Also in the original graphic: Notes: Whites include only non-Hispanic whites. Blacks include only non-Hispanic blacks. Hispanics are of any race. Don’t know responses are not shown.
Survey Date: February 16-March 14, 2007
What Works
This is one simple way to display data that is supposed to add up to 100%. It doesn’t work well when there are more than two categories, but I would rather see two categories like this than see two categories in pie charts. Two category pie charts often end up looking like pac man which could be particularly unfortunate when it is divorce data that is being displayed.
What Needs Work
I don’t understand why there are colors here. Shades of gray are just fine and would give the graphic a cleaner look overall. More importantly, I am unsure that it makes sense to portray age, race, and gender as the same kinds of data. From a strictly technical perspective, age is ordinal data here but race and gender are nominal data. More broadly, thinking that gender and race and age are having similar impacts on how people feel about divorce just doesn’t make sense.
Another thing that bothers me is the missing data. Sure, there’s a disclaimer than don’t know answers aren’t displayed, but I kept fixating on the fact that the numbers didn’t add up to 100 as they should. I would show those don’t know’s since not knowing how you feel about divorce seems like a piece of data to me, not just something someone forgot. I can forget a behavior (like whether or not I locked the door behind me this morning) but I can’t very easily forget an attitude. I have trouble, for example, forgetting how I feel about leaving an unhappy marriage. It’s also hard to use an “I forget” response when the question has been posed. If you’ve forgotten, now’s the time to remember! How about it, marriage forever or leaving if you’re pretty sure you’d be better off alone? The point is, saying “I don’t know” to this question is a key data point, not just a trivial lapse of memory about what a behavior.
Stimulus Package - Washington Post (Laura Stanton)
What Works
First, the paper allowed three different graphics to run – the overview provided by the bars at the top that show how the stimulus is divided by spending and tax cuts, the more granular breakdown of the pigeonholes for these dollars, and the time line that helps us understand when the money is going to hit the economy (and when we can expect all these transit programs to get going). Second, the main graphic does two things. It is both a fairly simple, readily understood cascading design that draws each category down to its constituent parts across the vertical access. It is also artful – when I look at this I see a sort of mobile hanging over head offering glassy baubles of funding to the madding crowds (ie the states). I’m not trying to insult the states here. In this economy, we’re all the madding crowds, but I really like the fact that the graphic incorporates mood and sensibility. Third, the timeline is a critical component of the stimulus package because there is so much anxiety about when this down turn will be ending. The stimulus money hitting the market is not a direct indicator that the downturn will end, but it is an indicator of when we can start looking for positive economic signs. Furthermore, the timeline could almost stand alone as both a timeline and a description of how the money was allotted. It is nice to be able to look at the package’s pigeon holes/piles of money in two different ways.
I also smiled when I didn’t see a map. Not every story can be visually summed up by the deployment of a shaded map.
What Needs Work
This blog isn’t wide enough to satisfactorily display the graphic so click through to get the whole story.
First, the elegant sophistication of this graphic is breathtaking. I love watching it and I have watched it for long enough to start asking questions about it. Maybe I am different than other people, an outlier of some sort, but in this case I don’t think so and that’s why my own fascination indicates a larger virtue of the graphic. If it draws people in and gets them asking questions, it is doing something right. Holding eyeballs in this media saturated world is a triumph in itself. Having answers to the questions that are posed is a secondary but even more critical step. To figure out what you’re looking at, here’s what the folks who made it have to say for themselves: “Cabspotting traces San Francisco’s taxi cabs as they travel throughout the Bay Area. The patterns traced by each cab create a living and always-changing map of city life. This map hints at economic, social, and cultural trends that are otherwise invisible. The Exploratorium has invited artists and researchers to use this information to reveal these “Invisible Dynamics.” The core of this project is the Cab Tracker. The Tracker averages the last four hours of cab routes into a ghostly image, and then draws the routes of ten in-progress cab rides over it.”
Second, they are right that just knowing where cabs go is more than knowing where cabs go. It’s knowing about urban space over time. It’s certainly knowing where the airport is (and that airports are far away). Looking at this we get to see the grid of the city and the longer stretch of highways and bridges bringing people in/out. It would be nice to see what this sort of ghostly cab mapping technique would reveal about cities I know a little better than San Francisco. Keep this site tucked in your back pocket for later this year, all you ASA meeting-goers.
