Visualization has been defined to be construction of images in mind using perception and human cognition. It has been described to be process with 4 stages (Figure 1: Stages in Data Visualization) (Information Visualization: Perception for Design)

. Collection and Storage

. Preprocessing Stage: Data is transformed to be used for manipulation. There is also data reduction and exploration involved at this stage.

. Mapping of Data to Visual Representation.

. Interpretation of Data using Perception and cognitive system.

The role of perception in Visualization is for interpretation of the Visual encodings and decode the information in graphs.

Cleveland McGill’s research in graphical perception is a classic study that had brought light on the perception playing the key role in interpreting the graphical data. There experimentation involved using the various visualizations in statistics and Human subjects to interpret the data. The study suggested recommendations on using the visualizations to ease Human interpretation when it comes to using the Visualizations (dot charts, dot charts with grouping and framed rectangle charts) to quantify the data. The authors proposed the theory of ‘Elementary perceptual tasks’ to identify the elements that can be used for perception(Cleveland and McGill, 1984). The 10 elements are: - as length, direction, angle, area, volume, curvature, shading, color saturation, position along common scale and position against non-aligned scales. Another theory proposed by the Authors is the ‘Ordering of Elementary Perceptual tasks’(Cleveland and McGill, 1984). The elementary perceptual tasks are ranked in the following order:

  1. Position along a common scale
  2. Position along nonaligned scale
  3. Length, direction angle
  4. Area
  5. Volume, Curvature
  6. Shading, color saturation

Numerous researches following the Cleveland McGill’s initial findings have replicated experiments that added more to the optimizing the Visualizations for ease of perceptions and looking to more attributes that could reduce errors and be producing more effective in quantification based interpretation of the data from the Graphs.

In Heer and Bostock’s paper, they have used the ‘Crowdsourcing’ method. The Amazons MTurks experimentation were used to gather the data for perception. Crowdsourcing is cost and time effective way of gathering data from replicate experiments of perception in visualization. Their research has shown that Luminance contrast, chart size and gridline spacing are also some of the attributes of visualization that increase perception(Heer and Bostock,2010).

The importance of color in data visualization is when the data type and color contrasting are correlated that adds to the Visualization perception. To differentiate between higher and lower values has been cited as an important for perception when the data is being compared across the average and it is suggested to use lighter tints for lower values and darker tints for higher values(Harrower and Brewer, 2003). Also, for comparison of two different visualization, the axes range plays a major role when comparing two visualizations generated from similar dataset(Cleveland, 1985).

If a person can make quick interpretations rather than looking at the table which is the source for the data, these visualizations must be optimized enough that allow any person with advanced skills to make simple quantifications accurately avoiding any simple error. If we look at the visualization in Figure 2, the data for Murders in USArrrests data, it is much easier to interpret for Murders by looking the shaded regions in the map and the scale given in the bottom of the visualization giving the higher rate as lighter shade and lower murder rate as darker shade. Also, looking at the the framed Rectangle Chart for Murder Rates(Figure 3) from Cleveland and McGill’s research it was determined to be easier for perception and quantification of data compared to the shaded map chart.

Thus, it can be concluded that graphical perception is dependent on ‘elementary tasks’ and different visualization attributes as outlined by numerous researches but essentially pioneered by Cleveland and McGill with their classic experimentation. A good visualization is dependent on the data type and ease of perception is dependent on the visualization made for the data. All these criteria that have been researched and experimented over the years have made for ‘ease of perception’.

References

Ware, Colin. Information visualization: perception for design. Elsevier, 2012.

Cleveland, William S., and Robert McGill. “Graphical perception: Theory, experimentation, and application to the development of graphical methods.” Journal of the American statistical association 79.387 (1984): 531-554.

Heer, Jeffrey, and Michael Bostock. “Crowdsourcing graphical perception: using mechanical turk to assess visualization design.” Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM, 2010.

Harrower, Mark, and Cynthia A. Brewer. “ColorBrewer. org: an online tool for selecting colour schemes for maps.” The Cartographic Journal 40.1 (2003): 27-37.

Cleveland, William S. The elements of graphing data. Monterey, CA: Wadsworth Advanced Books and Software, 1985.