Regression model - Correlation Between Age and Coronavirus Cases in Australia

Abstract

This study addresses one of the most important and topical issues of modernity - COVID-19. As part of the study, we tried to statistically describe and analyze one of the largest databases related to the virus - the case of Australia. The aim of the research is to raise an awareness about this still unknown phenomenon and to turn it into a kind of recommendation document for its practical implementation in the future.

More specifically, the main focus of the study is the correlation analysis of the number of people infected with the virus and their age groups, thus determining whether the relationship between these two variables exists, for which the T-test had been calculated using some of the Excel tools. In order to have a better visualization and maximum degree of reliability achieved of the results, we came up with another variable - gender and compared the cross-tabulations of infected men with the cross-tabulations of infected women.

The study found that there was no such correlation between age and COVID-19-infected cases (since we had to reject the null hypothesis at a 20% significance level, while for women the null hypothesis has been validated at both 10 and 20% significance). Therefore, we had to use the so-called "scatterplots", or graphs, after standardizing the data obtained and minimizing the statistical difference between the genders. Indeed, in both cases the line of reference between age and ‘coronavirus’ cases had a negative direction. And the best fit line had a negative slope too. Moreover, in both cases we obtained a linear scattering, which gives us legit reason to conclude and assume that there is a regular correlation between these two variables (age and COVID-19 cases).