Conceiving and Reasoning Visualization Design Options - Data Visualization: A Successful Design Process - page 81

As discussed in the previous chapter, learning about the physical properties of your
data gives you an important sense of the shape and size of your data. As you refine
your editorial focus you will develop an understanding of the data variables you
may seek to display graphically.
The quantity and nature of the variables you are using will have a significant
influence on reducing the range of suitable chart types you might be able to use
within the method family you have chosen. As discussed earlier, this process of
eliminating choices can only be of help to us as we move forward.
Referring back, once again, to our demonstration for the Olympics project, the data
we were looking to use for our final story was event year (quantitative interval-
scale), medal totals (quantitative ratio-scale), and country (categorical nominal). We
had a good sense of the range and distribution of values held against each variable,
we were just highlighting two countries and we wanted to show the full five-event
transition. The best solution, therefore, was to use the line chart as we have just seen.
In Chapter 5 , Taxonomy of Data Visualization Methods , we go into much more detail
about this taxonomy and the range of chart types that sit underneath each of these
five headings. You will see a gallery of some of the most common, contemporary
methods. Each example is accompanied by a description of the chart and an outline
of the data variables that each one can realistically accommodate. This will give
you a good sense of some of the common data representation techniques. It could
act almost like a creative menu for you to refer back to when seeking ideas and
potential solutions.
Determining the degree of accuracy in
interpretation
Now we start to step into the minefield. You can be assured that this section will
have been the most revised and rewritten across the entire topic.
Having identified the general visualization method and started to filter down further
to identify the most suitable chart types, we now have to consider another key issue.
This judgment gets to the very heart of the form/function or art/science fault lines
that exist in this field—to what degree of accuracy do you wish readers to be able
interpret values from your visualization?
You might ask in response, why would you ever not wish to maximize the precision
of interpretation? Surely, the mission is to deliver as much accuracy through our
representation as possible?