A Multilevel Spatial Model to Investigate Voting Behaviour in the 2019 UK General Election.

This paper presents a modelling framework which can detect the simultaneous presence of two different types of spatial process. The first is the variation from a global mean resulting from a geographical unit's 'vertical' position within a nested hierarchical structure such as the county and region where it is situated. The second is the variation at the smaller scale of individual units due to the 'horizontal' influence of nearby locations. The former is captured using a multi-level modelling structure while the latter is accounted for by an autoregressive component at the lowest level of the hierarchy. Such a model not only estimates spatially-varying parameters according to geographical scale, but also the relative contribution of each process to the overall spatial variation. As a demonstration, the study considers the association of a selection of socio-economic attributes with voting behaviour in the 2019 UK general election. It finds evidence of the presence of both types of spatial effects, and describes how they suggest different associations between census profile and voting behaviour in different parts of England and Wales.

Missing data as data.

Our "digified" lives have provided researchers with an unprecedented opportunity to study society at a much higher frequency and granularity. Such data can have a large sample size but can be sparse, biased, and exclusively contributed by the users of the technologies. We look at the increasing importance of missing data and under-representation and propose a new perspective that considers missing data as useful data to understand the underlying reasons for missingness and that provides a realistic view of the sample size of large but under-represented data.

Measuring the exposure of Black, Asian and other ethnic groups to COVID-infected neighbourhoods in English towns and cities.

Drawing on the work of The Doreen Lawrence Review-a report on the disproportionate impact of COVID-19 on Black, Asian and minority ethnic communities in the UK-this paper develops an index of exposure, measuring which ethnic groups have been most exposed to COVID-19 infected residential neighbourhoods during the first and second waves of the pandemic in England. The index is based on a Bayesian Poisson model with a random intercept in the linear predictor, allowing for extra-Poisson variation at neighbourhood and town/city scales. This permits within-city differences to be decoupled from broader regional trends in the disease. The research finds that members of ethnic minority groups can be living in areas with higher infection rates but also that the risk of exposure is distributed unevenly across these groups. Initially, in the first wave, the disease disproportionately affected Black residents but, as the pandemic has progressed, especially the Pakistani but also the Bangladeshi and Indian groups have had the highest exposure. This higher exposure of the Pakistani group is not straightforwardly a function of neighbourhood deprivation because it is present across a range of average house prices. We find evidence to support the view, expressed in The Doreen Lawrence Review, that it is linked to occupational and environmental exposure, particularly residential density but, having allowed for these factors, differences between the towns and cities remain.

A spatial analysis of variations in health access: linking geography, socio-economic status and access perceptions.

BACKGROUND: This paper analyses the relationship between public perceptions of access to general practitioners (GPs) surgeries and hospitals against health status, car ownership and geographic distance. In so doing it explores the different dimensions associated with facility access and accessibility. METHODS: Data on difficulties experienced in accessing health services, respondent health status and car ownership were collected through an attitudes survey. Road distances to the nearest service were calculated for each respondent using a GIS. Difficulty was related to geographic distance, health status and car ownership using logistic generalized linear models. A Geographically Weighted Regression (GWR) was used to explore the spatial non-stationarity in the results. RESULTS: Respondent long term illness, reported bad health and non-car ownership were found to be significant predictors of difficulty in accessing GPs and hospitals. Geographic distance was not a significant predictor of difficulty in accessing hospitals but was for GPs. GWR identified the spatial (local) variation in these global relationships indicating locations where the predictive strength of the independent variables was higher or lower than the global trend. The impacts of bad health and non-car ownership on the difficulties experienced in accessing health services varied spatially across the study area, whilst the impacts of geographic distance did not. CONCLUSIONS: Difficulty in accessing different health facilities was found to be significantly related to health status and car ownership, whilst the impact of geographic distance depends on the service in question. GWR showed how these relationships were varied across the study area. This study demonstrates that the notion of access is a multi-dimensional concept, whose composition varies with location, according to the facility being considered and the health and socio-economic status of the individual concerned.

The impact of community-based outreach immunisation services on immunisation coverage with GIS network accessibility analysis in peri-urban areas, Zambia.

BACKGROUND: Accessibility to health services is a critical determinant for health outcome. OBJECTIVES: To examine the association between immunisation coverage and distance to an immunisation service as well as socio-demographic and economic factors before and after the introduction of outreach immunisation services, and to identify optimal locations for outreach immunisation service points in a peri-urban area in Zambia. METHODS: Repeated cross-sectional surveys were conducted for two groups of children born between 1999 and 2001, and between 2003 and 2005.The association between immunisation coverage for DPT3 and measles, and access distance, child sex, female headed households, and monthly household income were assessed using logistic regression analysis. Optimal locations for outreach service points were identified using GIS network analysis and genetic algorithms. RESULTS: Before the introduction of outreach services, longer distances to the service points were associated with lower DPT3 and measles immunisation coverage (OR=0.24, 95% CI 0.10 to 0.56, p<0.01 for DPT3; and OR=0.38, 95% CI 0.17 to 0.83, p<0.05 for measles). However, access distances were not an impediment to immunisation coverage once the outreach services were introduced. The average distance to immunisation services could be decreased from 232.3 to 168.4 metres if the current 12 outreach service points were repositioned at optimal locations. CONCLUSION: Access distance to immunisation services was a critical determinant of immunisation coverage in a peri-urban area. Intervention via outreach services played an important role in averting the risk of missing out on immunisation. Optimal location analysis has the potential to contribute to efficient decision making regarding the delivery of immunisation services.

Using genetic algorithms to optimise current and future health planning--the example of ambulance locations.

BACKGROUND: Ambulance response time is a crucial factor in patient survival. The number of emergency cases (EMS cases) requiring an ambulance is increasing due to changes in population demographics. This is decreasing ambulance response times to the emergency scene. This paper predicts EMS cases for 5-year intervals from 2020, to 2050 by correlating current EMS cases with demographic factors at the level of the census area and predicted population changes. It then applies a modified grouping genetic algorithm to compare current and future optimal locations and numbers of ambulances. Sets of potential locations were evaluated in terms of the (current and predicted) EMS case distances to those locations. RESULTS: Future EMS demands were predicted to increase by 2030 using the model (R2 = 0.71). The optimal locations of ambulances based on future EMS cases were compared with current locations and with optimal locations modelled on current EMS case data. Optimising the location of ambulance stations locations reduced the average response times by 57 seconds. Current and predicted future EMS demand at modelled locations were calculated and compared. CONCLUSIONS: The reallocation of ambulances to optimal locations improved response times and could contribute to higher survival rates from life-threatening medical events. Modelling EMS case 'demand' over census areas allows the data to be correlated to population characteristics and optimal 'supply' locations to be identified. Comparing current and future optimal scenarios allows more nuanced planning decisions to be made. This is a generic methodology that could be used to provide evidence in support of public health planning and decision making.

Can fluorescence spectrometry be used as a surrogate for the Biochemical Oxygen Demand (BOD) test in water quality assessment? An example from South West England.

The fluorescence intensities of tryptophan-like, tyrosine-like and humic-like materials were determined using excitation-emission-matrices (EEMs) for a wide range of samples including natural surface waters, sewage and industrial effluents and waters that have experienced known pollution events from the South West of England (n=469). Fluorescence intensities reported in arbitrary fluorescence units (AFU) were correlated with standard five day Biochemical Oxygen Demand (BOD(5)) values which were used as an indicator of the amount of biodegradable organic material present. Tryptophan-like fluorescence, which has been found to relate to the activity of the biological community, showed the strongest correlation with BOD(5). Fluorescence analysis of the tryptophan-like peak (excitation/emission wavelength region 275/340 nm) is found to provide an accurate indication of the presence, and relative proportions of bioavailable organic material present (natural or anthropogenic). It therefore provides an insight relating to its oxygen depleting potential. Thus fluorescence spectroscopy is recommended as a portable or laboratory tool for the determination of the presence of biodegradable organic matter with intrinsic oxidising potential in natural waters. The novel application of Geographically Weighted Regression (GWR) to the data illustrates that strong local relationships exist between the two parameters and that site specific character may be a strong factor in the strength of the tryptophan-like fluorescence/BOD(5) relationship.

Geographically weighted visualization: interactive graphics for scale-varying exploratory analysis.

We introduce a series of geographically weighted (GW) interactive graphics, or geowigs, and use them to explore spatial relationships at a range of scales. We visually encode information about geographic and statistical proximity and variation in novel ways through gw-choropleth maps, multivariate gw-boxplots, gw-shading and scalograms. The new graphic types reveal information about GW statistics at several scales concurrently. We impement these views in prototype software containing dynamic links and GW interactions that encourage exploration and refine them to consider directional geographies. An informal evaluation uses interactive GW techniques to consider Guerry's dataset of 'moral statistics', casting doubt on correlations originally proposed through visual analysis, revealing new local anomalies and suggesting multivariate geographic relationships. Few attempts at visually synthesising geography with multivariate statistical values at multiple scales have been reported. The geowigs proposed here provide informative representations of multivariate local variation, particularly when combined with interactions that coordinate views and result in gw-shading. We argue that they are widely applicable to area and point-based geographic data and provide a set of methods to support visual analysis using GW statistics through which the effects of geography can be explored at multiple scales.