Key points are not available for this paper at this time.
African conflicts are highly represented in cross-national conflict event datasets, and their causes are increasingly investigated in quantitative analyses. Many of these datasets make use of international media reports to compile information on different forms of conflict events, ranging from violent armed conflict to forms of social unrest such as protests and riots. African studies scholars more commonly rely on local news sources to study these phenomena. In this research note, we investigate the effect of using different types of news sources to analyze conflict patterns in Africa by comparing the coverage of conflict events in Nigeria by three datasets: The Social Conflict in Africa Database (SCAD), the Armed Conflict Location and Event Dataset (ACLED) and a dataset developed by the authors, based on Nigerian sources and focusing on the period April 2014–March 2015. We investigate the overlap between the datasets and whether this is affected by event characteristics, including violent and non-violent events. The comparison shows empirically the important differences that can exist, in terms of both the absolute number and the sub-national distribution of reported conflict events. Such variations between datasets in terms of event, but also time and geographical coverage require careful reflection by users.
Demarest et al. (Tue,) studied this question.