Considering the plots and regressions results I have conceived, it is to note that with increase of both Internet Penetration and Internet Freedom, Democracy decreases. Now, looking at the regression results, by using Internet Penetration and Internet Participation can be predicted only 18% of the variance. When adding Internet Freedom to the predictors, the rate becomes 83%. The model also shows that
with the increase of Internet Participation, Democracy Index decreases by 2.58;
with the increase of Internet Penetration, Democracy Index increases by 0.016;
with the increase of Internet Freedom, Democracy Index decreases by 0.086.
Thus, the results are quite controversial as it is considered that in “free” societies Internet is also a free space.
As it can be clearly seen from the boxplot, countries with democracy regimes have higher Human Development Indices. Based on that, I can check whether HDI is associated with Democracy rates in a country. After checking for correlation via the scatterplot (which clearly shows that HDI is in a positive relationship with Electoral Democracy Index), I built a linear regression model with such predictors as
Electoral Democracy Index (ELD),
Liberal Democracy Index (LD),
Egalitarian Democracy Index (EGD)
Democracy Index (which required a few iterations to find the model which explains the maximum of variance)
In the end, I came up with a model with 52% of variance explanation. The most interesting findings are that with an increase in ELD and Democracy, HDI decreases by 0.756 and 0.03 respectively (keep in mind that it is not in the actual scale, meaning that I would need to convert it to count). On the other hand, With an increase in LD and EGD, HDI increases by 0.15 and 0.8 respectively.