Thanks to Craig Fratrik and Bryan Castañeda.
A few days ago I gave a talk at Microsoft Beijing titled “The Rise of Personal Science: Discoveries about Acne, Blood Sugar, Mood, Weight Loss, Sleep, and Brain Function.” (Thanks to Richard Sprague, who invited me.) The audience was engineers.
In response to a question, I said that the relationship between personal science and professional science resembled the relationship between professional science and engineering. Cause-effect statements (X causes Y) vary in their degree of plausibility anywhere from zero (can’t possibly be true) to one (absolute certainty). Engineers, professional scientists, and personal scientists tend to work at different places along this scale:
Engineers work with cause-effect relationships at the top of the scale, that are well-established. (For example, Newton’s Laws.) Relationships in which we have total confidence.
Professional scientists like to study cause-effect relationships that are in the middle of the scale of degree of belief: true and false are equally plausible. When both true and false are plausible, you can publish the results no matter what you find. If everyone already agrees that X causes Y, further evidence isn’t publishable — too obvious. If it is highly implausible that X causes Y, professional scientists cannot study the question because a test of whether X causes Y is too unlikely to pay off. If you find that X does cause Y you can publish it but that’s too unlikely. Finding that X does not cause Y is unpublishable (“we already knew that”).
Personal scientists can easily test ideas with low plausibility. First, because personal science is cheap. Many tests cost nothing. Second, because what other people think is irrelevant. (A professional scientist who takes seriously an idea that “everyone knows is nonsense” risks loss of reputation.) Third, because there is no pressure to produce a steady stream of publications. An example of a personal scientist testing an idea with low plausibility is when I tested the idea that standing causes weight loss. I thought it was unlikely (and, indeed, I didn’t lose weight when I stood much more than usual). But I could easily test it. It led me to discover that standing a lot improves my sleep.
Plainly we need all three (engineers, professional scientists, personal scientists). Has anyone reading this heard someone besides me make this point?
I have been shocked — I sort of continue to be shocked — how much I have been able to discover via personal science. But a high rate of discovery makes sense if personal science supplies a necessary ingredient — ability to test low-plausibility ideas — that has been missing.
Recently there was a competition for Tsinghua civil engineering majors. Whose structure can support the most weight? And so on. At the end of the competition, a professor handed out prizes to the winners. After the awards ceremony, the professor who had handed out the awards said to a colleague, “I don’t like this job.” His colleague was surprised: What was so bad about handing out awards? The professor explained that the students’ names sometimes included characters so obscure that he didn’t know them. Which was embarrassing.
A new BBC series Metalworks! is about the history of British metal working. My theory of human evolution says that decoration — more precisely, our enjoyment of it — evolved because it helped the most skilled craftsmen make a living. Long ago, technology evolved via massive amounts of trial and error, which required subsidy since payoff (discovery with practical value) was so infrequent. It was much easier to discover/learn how to make something that looked better than something that worked better, but the two sorts of discoveries were correlated: trial and error produces both.
The episode on ironwork (The Blacksmith’s Tale) makes explicit how desire for decoration made it easier for the most skilled iron workers to make a living:
[Expert, at 16:50:] “I think decoration entirely depends on the amount of money the patron wanted to spend on that particular object.” [Narrator:] By the end of the 15th Century, wealthy patrons, such as the Church and monarchy, were hand-picking known craftsmen at the top of their game to match a commission’s requirements. When King Edward IV commissioned the Cornish smith John Tresillion to make these Gothic gates at Windsor in 1497, he did so with good reason. . . . [Expert:] “No blacksmith, ordinary blacksmith who was used to making horseshoes, could dream of working to this standard of perfection.”
Quality of decoration is easy to see. It doesn’t matter but it correlates with something that does matter — amount of trial and error (more trial and error, more innovation). We reward decoration to increase innovation.
The Shangri-La Diet derives from a theory of weight control that emphasizes smell-calorie learning. Smell-calorie learning evolved for the same logical reason. Smells don’t actually matter for health. But they are easy to notice and they correlate with things that do matter for health, such as calories. Via smell-calorie learning we learn the correlations. After that the foods that smell best are the ones that contain more calories.
Jim Breed has been taking large amounts of Vitamin D3 (5000-10000 IU/day) since 2008. Yet when he switched to taking it in the morning, his sleep quickly improved. Here’s what happened:
I’m a married man, 230-240 pounds over the past 4 years, born in 1957, and I work as an engineer for the Department of Energy in an office. I try to do cardio for thirty minutes four times a week.
In 2008, I began taking 5-10000 IU Vitamin D3 daily. My blood tests:
October 2009 50 ng/ml
August 2010 65 ng/ml
May 2012 84 ng/mlMy doctor said to reduce my intake from 10000 IU to 5000 IU when it hit 84, as he prefers levels under 80.
