If you have read our recent post on dealing with difficult datasets, you will remember our efforts to create a visualization of human trafficking data taken from the Secretary of State’s Trafficking in Persons (TIP) Report.

I had sourced the data on different types of human trafficking from Richard Frank’s “Human Trafficking Indicators, 2000-2011: A New Data Set” and created an interactive visualization where a user could look at different human trafficking factors and see them on the map.

This is when we realized the potential of the visualization. Rather than just reporting information, we could combine secondary research to design with the data.

This lead me to an interesting dilemma — how can we add additional information to a dataset without manipulating it in an irresponsible way?

From this question, I ended up experimenting with the visualization and the creation of the “Badness Map” (for lack of a better name at the moment). This was a simple test to see if I could combine existing parts of the data set to reveal additional information. For this map, I combined in D3 several variables from the data set such as “Victim Protection”, “Victim Prosecution”, “Victim Services” as well as some assorted others and scored the severity of the variable on a 1-5 ratio. I realized that the most interesting revelations from the data came from visualizing more than one dataset at once and producing relative visualizations.

An easy to understand example would be to relate the dataset information about the number of laws preventing human trafficking in a given country to the quality of that country's enforcement of those laws. While it does look good if a country has a large number of laws to protect trafficked persons, what does it mean when none of those laws are enforced? Is to better to have less laws around human trafficking that are being enforced better?

Up until this point the data was been seemingly flat in it’s presentation of information. If a country has a high GDP and a stable government, they would often receive a high score in on the Human Trafficking Index. However by integrating additional data sets, I am beginning to make new and interesting inferences.

Some topics I am invesitgation are:

  • Does a lack of access to education drive the proliferation of forced domestic labor in the developed world?
  • Is the social taboo surrounding sex workers enabling human traffickers?
  • Are our instatiable desires for raw materials in the Western World driving child labor in the developing world?

Just two weeks left in class but this project is getting exciting!