Kids still fat

March 29, 2014 by Eugene Wei

Results of a recent study seemed to indicate a huge drop in obesity among kids, but only in the 2 to 5 year old band. This generated much puzzlement: what would cause just children in that age range to be less obese?

It turns out, according to several who've analyzed the study and the data, the news might just be a lot of statistical noise and fury, signifying not a whole lot. Emily Oster at the new FiveThirtyEight:

If you look at each age group separately, you’re likely to find large changes in some of them. That’s because as you select smaller and smaller groups within a data set, your problem with noise worsens. For example, say you survey 81 people who are 34 years old in 2003, and 85 people age 34 in 2011. What if you just happen to pick a particularly obese group in 2003 and a thinner group in 2011? With just 80-some people, it’s easy to see how this could happen, and yet you might inappropriately conclude that 34-year-olds in the broader U.S. population have become less obese.
One way to prevent yourself from drawing this erroneous conclusion is to apply a more rigorous t-test to the data, one that takes into account the fact that you looked at many subsets. In the JAMA paper, the authors did not do this. They tested for statistical differences within their smaller age groups, but interpreted their p-values relative to the standard significance level of .05. In the case of the 2- to 5-year-olds, the authors reported a p-value of .03. If this were an analysis of the overall NHANES, that p-value should be taken to indicate a significant change. But since the authors are slicing the data and doing multiple tests — they consider at least six different age groups — the statistics need to be adjusted to account for that. With that adjustment, they’d need a p-value of less than .008 to conclude an effect is significant.

Kevin Drum:

To recap: the CDC study was small and had large error bars; other, larger studies find only slight drops in obesity; and there's no indication of any behavioral changes that might have produced a dramatic weight loss. I'd add to that the fact that the CDC data showed no correlation between lower weight at ages 2-5 and lower weight a few years later at ages 6-11.
Bottom line: I hate to be such a buzzkill, but the CDC result seems highly likely to be nothing more than statistical noise. Childhood obesity has barely budged in the last decade.

Is giving directly best?

March 21, 2014 by Eugene Wei

I've written before about effective charity, specifically focusing on optimal strategies for doing the most good. GiveWell still ranks GiveDirectly as its top charity right now.

In the Stanford Social Innovation Review, Kevin Starr & Laura Hattendorf express their doubts.

But is GiveDirectly’s model, as Slate put it: “the best and simplest way to fight poverty”?
No. It’s an experiment—an important one, but an experiment nonetheless. We hope we’re wrong, but our hunch is that it is more of a 1-year reprieve from deprivation than a cost-effective, lasting “solution to poverty.”
Poor people are poor because they don’t have money, and so we think unconditional cash transfers should be judged primarily by how much money recipients are making a few years out from the windfall. GiveDirectly’s work in Kenya is too new to know that, but cash-transfer enthusiasts like GiveWell point to other studies of other cash-transfer programs and predict a slam-dunk.
We looked at those studies, and we’re puzzled as to why GiveWell’s analysts, who rightly prize impact and cost-effectiveness, chose GiveDirectly as one of their “Top Charity” trinity. In the most relevant, longer-term study that GiveWell cited—a program working with unemployed youth in Uganda—recipients of an initial $382 grant had a 41 percent greater monthly income 4 years out. This sounds like a big return on the donor’s dollar.
It’s not. Working out the cost-effectiveness of income-generating funding can be confusing, and we at Mulago find it useful to benchmark grants by calculating the amount of additional income over 3 years, divided by the amount of grant money it took to generate it: the income bang for the donor buck. Of the baseline income of those youth, 41 percent turns out to be $11 per month. By that calculation ($11 per month x 36 months ÷ $382), the unconditional cash grant produced $1.03 of additional income over 3 years per donor dollar, essentially a wash.

Chris Blattman responds with several points, two of my favorites being the following.

