What lurks beneath the tip of the iceberg? Exploring the “missingness” problem in cyber events data

Abstract With growing interest in and attention to cybersecurity, pundits, media outlets, and policymakers are interested in meaningful insights about the nature of cyber conflict. However, the challenges to rigorous inference in cybersecurity research remain daunting. Perhaps the most serious concern is the inability of researchers to observe the full universe of cases and behaviors under study. This paper addresses the critical problem of missing data in cyber incident data collection efforts. We begin with an overview of missingness as it relates to cyber conflict research. We pay particular attention to selection bias resulting from missing data. We apply a Heckman selection model to assess missingness in cyber data and recommend that scholars working with cyber incident datasets consider similar methodologies or strategies to address missingness-induced bias and enhance confidence in the identified relationships.