Conceiving and Reasoning Visualization Design Options - Data Visualization: A Successful Design Process - page 88

However, the radial diagram was the solution picked for the final piece. So why was
this? Aside from some printing-readability issues that undermined the matrix chart,
he commented: "from a visual point of view, it does not look very tasty." The radial
diagram looked more appetizing—more edible—and fitted more strongly with the
metaphor of a fruit bowl. The matrix chart solution looked more like fungi!
The decision to sacrifice certain qualities of interpretive enlightenment—offered
by the matrix chart—was justified by the designer's instinct to enhance the visual
metaphor emerging from the design of the radial diagram.
Referring back to the "wind map" again, here we saw a visual design that seemed
consistent with the subject matter: it looks and feels like wind would look like
in our imagination. In the piece that we saw in Chapter 2 , Setting the Purpose and
Identifying Key Factors , titled Iraq's Bloody Toll , this was a powerful story made even
more impactive through the subtle but emphatic visual metaphor created by the
arrangement of the chart and the color scheme.
These examples reinforce the value of Edward Tufte's message, from earlier, about
the importance of making judgments through our own lens and based on our own
word of authority.
Choosing the final solution
From the options and influences that we have just processed, we should be able
to narrow down and identify the right data representation specification for our
visualization.
It may be that this specification matches neatly with an established chart type and
we can essentially "pick it off the shelf" and deploy it. This could be considered our
top-down approach. In Chapter 5 , Taxonomy of Data Visualization Methods , we'll see a
gallery of contemporary solutions. From this selection you may identify a particular
solution that fits with your context both structurally and metaphorically.
Alternatively, we may wish to custom-build a solution from the bottom up, carefully
constructing a design one visual variable at a time, accommodating the range of data
variables we want to show, and the style with which we want to show it. Of course,
every chart type had to originate from somewhere and be invented at some point. It
may be that as we construct our unique solution, we end up slipping back towards
realizing that a tried-and-trusted option remains the best choice.