A1-KenrickKin

(diff) ←Older revision | Current revision (diff) | Newer revision→ (diff)

 Good Visualization

Newsweek, September 3, 2007

Explanation This is an attractive visualization of hip and knee replacements that uses a single diagram, nicely spanning the height of a full page. The information about each joint replacement is distinguished by a distinct color. There are snapshots of what our natural joints look like before a replacement, and with the aid of transparency, color, and dotted lines, the reader can see how the artificial joints are inserted into existing bone. At first glance, I did find the numbering a bit confusing - that the hip joint boxed in gray was designated "2" at the top of the page.

Deconstruction The data here is hip joint and knee joint, before and after. Color is used to denote the nominal data of whether a hip joint or a knee joint is being described. Numbered labels distinguish between before and after. Transparency and shading depict the nominal data of whether an object is artificial or not and if it's inserted into bone or not.

Time, August 27, 2007

Explanation I like that they use three separate graphs, one for each sport. This avoids the clutter of having six different lines that are categorized by both sport and whether it referred to percentage of schools or female head coaches. Since this is a wide graph, the horizontal lines aid the reader to estimate percentages. However, the extra tick marks between intervals of five years is unnecessary. The labeled percentages help to emphasize the the current low percentage of female coaches.

Deconstruction I don't know how large the data set is here, but it has data from 1978 to 2006. Quantitative data in the form of time is represented by the x axis. Quantitative data in the form of percentage of schools or percentage of female coaches is represented by the y axis. Whether the data is about percentages of schools or percentages of female coaches is nominal data represented by color. The three different sports depicted is also nominal data that is denoted by strings.

National Geographic, August 2007

Explanation It isn't obvious to me right away why there are two y axes. The right is clearly temperature, but what are those numbers on the left? All it says at the top left is "Hurricanes and tropical storms." Apparently they are referring to the number of storms. The two line graphs are entirely separate data that share the same space, requiring the reader to first understand that there are two axes in different measurements and then matching up the red line to the right axis and the blue line to the left axis. It doesn't help that "Sea-surface temperature" is left justified, while the corresponding axis is on the right. The graph does make sense once all the pieces fit together - there appears to be some correlation between sea-surface temperature and the number of storms.

Deconstruction Time is quantitative data expressed along the x axis, ranging from 1861 to 2001. Temperature and number of storms is quantitative data expressed along the y axis. Color is used to distinguish between number of storms and quantitative data.

Although it does not seem to be ideal to have two graphs with different y scales sharing the same x axis, this visualization does help convey that there is a correlation between the sea surface temperature and the number of storms. So sticking National Geographic's decision to use the same x axis, I wanted to make it very clear as to which line corresponded to which axis, which was not as apparent in the original design. Really all that was needed is some more color. By making Degrees Celsius the same color (in this case red) as the temperature plot, it clearly identifies that the y axis on the right corresponds to the sea-surface temperature. The color blue does the same for the storms plot. To further emphasize this point, I moved the Sea-surface temperature string more towards the right, to associate it with the y axis on the right. The "Storms" and "Sea-surface temperature" labels might not be necessary, as one can infer the plots based on the y axes. Although this might be a bit silly, I added "number" in front of "hurricanes and tropical storms" to clarify what metric the left y axis is using. I also preferred the two axes labels to appear above the measures (or on the side of the axes), rather than in line with the top most one.