What Needs Work
I just wish there were a simple way to say a little more about the cabbies themselves, who end up looking like infrastructure or phantoms, rather than actual people. In New York, 91% of the cab drivers are immigrants and only 1% are women (2006 Schaller Consulting). Is there a way that this cab-tracker could become a little more about the humans in the city?
Relevant Resources
Richards, P. and Schwartzenberg S. Snibbe S. and Balkin A. cabspotting San Francisco.
amazon.com, walmart.com, target.com, kmart.comCity Data
What Works
This is a graphic generated by one of google’s trend analysis tools. I simply typed in the web addresses I was curious about and google graphed their relative traffic patterns, using the first page I entered to set the scale. In their words, this is what the tool does: “Google Trends analyzes a portion of Google web searches to compute how many searches have been done for the terms you enter, relative to the total number of searches done on Google over time. “ If I were you, I would ignore the value of the scale and just keep in mind that it is relative. We’re measuring not total volume, but the volume of these four sites relative to one another.
Amazon clearly has far more traffic than the other three sites. Because walmart, target, and kmart rely on their physical stores, just looking at this web traffic does not tell you much about relative sales. I don’t who else is like me, but I often use amazon as a sort of loosely organized reference site, finding it faster to look their for publication dates of books than to go to my library’s site or fish the book off my shelf. I might be an outlier in this regard – most people don’t spend time every day wondering about publication dates – but there is probably a fair amount of traffic on amazon related to their product reviews that may not result in sales at amazon. All of this activity generates traffic, not sales. All three of the other retailers also feature customer reviews, by the way.
What works here is sort of unclear. On the one hand, just look at how similar walmart.com and target.com are. They track each other so closely they are visually difficult to distinguish. And just look at how important the holidays are to all these retailers.
The city data relies heavily on which website is input into the search field first. Seattle might not have even been included if I had put walmart.com first, but many cities in the south would have been. Minneapolis would be up there if I had put target.com first. kmart.com first motivates Philly to the front of the pack.
What Needs Work
My biggest critique of this sort of thing is that it’s unclear what the heck to take from it. If you are just trying to beat some competitor, having google show you their relative traffic is immensely useful. But what else is this good for? Anyone?
Let me just point out that this only works for large sites. Google can’t tell us much about the vast sea of smaller sites.
Open Access – Transparency
In the end, though, the move towards making data publicly available is fabulous. I can’t see how this particular instance is broadly useful to me – it’s fascinating, sure, could be good for marketing departments internal to these companies, but then what? My confusion just means that I am a short-sighted fool. Google should be applauded for creating a non-prescriptive tool to explore the data they have that is so basic it can be used by anyone for who knows what.
Relevant Resources
Benkler, Y. (2006) The Wealth of Networks: How Social production Transforms Markets and Freedom. New Haven: Yale University Press.
US population growth 1790-1990 {freeze frame at 1920} - University of Kentucky Appalachian Center
Link rot note
This post used to source a population growth animation from zachofalltrades.net but that website is no more. The University of Kentucky Appalachian Center is better, so count yourselves lucky if you missed the original post in favor of this update.
What Works
First, you must click through and watch the animation. Praise #1: yay for gifs.
Like the previous post that looked at China, this animation is trying to tell a story about population growth over time. The major difference is that the Chinese example was strictly demographic – looking at variables like gender and age but not at all concerned with geography. This one shows both geography and population growth though it does not include information about gender, age, race, etc.
What Needs Work
If this graphic were three dimensional, if density piled up, it would start to ‘feel’ heavier over time so that the same way that the westward expansion of the population just appears without you having to puzzle it out, the density of population in cities would be simply obvious. This is not meant to be a dig to the graphic’s creator. I just offer this critique as a way to think about just why and how ‘seeing is believing’. Watching the population move west is certainly a ‘seeing is believing’ moment because viewers do not have to think, they just have to watch. Realizing that the population of the US is now hugely larger than it was back in the 1880’s actually takes a little thought. You have to realize that not only did people move west, but they continued to live in the east in greater densities which is indicated by the size of the yellow circles, but would be even more obvious if the cities were like little hillocks on the landscape. Big yellow dots equaling density requires a move from the ‘seeing is believing’ to something else. If, however, the map grew in the third dimension as a more direct representation of the mass of humanity sitting on the face of the earth at these locations, we’d be back in ‘seeing is believing’ territory.
