From Knowledge Mediators to Disenchantment with Big Data analytics

New epistemic technologies are transforming knowledge work across organizational contexts. Their implementation in professional practice hinges on collaboration and coordination across different types of experts. While research has shown that representational objects and brokering practices help bridge knowledge and facilitate collaboration, we examine how they can create unexpected difficulties in the integration of an epistemic technology to professional practice. We investigate a two-year project where academic medical researchers hired data analysts to identify premature birth trends in the U.S. birth records using a pattern recognition algorithm. Our findings demonstrate a process of disenchantment, where the medical researchers shifted from the lure of big data analytics to disenchantment about their affordances for medical science. This emerged when visualizations of data intensive analyses and then the data analysts as mediators failed to convey and explain insights in a way that aligned with the medical researchers’ expectations and domain expertise. Our findings contribute to research on the integration of epistemic technologies in knowledge work, objects and actors as knowledge mediators, and data analysts as an emerging brokerage profession.