Statistically informed ideas on how to make the world work better.

Category: Insights

The irrationality of modern child safety

Soren in his car seat

Soren in his car seat

Like most new parents, I devote a fair amount of thought to keeping my baby safe, but at the same time want to know when I can cut corners to reduce my stress level without increasing the risk too much. For example, the other day we were driving back to Quebec from Boston, trying to make it to the kennel to pick up the dog before it closed at 5:30. It was our first trip with Soren. Of course, traveling with a baby is more complicated than as a couple, and we got to a point where it was pretty close whether or not we could make it back. And then Soren pooped, and needed to feed…

Our dilemma, then, was whether Ju-hong would attempt to deal with one or both of these things while the car was moving (illegal of course) or whether we pulled over and risked all the inconvenience of not making it back in time to pick up the dog: an extra day’s charges, an extra hour of driving to get the dog the next day (with Soren in tow, of course), lost work time, feeling bad for the kennel owners, etc. In the end we stopped, and made it back a bit late but still got the dog.

Most parents would say stopping was the safe thing to do. Conventional wisdom says that letting a kid in a car without proper restraint is tantamount to murder. But most parents and conventional wisdom are wrong in this case, as were we. Read the rest of this entry »


The costs of too much choice: How the science of evolutionary development justifies Obamacare

One of the more difficult and technical fields one could choose to study is Evo-Devo, or the evolution of development. Briefly, it is the field that studies how genetic programs determine the developmental process, how these programs evolve, and how the types of programs available constrain the directions evolution can take. For example, if humans were to evolve wings (an essential impossibility for many reasons), Evo-Devo lets us make the clear inference that we would not evolve them as sprouting from our shoulders like angels, but rather as modifications of our arms. Why? Because in all tetrapods (i.e., eptiles, amphibians, birds, and mammals) there is a developmental program to produce four limbs. Limbs can be lost (snakes, whales) and modified for flight (bats, birds), but they cannot be added.

One of the key insights to emerge from Evo-Devo is that developmental programs are highly organized. They have evolved ways to facilitate future evolution, called evolvability. They achieve this using mechanisms known as gene regulatory networks, compartmentalization, and canalization. While the details of these mechanisms are beyond the scope of this post, they have in common that they are ways to facilitate long-term evolution at the cost of flexibility. That is, they standardize the developmental process to give consistent results, but limit the forms that can be arrived at. Again, tetrapod limbs are a good example: if tetrapod limbs were not the result of a fairly standardized genetic module, we would be able to evolve them anywhere any time – the nose could become a hand, we could evolve rows of wings up and down our backs, etc. However, the  result would be chaos. It would be too easy for a minor mutation to mess up development, too easy for the final form to depend too heavily on what gene combinations one has (image if parents regularly “accidentally” gave birth to children with 6 or 10 limbs, just because of  how their genes got combined…), and too hard to control the evolution of limbs as the environment changed and a specific sort of form became necessary. In other words, we gave up flexibility for stability and predictability.

How does all this relate to Obamacare?  Read the rest of this entry »

Phil Garrity on the human costs of perfect analysis

Partners-In-Health logo

The Partners-In-Health logo

I often write here about some of the hard-nosed reasons to be wary of trying to measure things that are not easily measurable – risks of bias, misaligned incentives, and missing important information that is harder to quantify. But, as a follow-up to my last piece, I came across an excellent essay by Phil Garrity, who works as a Monitoring-Evaluation-Quality (MEQ) program assistant at Partners in Health. His job is to try to measure hard-to-measure things, and he makes an excellent case for a soft-nosed risk of trying to measure everything: that we lose a bit of our humanity. This is particularly poignant if you know that he’s a young guy who’s just returning to work after a fight with bone cancer. He has given me permission to post the essay, which is below:

Read the rest of this entry »

Optimized charitable giving, evidence-based medicine, and the risk of thinking we can measure everything

GiveWell logo, taken from there website.

The GiveWell logo, taken from their website.

I read an interesting blog post this morning on Wonkblog about how some people are getting jobs on Wall Street in order to save the world: the idea is to make as much money as quickly as possible, live on next to nothing, and then use the saved money to save the world more efficiently than one could by joining the Peace Corps or becoming a doctor.