A graphic that is a ‘seeing is believing’ creation is instantly legible and can free your brain to think about other things which is a good thing. On the other hand, a graphic that achieves a ‘seeing is believing’ mechanism will end up obscuring complexity. This is good when that complexity does not add to the ability to think through the next set of concerns, but can be a serious drawback. It is good to be able to get a diversity of people able to quickly grasp an argument, but there is a danger in presenting an hermetically sealed glossy image.
My apologies for failing to post for a few days. If you noticed, I’m flattered. I had some deadlines and limited access to the internet late last week. Unfortunately, March is a very busy month and I will likely have this problem again before the month is out. I’ll try to make up for it when I can.
What Works
I have always been a fan of the population by gender and age chart, even in the static form that you see before clicking through above. It is quite an achievement to clearly represent three different variables on a two dimensional graph. It helps immensely that gender here is a binary value. If it were tertiary or tertiary plus, this strategy would fall apart. Once you click through, you’ll see that the animation adds yet another variable, time. And time is a real kicker here. You can see how China’s population goes from having many young people and few old people to 2050 where the largest category is between 60 and 64 years old. Great way to take an old graphic technique – the static version – and animating it.
(I would love seeing this thing as population by age sticking married people on one side and unmarried people on the other as an animation.)
What Needs Work
The colors and overall treatment of the graphic as a designerly element. Red and orange makes it look a little like it’s yelling ‘Caution! Proceed at your own risk!” the whole time. But then, I guess we all have to worry about what is going to happen when the population pyramid becomes a slender pillar with an ionic capital.
Relevant Resources
United Nations (1999): World Population Prospects. The 1998 Revision. New York. Link to animation. [graphic credit to Heilig, G. 1999]
Death Penalty Costs in Maryland - The New York Times
What Works
As you may recall from last week’s post on the death penalty, the use of the death penalty is not a deterrent to murder. Today in the New York Times, an article by Ian Urbina focuses on the fiscal reality of the death penalty citing a study done by the Urban Institute along with proposed legislation to get rid of the death penalty to help states meet their budgetary goals. “The Urban Institute study of Maryland concluded that because of appeals, it cost as much as $1.9 million more for a state prosecutor to put someone on death row than it did to put a person in prison. A case that resulted in a death sentence cost $3 million, the study found, compared with less than $1.1 million for a case in which the death penalty was not sought.”
What works about the graphic is the combination of bars with numbers. Basically this is just a spreadsheet with some bars next to the costs. For those of you social scientists out there who have grown fond of your tables, think about adding bars with interval level data (like costs and population).
What Needs Work
The bars should also appear in the last row on the table where the totals are displayed if this bar-in-table trick is going to work. I can see that the graphic would have had to stretch to accomodate the $3m bar, but the visual effect of having the whole table stretched to fit that bar would have been powerful. As it is, the visual impact of the bar technique is not fully realized.
Relevant Resources
Roman, John; Chalfin, Aaron; Sundquist, Aaron; Knight, Carly; and Darmenov, Askar. (1 March 2008). The Cost of the Death Penalty in Maryland. Washington, DC; The Urban Institute.
This visualization is going to take a bit of explaining. Mapping the internet is a question that has intrigued folks who are worried about internet security, the digital divide, robustness, even artists who just wonder about all those bits of information flowing around us.
Remember The Matrix? Couldn't help but mention it here.
This visualization attempts to describe the structure of the internet as a network, not to map its black holes, censorship holes or describe actual geographic nodes like Akamai in yesterday’s post. This is a different sort of map and it requires some background reading. The authors set up a strategy for exploring the network terrain of the internet that generated these three areas – the central nucleus area consisting of the most highly connected nodes, a fringe around the edges of a whole bunch of pages that would be cut off completely if the nucleus were removed, and then a sort of spongy area in between these extremes full of nodes that could connect to each other if the nucleus were removed but not nearly as efficiently. Call it the peer-to-peer zone.
