Morals, morale and motivations in data fabrication: Medical research fieldworkers views and practices in two Sub-Saharan African contexts

Data fabrication, incorrect collection strategies and poor data management, are considered detrimental to high-quality scientific research. While poor data management have been occasionally excused, fabrication constitutes a cardinal sin – scientific misconduct. Scholarly examinations of fabrication usually seek to expose and capture its prevalence and, less frequently, its consequences and causes. Most accounts centre on high-income countries, individual senior researchers and scientists who are portrayed as irrational, immoral or deceptive. We argue that such accounts contain limitations in overlooking data collected in 'the field', in low-income countries, by junior researchers and non-scientists. Furthermore, the processes and motivations for fabrication and subversive practices are under-examined. Drawing on two separate ethnographies, conducted in 2004–2009 in medical research projects in sub-Saharan Africa, this paper investigates fabrication among fieldworkers using data from observations and informal conversations, 68 interviews and 7 Focus Group Discussions involving diverse stakeholders. Based on an interpretative approach, we examined fieldworkers' accounts that fabrications were motivated by irreconcilable moral concerns, faltering morale resulting from poor management, and inadequate institutional support. To fieldworkers, data fabrication constituted a 'tool' for managing their quotidian challenges. Fabrications ranged from active to passive acts, to subvert, resist and readdress tensions deriving from employment inequalities and challenging socio-economic conditions. We show that geographical and hierarchical distance between high-ranking research actors and fieldworkers in contemporary configurations of international medical research can compartmentalise, and ultimately undermine, the relationships necessary to produce high-quality data. In focusing on fieldworkers, we argue for the inclusion of wide-ranging perspectives in examinations of data fabrication.