More Data, New Problems: Audiences, Ahistoricity, and Selection Bias in Terrorism and Insurgency Research.

The quality of social science inquiry examining terrorism, insurgency, and nonviolent civil resistance has progressed rapidly in recent years. One reason for this advancement is the emergence of new datasets and the subsequent application of quantitative methods to the analysis of asymmetric political conflict between states and nonstate actors. Despite rapid development within the research paradigm, the use of new data has coincided with several methodological and conceptual challenges. This inquiry employs insights from qualitative social science methodology and organizational sociology to highlight and propose solutions to three shortcomings found in recent quantitative analyses of asymmetric conflict. The first problem arises from scholars' proclivity to ask research questions based on easily accessible categories of data rather than on theoretically significant puzzles in the literature. The second pitfall concerns limitations and "conceptual stretching" associated with static, nominal variables constructed to enable statistical inference. Finally, the third class of research obstacles arises from selection bias caused by underreporting of data. Each of these methodological problems potentially undermines theoretical claims made in recent work on insurgent organizations, terrorism, civil war, and nonviolent resistance. [ABSTRACT FROM AUTHOR]