The overarching aim for our methodological approach is to impose consistent rules across disparate data sources. In other words, the rules for how we source our data are loose enough to accommodate data in different forms while strict enough to fit them into a coherent dataset.
When there is an absence of data on the map, this does not reflect a declared lack of presence according to a data source; instead, it simply reflects the fact that there is insufficient information.
The map reflects countries that have full United Nations recognition. We made this decision for consistency across multiple data sources and for political neutrality.
We have five categories in our dataset, which are defined as follows:
Presence - A binary question asking if the U.S. military conducts operations, both hostile and non-hostile, on the ground or in the airspace of a country.
Troop numbers - Uniformed military personnel, not contractors or Department of Defense civilian employees.
Troop casualties - Deaths, not injuries, of uniformed personnel. The country associated with this category is the country where the incident leading to death occurred and not where the death actually occurred. For example, if a troop was shot in Afghanistan but died in a military hospital in Germany, the death would be counted in Afghanistan.
Adversary casualties - Deaths, not injuries, of enemy combatants.
Civilian casualties - Deaths, not injuries, of civilians not directly associated with U.S. military operations.
Finally, all of the data we use is publically available and accessible through Internet searches. We are simply presenting it in one place in a way that is easy to interpret.
The data we use can be found from the following sources: