With so many conflicts around the world today, concern for people forced to flee their homes has seldom been greater. Statistician-activist David Banks—a member of the ASA’s Committee on Scientific Freedom and Human Rights—highlighted the increasing number of refugees as one of the most pressing issues the statistical sciences can address today. Statisticians and data scientists in academia and government and at nongovernmental organizations are rising to the challenge by finding data, analysis, and statistical support for humanitarian agencies. This month, we look at ways to support refugees and other displaced persons.
The UN Refugee Agency, UNHCR, maintains a website with a wealth of data and other resources for researchers. The Data Finder can be queried and the data downloaded. Be sure to check out the methodology section for necessary background on their data collection and valuable instruction on best practices for D4G projects in general.
Another important UN resource is the International Organization for Migration. Its mission is to “help ensure the orderly and humane management of migration to promote international cooperation on migration issues, to assist in the search for practical solutions to migration problems, and to provide humanitarian assistance to migrants in need, including refugees and internally displaced people.” The organization features a data portal with publicly available data sets, interactive dashboards, and analytic reports.
When working in this area, it’s important to pay close attention to the language used. “Refugee” typically refers only to people who have crossed international borders, while some others have been internally displaced. The UN estimates there are millions around the world who are considered “stateless”: people who have been stripped of their citizenship, denied citizenship by their home country (e.g., Rohingya in Myanmar), or whose country of citizenship no longer exists. All these categories have a lot of overlap, with many people in multiple but not all categories. Refugees may or may not be forcibly displaced. Stateless persons might not be refugees—that is, not displaced internationally. United Nations agencies have data and resources for all categories, so it’s important to understand the language to know who is being included in a particular data set.
Combining learning from these resources with the lived experience of people affected will provide the best analytic results. For example, like many families, mine came to North America as refugees. In our case, that was 400 years ago (we still haven’t forgotten), so it’s important for me to be connected with a local immigration and refugee task force to do sound analysis.
In working with country-level data in Data for Good, a statistical method I have found useful is matched case control studies (MCCS). Most often, these studies take people with a particular concern, often a health issue, and match each one to the most similar person they can find who is not affected. Data is then mined to find consistent patterns distinguishing the affected and not affected groups.
MCCS is a well-established methodology in public health research, with many examples, webinars, and tutorials available. In the European Journal of Epidemiology article “Case-Control Matching: Effects, Misconceptions, and Recommendations,” Mohammad Mansournia, Nicholas Jewell, and Sander Greenland review the strengths and weaknesses of MCCS and how to address common problems.
Applying MCCS to country-level data begins with identifying a list of countries where the concern is most severe. This list serves as the affected group. The control group is created by matching each affected group country to the most similar country with a much smaller prevalence or severity of the problem. Country-level data is added for each country from multiple sources—some of the best are the Open Data website from the World Bank and the CIA World Factbook.
The humanitarian crises of refugees and other displaced persons can seem overwhelming. Despite the challenges, many statisticians and data scientists are getting involved with high-impact research, analysis, and statistical support for humanitarian organizations. There are so many opportunities for research that can make a critical difference in the lives of people today with Data for Good!
Getting Involved
In opportunities this month, check out how climate change is becoming a major driver of refugees and displaced persons today. The UN’s sustainable development goals intersect so many important issues, you are sure to find one to match your interests.
Also, the ASA’s Committee on Funded Research has a funding opportunities group that offers a daily digest of solicitations for research funding opportunities and a site with external funding sources
David Corliss
With a PhD in statistical astrophysics, David Corliss works as a data scientist in industry. He serves on the ASA Board as a Council of Chapters representative and is the founder of Peace-Work, a data for good nongovernmental organization.
This column is written for those interested in learning about the world of Data for Good, where statistical analysis is dedicated to good causes that benefit our lives, our communities, and our world. If you would like to know more or have ideas for articles, contact David Corliss.
Leave a Reply