Academic Horror Story (Duke University)

Seth Roberts5 Comments

Duke University officials have known since 2009 that there were serious problems with Anil Potti’s research — serious enough to believe it is fraudulent. Here is how one researcher put it:

The Duke investigators said their data showed that expression of a particular gene, ERCC1, correlated with response to some agents. However, the commercial microarray chip the Duke investigators said they used in their experiments does not include that gene. “I admit this is one for which I do not have a simple, charitable explanation,” [said] Dr. Baggerly.

Potti, you may remember, lied about having a Rhodes Fellowship. Duke’s first investigation found him innocent.

Later events caused Duke officials to reconsider. They are still making up their minds. This is a horror story because a clinical trial based on Potti’s research is in progress. A hundred cancer patients are getting treated according to Potti’s research — that is, according to research that is probably fraudulent. Duke has done nothing to warn the patients or stop the trial.

The whole thing reminds me of UC Berkeley researchers taking weeks to tell a woman she had a large lump in her brain. As if their legal liability were more important than her life.

The Thick-Fingered Surgeon

Seth Roberts3 Comments

Kim Øyhus, who has a proof that correlation is evidence of causation, told me this story of medical overtreatment:

At 16 I got glass splinters and sand inside my hand when a test tube broke because it was handled too hard. The small local clinic sewed the wound shut without close examination, so a glass splinter and sand remained deep inside. About 5 years later the glass splinter cut itself loose because of bowling, and for about 10 years made the hand problematic to use. It became swollen and partly numb each time I used it with force.

So, 5 years into this I decided to do something about it. My mother contacted the local clinic again. “Come back tomorrow, and we will look at it.” the doctor said to me at the first examination.
The next day I arrived to a ready operation table of the simplest kind, and just the doctor.

“I thought we should look at it today, not operate it,” I said.

“You know perfectly well that that means operating,” he said.

I took that answer as a hint that he might not be an honest person.

In addition he had nervous tics in his shoulders and arms, as well as big thick sausage fingers, as if he plowed hard soil every day.

So, there I lay on the hard mattress, arm outstretched while he plunged the local anesthetic needle hither and dither inside my hand, while my unease continued to grow. So when he took the scalpel and pointed it at my hand, I said “No. There is not going to be an operation today.” and rose from the guerney.

“You can’t just leave like that!” he said.

“It is my hand, so I decide what is to be done with it,” I answered, and left the room with my mother.

Having seen the entire ordeal silently, which is very atypical of her, she was visibly relived, and agreed entirely with my decision. She thought he was extremely nervous.

As we drove away, I saw the doctor sitting smoking on some wooden stacks outside, looking somewhat forlorn. I waved, and he waved back.

Fortunately, the needle had moved the glass splinter to a better place, so the hand was useful again for a few years after that.

When it started getting really bad again, I asked other doctors how stuff like that could be fixed, and they told me that hand operations are exceedingly difficult due to the delicate nature and lots of nerves, tendons, muscles, and so on everywhere tight together, so it requires a surgical team with a very good and experienced surgeon, long operating time, and often unconscious anesthesia. And so it was. They found sand inside nerves. I can tell you that is uncomfortable to have. The glass splinter I knew was there because I could feel it by poking hard with my fingers before the hand got swollen, was nowhere to be found.

The recovery took many years. It’s OK now. And eating omega-3s and dropping carbohydrates this last year significantly improved it. It became softer, more bendable, and more sensitive, less numb, even though it is decades old now.

First Day of Class

Seth Roberts7 Comments

Today was my first day of class at Tsinghua. I am teaching a seminar called Frontiers of Psychology. There was only time for about half of the 40-odd students to identify themselves, which included saying their favorite book. Three girls said their favorite book is Pride and Prejudice. Two said The Little Prince. One said Harry Potter. One said Rebecca by Daphne Du Marier (published 1938). One boy said he didn’t have a favorite book — reading books was a waste of time. One boy said his favorite book is Ulysses.

Most of them, perhaps 80%, chose a non-Chinese book as their favorite. One French, two German, the rest English (which they may have read in Chinese translation). At first I was surprised but then I realized it made sense. Chinese civilization was more advanced than European civilization for a long time but when Gutenberg invented the Western version of the printing press everything changed. In Europe, unlike China, books became cheap and literacy spread. With literacy came a book industry. A large number of Europeans have been reading books for 500 years. In contrast, the Chinese language, with thousands of characters (in contrast to 26 lower-case and 26 upper-case letters) made printing difficult. With reading material rare, so was literacy.

The Treatment Trap by Rosemary Gibson

Seth Roberts2 Comments

The Treatment Trap, a new book by Rosemary Gibson, is about the overuse of medical care — too much medicine. In this talk, Gibson tells how a woman getting a heart check-up overheard a conversation: “We’re only doing 9 bypasses a day, we need 14 a day to keep this place running.” The result of her check-up: She needed a bypass!

My encounter with too much surgery (and here). The Safe Patient Project is gathering stories of overtreatment, although it is unclear what they will do with them.

Unschooling

Seth Roberts10 Comments

Home schooling has a new name, or at least a new variety: unschooling, notable for the absence of textbooks.

When the conference [about unschooling] is over, Ms. Laricchia will return to collaborating on building an online business with her son, Michael, 13. Her daughter, Lissy, 16, is a photographer who was recently invited to participate in a show in New York. The oldest child, Joseph, has turned 18 and is no longer being actively unschooled. His mom happily admits that the change has had almost no effect on his day-to-day life.

Thanks to Anne Weiss.

reCAPTCHA and Self-Experimentation

Seth Roberts3 Comments

reCAPTCHA is the use of CAPTCHA security to read words that optical character recognition has failed to read. You see two words rather than one. The second word is the hard one. This 2008 article by its inventors (computer-science professors) says reCAPTCHA is a way that

“wasted” human processing power can be used to solve problems that computers cannot yet solve.

Self-experimentation like mine is similar. I did it in my spare (“wasted”) time. I was going to sleep anyway, I just recorded my sleep. And I found new answers to old questions, such as how to sleep better, that professional scientists had not yet found. You could say I solved problems that professional scientists aren’t yet capable of solving.

I believe that reCAPTCHA and self-experimentation like mine are two ends of what will be a power-law distribution of the use of “spare” human processing power. reCAPTCHA: many people, tiny amount of time per contribution. Self-experimentation like mine: Tiny number of people, large amount of time per contribution. Halfway (in log units) between reCAPTCHA and self-experimentation like mine is Wikipedia: middling number of people, middling time per contribution. Writing open-source software, to the extent that it’s unpaid, lies somewhere between Wikipedia and my self-experimentation.

Volunteer work is nothing new. Intellectual volunteer work is nothing new — most books are written essentially for free. What is new is cheap distribution of intellectual volunteer work. Which greatly increases the diversity of what can be done and the extent to which it can be cooperative.

Jane Jacobs and Food Trucks

Seth Roberts10 Comments

In this article about food trucks, Ed Glaeser doesn’t mention their educational value: They allow people with a new idea to test it relatively cheaply. If it works they can expand. I saw this happen in Berkeley. A food truck that sold stuffed potatoes eventually became a store that sold the same thing. Food trucks don’t merely create jobs, they can create the best kind of jobs: Those that provide new goods and services. Unlike jobs created by building dams or highways.

Any advanced economy needs a constant stream of new goods and services to replace the ones that are inevitably lost. It goes against the survival instincts of people in power (government officials) to help those at the bottom (e.g., potential food truck owners) because they seem so much less powerful than those at the top (e.g., restaurant owners) who are threatened by those at the bottom.

All this should be utterly obvious — as it is to anyone who has read Jane Jacobs on economics. But it isn’t. In science, too, every field needs a constant stream of new empirical effects (in experimental psychology, new cause-effect relationships) to replace the ones that have been studied to death. So every field needs a cheap way of searching for those effects, but no field, as far as I know, has such a way. In science, editors and reviewers are like government officials. They can discourage new ideas (food trucks) by enforcing “high standards” (regulations) whose costs they fail to understand.

Via Marginal Revolution. “[David] Westin’s biggest weakness [as head of ABC News] was that he lacked the entrepreneurial spirit to launch innovative and creative ventures.”

Dry Eye and Fish Consumption

Seth Roberts17 Comments

Let’s say that dry eye is caused by lack of omega-3. If you eat enough omega-3, you’ll never get it. Here is a recently-discovered association with tuna consumption:

Tuna consumption [1 serving was 113 g (4 oz)] was inversely associated with DES [Dry Eye Syndrome] (OR: 0.81; 95% CI: 0.66, 0.99 for 2—4 servings/wk; OR: 0.32; 95% CI: 0.13, 0.79 for 5—6 servings/wk versus =< 1/wk P for trend = 0.005).

If tuna were a good source of omega-3, eating 5-6 servings per week would completely prevent dry eye. But it doesn’t. Which supports what I have come to believe for other reasons: oily fish, in the quantities most people eat it, is a mediocre source of omega-3. Even if you eat tuna almost once/day, you don’t get enough. To get enough omega-3, look elsewhere.

Thanks to Brent Pottenger.

A Unified Theory of Japanese Food

Seth Roberts3 Comments

I used to like Japanese food because it was less fattening than other foods — I lost weight eating sushi. Now I like it because the Japanese eat so much fermented food: miso, pickles, yogurt, Yakult, umeboshi (pickled plums), natto, vinegar drinks, and alcoholic beverages. A Tokyo food court might have 20 types of pickles, 15 types of miso, and 10 types of umeboshi.

Abundance of fermented food isn’t the only way Japanese food is unusual. I see Japanese food as an outlier on three dimensions:

  • Use of fish. More fish-centered than any other major cuisine.
  • Beauty. More beautiful than any other cuisine.
  • Fermented food. More fermented food than any other cuisine.

As I’ve said, lightning doesn’t strike twice in one place for different reasons. If two rare events could have a common explanation, they probably do. I’ve discussed before why a fish-centered cuisine could lead to better visual design: Because cooks can’t use complex flavorings to show how much they care (it would make all fish taste the same), they take pains with appearance to convey this.

What about fermented foods? Here’s an idea: In the development of Japanese cooking, lack of complex flavoring of main dishes increased desire that other parts of the meal provide complexity, which is what fermented foods do so well. For example, Japanese meals often include pickles. We want a certain amount of complexity in our food, in other words. Most cuisines provide complexity via complex spice mixtures (mole sauce, harissa, curry powder); Japanese cuisine provides it with fermented foods. (I love Japanese curry, but it isn’t common.)

This explanation predicts that desire for complexity is like thirst: It grows over time and can be satisfied. Prediction 1: Eating one complex food will make a second one will taste less pleasant, just as drinking one bottle of water will make a second bottle of water taste less pleasant. Prediction 2: Over time, the pleasure provided by complexity grows. The same complex-flavored food will taste better at Time 2 than Time 1 if you haven’t eaten anything with a complex flavor between the two times.