Wednesday, December 29, 2010

In the holiday spirit

Lest Spittle-flecked Ire become entirely cantankerous, I'd like to point out one of the best articles I've ever read in the NYT Science section: A Scientist, His Work and a Climate Reckoning.

Lest Spittle-flecked Ire lose it edge, I'd like to point out my only complaint is that when the article states
Later chemical tests, by Dr. Keeling and others, proved that the increase was due to the combustion of fossil fuels.
it does not mention which chemical tests. CO2 produced by burning fossil fuels will have much less Carbon-14 as it comes from sources much older than 6000 years, so the atmospheric ratio of Carbon-14 to Carbon-12 will steadily decrease while [CO2] increases.

However, one must account for a countervailing production of Carbon-14 during the nuclear test era. And there is some neat stuff about relative Carbon-13 depletion involving evolving chemical processes in plants that I just learned about a few minutes ago.

It does help me see why Creationists and the "global warming is a hoax" crowd are natural allies.

Saturday, December 4, 2010

My Rationally Held View of Humans

I borrowed this quote from here.
It appears, therefore, that a swarm's scout bees do something sharply different from what humans do to reach a full agreement in a debate.  Both bees and humans need a group's members to avoid stubbornly supporting their first view, but whereas we humans will usually (and sensibly) give up on a position only after we have learned of a better one, the bees will stop supporting a position automatically.
I do not think that is true at all. Humans rarely give up on their positions*, and are frequently selectively skeptical or easily convinced by data that either confronts or confirms their stubbornly supported first view. 

This is a good short survey pointing to a couple studies of how humans really behave that includes a great quote on this subject by Bertrand Russell:

If a man is offered a fact which goes against his instincts, he will scrutinize it closely, and unless the evidence is overwhelming, he will refuse to believe it. If, on the other hand, he is offered something which affords a reason for acting in accordance to his instincts, he will accept it even on the slightest evidence.
My own favorite quote on this is from Max Planck:
A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.

You can just feel the ire.

* Positions, such as this human's view of how humans behave regarding their positions.

Yes, This Again.

The impetus behind this blog was a study that written about in the NYT about an AIDS vaccine. The study was completely inconclusive, yet was promoted as a major advance. Here we go again:
In the study, 2,499 men in six countries — Brazil, Ecuador, Peru, South Africa, Thailand and the United States — were randomly assigned to take either Truvada or a placebo and were followed for up to three years. For ethical reasons, they were also given condoms, treatment for venereal diseases and advice on safe sex. There were 64 infections in the placebo group and 36 in the group that took Truvada, a 44 percent risk reduction.
Two in the Truvada group turned out to have been infected before the study began. When the remaining 34 were tested, only 3 had any drug in their blood — suggesting that the other 31 had not taken their pills.
Different regimens, like taking the pills not daily but only when sex is anticipated, also need testing.
Also, many men in the study failed to take all of their pills, and some clearly lied about it. For example, some who claimed to take them 50 percent or 90 percent of the time had little or no drug in their bloodstreams.
The pills caused no major side effects, though men who began to show signs of liver problems were taken off them quickly. Some men stopped taking the pills because they disliked relatively minor side effects like nausea and headaches. Also, some stopped bothering once they suspected that they might be taking a placebo.
“People have their own reasons,” [Policy Director of amfAR Chris] Collins said. “People don’t take their Lipitor every day either.” 
Again, how hard is it to take sqrt(1250)? But then there's the feckless methodology! Truvada is also apparently effective regardless of the regimen. Imagine that. So is prayer.

I was wondering what is going on with these AIDS studies that keep showing comparably inconclusive results, yet promoting them as breakthroughs. So I went back and looked at the previous study. Then I found this released a month later calling into question that study. And then I realized: all three of these articles were written by the same author, Donald McNeil, and both studies were under funded via the National Institutes of Health infectious diseases director Dr. Anthony Fauci. But this was especially rich:
Putting several biostatistical analyses in a news release “would have confused everybody,” Dr. Fauci said, and suggesting that the researchers were engaging in a cover-up is “absurd.”
Everybody? Statistical analyses are not confusing if they are conclusive; if only one method out of several confirms your hypothesis, your data is not conclusive.

Wednesday, December 1, 2010

Glucose and Aggression ?

The major discovery is that there is a journal called Aggressive Behavior.

