Measuring the built environment: validity of a site survey instrument for use in urban settings
Introduction
As the evidence for inequalities in morbidity and mortality by occupational social class and wealth has become irrefutable (Acheson, 1998; Marmot and Wilkinson, 1999), attention has turned to geographical variations in health, and to the effects of context (MacIntyre, 1997; Ecob and MacIntyre, 2000). Despite the importance currently attached to building ‘healthy communities’ (Department of Health, 1999); National Strategy for Neighbourhood Renewal, 2000), it is still not known which aspects of the social, economic and physical environments have the greatest effects on health (Sloggett and Joshi, 1994; Lynch et al., 2000). This is partly because most previous research on the geographies of health has been based on studies of the aggregated socio-economic characteristics of people living in particular areas (measures of ‘social composition’), rather than ‘contextual’ characteristics of the places where people live (MacIntyre, 1997; Ecob and MacIntyre, 2000).
There is an extensive literature on the associations between poor housing and worse physical and mental health (Wilkinson, 1999; Dunn, 2000), but most of this is based on cross-sectional studies, which are unable to distinguish between causal associations and those due to social selection or recall bias. Furthermore, most previous research into the effects of housing on health has been concerned with tenure (Lewis et al., 1998; Weich and Lewis, 1998), and the effects of structural problems, such as damp or infestation (Platt et al., 1990; Hopton and Hunt, 1996; Smith and Mallinson, 1997; Acheson, 1998; Weich and Lewis, 1998; Marsh et al., 2000). There have been very few studies of the associations of the effects of the wider environment outside the home. One exception was a survey of a random sample of residents in West Central Scotland, which identified three ‘psychosocial benefits’ associated with the home: the home as haven, as a locus of autonomy and as a source of status (Kearns et al., 2000). Statistically significant associations were found between all three dimensions and respondents’ perceptions of the ‘area environment’ (such as litter, vandalism and crime), and the ‘people in the area’ (including noise, other types of disturbance, and ‘neighbourhood reputation’). However, only the latter remained statistically significant after adjusting for housing tenure and problems associated with the dwellings themselves, such as damp.
There is also evidence that those living in urban areas experience worse physical and mental health than those living in rural or suburban areas (Lewis and Booth, 1994; Meltzer et al., 1995; Dorling, 1997). The mechanisms underlying this association remain poorly understood, although there is now little support for the once-popular notion of geographical drift (Lewis et al., 1992; Verheij, 1996). It is not yet known, however, to what extent higher rates of morbidity and mortality in urban areas are explained by individual-level socio-economic deprivation, such as low income or unemployment (Sloggett and Joshi, 1994; Dunn, 2000).
One important aspect of life in urban areas is the built environment. Defined broadly, the built environment includes many characteristics of places that cannot be reduced to the characteristics of the people who live there, such as housing form, roads and footpaths, transport networks, shops, markets, parks and other public amenities, and the disposition of public space. The built environment is likely to affect important aspects of the environment in which people live in both direct ways, such as effects on traffic, noise and air pollution, and in less tangible ways, by influencing the sense of community and ‘social capital’ in an area (Dunn, 2000). Evaluating associations between the built environment and health is therefore highly challenging, particularly when attempting to elucidate the mechanisms that might link these. Two previous studies of the effects of urban regeneration on mental health assessed the impact of improvements in the built environment that were intended to increase security and community participation, even though both were primarily ‘bricks and mortar’ interventions. Although both treated urban regeneration as a ‘black box’ intervention, both found that improvements in the built environment were associated with lower levels of anxiety and depression (Halpern, 1995; Dalgard and Tambs, 1997). Both studies were, however, limited by measures of the built environment that relied on residents’ perceptions, and may therefore have been prone to recall bias.
The dearth of empirical research into the effects of specific features of the built environment on health may be partly due to the absence of reliable and valid ‘objective’ measures. One notable exception was the study of young married women living on a housing estate in south–east London (Birtchnell et al., 1988). In this study, Birtchnell and his colleagues rated several aspects of the built environment including housing form, density, accessibility, entrance type and position, and control over the ‘buffer zone’ between private and public space, using a 15-item scale of ‘design variables’. Although this scale was based on previously published research, the authors did not report its psychometric properties. Birtchnell and his colleagues (1988) found that, compared with non-depressed controls, depressed women were significantly more likely to be living in blocks with raised walkways than in brick or concrete houses, or in tower blocks.
The present study was conducted prior to an evaluation of the effects of an urban regeneration programme on the mental health of local residents in an electoral ward in north London. The main initial thrust of the regeneration programme was external refurbishment, and improving the quality of the built environment. We hypothesised that improvements in the built environment would be associated with a reduction in the prevalence of depression. The aim of the present study were twofold; first, to evaluate the psychometric properties of a set of ‘objective’ measures of the built environment, and second, to validate this by evaluating associations with the prevalence of depression, among residents in two inner city electoral wards in north London.
