A Money Media Project

Ming Li

March 22, 2020

Part 1: Data analysis

Please review the data and pitch three stories that you would want to write.

1. American Voting Age Population Decreases in Every Blue State

- A thesis/nut graf

The percentage of voting age population has decreased in all Democratic states in the US, census surveys have showed. Meanwhile, five Republican states have seen an increase of the ratio of voters in its population from the 2007-2011 survey to 2012-2016 survey.

- Two ideas for potential charts

* A bar chart of top 10 states with worst decrease of the share of voting age population.

* A small multiple charts with two maps. One is a US map of states colored with blue(Democratic)/red(Republican). Another is a US map of states also colored with blue (decrease in voters’ ratio) and red (increase in voters’ ratio). You will easily find that all Democratic states have seen the percentage decreased.

- A list of what additional information you would need in order to write each story. For example: Are there other datasets you should look at? Are there certain types of sources you would like to talk to?

* I’ve done the data analysis and found that the percentage of voting age population only decreased in five states.

I need another dataset of which US states are Democratic states / Republican states / swing states in history. Because the original data is from a 2012-2016 survey, so the second dataset should also be from the same period. For example, a 2016 Gallup survey.

* To make the story more comprehensive, we could interview political science expert about how the demographic change, if continues, could impact the upcoming election.

2. Asian population is the fastest growing racial group in US

- A thesis/nut graf

Data has showed that Asian population has become the fastest growing racial group in US from the census’s 2007-2011 survey to 2012-2016 survey. Hispanic population is the second fastest growing group.

- Two ideas for potential charts

* A bar chart of population growth rate of different racial groups, highlight Asian, make the rest grey.

* A scatterplot about the population growth of various subgroups (Chinese, Japanese etc.) in the Asian population. Each dot represents a subgroup. X axis is the percentage of the group’s population in all Asians. Y axis is the group’s population growth from 2007-2011 survey to 2012-2016 survey.

- A list of what additional information you would need in order to write each story. For example: Are there other datasets you should look at? Are there certain types of sources you would like to talk to?

* A more detailed dataset about subgroups of Asian population will be interesting to look at, so that the second scatterplot could have more dots thus become more comprehensive.

3. Uninsured Adult Americans by State

- A thesis/nut graf

On average, adult Americans in the southern states are more likely to have no health insurance compared to people in the north, a 2012-2016 census survey shows. Texas has the highest uninsured rate of 16% in the country, while Massachusetts has the lowest rate of 3%. In general, men are more likely to get uncovered than women.

- Two ideas for potential charts

* A US state map with all states colored by uninsured rate. (see below)

* A range plot about the gender gap of uninsured rate in each US state. (see below)

- A list of what additional information you would need in order to write each story. For example: Are there other datasets you should look at? Are there certain types of sources you would like to talk to?

* I need another dataset of the same 2012 – 2016 survey about health insurance coverage in each US state.

* To further explain why states are so different in health insurance coverage, it would be helpful to interview health care experts about the distinct health care systems in different states.

Part 2: Data visualization

1. A Map

How/Why Did I Make The Map

I used Pandas to clean and merge the dataframes. Then I used Qgis to create a US state map based on the nature earth spatial data. Finally, I put it into Adobe illustrator to make it prettier.

I chose to map because the data showed a very obvious geographical pattern, with the southern states on average have higher uninsured rates. Readers can also identify the uninsured level of each state from the map.

2. An Interactive D3 Chart

Source: US Census Bureau

By Ming Li

How/Why Did I Make The Chart

I used D3.js to make the chart. To make the chart interactive, I used d3.tip. When you hover over the circles, you can see the exact number the dot represents.

I chose the range plot because it is a very good way to show the difference in uninsured rates between men and women. It’s also easy to see how the gaps are different among states.

About the Project:

I used sawhorse to build the whole page.

Details of the datasets, codes and the original graphics are on the project's Github page.