Friday, March 22, 2013

Spring Break

Weekly roundup is on Spring Break...

I'm taking a few days to polish up a conference poster, revise a journal article, and catch up on rest and relaxation.
Source: Minnesota Daily News
See y'all again in about a week.

Friday, March 15, 2013

Data link roundup (week of March 15, 2013)

The week's top data analysis links...

This week's theme: transportation!


Passenger rail ridership is growing faster than any other mode of travel. "Amtrak ridership grew by 55 percent since 1997, faster than other major travel modes, and now carries over 31 million riders annually," according to a new report released by Brookings. Ridership is concentrated in the northeast and western corridors, and 90 percent of Amtrak's ridership is generated by major metropolitan areas.

Brookings' interactive map has more details.

Source: Brookings


A new planning text, Crime and Planning, suggests that transit stations are not nearly as crime-prone as people fear.


While airline fares have increased in recent years, they are still far lower, in inflation-adjusted terms, than they were three decades ago.
Source: Washington Post


On any given day there may be more than 19,000 flights over the United States. Aaron Koblin turns FAA data into a visualization of a day's air traffic patterns...

Thursday, March 14, 2013

Sunday, March 10, 2013

Wealth inequality

Last week I discussed a hands-on exercise to illustrate the unequal distribution of income using 100 pieces of candy distributed among students in the classroom.

If you are interested in an eye-opening explanation of income distribution without the sugar rush, this video may be useful...

Note: To drive the point home, and to correlate more closely with the video (above), the 100 candy exercise could be modified to select 1 student (at random) who receives 40 pieces of candy, while the remainder of the top quintile divides up the other 42 pieces, and the bottom 60 percent of the class has to share 5 pieces of candy among a dozen people.

Friday, March 8, 2013

Data link roundup (week of March 8, 2013)

The week's top data analysis links in honor of International Women's Day


The Economist compiled data on workplace gender equity across more than two dozen countries. Indicators include: higher education, female labor force participation, difference between median earnings for full-time male and female workers, women in senior management as a percent of total, and child-care costs as a percent of average wage.

The result: New Zealand's glass ceiling appears to be highest, while women in Japan and South Korea are likely to suffer frequent headaches as they bump their heads against the barriers of workplace gender discrimination.


New analysis of Jeopardy contestants show that men and women "uptalk" differently. Uptalk is characterized by rising intonation at the end of a sentence that makes the sentence sound like a question (think Valley Girl).

Findings, published in the journal Gender and Society show:
"The more successful a man is, the less likely he is to use uptalk; the more successful a woman is, the more likely she is to use uptalk."


Source: National Library of Australia
As discussed previously, despite media headlines, older mothers are not a new phenomenon.


Despite the fact that amniocentesis testing cannot be done until weeks 14-20 of a pregnancy, the state of Arkansas passed a law to prohibit the termination of a pregnancy after week 12. Amniocentesis tests alert parents to severe birth defects including neural tube deformities such as spina bifida.

In short, by prohibiting the termination of a pregnancy before amniocentesis can be performed, the new Arkansas law will force families to continue a pregnancy even if the baby cannot and will not survive birth.


The Freakonomics folks explore gender similarities and differences in a series of interviews, blog posts, and podcasts... (I may not concur with all views expressed in the podcast, but it is fascinating.)


From the Data Insights archives...

Wednesday, March 6, 2013

Classroom exercise: making data real

Over the past several years, I've tried to explain the (un)equal distribution of wealth to my class.

I've used visualizations like this "champagne glass" chart of income distribution by population quintile:

Source: Sociological Images

And this chart comparing Wall Street bonuses to the salaries of the average worker:
Source: The Economist

But the concept does not always sink in...

A colleague explained a "cookie sharing" exercise that she uses in her class, so I modified this demonstration for my own classroom using candy. (Individually-wrapped candy is cheaper to buy in bulk and is easier to distribute without a crumbly mess than cookies are.) To date, nothing has worked better to illustrate the unequal distribution of wealth than 100 pieces of candy...


