新型コロナウイルスの空間的相関に関する分析 : 第1波，第2波に関する分析

In this paper, we apply spatial statistical methods such as Moran’s I to examine the geographical pattern of the spread of COVID-19 from metropolitan to rural areas. We first characterize the geographical distribution and spread of COVID-19 in time series data from January 28 to February 1, 2021, and then test the statistical significance of the spatial correlation patterns (regional correlation using Moran’s I). In light of the p-value of Moran’s I, our results suggest the number of infected people was inversely proportional to the power of regional transmission of COVID-19 until June 23, which is in line with theoretical expectations. However, from June 30 to August 31, the p-value of Moran’s I statistic and the number of infected people became proportional. There is a possibility that the spread had become chronic in the surrounding cities and that therefore the spatial correlation was weakened to statistical insignificance.