The complex and uncertain environment of the humanitarian response to crises can lead to data bias, which can affect decision-making. Evidence of data bias in crisis information management (CIM) remains scattered despite its potentially significant impact on crisis response. To understand what biases emerge in complex crises and how they affect CIM, we conducted a combined interview and document analysis study. Focusing on the largest humanitarian crisis in the world, i.e., the conflict in Yemen, we conducted 25 interviews with managers and analysts of response organizations, and assessed 47 reports and datasets created by response organizations in Yemen. We find evidence of a cycle of bias reinforcement through which bias cascades between field, headquarters and donor levels of crisis response. Researchers, as well as practitioners, need to consider these underlying biases and reinforcement loops because they influence what data can be collected when, by whom, from whom, and how the data is shared and used. To the CIM literature, we contribute an in-depth understanding of how four types of data bias emerge in crises: political, accessibility, topical, and sampling bias.@en

A crisis requires the affected population, governments or non-profit organizations, as well as crisis experts, to make urgent and sometimes life-critical decisions. With the urgency and uncertainty they create, crises are particularly amenable to inducing cognitive biases that influence decisionmaking. However, there is limited empirical evidence regarding the impact of cognitive biases on estimation, judgment, and decision-making tasks in crises. Possible biases occurring in crises are: (1) to be influenced by how information is framed (i.e., framing effect), (2) to overly rely on information that confirms rather than opposes preliminary assumptions (i.e., confirmation bias), (3) to rely heavily on a skewed informational cue when making estimations (i.e., anchoring bias), and (4) to see the own decision-making as less biased than decision-making of others (i.e., bias blind spot). We investigate these four cognitive biases using three online survey experiments targeting crisis-affected people of the general public (n = 460, mTurk workers), governmental and non-profit workers (n = 50, mTurk workers), and crisis experts (n = 21, purposefully sampled). Our findings show that crisis experts are the least biased group but are still significantly affected by anchoring, framing, and bias blind spot. Crisis-affected people from the general public showed the strongest susceptibility to all four biases studied. The findings have implications for future research on crisis information systems (IS) design. As crisis response is increasingly facilitated through IS, we propose debiasing functions that account for biased user behavior in crises. @en