Prior to beginning the morning D, I took one 5000 IU gel cap two times during the day. I usually took my supplements at breakfast, lunch, and dinner. I might take 5000 at lunch and 5000 at dinner. Since I started taking Vitamin D in 2008, I have not missed a day of work due to illnessk. This is unusual. Throughout grade school, high school, college, and my work for the Federal Government, I have never had a complete year where I wasn’t sidelined by some cold, flu, what have you for an unscheduled absence due to sickness
In November 2011, I started reading posts on your blog about morning D affecting sleep. I tried taking my Vitamin D in the morning — upon awakening, which is usually 6:30 am. Definitely by 8:30 am. Within a week, I noticed two things:
1. I started sleeping all night without waking in the middle of the night. For years, I have had trouble sleeping throughout the night. I usually slept with earphones so I could listen to the radio around 2 am for an hour until I fell asleep again. Once I woke up, it took a lot to get back to sleep. Now, I may stir for a few minutes or get up to use the toilet, but I fall back to sleep very quickly.
2. My bed times became more consistent and earlier. I married my wife in 2005. It had been a consistent source of tension between us that she liked to go to bed earlier than me (10:30 pm for her 11:30 for me). I wasn’t tired. Midnight would have been better for me. Since the morning D, I have consistently beaten my wife to bed. I am just done for the day and ready to go to sleep by 10:30. This is truly a change for the better.
I did not expect these results. I reasoned that since my blood levels were so high, when I took D would not matter. I was surprised to be wrong.
I also found that an unexpected reduction in my dosage screwed up my sleep. This was really exciting. My wife and I traveled to Livermore, CA from our home in KS. She gave me some travel supplement packs that she makes up and it did not have our usual brand of D. Here is how the trip went.
- Friday-fly to CA 10,000 iu at waking; great sleep in hotel room
- Saturday-travel D; great sleep in hotel room
- Sunday-travel D; great sleep in hotel room
- Monday-travel D; great sleep in hotel room
- Tues-travel D; restless sleep in hotel room. I heard the A/C unit. The bed was uncomfortable. I fell asleep during a meeting with the lights down at work
- Wed-travel D; terrible sleep in hotel room. In a meeting at the lab I make a mental note to ask my wife what the travel dosage is.
It turns out that the travel dosage was 1000 IU. I thought it was just a different brand. Definitely, 1000 IU was not enough. I immediately went to Walgreens and got some 5000 IU gelcaps and began taking the 10000 IU upon waking. It took about a week to re-establish my previous restful sleep.
Reducing my dosage from 10000 IU to 5000 IU did not disturb my restful sleep patterns- When my doctor had me cut back to 5000 IU at the beginning of May, I didn’t notice any decrease in restful sleep. If anything, it seems to be better than when I was taking 10000 IU.
Thanks to Jonathan Graehl.
Thanks to Tim Beneke.
In a comment on an article in The Scientist, someone tells a story with profound implications:
I participated in 1992 NCI SWOG 9005 Phase 3 [clinical trial of] Mifepristone for recurrent meningioma. The drug put my tumor in remission when it regrew post surgery. However, other more despairing patients had already been grossly weakened by multiple brain surgeries and prior standard brain radiation therapy which had failed them before they joined the trial. They were really not as young, healthy and strong as I was when I decided to volunteer for a “state of the art” drug therapy upon my first recurrence. . . . I could not get the names of the anonymous members of the Data and Safety Monitoring committee who closed the trial as “no more effective than placebo”. I had flunked the placebo the first year and my tumor did not grow for the next three years I was allowed to take the real drug. I finally managed to get FDA approval to take the drug again in Feb 2005 and my condition has remained stable ever since according to my MRIS.
Apparently the drug did not work for most participants in the trial — leading to the conclusion “no mnore effective than placebo” — but it did work for him.
The statistical tests used to decide if a drug works are not sensitive to this sort of thing — most patients not helped, a few patients helped. (Existing tests, such as the t test, work best with normality of both groups, treatment and placebo, whereas this outcome produces non-normality of the treatment group, which reduces test sensitivity.) It is quite possible to construct analyses that would be more sensitive to this than existing tests, but this has not been done. It is quite possible to run a study that produces for each patient a p value for the null hypothesis of no effect (a number that helps you decide if that particular patient has been helped) but this too has not been done.
Since these new analyses would benefit drug companies, their absence is curious.
Shortly before Obama took office, many American banks, including the largest ones, were given a huge amount of money by the Federal government (“bailed out”). Why? Because Secretary of the Treasury Henry Paulson, Chairman of the Federal Reserve Ben Bernanke and other economists (not necessarily independent of Paulson and Bernanke) predicted a second Great Depression if they weren’t. I didn’t believe Paulson et al. — their track records of prediction were terrible. They hadn’t foreseen the crisis. Why should I think they knew how to fix it? I believed their predictions of disaster were too confident.
At the time I didn’t know this bit of history:
The blood-curdling threats [now] being issued by Eurocrats should sound familiar to British readers. We went through precisely the same experience 20 years ago, when we were stuck with an over-valued exchange rate in the Exchange Rate Mechanism.
As in Greece, our leaders – all the main parties, the CBI, the TUC, the Bank of England – assured us that leaving the ERM would be disastrous. On September 11, 1992, John Major solemnly told us that withdrawal was ‘the soft option, the inflationary option, the devaluer’s option, a betrayal of our country’s future’.
Four days later, we left the system, and our recovery began immediately. Inflation, interest rates and unemployment started falling, and we enjoyed 15 years of unbroken growth
Those who don’t know the past are doomed to over-trust experts.
Thanks to Song Chen and Edward Epstein.