Is the spread of COVID-19 across countries influenced by environmental, economic and social factors?

Mohammad Alamgir Hossain

The copyright holder for this preprint . https://doi.org/10.1101/2020.04.08.20058164 doi: medRxiv preprint 11 realization of the process through which data is generated 11 . The realized value is a random selection from a range of possible values that could be generated by the interaction of all factors. Though it is not easy to reveal the true characteristics of the underlying data generation process, the realized or observed values can be utilized to draw inferences about the process. As a stochastic process the important determinant of the number of cases of a day is its own past values, implying that Yt, the number of cases on a day, depend on the past values, i.e. Yt-1, Yt-2, Yt-3 etc. Hence, an autoregressive model may be suitable to approximate the data generation process of the cases of infection on a day. Nevertheless, our aim in this study is not to find the data generation process of total number of cases of a country on a specific day. Rather, our objective is to examine data to investigate any possible linkage between country level severities of infection with few selected environmental, economic and social variables. For this purpose, we first simply regress total number of cases of infection per one million people by countries reported on a recent day (03 April 2020) on our selected explanatory variables. Specifically, we estimate the following regression-= + 1 1 + 2 2 + 3 3 + 4 4 + 4 5 +