Ways of thinking about data
Judith J. Lambrecht, Stephan Paul Flister · 1998
Within a larger frame concerning how ordinary people learn to use complex tools, database software was selected as the archetypical tool of the knowledge worker in the Information Age. As non-technical professionals, how do knowledge workers learn the complex technology of databases? The disciplines of education, software engineering, and cognitive anthropology grounded the study in a guiding perspective of the learner, seeking ways to improve how tools can support learners' efforts to use them, as opposed to how learners can be fixed to better use the tools. The evolving methodology of phenomenography (Marton & Booth, 1997) makes the basic assumption that the conceptions people hold differentially affect how they approach tasks and the outcomes they achieve. Accordingly, the study sought to clarify the conceptions people hold around learning, data, and database software. Given a theoretical framework of situated and distributed cognition, these conceptions are better studied under conditions of work rather than laboratory. So, evidence was gathered from an ongoing work setting, a class of students learning how to teach others about business software, and from the tools themselves. These considerations were brought to bear on the operational research question: what ways of thinking about data-related concepts do practitioners exhibit that may affect their design of work activity? Using concept mapping as a primary data transformation tool, the study found that data can be seen as a hierarchy of conceptions--content, attributes, structure, and purpose--the thrust of which is the development of meaning on a foundation of content, which is a parallel construction with knowledge. These findings are used to suggest that business education curriculum should thoroughly encompass purposes. Content classifications are suggested that go beyond the practice of teaching how to activate the decontextualized functional aspects of the software. The findings suggest that software tools should support a distributed cognition view of the workplace. A development methodology is proposed that casts software development as the relocation of cognitions in a cognitive network. One possible direction is to develop a process metaphor to expand the now standard desktop metaphor for software interfaces.