Teaching Information Visualization - Information Visualization: Human-Centered Issues and Perspectives - page 83

Teaching Issues in Information Visualization at Dagstuhl: Idon'tknow
the content or culture of InfoVis or Computer Science education. This is one of
the diculties associated with having interests that lie between rapidly devel-
oping disciplines. But the Dagstuhl discussions and survey suggest that many
of those involved are facing similar issues with which I am familiar. The sur-
vey draws attention to a range of approaches, topics, academic disciplines, text
books involved.
There is great variation in the approaches and syllabuses of the courses re-
ported in the survey. 'Information Visualization' is the most popular title, but
course names that begin: 'Models and Algorithms for . . . '; 'Data Visualization';
'Computer Graphics and ...'; 'GeoVisualization' are evidently situated in dif-
ferent ways and will require different emphases. This is evident in Figure 1—a
tag cloud showing relative occurrences of topics listed in courses included in the
Dagstuhl survey.
Despite this texture, some consistent themes emerge from the survey and
particularly the discussion at Dagstuhl. Three particular themes resonated with
me and related to my experience. They are also evident in a report produced
following an open workshop at IEEE Visualization 2006 in which I graphically
depict the relationships between related courses with an Information Visualiza-
tion emphasis [56]:
1. a trend/desire for evaluative and critique-based learning that can draw from
a range of related disciplines;
2. an emphasis on 'learning through doing' as opposed to a transmissive ap-
proach to learning;
3. some concern about developing appropriate exercises and assessment for vi-
sualization classes.
I'll briefly consider these key issues and reflect upon their relationship with
my teaching. Doing so may provide some synergies in the cross disciplinary
educational mash-up that is developing in Information Visualization.
Critiquing: Critiquing involves students applying and developing their knowl-
edge through evaluation and review. Robert Kosara argued convincingly for a
critique-based approach to Information Visualization at Dagstuhl. He describes
this as a 'highly interactive and human-centered way of designing things' [40].
An approach that evaluates graph drawing algorithms and evaluation criteria is
documented in the Dagstuhl survey and described in Section 2.3 above.
As a learning device, critiquing may involve evaluating existing work, soft-
ware or graphics, but importantly should involve the practical application of
theory. I ask students to critique existing graphics, those that they have devel-
oped and software systems using a number of criteria, such as:
- Graphical integrity, graphical excellence and Tufte's 'theory of data graph-
ics' [69]
- Using appropriate symbolism [6, 10]
- Map symbolism [45, 60]