Flabby Coverage

In this month’s Popular Science magazine, Rebecca Skloot takes issue with how the news media have completely misunderstood a recent CDC report. In spite of headlines like “Fat May Be Good”, the report supports what scientists (and the public) have known for years: obesity kills.

The study is titled “Excess Deaths Associated with Underweight, Overweight, and Obesity.” How anyone could read that and reduce it to “Studies Show: Being Fat Is Not So Bad” is beyond me. These results corroborated an overwhelming body of research: Obesity is linked to deadly diseases.

The CDC did find that fewer people died in 2000 from obesity-related causes (111,909) than had been previously estimated (365,000).

But as Skloot points out, it is difficult to estimate obesity deaths. For one thing, many deadly diseases also cause severe weight loss. The CDC study also used the “unreliable” body mass index (BMI) to evaluate obesity.

Major-media coverage […] tended to compound the problem with fuzzy math, often reporting that 25,814 Americans died from obesity, though the actual number was 111,909. Because the CDC study documented fewer deaths in the “overweight” category than in the “normal” category, the media subtracted the number of overweight people who didn’t die from the number of obese people who did—as if deaths that don’t happen somehow cancel out deaths that do.

Taking a trip down memory-chip lane

The New Scientist carries an article about retro-computing: recovering and rebuilding old computers.

“I think everyone remembers their first machine,” says Williams. “If you’re a nerd, it makes a really big impression on your life.”

“My personal favourite is a white Siberian copy of a Sinclair Spectrum,” says curator Tilly Blyth. “But then I’m an eighties girl.”

I obviously picked the wrong research topic

I can’t believe that someone funded this research. In a series of experiments that required volunteer couples and a giant PET scanner, the diligent researchers discovered that many areas of a woman’s brain turn off during “intimate” moments. At those times, “women do not have any emotional feelings,” says Gert Holstege of the University of Groningen.

Hmm… That’s not what they say.

In addition, Holstege helps resolve an important and long standing question for men everywhere: How to tell if a woman is … shall we say, less than sincere in her reactions. Unfortunately, home PET scanners are still prohibitively expensive, and are not commonly built into headboards.

They also include the following cruel footnote:

The team has already done a similar study involving 11 men, […] [however] the results are probably unreliable and need to be repeated. The problem is that PET scanners measure activity over two minutes – and in men it is all over in a few seconds.

And to think I spent all those hours in grad school with only a soldering iron in hand.

Learn computer programming and get a better job?

The AP reports fewer and fewer graduates in engineering and computer science will actually work in the high-tech industry. Programming Jobs Losing Luster in U.S. Instead, they are heading to careers in management consulting and marketing.

As tens of thousands of engineering jobs migrate to developing countries, many new entrants into the U.S. work force see info tech jobs as monotonous, uncreative and easily farmed out — the equivalent of 1980s manufacturing jobs.

“U.S. graduates probably shouldn’t think of computer programming or chemical engineering as long-term careers” according to Albert C. Gray.

At Stanford, career experts […] suggest students develop foreign language skills to land jobs as cross-cultural project managers — the person who coordinates software development between work teams in Silicon Valley and the emerging tech hub of Bangalore, India, for example.

Yeah, that’s the ticket – after all, there aren’t any good project managers in Bangalore, right?