Affordances of Narrative and Numerical Data: A Social-Semiotic Approach to Data Use

• In Event: The Complexity of Data-Based Decision Making: Evidence From the United States, Norway, New Zealand, and Belgium

Abstract

Purpose and theoretical framework
Data-driven decision making in education (DDDM) is a sensemaking process in which practitioners and researchers interact with different systems of meaning such as anecdotes or spreadsheets. Despite its proliferation in the last decade, research on DDDM is characterized by the pervasiveness of empirical studies and corresponding lack of theorization (Prøitz et al., 2017). These different traditions influence what counts as data, how data use is perceived by users, and whether multiple forms of data are perceived to have equal weight in decision-making (Schildkamp et al., 2013).The representational qualities of data and their influence on practice are critical but less well-discussed aspects of data use. Drawing on social semiotics, this theoretical article proposes that data should be discussed in terms of narrative and numerical modes of representation.

Method
In social semiotics, the term affordance is used to describe the characteristics of sets of sensemaking resources recognized in a community (i.e., ways of representing used within a community and the strategies used to understand these ways of representing) (Jewitt et al., 2016).

Data sources
Data have a physical aspect (i.e., they are associated with technologies of production and reproduction of information) and a semiotic aspect (i.e., they carry meaning and value within and across cultural contexts). The physical and semiotic aspects of data are closely related. Here, we distinguish between narrative and numerical data. Narrative data typically consist of protagonists and actions organized in a temporal structure (e.g., anecdotes), while numerical data typically consist of mathematical notations and visual representations such as graphs and figures (e.g., survey scores). We argue that the representational properties of these two modes affect how data are interpreted and acted upon.

Results and conclusions
We present two contrasting cases from New Zealand and Norway of how affordances affect teachers’ data use processes. The New Zealand case elaborates on the Learning Schools Model, in which both numerical and narrative sources are critical for identifying problems (numerical) and understanding why problems might exist (numerical and narrative). In the Norwegian case, a national literacy initiative yielded an abundance of narrative data but relatively less numerically configured information. Teachers saw numerical data as less relevant than narrative data gained from firsthand experience with students (Mausethagen et al., 2017).

Scientific significance
We discuss five challenges arising from our theorization about the affordances of data: 1) affordances shape the flow and aggregation of data through school systems, 2) numerical data may be misperceived as neutral or objective; 3) silo problems may be exacerbated when data are represented in multiple modes; 4) empathy resides in narratives, not in numbers; and 5) simplistic interpretations of the representational qualities of data sets should be challenged.

Authors

• Henning Fjørtoft, The Norwegian University of Science and Technology

• Mei Kuin Lai, The University of Auckland