postcode details of all leukaemia cases in Wales between 1991 and 2001.This paper describes a project that aims to examine the extent to which this dataset exhibits clustering tendencies. The exploratory techniques developed can be used to highlight areas where further investigation is needed in order to account for such clusters.Building upon previous research, the Cluster Location Analysis Procedure (CLAP) has been developed. CLAP incorporates the best attributes of other cluster detection tools (exhaustiveness, independent analysis), while introducing novel qualities of its own. Requiring no previous knowledge of the dataset, CLAP independently interprets the information in order to produce accurate results. CLAP is efficient and unlikely to disregard potential clusters. While CLAP was inspired by the need to analyse the leukaemia data provided, it is applicable to any type of point dataset containing relatively few cases.",

author = "Ian Wilson and Paul Jarvis and Gary Higgs and Andrew Ware and David Kidner",

year = "2003",

language = "English",

}

TY - CONF

T1 - The use of exploratory data techniques in the investigation of potential clusters of leukaemia in Wales

AU - Wilson, Ian

AU - Jarvis, Paul

AU - Higgs, Gary

AU - Ware, Andrew

AU - Kidner, David

PY - 2003

Y1 - 2003

N2 - Spatial data has particular characteristics that allow its distribution to be inspected using analytical techniques. These techniques can be used to determine whether point datasets display a tendency towards spatial clustering. Locations of any clusters can then be ascertained. The Welsh Leukaemia Registry has provided postcode details of all leukaemia cases in Wales between 1991 and 2001.This paper describes a project that aims to examine the extent to which this dataset exhibits clustering tendencies. The exploratory techniques developed can be used to highlight areas where further investigation is needed in order to account for such clusters.Building upon previous research, the Cluster Location Analysis Procedure (CLAP) has been developed. CLAP incorporates the best attributes of other cluster detection tools (exhaustiveness, independent analysis), while introducing novel qualities of its own. Requiring no previous knowledge of the dataset, CLAP independently interprets the information in order to produce accurate results. CLAP is efficient and unlikely to disregard potential clusters. While CLAP was inspired by the need to analyse the leukaemia data provided, it is applicable to any type of point dataset containing relatively few cases.

AB - Spatial data has particular characteristics that allow its distribution to be inspected using analytical techniques. These techniques can be used to determine whether point datasets display a tendency towards spatial clustering. Locations of any clusters can then be ascertained. The Welsh Leukaemia Registry has provided postcode details of all leukaemia cases in Wales between 1991 and 2001.This paper describes a project that aims to examine the extent to which this dataset exhibits clustering tendencies. The exploratory techniques developed can be used to highlight areas where further investigation is needed in order to account for such clusters.Building upon previous research, the Cluster Location Analysis Procedure (CLAP) has been developed. CLAP incorporates the best attributes of other cluster detection tools (exhaustiveness, independent analysis), while introducing novel qualities of its own. Requiring no previous knowledge of the dataset, CLAP independently interprets the information in order to produce accurate results. CLAP is efficient and unlikely to disregard potential clusters. While CLAP was inspired by the need to analyse the leukaemia data provided, it is applicable to any type of point dataset containing relatively few cases.

M3 - Paper

ER -