I am an Assistant Professor in the Department of History and Politics at Drexel University. In 2011, I completed my Ph.D. in Political Science at Yale University. I work as a consultant to the Human Rights Data Analysis Group. I have also worked with groups like UN Action Against Sexual Violence in Conflict as a statistical consultant. My dissertation considered variations in the types of violence that armed groups use against noncombatants, showing that political education programs can limit repertoires of violence against civilians during armed conflicts. More generally, I am interested in how information, ideologies and behavioral norms are transmitted through hierarchical organizations (see my Research page and my CV for more information).
14 June 2013: Open Letter and a New Book Chapter!
Recently, University of New Mexico Associate Professor Geoffrey Miller (who is currently visiting at New York University) wrote a scary and offensive tweet claiming that "obese PhD applicants" would not be able to finish dissertations because they lack "willpower." I responded to the tweet with an open letter calling on Miller to recuse himself from positions in which he might evaluate students or colleagues, particularly "peer review" processes such as tenure and promotion committees. Obviously, both the tweet and my response fall outside my academic research, but appearance bias is an important issue for all of us in academia. In addition, I have a chapter in an excellent new edited volume, Counting Civilian Casualties, which is just out from Oxford University Press. In the chapter, we outline a number of instances in which multiple convenience datasets disagree about basic patterns of violence, such as trends over time and space.
11 March 2013: Data-blogging!
I've been doing a lot of writing, lately, about the ins and outs of human rights data. Today, my colleagues at the Human Rights Data Analysis Group released the first entry in a five-part series of frequently asked questions on Multiple Systems Estimation. Today's entry is the bare bones, including (importantly!) a definition of statistical inference. Over the next two weeks, my posts will cover data collection, cleaning and canonicalization, matching across datasets, stratification, the estimation process and the validity of the technique. In a similar vein, last month I published a guest post about sexual violence data for the Women Under Siege blog, dramatically titled "The Devil's in the Data". It explores how elements of rape culture are reproduced in rape statistics -- and why we should be wary when making inferences from those data.
8 February 2013: New USIP Special Report
Today, the USIP released a Special Report on wartime sexual violence that I co-authored with Dara Kay Cohen and Elisabeth Wood. The report, titled Wartime Sexual Violence: Misconceptions, Implications and Ways Forward, discusses ten key misconceptions that members of the public and policy-makers tend to hold about wartime rape and other wartime sexual violence. Drawing on research findings from all three co-authors and several others, we argue against several "conventional wisdoms." Particularly important, I think, are our findings that (1) state armed groups are more likely to be reported perpetrating sexual violence than rebel groups, (2) widespread sexual violence does not necessarily imply that sexual violence is used "as a weapon of war," and (3) that the evidence for a global increase or decline in sexual violence is unsound, and we should focus on local variation.
16 December 2012: Reflections on the Shooting in Newtown, CT
Having received a few requests to reflect on the tragedy in Newtown, CT, that occurred on December 14, I wrote a short piece about whether, and how, my research on wartime violence speaks to peacetime mass killings (and/or peacetime killing more generally). My effort---which is not intended as an academic exercise but does include references to a few key findings---is here. From my vantage, which is essentially "What can research tell us about policy responses to gun violence?", the main point to consider is that highly visible crimes are often highly unrepresentative. It's right that we should be shocked at this violence, that we should mourn it and look toward solutions to mass killing, but we need to remember that this is not representative of gun violence in this country.
14 December 2012: AGID Crowd-Funding Experiment: Success!
We reached our goal of $3,607 (25% of research assistant Arhama Rushdi's six-month salary) with hours to spare today, and went on to raise a total of $3,790 from 71 different donors. Thank you to all the donors who made this possible! Their names, excepting those who wished to remain anonymous, are listed on our Thank You, Donors page.