The post discussed a website/organization called GiveWell that takes a very hard-nosed, analytical approach to how we should most efficiently use our charitable dollars to do good in the world. The ballet or the symphony is nice, but by buying bed nets to prevent malaria you could be saving children’s lives for very little money, so guess which GiveWell recommends you to donate to? They choose a small number of top charities among a large number they review, and they are very careful not to make claims that the non-top charities are not useful, only that there is very good evidence that the top charities are useful. I am truly impressed with the thoughtfulness of the approach and the quality of the research they seem to have done.

But – and there’s always a but – it struck me that there is a limit to this approach to charitable giving, and it is strikingly similar to a limitation of evidence-based medicine that I’ve been bumping into recently. Read the rest of this entry »

How conservatives should have argued in court that gay marriage harms traditional marriage

korean traditional wedding



I believe strongly in gay rights: I believe that whatever rights straight people have, gay people should have too. But, as regular readers of this blog know, I try to be honest about the evidence for my positions, and I point out when people I agree with on the big picture are wrong about the details or the rationale. Such is the case for the idea that gay marriage has no impact on traditional marriage.

On its surface, the idea that the gay couple next door will destroy my heterosexual marriage by tying the knot is laughably absurd. It’s so absurd that opponents of gay marriage couldn’t even find any rationale for it when arguing it in court, first before the California state supreme court, then before the US supreme court. When pressed by the justices, the lawyers ummed and ahhed and sputtered, a rarity in court.

However, this absurdity is only when we consider individual cases – there is substantial truth to the idea that gay marriage undermines traditional marriage at the societal level. American conservatives couldn’t make this argument because they are so ruggedly individualistic that more communal arguments didn’t occur to them (even if they believed them in their heart of hearts). Here’s how it works:

Read the rest of this entry »

What crime statistics, standardized tests, and scientific researchers have in common

I had thought about calling this post “Cohen’s law for predicting distortions in incentivized systems.” Tongue-in-cheek of course – it’s approximate, and thus not really a law. And I don’t like the self-aggrandizing habit of naming a law after oneself. And it would have been a dry title, and you probably wouldn’t be reading this. Nonetheless, this post is about the single most important thing that everyone designing public policy should understand. It is about the principle that makes most public policy fail (or work less well than intended).

Most public policy is designed to achieve certain goals – lowering crime, improving education, advancing scientific knowledge, improving health care etc. And most of the time, these goals are achieved by trying to get the right people to do the right things: police to arrest criminals, teachers to teach well, researchers to perform well, doctors to treat patients well, etc. In order to encourage this, most policy incorporates some form of incentives: tax structure, salary scales, rewards for good performance, and so forth. Police departments are judged by their crime statistics, and in turn find ways to pressure their officers to deliver these stats. In US education policy, No Child Left Behind was supposed to implement standards to encourage schools and teachers to perform better. Researchers who are productive are more likely to get funded for their next research grant. And so forth.

Read the rest of this entry »

What evolution can teach us about breastfeeding and natural childbirth

IMG_0109This little guy is my son, Soren, born March 13 (and the reason I haven’t posted for a while). In the photo, hours after birth, he is hooked up to electrodes, a blood pressure cuff, and various paraphernalia of the modern medical establishment, a result of having to undergo CPR to restart his heart and breathing as he emerged from a cesarian. He is doing great now, but it is clear that modern medicine saved his life (and possibly that of my wife), probably 2-3 times over between the long labor, the cesarian, and the resuscitation. (He is doing great now.)

Like many in our generation and social class, especially here in Quebec, we had wanted to have a “natural childbirth,” i.e., to see if we could deliver without any anaesthesia or other interventions of modern medicine. Unlike many of our friends, we had not signed up with a midwife or birthing center: the public hospital here in Sherbrooke has an excellent maternity ward, and nurses are trained to help with natural or medicalized childbirth, or any combination, depending on the wishes of the parents.

Despite my preference for “natural,” I was also acutely aware that childbirth is different than other aspects of nature: in this case, natural implies very high levels of both infant and maternal mortality. In contrast to breastfeeding, where all the evidence points to breast milk being superior to any technologically developed infant formula, “natural” childbirth does not always equate to good childbirth. As is often the case, this is clearer when seen in the light of evolution:

Read the rest of this entry »

The benefits of a Mediterranean diet: thoughts on the new study

olive oil

A new study in the New England Journal of Medicine claims that a Mediterranean diet (lots of nuts, fish, olive oil, and fruits and vegetables, not too much dairy, red meat etc.) can dramatically lower cardiovascular disease events and mortality. My opinions on this study are a bit schizophrenic – it confirms what I’ve been saying for a while, but I don’t trust the methods. In the end, I think the study is largely correct, but somewhat by luck.

