CS-5630/6630 | Visualization | Fall 2014
Week 14: Income Change in U.S. and Other Advanced Countries - Printable Version

+- CS-5630/6630 | Visualization | Fall 2014 (http://www.eng.utah.edu/~abigelow/bulletinBoard)
+-- Forum: CS-5630/6630 | Visualization | Fall 2014 (/forumdisplay.php?fid=1)
+--- Forum: Design Critiques (/forumdisplay.php?fid=3)
+--- Thread: Week 14: Income Change in U.S. and Other Advanced Countries (/showthread.php?tid=108)

Pages: 1 2


Week 14: Income Change in U.S. and Other Advanced Countries - lzhang - 12-04-2014 03:29 PM

The link for the visualization is at
http://www.nytimes.com/2014/04/23/upshot/the-american-middle-class-is-no-longer-the-worlds-richest.html?abt=0002&abg=1&_r=0

There are three figures on that page. The critique is for the second figure (from top to bottom). A screenshot of the visualization is provided at the end.

The author(s) want to use this visualization to illustrate that the middle class of U.S. are less richer than middle class in other countries nowadays compared to decades ago.

The dataset for this visualization are time series of per capita income (after inflation and tax adjustment). These time series are classified by two hierarchical attributes: country and the percentile of income level within a country.

This visualization compares U.S. to 10 other countries: Canada, Britain, Germany, Norway, Netherlands, France, Sweden, Finland, Ireland, and Spain. Within each country, 11 classes of income level are considered: 5th, 10th, 20th, 30th, 40th, 50th (the median), 60th, 70th, 80th, 90th, and 95th percentile.

This visualization uses small multiples to present the time series according to their corresponding income percentile. Within each view (fixed percentile), each time series for a country is represented by a line chart (spatial encoding). Color encoding is used to differentiate U.S. (purple) and other countries (gray in normal case and black when highlighted). The color choice is very nice in the sense that the contrast is very sharp. These encodings are the same across all views.

Although there are a total of 11 countries, only 3 colors are used to encode them. The 10 countries other than U.S. are aggregated into 1 class--``other countries'', since the main purpose is to compare U.S. to other countries. This is a very smart choice. If one uses distinct colors for countries, given that we have 11 countries, the number of colors needed is closed to the upper limit--12--of the suggested number of distinguishable colors for comparison, thus might overwhelm readers and distract the attention of comparing U.S. to other countries.

The visualization also enables comparison between U.S. and a specific country by highlighting: when the mouse is moved over a line corresponding to a country, that line pops out by changing color from gray to black and a legend indicating the name of the country appears in the bottom-right region of the figure. Moreover, this is a linked highlighting that all line charts corresponding to the same country across the views will pop out.

There is another interaction which is a checkbox (located below the title) that allows readers to switch between smoothed line charts (the first attached figure) and unsmoothed line charts (the second attached figure). The unsmoothed line charts corresponds to the raw data and contains wiggled lines. The smoothed line charts corresponds to the transformed data that is obtained by smoothing the raw data, and it contains straight line. The smoothed data is presented by default. The transformation from raw data to smoothed data emphasize the purpose of the visualization is to compare the trends of income evolution between U.S. and other countries rather than the detailed income evolution itself. Another way to interpret these two datasets is that the visualization for the raw dataset may be constructed for the purpose of discovering trends, patterns, and relationships; and the visualization for the smoothed dataset may be constructed for the purpose of presenting the findings--smoothing emphasizes the findings and suppressing the noises.

I like this visualization very much.
  • It perfectly serves for the purpose of presenting the relationships of the trends of income evolution between U.S. and other countries.
  • The encodings are very precise and succinct thus very expressive, and spatial encoding for time series and color (hue) encoding for categories thus very effective.
  • The visualization nicely and naturally fits into the space of the paragraph such that the whole page is visually appealing.
  • Finally, the most significant reason I like it is that it not only presents the author's point, but also illustrates something that is not mentioned nor intended by the author: we can observe from the slopes of the line charts across the views that the income of richer classes increases much faster (steeper lines) than poorer classes (flatter lines). Moreover, we can observe (from the smoothed data) that the increment of income is almost an increasing function of the income level (percentile), and this pattern holds for all countries studied. We can easily draw a conclusion from this observation: the wealth gap is increasing!
    This is an excellent example that visualization not only helps us to find intended result, but also can lead to unexpected findings.

The only suggestion I have for this visualization is that it might be helpful to provide legends for countries such that we can highlight the desired country by click the corresponding legend. In the original visualization we can only select lines without knowing countries. By these amendment, we can select in both directions and increase flexibility.

