Posts
Award given to article challenging Iraq mortality methodology
by Scott Gilbreath ~ December 15th, 2008
The Journal of Peace Research has given its 2008 Article of the Year award to Neil F. Johnson, Michael Spagat, Sean Gourley, Jukka-Pekka Onnela & Gesine Reinert for “Bias in Epidemiological Studies of Conflict Mortality”. The article challenged the methodology utilised by Gilbert Burnham et al. in their study of deaths in Iraq, published in The Lancet. Johnson et al. showed “convincingly” that the Lancet study’s methodology entailed a significant upward bias and, therefore, that its mortality results may be seriously overestimated.
The third JPR Article of the Year Award goes to Neil F. Johnson, Michael Spagat, Sean Gourley, Jukka-Pekka Onnela & Gesine Reinert for ‘Bias in Epidemiological Studies of Conflict Mortality’, Journal of Peace Research 45(5): 653–663.
A jury consisting of Lars-Erik Cederman (ETH Zürich), Jon Hovi (University of Oslo) and Sara McLaughlin Mitchell (University of Iowa) has awarded the third Journal of Peace Research Article of the Year Award to Neil F. Johnson (University of Miami), Michael Spagat (University of London), Sean Gourley (University of Oxford), Jukka-Pekka Onnela (University of Oxford and Helsinki University of Technology) and Gesine Reinert (Oxford University). In its assessment of all research articles published in volume 45 of JPR, the jury paid attention to theoretical rigour, methodological sophistication and substantive relevance. According to the jury, the prize-winning article, ‘Bias in Epidemiological Studies of Conflict Mortality’, Journal of Peace Research 45(5): 653–663, provides an important advance in the methodology for estimating the number of casualties in civil wars. The authors show convincingly that previous studies which are based on a cross-street cluster-sampling algorithm (CSSA) have significantly overestimated the number of casualties in Iraq. Using a model that divides the underlying population in two groups (those that can and those that cannot be sampled under the CSSA methodology), the authors estimate the exposure to violence for each group, and quantify the potential bias resulting from the CSSA. Furthermore, they suggest and justify plausible values for the model’s parameters, provide a sensitivity analysis on the parameter values to help readers form their own judgements, and derive a simple formula that can be used to gauge and adjust for the bias. The prize-winning article is highly policy relevant and provides a sophisticated and innovative analysis of casualty estimates that have already attracted attention from policymakers.
The abstract of the award-winning article is posted here.
Prof Michael Spagat, one of the authors of the award-winning article, has a website on “Conflict Mortality Surveys” at Royal Holloway College, University of London.
Here are links to some posts on the Lancet study at my former blog Magic Statistics:
- Lancet study of Iraqi deaths is statistically unsound and unreliable
- A specification of the Lancet study’s “Main street bias”
- In the spotlight again: Lancet study of Iraqi deaths
- Lancet study: Serious ethical lapses, data quality problems
Print This Post