Taylor & Francis Group, Statistics and Public Policy, 1(1), p. 69-77, 2014
DOI: 10.1080/2330443x.2014.928245
Full text: Download
This study uses disease surveillance cluster analysis methods to identify clusters for the three most common pediatric cancers in Florida (brain tumors, leukemia, lymphoma). In addition to a univariate purely spatial analysis, a space–time analysis is done, followed by a nonparametric permutation test for space–time interaction. This is followed by a multivariate spatial analysis for the three pediatric cancer types together. The main findings include a significant spatial cluster for each of the three cancer types covered in this study, in addition to identifying two significant clusters for all pediatric cancer types combined.