The field of statistics isn't usually thought of as helping to bring down dictators, but the work of Jana Asher, a doctoral student in Statistics, may have done just that. Asher, along with Patrick Ball, a statistician with the American Association for the Advancement of Science (AAAS), and researchers from the University of Chicago and ABA/CEELI (the American Bar Association's Central and East European Law Initiative), helped to analyze data from Kosovo from the Spring of 1999, as part of the war crimes trial of Slobodan Milosevic, the former leader of Yugoslavia.
The goal of the researchers was to examine the hypotheses of the International Criminal Tribunal for the Former Yugoslavia concerning mass migrations of ethnic Albanians during the months of March to June. The three hypotheses of the tribunal dealt with NATO bombings, actions of the Kosovo Liberation Army, and a systematic attempt by the government to purge the country of ethnic Albanians. Ball's team had to see if their data was consistent with or contradictory to any of these hypotheses.
Using a technique called Multiple Systems Estimation, Ball's team discovered that the data that they had was consistent with the third hypothesis, that a systematic campaign of massacres and expulsions was responsible for the exodus of the ethnic Albanians in question. "We were pretty sure Milosevic had done this," says Asher. "But as researchers, we had to stay pretty impartial."
The team used four separate lists to determine how many Albanians had died during this period. Asher used the overlap between the lists to estimate the total number of deaths. Once the team knew how many had died, they could begin examining when many of the peopled died, and attempted to coordinate the patterns of deaths with the exodus of Albanians. Interestingly enough, the two factors peaked almost completely in sync with each other during the four months in question.
"She is a terrific data analyst," Ball says of Asher. "Last fall, we had a particular data problem, and there was an obvious way to look at it, and a right way to look at it. Jana thought of an important innovation that allowed us to use the method [multiple systems estimation] in what we believe is the right way."
Ball presented these findings at the trial of Milosevic on March 13, where he was the lone statistician in the midst of many Kosovar Albanian witnesses, many of whom were peasants who had experienced the exodus. "Patrick came back and said that Milosevic was just [completely intimidating] the peasants," reports Asher. "But he came in and gave quantitative data of the overall picture for the country, and I think that was very compelling."
Asher had worked with Ball before, as part of her Ph.D curriculum. She assisted him with data analysis for "Country X," and, following that, Ball asked her for her help with analyzing the Kosovo data. Asher had always been interested in doing human-rights work in the field of statistics, and this offered her the perfect opportunity.
Asher enrolled as an undergraduate at Carnegie Mellon University in 1996 to become an actuary to support her family. A year later, she was accepted into the graduate program in Statistics, where she earned her master's degree. Following that, she worked for the United States Census Bureau as a mathematical statistician, but returned 18 months later "to pursue a Ph.D in a field that I have grown to love."
"I started out as this single mom who desperately needed a job, and was lucky enough to get into grad school. I find this work very, very, fulfilling," Asher says. Currently, she's involved in other human rights projects, including one for the National Commission for Human Rights in Mexico, where she is measuring human rights violations in the country, and examining if Mexico's record on human rights is improving or not. She also plans to continue collaborating with Ball.