Introduction

Female representation in national parliaments is a topic I found interesting to explore this year. I find it particularly interesting to see whether female representation in parliaments around the world align with my expectations or not. With the recent update of the US vice president-elect being not only the first woman in her position but also a woman of African-American and South Asian descent, this topic just became even more relevant. “Women In Parliament Around The World” is a report I have previously created on this topic. With this study I am exploring a different tool to determine whether it better represents my data and the information I want to present.

Process + Tools

The dataset used was the same one used for my previous report, Proportion of seats held by women in national parliament found on UNdata. It shows the proportion of seats held by women in each country as a percentage over 9 years (1990, 2000, 2005, 2010, 2015-2019).

The tool used to create the map is Carto, a web-based platform used to visualize and analyze geospatial data. Unlike the tool used for my previous report, Tableau, Carto required a shape-file to determine the shape of each country on the world map. Initially I used a world map found in Carto’s own database, however having come upon a glitch I was unable resolve, I had to source the world map from elsewhere. I ended up using Countries WGS84 found on ArcGIS.

Using Carto, I visualized my data in the form of a cloropleth, keeping colors consistent with my previous work on this topic. I added two “widgets” which work as filters to pick particular sections of the timeline and particular countries.

Results

Click here to view the map I created.

The Carto basemap used to create this visualization was chosen as such because the dark blue makes a strong contrast against the pink color palette of the data represented on it. This makes it easier to view and map and make differentiations. It was also chosen based on the fact that countries are not labeled, keeping it clean and making it easier to focus on the color contrast between each country. The Timeline widget was chosen because the Carto histogram feature did not allow for years to be separated and labeled individually. The years were automatically put into a number of “buckets”. Ideally, a user should be able to pick a year and view the map for that year. The Countries widget was chosen so that users can pick a country and view it on the map as well as find their female representation data. Putting the Timeline before the Countries widget also allows users to view the countries with top representation based on the selected time period.

The results I got from plugging my dataset into Carto looked similar to but not exactly like that on Tableau. For example, when the rightmost section on the Timeline is selected (which represents the most recent time period) the Countries widget shows countries with the most female representation in parliament. The results look similar to a visualization I created of the top 10 countries with female representation in parliament previously. As shown in the screenshots below, Carto puts Mexico in third position on the list while the Tableau does not show Mexico at all.

I deduce that this misalignment may be because of the different ways the two tools are analyzing the data. Carto may be showing the highest percentage reached by each country during the most recent period of years, while Tableau shows the exact value from the selected year. However, upon double checking with another visualization I created on Tableau, I found that Tableau’s top 10 list did not align with the original data. As shown in the graph below, Mexico does fall into third place according to the most recent data, proving Carto’s analysis correct.

As shown in the GIF below, Carto’s map represents how female representation spread over time and increased overall, however it does not show the progress of each country.

It seams that the coloring of each country shows proportion of female representation in relation to the rest of the world rather than to its own progress. The images below show this. The image on the left shows earlier data for the US, which has a smaller percentage of female representation. The image on the right shows more recent data, where the percentage is higher. However, the color of the US on the left is darker than on the right. As the colors don’t correlate with the numbers, I deduce that the coloring of each country on the map is in relation to the other countries instead of its exact value.

Reflection

Using Carto to create this map allowed me to understand which tool would work better to represent the data I am interested in. The quality of the map and the styling options for the map were two features I appreciated about Carto. Being able to select every color for the legend and the number of colors made it easy to control the data representation. This is not a feature that Tableau offers. However, the widgets on Carto did not have a way to be customized, making the process of understanding and selecting data difficult. Users are unable to tell what years they have selected on the timeline, they can only assume the rightmost bar is the most recent based on the widget title. Another limitation I found with Carto was with that the Countries widget did not have a dropdown menu option beyond the 5 countries it displays. It had a search function, but this is not the easiest to use as country names officially change over time and may not align with what a user expects (For example “Republic of Korea” instead of South Korea).

When it comes to the coloring of the map, I prefer the way Tableau does it over Carto. The cloropleth is the visualization I want to prioritize with my research on this topic. Having it represent not only the worldwide growth in female representation in parliament but also the growth of each country over time is important. The way Carto does it, the coloring compares countries to one another for selected time periods but does not show each country’s growth. For this reason, and also for being able to create multiple filters and visualizations using one dataset I think Tableau would be a more suitable tool for my research.