Cognitive Artifacts in Complex Work - Ambient Intelligence for Scientific Discovery - page 164

defined periods of research work. The diary was a standard Word file designed
to collect detailed responses on a daily basis, with each response associated with
an information episode, using 6-10 open-ended prompt questions per response
day. Questions were varied each day, to elicit a wide range of information tasks
over the task diary period.
The field research sessions consisted of one-week of full-day visits to the
scientists' laboratories and work locations. Two researchers conducted observa-
tions and interviews, recording sessions with digital video and/or audio. During
the field period, 2 one-hour semi-structured interviews were held with principal
investigators (PIs) and selected researchers. The interview protocols for PIs col-
lected information use data on research projects, grant and manuscript practices,
and program development. Researcher protocols gathered data on experiments,
computer and information service use, use of journals and articles, and specific
information tasks. Cognitive walkthroughs of specific information services used
by these researchers were conducted to capture and evaluate actions, decisions,
and resources used.
Research practices and associated information activities were reconstructed
using a research information life-cycle diagram. This analytical instrument was
designed to capture a temporal model of information activities within the project
unit of analysis. Participants further described research project events, decisions
and processes, experiments, and other associated research tasks. After describing
a complete project history (supported by searches, papers, posters, and other
documentation) we recorded information tasks associated with each event. While
based on a retrospective method, we attempted to maintain construct validity
relative to actual and current information tasks by specifying current or recent
research projects. We related the diagram's data to diary and interview data to
compare task descriptions and information services used. In some cases, partici-
pants gathered supporting documentation such as printouts of searches, reference
lists, abstracts, and other materials in advance of the field visit. This method
enabled participants to describe project and experimental activities, related in-
formation tasks, and collaborative tasks along a defined time line representing
the process of a current experimental research project.
4.2 Discovery as an Information Practice
Studying the work of scientists from task diaries, interviews and observations
shows multiple projects, competing demands, and the unplanned events of re-
search practice. Faculty and graduate students conducted experiments, orga-
nized projects, and sought information through searches, artifacts, and requests
to colleagues. We identify seven continuous practices; tasks that refer to cognitive
artifacts are shown in italics.
- Managing research: Organizing schedules , activities and commitments for
research
- Conducting research: Preparing equipment and materials for experiments,
running experiments, documenting experiments and procedures, sharing
findings