Section snippets
Methods
The present study was conducted as a preliminary to a prospective cohort study, comparing changes in mental health over three years among individuals living in two electoral wards in north London. One of these wards was the site of a programme of improvements to the built environment, as part of the Capital Challenge scheme, while the other was selected as a control, for the purposes of evaluation. The control ward was chosen on the basis of its similarity to the intervention ward in
Inter-rater reliability of BESSC items
Table 1 shows the degree of inter-rater reliability for the full list of categorical items in BESSC, across the 11 housing areas surveyed for the reliability study. Low kappa coefficients were found when comparing the rankings used to rate the division of space within each housing area (BESSC item 7, Appendix A), the highest of which was 0.46 (SE 0.23) (p=0.02), for shared garden/open space. Table 2 shows the intra-class correlation coefficients for the estimated distances from the centre of
Discussion
It is still not known whether, or how, the built environment affects health, independent of individuals’ material circumstances. The present study represents a preliminary attempt to develop a set of reliable measures of the characteristics of the built environment in urban settings, with which to further address these questions. There was relatively little published empirical evidence to guide the choice of items, and consequently we opted for items with the greatest face validity, and those
Conclusions
Our findings suggest that it is both possible and feasible to rate the characteristics of the built environment in an urban setting independent of residents’ subjective perceptions. Most items, excluding those that required researchers to rank the use of land, showed at least moderate inter-rater reliability. Although further work is now required to refine and validate the measures described in this paper, in other independent areas, they may well prove useful in the further study of the
Acknowledgements
We would like to thank Ken Brodie and Juliet Matthews for collecting the site survey data, and Haroula Baladimou for collating it. We are also grateful to David Walker, of the London Borough of Camden, for providing us with demographic data about the two study wards.
References (41)
- et al.
The beauty and the beastsome preliminary comparisons of ‘high’ versus ‘popular’ residential architecture and some public versus architect judgements of the same
Journal of Environmental Psychology
(1989) - et al.
Small area variations in health related behaviours; do these depend on the behaviour itself, its measurement, or on personal characteristics?
Health and Place
(2000) - et al.
Schizophrenia and city life
Lancet
(1992) - et al.
Socioeconomic status, standard of living, and neurotic disorder
Lancet
(1998) The Black Report and beyondwhat are the issues?
Social Science and Medicine
(1997)- et al.
The phyiscal environment of street blocks and resident perceptions of crime and disorderimplications for theory and measurement
Journal of Environmental Psychology
(1992) Explaining urban-rural variations in healtha review of interactions between individual and environment
Social Science and Medicine
(1996)Independent Inquiry into Inequalities in Health
(1998)- et al.
Normality, deviance and minor psychiatric morbidity in the community. A population-based approach to General Health Questionnaire data in the Health and Lifestyle Survey
Psychological Medicine
(1993) - et al.
Statistical Methods in Medical Research
(1994)
Criterion validity of the Center for Epidemiologic Studies Depression scaleresults from a community-based sample of older subjects in the Netherlands
Psychological Medicine
Responsive EnvironmentA Manual for Designers
Depression and the physical environmenta study of young married women on a London housing estate
British Journal of Psychiatry
Utopia on TrialVision and Reality in Planned Housing
Urban environment and mental health
British Journal of Psychiatry
Death in Britain How Local Mortality Rates have Changed: 1950s to 1990s.
Housing and health inequalitiesreview and prospects for research
Housing Studies
Prevalence of depression in Los Angeles County
American Journal of Epidemiology
Cited by (82)
How do schoolchildren perceive litter? Overlooked in urban but not in natural environments
2022, Journal of Environmental PsychologyCapturing reality: Validation of omnidirectional video-based immersive virtual reality as a streetscape quality auditing method
2022, Landscape and Urban PlanningA systematic review of the relationship between objective measurements of the urban environment and psychological distress
2016, Environment InternationalCitation Excerpt :Typically, a census tract in the US has 4000 to 6000 people, and a CCD has an average of 255 dwellings. Seven studies defined their own neighbourhoods using a variety of different definitions, including: (1) a housing area that was geographically bounded and in which the majority of the housing was homogeneous in form and character (Weich et al., 2001; Weich et al., 2002), (2) buffer zones, such as a 0.25 km (Downey and Van Willigen, 2005), 100 m, 500 m and 1000 m (Berke et al., 2007), 1 km and 3 km (Maas et al., 2009) radius around a participant's home), and 3) street block, such as approximately ten small contiguous streets (Araya et al., 2007) or another definition requiring four streets and eight street sides (Mair et al., 2010). Two methods were used to measure the urban environment objectively; namely independent observational measures and Geographic Information System (GIS).
Urban Health Systems: Overview
2016, International Encyclopedia of Public HealthUnderstanding Urban Planning Outcomes in the UK: Practitioner Perspectives in Outcome Assessment
2023, International Journal of E-Planning Research