Week 6 – Perception & Attention



* in-class lab #3


One thought on “Week 6 – Perception & Attention

  1. hsssajx 03/05/2013 at 6:05 PM Reply

    Comments on Cleveland and McGill, 1984:
    This work is about how to improve the visualization of quantitative data through human graphical perception. The author shows how elementary perceptual tasks that are used to extract the quantitative information on a variety of common graph forms and designs an experiment to rank the relative accuracy levels of different elements used in different visualization strategies.
    Sample distribution function plots, bar charts, pie charts, divided bar charts, statistical maps with shading, curve-difference charts, Cartesian graphs, triple scatter plots, volume charts and juxtaposed Cartesian Graphs are the ten ways that graphically visualize data. In accordance to the studies about accuracy of extraction in previous studies, the author ranks the accuracy from most to least as follows: position along common or nonaligned scales, length judgments, area judgment, volume judgment, shading and color saturation judgment. Power law in visual perception, Weber’s Law and many results from other famous psycho-physical experiments have been quoted to support the ranking above and also the following two experiments’ design in the measurement of accuracy. With the previous rank as assumption, the author conducts a position-length experiment and a position-angle experiment that ask subjects to quickly indicate which and what percentage of the two segments is smaller in bar charts or pie charts. The results suggest that: a grouped dot chart is preferred than a group bar chart; curve-difference charts could show differences more directly, framed-rectangle charts should be a good replacement for statistical maps with shading, etc.
    Although elementary perception task could give very concise comparison between the human’s perceptions on different graphical elements, human perception does not mean viewers split those basic elements like shape, position or angle before they make judgment about the whole graph. As more complicated combinations of elements occur, we cannot easily conclude that the most easily perceived elements’ combination also forms a best one. Moreover, this kind of perception studies could not provide researchers with guideline to construct a graph, not only because human’s reaction in experiment could be quite different from their usual ways, but also because the comparisons in those experiments are not based on the variation of difference kinds of data. That is to say, elementary perceptual studies could only tell you what might be better to be used rather than should be used.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: