From CS294-10 Visualization Fa08

Jump to: navigation, search

Good Visualization

Source: Scion Car Ad, Waxpoetics Magazine, Vol. 29, Page 44, Published by Wax Poetics Inc.

Description: I chose this visualization not due to its great proficiency in presenting information, but interesting point of view in creating an effect for the user in the context of being used as an advertisement. The left side of the page shows simple directions to create a paper throwing star, segmented into 10 steps shown spatially. The right side of the ad (not shown) is your basic picture of a hyped up car ad in all its ridiculous brim stone glory. We talked about how in class that one of the finer points of designing a visualization is catering to the finite attention span one is willing to give. Nowhere is this sensitivity more crucial than advertisement. Knowing this, the graphic designer went the minimalistic approach creating a simple narrative sequence of how to create a throwing star. Its unlikely that the user will actually create the star, but its important that he or she be able to follow visually how to create it so they can create on in their head. By using high contrast colors, minimal information, and familiar symbols, the user is able to be satisfied with having quickly attained the knowledge of making a throwing star. They can then attribute this satisfaction and "aggressive qualities" of the throwing star to the the car.

Deconstruction: This visualization encodes the instructional data of the physical actions necessary to fold a sheet of paper into a throwing star. The data set encapsulates 10 spatial positions. Within each of these spatial positions is additional data indicating motion required to get to the next spatial position. The ten positions are shown in an ordinal manner, explicitly sequenced from 1-10. This sequence is encoded with the both actual numerical labeling, and x-y positioning (sequenced left to right, top to bottom consistent with western reading style). The position of the paper within each step is encoded as quantitative ratio data with both x-y positioning, and shape. The outlining box of each step is kept the same size to allow the relative in sizes and position to be compared from step to step. Ordinal labels of dotted lines and connected lines are used to indicate whether the paper is to be folded or cut. The motion of the action required to reach the next step is shown using familiar ordinal labels of scissors and an arrow. They are encoded with position (where the fold or cut is) and length (how long the fold or cut is). Explicit directions are typed when the simple labels don't fully suffice. Additional design elements include using high contrast colors to distinguish the paper from the background and clear all capitalized font to demand attention.

The ad is has a slew of gratuitous "uninformative elements" but few that actually are a component of the throwing star instructional visualization. The spiky monster indeed does detract a bit from the users attention to the sequencing. Once could argue that showing the monster holding the star gives a sense of perspective of how big the star should be. However that is stretching it. Most of these uninformative elements are simply to reinforce to notion that that car is aggressive and hip. The data presented in the visualization that actually pertains to folding the star is very minimal and rather tightly packaged. The numbering is slightly stylized using underscores and redundant with spatial sequencing, yet is not that obstructive.

Bad Visualization

Source: Tao of Jeet Kune Do by Bruce Lee, Page 186, Published by Black Belt Communications LLC.

Description: I understand that Bruce Lee probably did not have the most time to develop clear visualizations (amidst all the butt kicking), but this "Body Blow" instructional visualization still makes me cringe. The diagram you see here was essentially a stand alone page to describe the two basic moves. Pardoning the lack of detail in the drawing (permissible, since the outlines of body positions is the only relevant component), the composition of the bodies are thoroughly confusing. He shows bodies in three different sizes, which do not correlate to significance of motion shown. This gives the user an obscured sense of perspective, which is just jarring. Secondly, the bodies are often places so close together they are overlapping, just creating a lump of bodies. Thirdly, multiple arrows are shown per body. This gives an unclear sequencing and sense of time. The positioning of bodies in general all show a complete lack of sequencing. Essentially a jumbled mess of still fight positions. The text adds nothing to the diagrams, but just assigns a label. If anybody could learn a move from this, I would be in shock... sorry Bruce.

Deconstruction: This visualization attempt to encode the sequencing of the physical actions necessary to complete two martial arts moves entitled the basic body blows. The data set consists of the spatial positions of 14 bodies. Additionally, within each of these positions, labeling data is used to signify motion. The spatial positions lack a clear encoding for sequencing. It is hard to determine whether the positions are meant to be completed in order from left to right. It can be inferred that proximity in position between two bodies indicate that the members are the attacker and defender. However this encoding becomes inconsistent and unclear when four bodies are lumped close to each other in the bottom right corner. The change in size of the bodies seems to project that there is a relationship between sequence or importance of the position and the size of the body, but this is also unclear. The actual body positions are encoded as quantitative ratio data, using x-y position and shape as primary means. The details almost give a sense of a third dimensioning of positioning, however 2 can only be clearly comprehended. Within each of these positions, motion is encoded as quantitative ratio data using arrows. The arrows use x-y position to show what body part moves, length to possible show the strength of the movement and shape (connected versus unconnected) to possibly show time. However, for the most part, prescription of order, time, and amount are lost in this visualization.

There are a few design elements that could be considered "uninformative". The details in the bodies give a hint at their three dimensional positioning, however is inconsistent and more confusing than anything. Without a sense of sequencing, the arrows are largely uninformative. Their encoded variables are too ambiguous to be put to use. The layering positioning of bodies is jarring and unnecessary. The text serves as a label but does not add to the comprehensiveness of the instructions whatsoever.


With this redesign, improving the encoding of sequencing was of paramount concern (following the completely ambiguous composition of the original). The original document showcased an overloaded number of moves with no clear methodology of sequencing. The move itself is a complex process with many options that cannot be broken down to a linear sequence. This flow diagram was found to be the best way to instruct a student on the methodology necessary to complete the action given a particular circumstance. Using the principle of small multiples, each motion is given its own individual diagram to minimize confusion and overloading of sequencing. The relative difference between positions should be large enough to use small multiples. The sequencing is encoded using x-y positioning along with arrows and labels to direct the process. Information for geometric body form is encoded again as quantitative ratio data using x-y position and shape. To improve clarity of the body form, consistent size was used kept in a 2-d environment. Additionally color was used to distinguish between opponent and student, and L and R are used to distinguish left and right hand (may be hard to see in small image).

[add comment]
Personal tools