Academic journal article URISA Journal

Making Area-Based Composite Indices Comparable across Time-The Role of Standardization and Outlier Removal

Academic journal article URISA Journal

Making Area-Based Composite Indices Comparable across Time-The Role of Standardization and Outlier Removal

Article excerpt

INTRODUCTION

Wellbeing Toronto, the Peg (city of Winnipeg), and the London Ward Well-being Scores (London, United Kingdom) are attempts by municipalities to provide publicly accessible indicators of community well-being. In particular, the city of Toronto defines its Wellbeing Toronto tool as an "interactive, evidence-based information system," aimed at providing users with the accurate data necessary to facilitate informed decision making (City of Toronto 2013a). By allowing users to both choose multiple indicators, and to add importance to the chosen variables in the form of weights, the tool enables users to build their own composite indices based on indicators they deem important for defining neighborhood well-being (City of Toronto 2013a). Such well-being measures are arguably required to improve government understanding of the effects of policy and public services, aid in the distribution of funds, and advise local decision-making processes (Greater London Authority 2012).

Though the exact role played by area-based indicators in the domain of policy and decision making often is vague, some studies do exist. Wells (2006) investigated the uses of community indicators through an examination of the Boston Indicators Project, a set of 150 indicators examining change and environments in the city of Boston and the surrounding region. While the majority of responses pertained to the environmental sector, the authors indicated a range of concrete uses for the information gathered (Wells 2006, The Boston Indicators Project n.d.). Documented examples included budget justification, the enhancement of credibility in the call for new plans, proposals and the forming of coalitions, and the use of indicator data to support grant and report writing by nonprofit sector workers (Wells 2006). In one specific example, the Environmental Protection Agency (EPA) used the data to illustrate a disparity in representation of visible minorities in leadership roles in the EPA compared to local-level administrations. In response to this recognition, focus was placed on diversifying the internship program within the organization (Wells 2006). In addition, Besleme et al. (1999) provide an example of a planning commission raising awareness through the distribution of memos where development proposals impacting quality-of-life indicators were adopted as part of their regional plan. Examples include bike paths, affordable housing, and casino complexes. Finally, Warner (2006) references the success of a community-indicators project in Jacksonville, Florida, in which indicators were integrated into the decision-making and information dissemination processes of the City Council, Sheriff's Office, and Human Services Council.

Composite, area-based measures of well-being are calculated using the concepts and techniques of multicriteria decision analysis (MCDA) in Geographic Information Systems (GIS) (Malczewski 1999, Malczewski and Rinner 2015). Individual indicators correspond to decision criteria and the resulting composite measure (index) is equivalent to an MCDA score. This research uses the Wellbeing Toronto tool as a platform to examine one aspect of GIS-based MCDA, indicator (criterion) standardization, in a situation where a composite index is generated for multiple points in time (spatiotemporal MCDA). Two approaches to standardization are used: separate and joint score-range transformation. The separate approach standardizes indicator values separately for each time. This approach, therefore, is expected to produce results that are comparable within each moment or interval (e.g., census year) but not across time. The joint approach standardizes indicator values on the basis of their joint distribution across multiple times. Therefore, this approach is expected to support comparison of results between different time stamps, while its validity within an individual moment or interval may be limited.

Our investigation examines the comparability of neighborhood rankings in the Wellbeing Toronto case study, when using separate (single-time) and joint (multiple-time) score-range transformations. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.