Week 2 Paper Reading
Paper:
- Graphical criticism: some historical notes. Hadley Wickham. Journal of Computational and Graphical Statistics, 2012.
- Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Methods. Cleveland and McGill. Journal of the American Statistical Association, 1984.
Response:
In the first paper, Hadley gave an introduction to data visualization history. There are about three main milestones, the first one gave recommendations on data visualization, I prefer suggestion keep symbols to a minimum. The authors of the second paper also insist on a simple graph. However, this makes contradiction to the third milestone in 3d visualization. Although the 3D plot is complex, in my opinion, it can show readers information in one more dimension, and alleviate data points congestion. The second milestone is the theory of color, and the third one is the development of the 3D plot.
And he thought the quality of the best graphs did not improve much. And he posts multiple problems about how to make better visualization task, why people make complex visualization plot than before, and why people favor flash over content.
In the second paper, the authors gave 10 elementary tasks ordered from most to least accurate:
-
Position along a common scale
-
Position along a nonaligned scales
-
Length, direction, angle
-
Area
-
Volume, curvature
-
Shading, color saturation
And after the experiment, the authors gave an application to redesign several much-used graphs:
-
Dot charts and Bar charts as a replacement for Divided Bar charts and pie charts; Grouped Dot Charts and Grouped Bar Charts as a replacement for divided Bar charts;
-
Showing Differences directly for Curve-Difference Charts;
-
Framed-Rectangle Charts as replacements for Statistical Maps with Shading;
-
Graphs support data analysis.
These replacements are based on previous elementary tasks. Take Pie Chart, for example, people hard to tell quantitive from an angle, even the wholeness. But bar charts or dot charts can great display quantitive greatly through position.
The experiments of proving the order of basic elements are preciseness. Subjectives gave different graphs and were asked to tell which segment is the largest and what percent the small is of the larger one. Through these experiments, the author could tell the importance of position, angle, and length.
Excepts for technology experiment, the author pointed out that a better plot should lead to more accurate judgments than others form.
This is some text!