Brower, A. M., & Carroll, L. (2007). Spatial and temporal aspects of alcohol-related crime in a college town. Journal of American College Health, 55(5), 267-75.

This paper addressing the relationship between drinking and student crime in Madison (Wisconsin), making two major arguments:

1. More serious crimes occur at the intersection between high density student areas and high density drinking outlets.
2. Non-student residents living near students submit noise complaints at around midnight when they are going to sleep and student parties are still continuing.

Methodology:

The methodology was very basic – effectively mapping reported crimes by time onto a basemap in ArcGIS.

Results:

1.  There was a peak in serious crimes  when bars closed downtown, which was identified as a high density cluster for drinking and crime incidents.
2. Noise complaints, as expected, were highest between 9pm and midnight. 

Although the paper lead to genuine zero-tolerance policies by the University and the city, the authors underestimated the statistical potential of GIS, as they presented it as a merely descriptive tool. As such we assigned the paper a mark of 5/10.

Rosero-Bixby, L. (2004). Spatial access to health care in Costa Rica and its equity: a GIS-based study. Social Science & Medicine, 58(7), 1271-1284.

Kevin presented a paper that used GIS analysis to investigate differential access to health care in Costa Rice. The data consisted of census data and health care provision data, with missing data acquired through the use of telephone interviews. The author also conducted random field visits to hospitals and did some groundtruthing for the GPS coordinates.

The author built a model to simulate distance to the nearest healthcare facility, mapping proximity to outpatient facilities. The smallest distances were clustered around San Jose but were otherwise very sparsely distributed. The model also accounted for visitor’s satisfaction and the choices they make when selecting between healthcare facilities.

The results indicated that 12-14% of the population of Costa Rica were underserved. On a temporal scale, inequity declined by up to 30% as a result of reforms in 1995 and 1996, although this number has been steadily increasing in recent years.

Overall, Kevin gave this paper a 9/10 because the models were well designed and the data was carefully validated. However, some decisions (i.e. cut-off values in the models) appeared to be selected arbitrarily.

Gilbert, A., & Chakraborty, J. (2011). Using geographically weighted regression for environmental justice analysis: Cumulative cancer risks from air toxics in Florida. Social Science Research, 40(1), 273-286.

This paper draws on previous literature about inequality in health geography to investigate the risk of cancer risk  from ambient exposure to air pollutants between different socioeconomic groups across Florida, focusing on the following 2 research questions:

1. Are cumulative lifetime cancer risks from ambient exposure to multiple sources of hazardous air pollutants distributed inequitably with respect to race/ethnicity and socioeconomic status across Florida?
2. How does the direction and significance of statistical relationships between cumulative lifetime cancer risk and race/ethnicity or socioeconomic status vary across the state of Florida?

Methodology:

Using modelled cumulative air cancer risk and socio-economic factors such as racial structure of communities, the authors conducted two forms of regression analysis. Ordinary Least Squares was used to acquire a global fit and Geographically Weighted Regression was used to determine whether the difference in the fit between the explanatory variables and the overall cancer risk changed differed across the state.

Results:

1. There is a positive and significant relationship between the proportion of Black, Hispanic and Asian residents and overall cancer risk and a negative and significant relationship between population density and number of people below the poverty line and overall cancer risk.

2.  In comparison to the OLS model, the GWR model fit the data disproportionately across the           map, with hotspots in urban centres like Miami. 

This paper makes a convincing case for GWR and the authors use GIS software to produce clear and legible maps, but more care could have gone into the selection of model data. The authors elaborate on the theoretical element of GWR but this soon becomes repetitive. As such, this paper has been given a rating of 7/10.