Science Daily has reported on a recent work that "Sweetened blood cools hot tempers." The claim is that people who drink lemonade sweetened with sugar show less aggression towards strangers within a short time window after the drink. The experiment is exquisite :

In the study, 62 college students fasted for three hours to reduce glucose instability. They were told they were going to participate in a taste-test study, and then have their reaction times evaluated in a computerized test against an opponent.

Half of the participants were given lemonade sweetened with sugar, while the others were given lemonade with a sugar substitute.

After waiting eight minutes to allow the glucose to be absorbed in their bloodstream, the participants took part in the reaction test.

The reaction test has been used and verified in other studies as a way to measure aggression. Participants were told they and an unseen partner would press a button as fast as possible in 25 trials, and whoever was slower would receive a blast of white noise through their headphones.

At the beginning of each trial, participants set the level of noise their partner would receive if they were slower. The noise was rated on a scale of 1 to 10 -- from 60 decibels to 105 decibels (about the same volume as a smoke alarm).

In actuality, each participant won 12 of the 25 trials (randomly determined).

Aggression was measured by the noise intensity participants chose on the first trial -- before they were provoked by their partner.

Results showed that participants who drank the lemonade sweetened with sugar behaved less aggressively than those who drank lemonade with a sugar substitute. Those who drank the sugar-sweetened beverage chose a noise level averaging 4.8 out of 10, while those with the sugar substitute averaged 6.06.

I can not dispute that a low blood sugar contributes to irritiability (and by extension, aggression.) Just last night, I waited too long between lunch and dinner and was becoming irritated with my surroundings until I was able to eat. But, I have just had a glass of extra-sweetened lemonade and I am still annoyed by the presentation of the study results. The comparison of the means of two populations is meaningless without an error bar on the mean. The numbers of members of each population is only 31, so the error on the mean may be significant relative to the difference of 1.26. Error bars please ! I don't want to pay to see the PDF of this article, so I am not sure if the fault lies with the authors or Science Daily. Did the authors measure their blood sugar? The article's abstract states :

Self-control consumes a lot of glucose in the brain, suggesting that low glucose and poor glucose metabolism are linked to aggression and violence.

Knowing this, it is imperative to know how well each participant metabolizes glucose. Probably more important would be to know the number of men and women in each population. The abstract goes on :

Study 1 found that participants who consumed a glucose beverage behaved less aggressively than did participants who consumed a placebo beverage. Study 2 found an indirect relationship between diabetes (a disorder marked by low glucose levels and poor glucose metabolism) and aggressiveness through low self-control. Study 3 found that states with high diabetes rates also had high violent crime rates. Study 4 found that countries with high rates of glucose-6-phosphate dehydrogenase deficiency (a metabolic disorder related to low glucose levels) also had higher killings rates, both war related and non-war related. All four studies suggest that a spoonful of sugar helps aggressive and violent behaviors go down.

Regarding Study 3 : I wonder what the relative levels of poverty are in these states.

Monday, November 22, 2010

Regress = Progress

In the generally annoying vein of journalists writing about science comes this. All of the items Mr. Horgan, apparent journalism school graduate, cites as regress are signs of progress. I think this points to a huge gulf between science and journalism. In journalism, realizing your proposed narrative is wrong means that you're back to where you started; in science, realizing your proposed narrative is wrong means you've made a giant leap forward in understanding.

Just think: how often do we see articles about, say, city government where the author has as its main point that  there is no narrative?

The key point is that frequently in science knowing more means we realize we know less.

I actually see one of Horgan's observations:
What I found fascinating was the issue's overall tone of caution rather than the traditional boosterish enthusiasm.
as progress in itself. "Boosterish enthusiasm" is, in my view, bad for scientific progress.

On to the examples ...

