Reaction: Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Methods

The paper introduces a set of elementary perceptual tasks which the terminology is confusing. When talking about tasks, it is mostly about the action what the user would do such as find the paths, draw the lines, or pick the shape. Later on, the authors admit that they could have used another terminology such as elementary graphical encodings, but I wonder why they have left it as it is and not change it.

Figure 16 is not understandable as one cannot differentiate type 1,2, and 3 which only has position on the right parentheses. Same to type 4 and 5. A good visualization needs to include the details so that one can understand it without reading text.

I agree that the frame of a framed rectangle increases accuracy(Weber's Law). The contrast helps to recognize and detect the difference easier.

With the experiment, they have set for all cases, subjects would judge the position-angle graphs first. During the experimental design, counterbalancing would be necessary to avoid the effect or impact on learning processes.

The paper provides that they have correctly predicted by the theory that the accuracy orders by position>length>angle.