1. Victory! If it’s becoming standard to judge interventions by their cost effectiveness, then I can’t be more thrilled. Same goes for GiveDirectly. You can think of cash transfers like the index fund of development (making GiveDirectly the Vanguard). If the NGOs (money managers) of the world can outperform the index funds, then the world becomes a better place.
...
3. Scalable? Whether these other interventions prove as scalable or replicable as cash is another question. Too many NGOs search for solutions to help 1,000 people a year not 1,000,000. But I’m confident some alternatives to cash will prove promising. Some already are, from vaccines to election monitoring, if only because they solve the problems cash cannot. I’m more skeptical we’ll see better alternatives for pure poverty-alleviation, but we’ll see.

I'd never thought of cash transfers as the index fund of charity, but it's a useful analogy. If you can't do better than giving someone cash, then give someone cash.

In Asian cultures, the most common wedding gift is cash. When I was younger that seemed like a gift for the creatively deficient, but at its heart is also a genuine pragmatism that signals the unselfish nature of the giver.

Beware the opposite sex during high school

March 21, 2014 by Eugene Wei

...the share of opposite gender friends has a sizable negative effect on high school GPA. The effect is found across all subjects for students over the age of sixteen, but is limited to mathematics and science for younger students.

From a research paper (PDF) by Andrew Hill at the University of South Carolina.

My mother was always wary of me chasing after girls in high school because she was worried it would interfere with my studies. Perhaps she was on to something.

(h/t Marginal Revolution)

The elevator

March 21, 2014 by Eugene Wei

For most city-dwellers, the elevator is an unremarkable machine that inspires none of the passion or interest that Americans afford trains, jets, and even bicycles. Wilk is a member of a small group of elevator experts who consider this a travesty. Without the elevator, they point out, there could be no downtown skyscrapers or residential high-rises, and city life as we know it would be impossible. In that sense, they argue, the elevator’s role in American history has been no less profound or transformative than that of the automobile. In fact, according to Wilk, the automobile and the elevator have been locked in a “secret war” for over a century, with cars making it possible for people to spread horizontally, encouraging sprawl and suburbia, and elevators pushing them toward life in dense clusters of towering vertical columns.

As the world's populations grows and congregates towards cities, we have to build upward towards the skies, so perhaps it's high time we give the elevator its due.

I liked this bit about how the elevator turned the top floor of buildings from the least to the most desirable level.

The arrival of the elevator upended more than urban planning: It changed the hierarchy of buildings on the inside as well. Higher floors had once been distant, scrubby spaces occupied by maids and the kind of low-rent tenants who could be expected to climb six flights of stairs. The more important people climbed at most one or two flights, which gave brownstone-style homes, for instance, their high-ceilinged parlor floors. While the arrival of elevators didn’t change this right away—the top floor of Henry Hyde’s building was occupied by the in-house janitor—the upper reaches of buildings eventually became desirable. The elevator ushered in the end of the garret and the beginning of the penthouse, as lawyers and businessmen came to appreciate the advantages of having beautiful, bird’s-eye views and respite from the loud noises of the street. Hotel owners, meanwhile, started turning their top floor rooms into their nicest ones. They could even rent out their roofs for garden parties where guests could survey the glittering new city, all without doing a bit of work to get there.

And I wasn't expecting a discussion of serendipity and elevator systems.

For elevator fans like Bernard, Wilk, Gray, and Carrajat, this mixing of worlds is one of the main things that makes elevators so important. And the more opportunities modern life gives us to separate ourselves from others—by getting into our cars and escaping into our suburban homes, by hiding in our cubicles and burying our heads in our social networks—the more the elevator matters as a place that squeezes us together for a moment and forces us to grapple with one another’s existence.
Sadly, there is cause to worry about the future of these moments. The next big leap in elevator technology, already active in large new office towers, is something called “destination dispatch,” which groups people who are going to similar floors together in order to get them where they’re going more quickly. Such a system, Wilk points out, is more efficient in terms of both time and energy, but it also makes it so that people who work on far-flung floors are less likely ever to run into each other. More specifically, it may reduce the chance that someone high up in a company’s hierarchy would share an elevator ride with someone who works down below. Serendipity, in this scenario, begins to recede.