Continuing from the previous posts, here I would like to share the findings and summary after using a simple Python code to manage the data.
After cleaning and checking available data from Gapminder, I decided to use the data from year 2014 with a total of 179 countries/observations. In that year, there were fewer missing data for all 7 variables (5 independents and 2 dependents), so I could get the most number of rows. The dependent variables are electricperperson and co2emission, while the independent ones are employrate, femaleemployrate, incomeperperson, urbanrate, and internetuserate (the explanation for each variable can be read in this post).
The histograms for each variable show that electricperperson, co2emission, incomeperperson, and internetuserate are having skewed-right/positive distribution, while the rest are in normal distribution. The positive distribution in electricperperson and internetuserate depicts the sad reality that a lot of people in this world are still having limited access to electricity and internet. Same thing with incomeperperson, there are way more people in low income and poverty globally.
We then are able to check the relationship between variables. From the graphs below, we can say there is strong positive correlation between Electric per Person and Income per Person, Electric per Person and Urban Rate, and Electric per Person and Internet User Rate. The more people earn money, the more they use electricity in their daily lives.
The same relationship is observed for CO2 Emission (see figure below). Higher income, urban rate and internet-user result in more carbon footprint. Meanwhile, employment rate and female employment rate seem not to have any effect towards electricity consumption and emissions. Thus, answering the research question that the urban lifestyle with higher income economy and internet usage tend to give impact towards the rise of electricity consumption and CO2 emission.
