“Reflections on How Designers Design With Data”, 2014-05-27 (; backlinks; similar):
In recent years many popular data visualizations have emerged that are created largely by designers whose main area of expertise is not computer science. Designers generate these visualizations using a handful of design tools and environments. To better inform the development of tools intended for designers working with data, we set out to understand designers’ challenges and perspectives.
We interviewed professional designers, conducted observations of designers working with data in the lab, and observed designers working with data in team settings in the wild.
A set of patterns emerged from these observations from which we extract a number of themes that provide a new perspective on design considerations for visualization tool creators, as well as on known engineering problems.
…Patterns: In our observational studies we observed all of the designers initially sketching visual representations of data on paper, on a whiteboard, or in Illustrator. In these sketches, the designers would first draw high-level elements of their design such as the layout and axes, followed by a sketching in of data points based on their perceived ideas of data behavior (P1). An example is shown in Figure 3. The designers often relied on their understanding of the semantics of data to infer how the data might look, such as F1 anticipating that Fitbit data about walking would occur in short spurts over time while sleep data would span longer stretches. However, the designers’ inferences about data behavior were often inaccurate (P2). This tendency was acknowledged by most of the designers: after her inference from data semantics, F1 indicated that to work effectively, she would need “a better idea of the behavior of each attribute.” Similarly, B1 did not anticipate patterns in how software bugs are closed, prompting a reinterpretation and redesign of her team’s visualization much later in the design process once data behavior was explicitly explored. In the time travel studies, T3 misinterpreted one trip that later caused a complete redesign.
Furthermore, the designers’ inferences about data structure were often separated from the actual data (P3). In brainstorming sessions at the hackathon, the designers described data that would be extremely difficult or impossible to gather or derive. In working with the HBO dataset, H1 experienced frustration after he spent time writing a formula in Excel only to realize that he was recreating data he had already seen in the aggregate table…Not surprisingly, the amount of data exploration and manipulation was related to the level of a designer’s experience working with data (P4).