Thread Closed 
Week 13: In Climbing Income Ladder, Location Matters
11-20-2014, 06:30 PM (This post was last modified: 11-20-2014 06:36 PM by holtvg.)
Post: #1
Week 13: In Climbing Income Ladder, Location Matters
Link: http://www.nytimes.com/2013/07/22/busine...tters.html

This visualization shows the chance that a child raised in the bottom fifth of the income gap will rise to the top fifth of the income gap when grown up in the United States. The top fifth of the income gap was assumed to be a mid-high salary of over $70k per year at age 35 and $100k per year at age 45. The visualization is very effective in encoding color so it's easy for the viewer to compare income gaps across various locations in the US as well as the highlight interactivity which helps see the percentage in various areas across the states.

The use of a legend helps the viewer determine what each color represents. I was surprised that Salt Lake City and the Vernal area had one of the highest likelihoods of a child climbing into the top income gap. I think some improvements that could be made could be to show the changing income gap over a time frame of several decades as well as some more markers to show the income probabilities in more cities. Judging by this data I don't think it's changed much in the last year as it's from 2013 but you never know.
Find all posts by this user
11-22-2014, 02:08 PM (This post was last modified: 11-22-2014 02:08 PM by lzhang.)
Post: #2
RE: Week 13: In Climbing Income Ladder, Location Matters
The critique given by holtvg is thorough and precise.

This visualization tries to illustrate the relationship between geographical location and the chance to rise over the income ladder. It is very natural that the author use a map to represent the geographical locations. The chance for rising over income ladder is encoded by color. Instead of treating the chance as a continuously changing quantitative variable, the author categorized this variable into bins: below 4%, between 4% and 10%, ..., and between 20% and 35%. Then the author uses a colormap for categorical data to encode them.

This visualization seems really expressive and effective in the sense that it makes the contrast between low chance and high chance very sharp. There is some subtlety here.

First by barely examine the legends, it seems that the colormap is divergent while the data is sequential. Although this may suggest the colormap used here is inappropriate, it is actually very effective if the author's purpose is to contrast the regions with low chance and regions with high chance.

Second, the category bins are not of the same size, that is, they span different size of intervals. This is something I don't understand. The middle three are regular in the sense that there is an increment of 5%. The first is slightly off with a gap of 6% while the last has a big gap. This division seems artificial. One explanation is that there is no observation in the bins spanning from 25% to 30% and the author aggregates this bin with the previous/next bin. However, by examining the data (through interactivity) carefully, I find that there are observations between 25% and 30%, for example, Plentywood area from North Dakota has a chance 26.4%.
Find all posts by this user
11-23-2014, 04:03 PM
Post: #3
RE: Week 13: In Climbing Income Ladder, Location Matters
I thought the critique that holtvg was very accurate. He mentioned both the pro's and con's of the visualization, plus a great description of what the visualization is trying to explain.

I agree with lzhang's comments on the legend. It doesn't make sense to have the first two bins have a difference of 6%, some with 5% and others with 15%. It seems really inconsistent, data can be misconstrued in this way.

I did like that the author choose to used a map of the United States ( as I would expect it to be). The colormap I thought was also helpful, but I think different colors could have been selected, such as yellow to red, or blue to red.
Find all posts by this user
11-24-2014, 02:07 AM
Post: #4
RE: Week 13: In Climbing Income Ladder, Location Matters
I completely agree with lzhang and holtvg description of the merits and shortcomings about the visualization.

The designer has made use of combination of multiple hues color schemes.(YlOrBr,PuBuGn) etc in order to represent the different percentages. Using this approach significantly improve the visualization look , but at the same time it becomes difficult to interpret the percentages unless we use mouse overs to manually determine the percentages.

For eg the Bakersfield area in California shows 12.4 percent and is shown by slightly lighter yellow color which is assigned to 10% in the legend. In comparison to Provo area (14%) which is close to the shade of light green. In my opinion the legend should completely describe the color scheme used in the map . It could have been a better approach to follow 1 color scheme between 2 ranges.
Find all posts by this user
11-24-2014, 04:18 AM
Post: #5
RE: Week 13: In Climbing Income Ladder, Location Matters
I think holtvg's critique was a bit short, but very accurate. There isn't very much to this visualization. It is just a simple color map with qualitative information overlayed on top of it. I don't think that the color scheme used was very good for the information type. The colors go from dark red, to yellow, and then to dark blue. This type of color scheme would be good for categorical data, but I don't think it works very well with quantitative data.
Find all posts by this user
11-24-2014, 10:10 PM
Post: #6
RE: Week 13: In Climbing Income Ladder, Location Matters
I completely agree with holtvg's short but precise critique. The color are simple and distinct. The designer has done good job of using multiple hue color combinations to represent the data.The rollover feature of the visualization is also useful to highlight the minor differences in percentages. Overall I think its a fairly simple and straightforward information
Find all posts by this user
11-24-2014, 10:49 PM
Post: #7
RE: Week 13: In Climbing Income Ladder, Location Matters
I agree with the critiques in general. The use of color in the visualization is good for showing the contrast between high and low probability areas. The legend is easy to understand and informative, though I'd like to know what the income of the bottom 5th is. Also, some of the 10-15% areas seem rather large, I think it would have been more informative if those areas had been split up, or some kind of representation of population had been presented.

The article has some other interesting and interactive visualizations that I think complement this visualization rather well.
Find all posts by this user
12-01-2014, 05:10 PM
Post: #8
RE: Week 13: In Climbing Income Ladder, Location Matters
Excellent summary, holtvg. The map is definitely a natural choice for the information presented within as holtvg stated. I would also agree that the map does make the contrast between the chances of climbing the income ladder apparent. In all honestly, not sure what you mean by the colormap being divergent versus sequential data. The separation of the bins also is not clear in why the author created them non-uniformally. There is an adequate amount of interactivity with the visualization as well. All-in-all, it is an effective visualization.
Find all posts by this user
Thread Closed 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)