[attachment=30][attachment=31]


RE: Week 14: Income Change in U.S. and Other Advanced Countries - ankit - 12-04-2014 07:02 PM

I cannot agree more with the review about this visualization.
For the purpose, the visualization looks really good.
Good things about the visualization:
1. Mouse hover feature helps differentiate each country (compare with that of US)
2. Color contrast is good. (Grey, black and purple)
3. The option of removing the smoothing of the curve (to show the actual curves)
4. The space utilization for the visualization is not very high
5. I agree that the slope for above median is higher (steep) than compared to the slope for below median (flat)

Suggestion:
1. Idea of selecting the country from the legend helps eliminate the permutations to find comparision with a particular country
2. I know that the goal was to do a comparision with US but it would have been nice to make a comparision between any 2 countries. (Interactivity to select any 2 countries from the legend)


RE: Week 14: Income Change in U.S. and Other Advanced Countries - citou - 12-05-2014 01:24 AM

The visualization is simple but clean and effective. It's rare that we can choose different curves: smooth or shape to show the trend, although I don't think it's necessary to have two curve styles, it's a good idea which is inspiring.
Normally, we'd like to choose various colors to represent different countries. The author just use gray and black to represent other countries that we can distinguish clearly. But the limitation is also obvious that we could just select one country each time. Also, we don't know which line represents which country. The only way finding a specific country which we care is testing one by one. Considering the optical limitation of color discrimination, maybe the author could give more choice: more countries could be shown at one time, countries in the same region such as Northern Europe or Southern Europe could be shown by different color hue with their names for example.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - u0869331 - 12-06-2014 09:59 AM

I think this is a very effective visualization. I like that it uses small multiple to represent the information for different categories of income over time. It shows a lot of comparable information in a single glance without any need to scroll the page.
I like the clean color coding used to compare the plots for all countries with USA. The option to apply smoothing of the curves or not is an added bonus. The smoothed curves a nice overview of the rate of change in the income of the income category over time.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - holtvg - 12-06-2014 08:03 PM

I think the visualization does an ok job of showing the data through a line graph but I think it takes some time to understand such as what the percentile means. It can be a little difficult to see the relationships between all the grey lines below since there's a lot of overlap so some color coding can help as well. I do however like the mouse over effect so you can see a clear overall trend between each country in comparison to US in middle class income.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - Chaomeng - 12-07-2014 02:13 AM

I think this is a pretty good visualization because it pretty much serves its purpose to give a clear view to support this article’s point. The visualization itself is clear and simple. It uses smalls multiple to represent the information of different income level categories over the time.
Two interactions are provided in it, both of which are simple but useful. The first interaction allows user to move the mouser and highlight the country line and compare to that country with U.S directly. The other interaction allows users to choose to display unsmooth line with raw data. But I think the smooth line is more useful because you can find the income increasing/decreasing tendency with the time.
I think this visualization really support the article’s idea that the US middle class income is decreasing compare to other countries, and here I agree with the poster that we can find that: the higher percentage of income level is, the higher the increasing rate is. That really mean the income gap between low level and high level in increasing.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - Khalid - 12-07-2014 11:35 AM

I think if they have computed the income difference between different countries and the U.S and then plot the results that would make comparisons easier. And they can keep the original data available by showing it upon mouse hover.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - Prashant - 12-07-2014 11:59 AM

There is no doubt that this is one of most coolest visualization. The amount of data being represented in a simplistic way is truly remarkable .

Pros :
1. Using summary line charts to indicate comparison of house hold incomes across multiple countries is quite effective.

2. The hover effect to perform comparison between 2 countries is really great.

Overall it is a really great visualization.

Critique Points:

1. I have always wondered if shortening labels is good or bad way to represent data. Majority of times people opt to represent all the information even if it add more content to the visualization. But I still feel it is debatable whether to include complete year details when performing year comparisons which involving different millenniums . Shortening 1980 to 80 and 2010 to 10 in my opinion adds an additional effort for the reader to understand the data.

2. It would have been nice if the designer could have used roll overs to indicate exact values by which the two countries differ while scrolling across different graphs between 1980 to 2010.

3. Also I felt labelling used to indicate american households is slightly confusing. At first glance when I saw 5th percentile I couldn't exactly determine what that label represented.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - amirali1985 - 12-07-2014 10:26 PM

1) The visualization is sparse, and a little dull especially for the first figure. While it conveys information, I believe for a popular article (and not a research paper), something more eye catching would have been better.

2) For the second figure, I thought the parabola effect was striking. It's interesting to see the increasing upslope at the higher incomes, and I think it conveyed the narrowing gap.

3)Finally, I thought the methodology was slightly flawed in only comparing income. Adjustment should and could have been made, in my opinion, for the level of government benefits and subsidies offered in each country. There's a large disparity between say, Norway and the US on that front, and some attempt to quantify this (admittedly challenging) confounding variable would have given a more realistic picture of the comparative standard of living.

I will say though, that the second figure wasn't terrible, and I liked that one could mouse over to identify which country was being discussed.


RE: Week 14: Income Change in U.S. and Other Advanced Countries - padmashree - 12-08-2014 12:01 AM

A nice piece of visualization.
The first visualization compares US's middle class income with that of other countries like France, Canada, Mexico, Ireland among others between 1980 - 2010. The choice of colors is good but not appropriately matched I feel. It would have been good if the color for US would have been purple instead of grey since it takes the attention away from US in this case. Also the gap between the incomes of US and that of the other countries keeps widening as we move towards the right. This gives an illusion that US middle class is faring better than other countries which is actually what the writer is trying to write against.
The second visualization is good. As Izhang mentions, the option to choose between smoothened line charts and unsmoothened line charts is made available. The default choice of presenting in unsmoothened line chart makes the comparision easier and less cluttered in some ways. Also there are small dotted lines which represent the percentiles which makes observation a little easier. I would have liked it better if we could choose multiple countries at once and compare them with that of US though. Choice of black on rollover is good.
All in all a good visualization