Making Reliable Data: Enacting and Negotiating Data Quality through Data Work

TY - ABST

T1 - Making Reliable Data

T2 - DASTS 2022

AU - Pedersen, Asbjørn Malte

PY - 2022/6

Y1 - 2022/6

N2 - In this paper, I explore how data quality is enacted (Mol, 2002) and negotiated throughcollaborative data work (Bossen et al., 2019) in a healthcare business intelligence unit (HBIU).High data quality is critical to support decision-making, quality care, and management of healthcare. Nonetheless, it poses as a challenge to both business intelligence (BI) and healthcare in general, as accurate, timely, and reliable data is needed to develop meaningful BI products as well as data-driven healthcare. While discussions often focus upon data collection processes and how to improve these to ensure high data quality, this paper offers a new understanding of how data quality also come into being or fall apart through the oft-invisible, but necessary data work in healthcare business intelligence.Drawing on insights from an ethnographic field study of a HBIU and semi-structuredinterviews, I trace how unreliable data is transformed into being reliable through different data work practices of data collection, processing, and visualization. Here, data quality is negotiated, contested, and maintained, when BI employees develop and deploy new BI products. Thus, one crucial aspect of the BI developers’ data work practice is to work with data as ‘registered reality’ which must approach the healthcare staff members’ ‘experienced reality’. Meanwhile, they must collaborate with healthcare staff members to overcome challenges like changes and differences in registration practices, data structure, and organization to do so. Hence, data quality, I suggest, is enacted, negotiated, and contested through collaborative data work practices. This study will contribute to our understanding of datafication and data work as socio-material practice.

AB - In this paper, I explore how data quality is enacted (Mol, 2002) and negotiated throughcollaborative data work (Bossen et al., 2019) in a healthcare business intelligence unit (HBIU).High data quality is critical to support decision-making, quality care, and management of healthcare. Nonetheless, it poses as a challenge to both business intelligence (BI) and healthcare in general, as accurate, timely, and reliable data is needed to develop meaningful BI products as well as data-driven healthcare. While discussions often focus upon data collection processes and how to improve these to ensure high data quality, this paper offers a new understanding of how data quality also come into being or fall apart through the oft-invisible, but necessary data work in healthcare business intelligence.Drawing on insights from an ethnographic field study of a HBIU and semi-structuredinterviews, I trace how unreliable data is transformed into being reliable through different data work practices of data collection, processing, and visualization. Here, data quality is negotiated, contested, and maintained, when BI employees develop and deploy new BI products. Thus, one crucial aspect of the BI developers’ data work practice is to work with data as ‘registered reality’ which must approach the healthcare staff members’ ‘experienced reality’. Meanwhile, they must collaborate with healthcare staff members to overcome challenges like changes and differences in registration practices, data structure, and organization to do so. Hence, data quality, I suggest, is enacted, negotiated, and contested through collaborative data work practices. This study will contribute to our understanding of datafication and data work as socio-material practice.

KW - Dataarbejde

KW - Business Intelligence

KW - Data

KW - Sundhedsvæsen

KW - STS

KW - Etnografi

KW - data work

KW - Business intelligence

KW - Healthcare

KW - Data

KW - Ethnography

UR - https://events.au.dk/dasts2022/about

M3 - Conference abstract for conference

Y2 - 2 June 2022 through 3 June 2022

ER -