Here’s how the authors described the process that generated the three classes of nodes:
First, we decompose the network into its k-shells. We start by removing all nodes with one connection only (with their links), until no more such nodes remain, and assign them to the 1-shell. In the same manner, we recursively remove all nodes with degree 2 (or less), creating the 2-shell. We continue, increasing k until all nodes in the graph have been assigned to one of the shells. We name the highest shell index k max. The k-core is defined as the union of all shells with indices larger or equal to k. The k-crust is defined as the union of all shells with indices smaller or equal to k.
We then divide the nodes of the Internet into three groups:
1. All nodes in the k max-shell form the nucleus.
2. The rest of the nodes belong to the (k max − 1)-crust. The nodes that belong to the largest connected component of this crust form the peer-connected component.
3. The other nodes of this crust, which belong to smaller clusters, form the isolated component.
Even if you don’t spend your days dividing networks into k-shells, I hope you now understand that this model’s strength comes from the fact that the structure was generated rather than imposed by initial assumptions. There were no initial assumptions.
What Works
Success here is that people who do not study networks can understand what these researchers did at all. Most highly specialized research (and pretty much all research is highly specialized) only makes sense to the people occupying the sub-sub-discipline actively working on those questions, equipped with the right language, fully immersed in the discourse of the niche. That would have been true if I had just tried to read this article without the accompanying image.
I also think it helps immensely to see the sketchy, comparatively unglossy schematic along with the polished final image. The glossy version adds in enough detail that I might have missed the big picture without having the schematic there to remind me that it isn’t about color or distance – that the contribution is all about the three types and their relationship to one another.
What Needs Work
Similar problem with this image as I had with yesterday’s image: the final image is so glossy and sealed that I feel like it’s hiding something. The more gloss on an image, the more it becomes impenetrable to critique. It presents itself as hermetically sealed – how can anyone get under the skin and assure themselves that this is a trustworthy image? This glossiness of the final image is probably why the schematic has so much appeal. It’s easier to see how the two were put together and *why* it is the way it is.
Aesthetically, I am not sure I like the colors and I think I would have tried to achieve the look of a solid core, a very fringe-y outer layer that has more volume but is almost insubstantial in its lacy-ness, and then a middle layer that sort of looks like a network made of jello. It is so easy to say these things when you don’t have to kill yourself in photoshop and illustrator making them happen.
Note
[There is another post on Graphic Sociology about mapping the internet about visualizing the map of an individual site which is here.]
Akamai Internet Traffic - Click Through for Interactive Graphic
Internet Traffic
This week we’re going to have a look at the internet. Here are two reasons why:
1. The not entirely superficial reason is that there are many great visualizations out there dealing with the internet, internet traffic, internet usage patterns, and so on. Many are interactive so you can play around with them yourselves.
2. The larger theoretical question about studying the internet and online behavior goes something like this: How much is people’s online behavior reflective of their offline behavior? Are people role-playing when they’re online, trying out personas they may not fully embrace offline (see Sherry Turkle)? Or is online behavior seamlessly integrated with offline behavior? We IM the people we’re about to have dinner with indicating that the people we talk to online are just about the exact same people we talk to offline? And if the relationship between online and offline behavior is somewhere between these two, how can we figure out just what is going on?
What Works
The graphic above is just a screen capture from Akamai’s site. In order to get the full impact, you have to click through and play around with it. Akamai has a slew of other visualizations you can play with that deal with network attacks, latency/network failure, retail data, news traffic, and so on.
Just to be clear, Akamai is a private company providing web-optimization services. In their shareholders’ quick facts, they say they serve up 10-20% of global internet traffic. What does this mean? It’s easy to forget that the internet requires physical structures, but this is part of what Akamai does. They maintain “40,000 servers in 70 countries within nearly 950 networks” all over the world slurping up electricity and information at about equal rates. The reason they do this is because if you are, say, a blogger in New York and you store your files on a server just down the hall (which is unlikely, but play along), if someone in Singapore wants to read your blog, the request is going to have to come all the way from Singapore to the server down the hall from you in New York and then the files will have to be sent all the way back to Singapore. This takes time, there might be network congestion along the way and if you are serving your readers in Singapore something a bit more bandwidth intensive than text (say a little clip of a new car racing around a track or a high quality music download) the person in Singapore may just lose interest before they even get the whole file. Akamai gets around this in part by duplicating files and storing them on servers all over. So if your reader in Singapore wants to access your site and you’re an Akamai customer, they will end up pulling those files from a server much closer to them, maybe in Singapore, but at least somewhere much closer than New York. Akamai’s clients tend to be Fortune 500 companies with global client bases and companies that rely on being able to transfer heavy files reliably and quickly (like music and software downloads). They do more than just the physical infrastructure, they mobilize their resources to detect net attacks, congestion, and then to re-route and avoid those things. The bottom line for us is that they make some of their knowledge of the ‘net available in these visualizations like the one above.
