Changes in the world people's average living standard and demographic structure in the Past Two Decades (2000-2020)
[WDI_longer]
[WDI_wider]

Figure-1

World Population Map

      In our project proposal, we asked questions about the world populations, gender, age distributions, level of economic standard, and how this variables correlated with each other. As the first visualization, we chose to plot a "population world map" which can give us an overview of some of the most important vairables that we care: population by country, female population, and male population.
      We designed a hybrid plot which was in the format of "scatter plot + pie chart" to achieve this goal. As shown above, the basis of this visualization is a world map. For each of the country, there is a pie chart showing the population information. The size of the pie chart represents the population of the country, and the location of the pie chart represents the location of the country (which maps with the country contour in the world map). Each pie chart has two slices: the blue slice represents the male proportion of the population and the red slice represents the female proportion of the population. This way we condensed 250+ country location, 250+ country population, 250+ x 2 gender population, 250+ x 2 gender population percentage into one single plot so that viewers can get some basic idea about these data.
      In our original plan, we are thinking about generating a static plot, and we want to make the pie charts' sizes linearly correlated with the countries' populations. However, in practice, we realized that if we do so, viewers will only be able to ge very limited information from this plot; and since there are countries having very huge populations such as China, India, etc. and there are also countries that have very small populations such as New Zealand, Mongolia, etc., making the pie chart size linearly correlated with the country population will make some small countries barely visible. Therefore we decide to make an interactive plot where viewers can hover mouse over the pie chart slices to get more information. And we did a square root transformation on the population to make the plot looking much more reasonable. In our final design, when hovering mouse over te oie chart, the name of the country, female/male population, and proportion of female/male to the whole population will be displayed, and each country have a reasonable-sized pie chart.
      With this visualization, one can easily spot some patterns of the data: China has the largest population in the world, India has the second largest population; the female to male ration for most of the countries are very balanced, etc.. This fullfills our expectation of our first plot to provide an overview of some critical variables that we are interested in our project.

Figure-2

Age Structure

      The age structure of a population refers to the proportionate number of people in different age groups in a given population over a given time period (2000-2020). It is a natural characteristic of a country's or region's population. The research of age structure and distribution is critical for determining potential school-age population, labor force, army, and electors, among other things. It can help people in the forecasting of population changes and population planning.
      From the first proposal to the midterm submission, and then to the final result, there were three changes to the Age Structure part.       In the proposal, a pie chart is used to roughly divide the age of the global population which allows the viewer to see the age group population comparison at a glimpse, allowing them to do an immediate analysis or quickly comprehend details.
      In the midterm part, the hand-painted pie charts of the proposal became reality, and more detailed age division was made. By generating interactive pie charts, we can see the total population in each age segment. There is button allows the specified age group to be hidden, emphasizing the contrasting age structure of the world population.
      But with a deeper understanding of our theme and a desire to provide users with more information, an age-sex pyramid (Figure-2) was built. The reason for the change to population pyramid was that pie chart by age group alone would not provide much useful information and would not support the entire story. And a population pyramid can give detailed information about the age and gender of a country's population. It is a simpler and easier method to give an overview than pie chart.
      In addition, it can be better related to the global male and female distribution of Figure-1 through gender differentiation. The colors of male and female correspond to the colors in Figure-1, and the color gradient is made for different population numbers, so that users can get a more concise and clear change information. It also gives us information about birth and death rates as well as life expectancy. A population pyramid tells us how many dependants there are which relates to the next part "New Birth in Five Continents - (Figure-3)". In figure-3, each horizontal line represents the proportion of male and female population of each age group in the whole population, each bar represents the size of population. It can be determined from the graph that which age group has the largest population, and which has the smallest. The top of the pyramid is very narrow, but it is evident that the aging population has started aging. A low birth rate is the main underlying reason.
      In the Population Pyramid process, female and male at different ages are extracted by reprocessing the data. Because the number of people in each age group is too large, the population data are processed in percentage. It can be plot directly when the statics are available for each age groups. The original intention was to generate an interactive image, but through various searches, the image generated by the package "Plotrix" in R Studio could not generate an interactive image, and the image generated by "Ploty" was not very successful and beautiful. In the end, plotrix was used to generate a picture, which could clearly express useful information and support the whole theme. Visualize the size of each age group and determines the age structure type, allowing the prediction of population growth and helping to understand the past, current, and future situations of a given population. It is an essential component in population control and policy making.

Figure-3

New Birth in Five Continents

      According to the proposal,to help policymakers predict and plan for population change, we need to reflect in detail on the growth trends of world population(Figure-4), Age structure changes(Figure-2), and population growth trends in different regions for the period 2000-2020. Therefore, to meet the third visualization requirement, we chose area maps to reflect population growth trends in the five continents. At last, Figure-3 was built up.
      The goal of the area map is to reflect the combined effect of fertility rate and population base on population growth (Birth rate information in Figure-5 and population base information in Figure-1).The Y-axis shows the number of new population, the X-axis shows the year, and the color is used as a symbol to distinguish the new population of the five continents, the product of XY-axis is the total new population of the five continents from 2000 to 2020.
      There were two significant changes from the first proposal to the final submission. During the sketch phase, it was decided to use different colors to distinguish the new population in the five continents. During the mid-point stage, it was found that the default colors did not have enough contrast and therefore reduce the readability, so a high contrast color scheme was chosen. In the final stage, the five continents were laid out in ascending order of the number of new birth people, making it easier for viewers to visualize the differences in population additions across the five continents.

Figure-4

World Population and GDP per Capita

      According to Proposal, it is necessary to visualize the world population trend from 2000 to 2020 and the GDP per capita trend to reflect the relationship between the two, and to explore the extent of the world population's contribution to the world economy to help governments designate appropriate economic development policies based on population development. After discussion within the group, it was decided to use a biaxial bar chart(Figure-4) to meet the above requirements.
      The biaxial bar chart, with year on the X-axis, world population on the Y1-axis, GDP per capita on the Y2-axis, and additional trend lines to show the trend of world population growth and GDP per capita growth. To make it easier for the viewer to identify the different measures, we decided to use colors to distinguish between GDP per capita and world population, with world population in orange and GDP per capita in blue.
      The figure went through two significant adjustments from the initial sketch to the final submission. In the sketch phase, it was decided to overlap the bars of world population and GDP per capita to show the trend of both. In the mid-term stage, it was found that overlapping the bars would obscure the trend of the world population and reduce the image's readability, so it was decided to adjust the transparency of the bars to improve the bar obscuration. In the Mid-term phase, it was decided to eliminate the world population bars and keep only the world population trend line to ensure visual simplicity.

Figure-5

GDP and Birth Rate in Developed & Developing Countries

      Based on our proposal, our goal is to figure out how the level of economic standard and development of countries around the world relates to demographics structure in the past twenty years. The most classic variable to describe a country's level of economic development is the total value of GDP. So, in this graph, we chose to combine GDP with the birth rate to conduct our research.
      Compared to our initial proposal, we chose to change the image style from a stacked linear image to a dynamic bubble image. This is because we later found that for the characteristics of our desired data, the bubble chart can better segment the feature points to show them on a country-by-country basis, and the dynamic effect is better to show the trend of GDP, birth rate, and the relationship between them over time.
      In contrast to the same plot in the mid-point submission, we re-logarithmically processed the GDP values in order to avoid blistering that could affect the visualization and to make the contrast more visible.
      We used population, GDP, country, year information in the "WDI_wider" dataset. The total population and GDP per capita of each country were further calculated from the original data, and the developed/developing labels of each country were confirmed based on these two parameters. We also changed the unit of GDP from one dollar to ten thousand dollar and converted the unit of population to per 10,000 people. Originally, we wanted to study the distribution, comparison and relationship between population and GDP in developed/developing countries, but found that the dispersion of population values was not uniform, and it appeared that most of the data piled up in a certain range, and a few data points reflected too large values, resulting in an unclear view and no observational meaning. So later we chose the birth rate, which is more statistically valuable and practically meaningful, as the comparison parameter.
      This bubble chart shows how each country's total GDP and birth rate changed over time from 2000 to 2020, and the size of the bubbles represents the size of that country's total population relative to other countries in that year. By clicking the "play" button, users can dynamically observe the trend of all indicators over time. And by combining GDP and birth rate, you can explore whether there is a correlation between them. When you mouse over each bubble, you can see information such as total GDP, total population (ten thousand people), birth rate, year, country name, and developed/developing country labels.
      From this graph we can learn that the difference in the value of total GDP in most developing countries is small and largely lower than in developed countries. Most of the countries with high birth rates are developing countries, and developed countries generally have lower birth rates. There are almost no countries with high birth rates and high GDP, and most bubbles are concentrated in low birth rates and low GDP.
      It is worth noting that the total GDP of developing countries is not necessarily less than that of developed countries. China's total GDP, for example, is outstanding and is accompanied by a huge population size.
      The temporal trend shows a decreasing trend in birth rates and a tendency for total GDP to increase in all countries globally.