Leslie Curry (1923–2009): Expounder of the Random Spatial Economy and Spatial Autocorrelation

Spatial autocorrelation mixtures in geospatial disease data: An important global epidemiologic/public health assessment ingredient?

“…In addition, regarding meaningful prominent details, the estimated PSA increases from 0.79 to 0.90 (i.e., a magnitude in much closer agreement with that for most remotely sensed data sets), and the implicit default value of zero NSA increases in degree to −0.32. With regard to hypothesis testing, medical geography conceptualizations conjecture a negative relationship between distance from a toxic environmental source and medical outcome responses (e.g., deaths from cholera); Snow's data imply a negative exponential decline in death density with distance from the Broad Street pump when augmenting the foregoing pure PSA specification with a distance covariate, its distance decay parameter estimation being −0.37, with an accompanying reduction in its residual PSA estimate to 0.54 (as expected, given arguments advanced by Curry [Griffith, 2023]). A shift in the inferential basis ensuant to the PSA-NSA respecification here has a distance decay parameter estimate of −0.27 with a 20% increase in its standard error, a revised PSA estimate of 0.79 (down from 0.90), reflecting part of this pure SA being confused through convolution with a distance decay geographic trend in Snow's landscape, and a slightly increased NSA estimate of −0.37.…”

“…These data were derived, in part, from the following resources available in the public domain: https://geodacenter.github.io/data-andlab/snow1/; https://www.cdc.gov/tb/statistics/default.htm; https://dataverse.harvard.edu/dataset.xhtml?persisten-tId, doi:10.7910/DVN/MR5IJN. How to cite this article: Griffith, D. A. (2023).…”