How are Realities Enacted by Data? Analysis of the Practices and Knowing of Professionals Working with Data in Light of the Actor-Network Theory

Purpose  In the context of growing interest in the use of data to create mathematical models that can predict future behavior and assist in business decision making, this theoretical-empirical article aims to understand how realities are enacted by data, identifying and describing the practices and knowing of two professional groups working with data.

Theoretical framework  Actor-Network Theory and Practice Theory.

Design/methodology/approach  This study was carried out within the methodological framework of Actor-Network Theory, from November 2017 to September 2018, including interviews and observations of daily work.

Findings  The findings suggest that realities portrayed in predictive models are constituted of arrangements of human and non-human elements, situational and emerging, contemplating data, technological potential and constraints, and political choices that permeate these configurations.

Practical & social implications of research  The research demonstrates that knowledge and predictive models generated by data cannot be understood without taking into account their contexts of origin, which makes it possible to question their supposed neutrality and objectivity. Given these compositions, models may lead to process optimization, but also to unexpected effects, such as errors in scenario forecasting.

Originality/value  The research contributes by making visible the configurations of human and non-human elements, situational and emerging, through which organizations and realities are enacted, based on data practices.

Keywords:
Professionals working with data; predictive models; enacted realities; actor-network theory; practice theory