Process and Pitfalls in Writing Information Visualization Research Papers - Information Visualization: Human-Centered Issues and Perspectives - page 139

Plaisant's discussion of the diculties of evaluating infovis is thoughtful and
germane [29]. In retrospect, a better, albeit wordy, name for this category might
be Summative User Studies , since the goal is to examine the strengths of
a system or technique. Evaluation is far too broad a term - because all papers
should contain some sort of validation. Even User Studies would not be the best
choice, because formative studies are probably a better fit for the Design Study
category, where ethnographic methods are often appropriate in the task analysis
to determine design requirements or iteratively refine a design. However, these
lines are not necessarily crisp. For instance, the MILC approach advocated by
Shneiderman and Plaisant [35] could fit into either a Design Study framework, if
the emphasis is on the formative ethnographic analysis and iterative design, or
a Summative Evaluation framework, if the emphasis is on the longitudinal field
study.
2.6 Model
Model papers present formalisms and abstractions as opposed to the design
or evaluation of any particular technique or system. This category is for meta-
research papers, where the broad purpose is to help other researchers think about
their own work.
The most common subcategory is Taxonomy , where the goal is to propose
categories that help researchers better understand the structure of the space of
possibilities for some topic. Some boundaries will inevitably be fuzzy, but the
goal is to be as comprehensive and complete as possible. As opposed to a survey
paper, where the goal is simply to summarize the previous work, a taxonomy
paper proposes some new categorization or expands upon a previous one and
may presume the reader's familiarity with the previous work. Good examples
are Card and Mackinlay's taxonomy of visual encodings [3] and Amar et al. 's
task taxonomy [1].
A second subcategory is Formalism , for papers that present new models,
definitions, or terminology to describe techniques or phenomena. A key attribute
of these kinds of papers is reflective observation. The authors look at what is
going on in a field and provide a new way of thinking about it that is clear,
insightful, and summative. An influential example is the space-scale diagram
work of Furnas and Bederson [7], and an interesting recent example is the casual
infovis definition from Pousman et al. [31].
A third subcategory is Commentary , where the authors advocate a position
and argue to support it. Typical arguments would be “the field needs to do more
X”, “we should be pushing for more Y”, or “avoid doing Z because of these
drawbacks”. A good example is the fisheye followup from Furnas [8]. These kinds
of papers often cite many examples and may also introduce new terminology.
Model papers can provide both a valuable summary of a topic and a vocab-
ulary to more concisely discuss concepts in the area. They can be valuable for
both established researchers and newcomers to a field, and are often used as
assigned readings in courses. I think this category name is appropriate and do
not suggest changing it.