What Needs Work
I would love to have more granularity and access to the actual numbers and the methodology. All these shiny interactive graphical toys run the risk of being too glossy, not data-transparent enough.
Not as Shiny, Quite Helpful
Internet Global Penetration Rates - Internet World StatsGlobal Distribution of Internet Users - Internet World Stats
These two graphs give a quick overview of who is using the internet by geographical location. You’ll see that rates of traffic can be a bit misleading – not all continents have the same population. That’s why I included the rate of internet penetration within the continents. A low rate of penetration tells you a lot about how the digital divide which is a very real problem. More on that later this week when we will address the digital divide directly. For now, it’s enough just to notice the difference in looking at the flashy, glossy Akamai graphic and the simple bar graphs. I don’t know about you, but I quite enjoyed playing with the Akamai graphic and encourage interactivity. Still, the combination of these two bar graphs above gave me a clearer answer to the big question about who in the world has access to the internet in the first place.
For those of you who aren’t watching the Oscars (or, in fact, maybe especially for those of you who are), I send some statistics your way on a Sunday evening. It doesn’t fit with a theme and there’s no way it’s going to be as popular as the blog about marijuana arrests in New York City. (Note that like any curious person, I fully intend to test my hypothesis that writing about drugs is more popular than writing about sex. Coming soon is a blog about measuring marital infidelity, an historically slippery subject that has generated competing statistics and tends to say more about survey methods than about sexual habits.)
But for tonight, I am sending you to a surprisingly emotional essay by Stephen Jay Gould on the trouble with reducing statistics to the central tendency. Yes, I said emotional. And then I said ‘central tendency’. What, you may wonder, can get your cold hearts pumping while talking about how to measure the central tendency? In a word? Cancer. In a few more words? A life expectancy delivered in terms of a right skewed median of 8 months.
He uses his own biography to make a broader point about the general tendency to divorce the intellect from the emotions: “Many people make an unfortunate and invalid separation between heart and mind, or feeling and intellect. In some contemporary traditions, abetted by attitudes stereotypically centered on Southern California, feelings are exalted as more “real” and the only proper basis for action – if it feels good, do it – while intellect gets short shrift as a hang-up of outmoded elitism. Statistics, in this absurd dichotomy, often become the symbol of the enemy.”
logic + love = the well-lived life? These are the questions I don’t even try to answer, it’s why I do sociology, not philosophy.
Seeing Skew
Skew Graph Examples - No Skew, Left skew, Right Skew (which is closest to Gould's case)
Just to refresh your memory on skewness, here’s a visual reminder of what’s at stake. Refer back here when Gould talks about the many people who aren’t diagnosed with the type of cancer he had until they die, stacking the left side of the graph high with cases of life expectancy equal to zero and creating a right-skewed life expectancy.
Epilogue: Gould is no longer alive, but he didn’t die of the cancer in this essay. He lived for another 20 years and died of a different cancer at age 60.
Relevant Resources
Gould, Stephen Jay. (1985) The Median Isn’t the Message currently reposted all over the blogosphere, but originally published in Discover Magazine in 1985.
About Graphic Sociology
Analyzing the visual presentation of social data. Each post, Laura Norén takes a chart, table, interactive graphic or other display of sociologically relevant data and evaluates the success of the graphic. Read more…
![US population growth 1790-1990 [freeze frame at 1920] - University of Kentucky Appalachian Center](https://thesocietypages.org/graphicsociology/files/2009/03/1920_us_population_growth1.png "US population growth 1790-1990 {freeze frame at 1920} - University of Kentucky Appalachian Center")