In a class of 20* students, I ask students to count off by 5.

With the champagne-glass income distribution chart on the screen, I call the 2s to the front of the room and hand out 12 pieces of individually-wrapped candy.

Each student walks away with 3 pieces.

Then I call the 3s, 4s, and 5s to the front of the room... Twelve students stand around while I count out 5 pieces of candy.

They complain. Occasionally one student dives in to snatch a chocolate before all of the pieces are taken. (Competition for scarce resources, perhaps?) I explain that they'll have to find a knife and divvy slivers of lollipop among the group.

I ask this group, the "lowest 60 percent"of the wealth distribution, to remain standing.

Then I call the 1s - the top quintile - to the front of the room. With some fanfare, I spread the remaining 82 pieces of candy across a desk.

The four students, if they choose to, can walk away with a hefty haul of 20 pieces each.

Usually students in the top quintile realize they do not each need 20+ pieces of candy. Occasionally a student will suggest that they can share their loot with the deprived 3s, 4s, and 5s. This can open up useful discussions of redistribution programs and of fairness in the distribution of resources.

Occasionally someone will suggest that, perhaps, the lowest 60 percent of the class did not work hard enough, and that, perhaps, the top 20 percent earned their diabetes-inducing haul of sweets. This can be an excellent opportunity to discuss how society defines inequality as fair. (I find that an effective response to the "fariness" issue is to pose the question "Do factory workers in China, who may work 10-14 hour workdays, 6 days a week, work less hard than we do?")

And because I'm trying to prove a data point, not start a riot in the classroom, I usually keep a spare stash of candy available to make sure, at the end of the lesson, that 3s, 4s, and 5s each get a piece of candy before they return to their seats.

*Note: This exercise works quite well in class of any size, but the number of pieces of candy should be adjusted accordingly... 200 pieces for a 40-student class, 50 pieces for a 10-student class, etc...

Monday, March 4, 2013

New volume released on the benefits (and burdens) of the American Community Survey

Last summer the National Academy of Sciences Committee on National Statistics hosted a workshop on the Benefits (and Burdens) of the American Community Survey.

For those who could not attend, the workshop summary is now presented in a new volume: Benefits, Burdens, and Prospects of the American Community Survey available for download (for free) from the National Academies Press website.

Friday, March 1, 2013

Data link roundup (week of March 1, 2013)

The week's top data analysis links...


The newest incarnation of SimCity allows players to build fascinating infrastructure like sewers and transportation networks.
Speck, the city planner, noted that this type of model is actually more advanced than what most traffic engineers use in real life. In fact, if the calculations the Sims citizens are making were even more complex, the game could have real-world applicability. “They need to consider that if they drive it’ll take them this long, and that they’ll have to pay this much for parking, and that they can’t work in their car but that the train has Wi-Fi, and so on,”
What the author misses in this piece is that transportation engineers, land use modelers, and urban planners ARE moving toward micro-simulation models like UrbanSim, PECAS, and Activity Based Transportation models that function much like the individual-level decision making that occurs in SimCity.

The models that technicians use just aren't as visually appealing as the video game. This aesthetic difference occurs, in part, because computing power goes toward calculating the tiny and multitudinous decisions about parking cost and travel time that Jeff Speck requests (above) rather than toward slick graphics...

(Also, I have a minor quibble with Jeff Speck's quote in The Daily Beast article: There is no such thing as a "traffic engineer." No one engineers traffic, they're transportation engineers.)


The Freakonomics team wants us to believe that the rise in obesity can be attributed, at least in part, to reductions in cigarette smoking.
Image source: Nationaal Archief
It's an interesting premise, but I'm not sure I buy it...


This map of paid maternity leave policies around the world is either a sad commentary on U.S. family policy, or positive commentary on family-friendly workplaces in 180 other countries...
You be the judge...

Source: New York Times
(Special thanks to Balancing Jane for pointing out the New York Times / xoJane graphic on maternity leave.)