Read the rest of this entry »

Prediction and truth: two ways to measure scientific progress

Samuel Arbesman has a really interesting piece in Slate today about scientific progress and computers that may make discoveries that humans can’t understand. One of the issues he raises is that science sometimes overhauls what is considered as “truth,” but nonetheless we continue to make progress, despite the fact that everything we believe today may be considered false tomorrow.

Perhaps the classic example of this is Newton’s laws of motion, which were overturned by Einstein. Einstein showed that Newton’s laws were good approximations at low speeds, but that as objects approached the speed of light they broke down; Einstein’s special relativity theory proposed equations that are valid at all speeds. Einstein not only proposed better equations; those equations implied a different understanding of the universe, an understanding that allowed scientists to pursue new avenues of inquiry.

This example is famous, and shows two important principles. First, as science progresses, prediction gets better and better. In terms of predictive power, Einstein’s contribution was an incremental one, not a revolutionary one. Newton’s laws still apply nearly perfectly at most speeds experienced in daily life, though many cosmological and sub-atomic phenomena can only be predicted with Einstein’s equations.

Second, in terms of our fundamental understanding of what the universe is like, Einstein’s theory was revolutionary. But there is always the possibility that it will be completely overturned by the next revolutionary theory. In this sense, the paradigm shift of Einstein is perhaps most useful in that it opened up new avenues for research, not in that it is a better approximation of the truth itself.

A relativist or a critic of science can always claim, correctly, that the current scientific consensus risks being overturned by the next big discovery, and that therefore scientific claims to understand the truth of the universe are weak. What is harder for a relativist to critique is the improvement in predictive capacity that has been achieved over the course of science history. That should count for something, even if the “truth” of science is at risk of being overturned.

However, the discussion above is predicated on a model of science that originates largely in physics. It is striking to me how many lay discussions of the philosophy of science assume that all science is like physics, that the best scientific design is always a controlled experiment, and that Karl Popper’s idea of falsification of hypotheses is supposed to be the one true scientific method (as mentioned in the Arbesman article, to my chagrin).

In fact, there has been no major paradigm shift in biology since Darwin, despite many in physics. And unlike in physics, where it seems the next great theory might overturn everything we think, in biology it is clearer and clearer that the work that remains is to iron out the details, not find the next grand theory. Yes, evolutionary theory has been gradually refined, and we have largely rejected some spurious ideas such as widespread group selection. Yes, biology is still informed by theory, and sometimes a creative new theory can change a field. But the fields that get changed are narrower and narrower, as the high level theories become more strongly confirmed.

Likewise, controlled experiments are rare in fields such as ecology and evolution, where it is generally impossible to rewind history: most of our knowledge of these fields is based on observational data and on a gradual accumulation of evidence rather than clear yes-no, up-or-down tests of hypotheses.

We never know what tomorrow will bring, and in a formal sense no aspect of scientific theory can be considered to be 100% proven beyond all doubt. (Perhaps God is faking our data for reasons unknown to us, after all…). But it seems less and less likely that there are any impending scientific revolutions outside physics. So, while we can certainly measure scientific progress as an increase in our predictive power (regardless of the underlying truth), it is also probable that in many domains we are not too far off from an accurate description of “truth,” even if we will never know for certain.

Thoughts on science, truth, and prediction? Leave them in the comments and I’ll respond…

Public access to scientific findings: a mixed blessing



The Washington Post is reporting that the Obama administration is ordering greater public access to publicly funded research. This sounds good, but what exactly does it mean, and why wasn’t it done already? In fact, most research funded by the US government through grants to independent researchers (e.g. such as me, at universities) is already required to be made available to the public within 12 months of publication. The same is true for health research here in Canada, via the Canadian Institutes of Health Research (CIHR).

