Information Systems (IS) phenomena have become increasingly volatile, complex and fast changing. Capturing their essence is an increasingly daunting task. Data science have emerged in awe to predict future outcomes. Decision-making thus becomes faster while data become bigger. Yet, in the wake of this promising path, many of these predictions lack accuracy due to the unpredictability of complex phenomena. That is why researchers promote the importance of thick qualitative data analysis as a way of seeking explanations of the generativity underlying complex phenomena. This approach is (in comparison) slow, but can answer why events occurred. Thus, we argue that sound accounts of complex IS-phenomena must come from a combinatory approach of fast predictions with slower accounts. Predictions apply laws theorized as causal mechanisms. When these outcomes do not arise, we suggest applying explanatory accounts that apply a different form of causality - generative mechanisms. Generative mechanisms can explain unpredictable outcomes, but can only be inferred through longitudinal qualitative studies. This paper opens up a research agenda for combinatory approaches of fast mechanistic predictions from big data and slower generative explanations from thick data. This combination will help capturing the essence of complex socio-technical phenomena in our capricious digitalized world.
|Title of host publication||ECIS 2019 Proceedings|
|Number of pages||9|
|Publisher||Association for Information Systems. AIS Electronic Library (AISeL)|
|Publication status||Published - 2019|