Here’s how the system works now: As a researcher, I want to publish my results. This helps other researchers (and the public) learn what I’ve done, and it helps my career. I do so by submitting an article to a peer-review journal, of which there are many. I make the choice based on how important I think my findings are (and thus how ambitious I can be in shooting for a top journal, which will be more widely read and look better on my CV). The journal may or may not accept my article, and will probably require extensive revisions. The process generally takes about 6 months from the initial submission, assuming the first journal accepts, and it requires a lot of time in terms of revising, formatting, submitting, proofreading, and so forth. This is excluding the time to prepare the original article.

Once my article is published, it is available online at the publisher’s website or in print at any library that subscribes to the journal. Most people now read online. Anyone can access the article’s abstract and some other data about it, but for most journals, only individuals affiliated with a subscribing university can access the full article. Anyone not affiliated can access for around $30, an exorbitant fee. Nearly all of the fees associated with publication are sustained by university library subscriptions. Publishers make enormous profits because they have a monopoly on each journal, and those who pay for the journals are not those who publish in them, removing the normal market forces from the system.

Some journals, however, are open-access, which means anyone anywhere can read the article online for free, but the researcher must generally pay a publication fee of $1000-2000. Most journals that are not open-access offer an option to publish articles as open-access, for a higher fee.

Since the public pays for research through government grants, the argument goes that the public should have access to the results. As a consequence, many funding agencies have started requiring that their grantees make their results public within a year of journal publication. This sounds great, and it is definitely an admirable goal, but the transition is difficult, as illustrated by my recent experience.

I had just had an article accepted in Mechanisms of Ageing and Devleopment (MAD), a good journal in my field and the best one for this particular article. It is a rather technical article unlikely to be of interest to people outside academia, but important as a base for my future research and to describe a method many others are likely to want to use. I had just become aware of CIHR’s new policy that the results had to be publicly available within 12 months, and I thought this was a good thing. So I started looking into what this meant for my upcoming publication.

MAD is owned by Elsevier, the largest academic publishing house. I started poking around their website, and I found that I could pay $3000 (from my grant money, not my pocket) to publish open access. It was not clear if I had the right to make my article public in some way without paying this. (It’s a lot of money, enough to support a summer research student who could complete a project and learn a lot, a much better use in this case than granting theoretic access to a public unlikely to read the article.) There was some information about different levels of access, “green,” “gold,” “white,” etc. It seemed that MAD was green, which meant that I was allowed to submit a non-formatted version of the article to a public archive after 12 months. However, another webpage said that Elsevier had no agreement with CIHR, which meant that I could not submit to an archive. It wasn’t clear.

So I wrote to both CIHR and Elsevier asking for clarifications. They both sent me links to various websites which were not very helpful or which I’d already seen. I finally talked by phone to someone at CIHR, who was very friendly and helpful but did not seem to know what to do. After repeated conversations with her, she suggested that I should try to negotiate the publication contract each time I published an article, an obvious non-starter since I have no leverage in such a negotiation, and since whatever journal functionaries will just want to get rid of my questions as quickly as possible.

Elsevier in fact responded in just this way. I was never able to get more out of them than an effective “just pay the $3000 and stop bothering us.” Eventually, I gave up and paid because I had little hope of finding something else out with more time, and I’d already put a lot of time into the question. But things won’t be any clearer next time around…

I strongly believe that all research findings should be publicly available, and that there should not be any private publishing houses such as Elsevier. Financing for publication should come from an international consortium of governments that agree to pay proportional to the amount their countries produce. It could even be that every researcher is required to pay open access fees for every article from grant money, and that these fees are slightly elevated in order to allow individuals without grant money to publish for free.

But we are a long way from this system. In the meantime, MAD was the journal I needed to publish my article in. Elsevier had a monopoly on MAD, and extorted $3000 from me (much more than the real publication cost, I’m sure). CIHR’s rules, implemented with good intentions but without the necessary agreements with publishers, meant that $3000 of their money (and mine) got wasted. And we are not, in practice, much closer to a true open-access world. In the meantime, I have one more expense and more administrative headaches that will slow down my scientific productivity…

One other quirk reported by the Washington Post: the Obama administration will require government researchers to publish their findings. This also sounds good, but is not workable in practice. Publishing takes an enormous amount of time and effort, and I publish perhaps one tenth of what I could or what I would like to based on time constraints. I have lots of old results sitting in folders that I don’t have time to write up. Forcing me to publish all my results (even the ones I deem less important) would slow down my productivity in generating the truly important results.

So, while well-intentioned, the new rules seem as likely to do harm as to do good….