The end of infectious disease
We used to think this was a) possible, and b) if possible, economically feasible. We've come to realize that evolution can be more powerful than our technology.
Space colonization
This will probably still happen soon on the scale of the lifetime of the solar system with our currently main sequence star, so don't fret, Trekkies. But if you can tell me what the point of the ISS is beyond a giant subsidy to Boeing to maintain some of our space infrastructure, I will give you a thousand dollars. I have never seen a peer-reviewed paper come out of the ISS; nearly all of the science is done by unmanned probes. 
But the real problem I have is that colonization is not science. It being nearly Thanksgiving, a seasonal example: the Mayflower was not a scientific voyage. 
Supersonic transport
The science behind supersonic flight happened before 1947. The fact that it is not economical to send hundreds of people at a time to business meetings and/or tropical islands faster than the speed of sound, but is economical to send one person and thousands of pounds of bombs is not a scientific fact.
Commercial fusion power
This turned out to be much harder than we thought; I myself was discouraged by my advisors from pursuing this field back in the 1990s. However missing a 20 year guess at commercial viability 30 years ago does not mean the field is going backward. It is still moving very, very slowly forward. Additionally, arguments against fusion's commercial viability are themselves due to progress in the field. We understand the problem better, and therefore understand that it is not likely to be economical anytime soon.
The origin of life
Keep up with the research buddy. Most of the problems with Urey-Miller type experiments are due to the fact that we have no idea what the chemical composition of the Earth's atmosphere was at the time. This is another case where we know more now and realize we know less.

Monday, November 1, 2010

Truth in data

Where did this image come from? I first noted it on Science Daily. Ignore for the moment the bombast accompanied by most any Science Daily post. Well, wait don't ignore that :
  • Science Daily : "Brilliant X-ray Explosion."
  • MAXI : "faint X-ray transient."
Hmmm. Which is it then?
Perhaps the Science Daily author was completely blown away by the accompanying image. After all, in 5 days, this object in the center of the field-of-view became significantly brighter. A close look at the two images shows that all objects in this field-of-view are brighter on Oct. 17. Why were these data sets not normalized to each other? From just these images, I think it is obvious that the relative brightness of the central object is greater on Oct. 17th. I can only guess that the two different normalizations were included to produce a more dramatic image. Unfortunately, this drama comes at the price of honesty.
There is much effort to bring science results to non-scientists. This effort is a cross between public outreach (i.e. "please keep letting the NSF, NIH spend your tax dollars") and public education (i.e. "this here galaxy is 12 billion years old. [12billion much greater than 5000]".) Some of these efforts are official, and others are part of a greater media.
The short time-scales of today's mass media are unsuited to science reporting. Science Daily is "Updated several times a day with breaking news...". I doubt it. Science is slow and painstaking work. To give the public the impression that there is such a thing as "breaking news" in science is potentially damaging to research. To generate "breaking news", a researcher may be tempted or coerced to scale the false color in an x-ray image at the behest of an editor or PI. I wish Science Daily included a minimum of technical detail in their articles and captions. A lay reader may learn a little more than what someone said. And, the editors may have to think a little more critically about the content they are including.

Friday, October 1, 2010

On separating religion and science

I've long tried to put my finger on exactly why I dislike philosophy. One reason is a certain amount of rectally disseminated speech on the subjects of how science works or the meanings of particular theories (especially quantum mechanics). I was reading this piece by atheist/philosopher Tim Crane and realized philosophers tend to do this exact same thing in fields outside the natural sciences (here, political science and history).

I am wondering if this falls under a higher heuristic that motivates philosophers I will call the preeminence of human thought. Included under this heuristic are claims that humans engage in rational thought divorced from evolutionary pressures (biological or cultural) and a lack of knowledge of -- or desire to perform -- empirical studies.

Crane states:
Religions do make factual and historical claims, and if these claims are false, then the religions fail.

There is no evidence for this. In fact, from political science, there is evidence against it. People have this weird tendency to discount empirical data from their political opinions. I know this from a great deal of firsthand experience talking to my dad. We feel like we're thinking about our political opinions, but facts never seem to trump political party affiliation. And there is no reason why this behavior should not extend to religious beliefs.

Crane's actual intent behind that statement was to differentiate religious claims from scientific claims: the former do not fail (in the eyes of believers) under rational scrutiny because they contain mysteries. However, switch the argument from religious to political claims, and this sounds ridiculous. Political positions do not contain mysteries in the religious sense of the word, yet political positions unsupported by evidence also do not fail (again, in the eyes of believers) under rational scrutiny.

This is where we come to history. Crane narrowly focuses on the Christian concept of mystery, and how it is a fundamental aspect of religion that allows it to "avoid" the rigors of rationalism almost as a side effect. Actually, mysterium fidei arose as a counter to the reintroduction of logic and the decline of Christian mysticism (to which mystery originally referred, not in our modern sense of unexplainable) around the time of the Christian conquest of Arab Spain. Arab translations of ancient Greek texts on logic and science began to be reintroduced to "dark ages" Europe. Peter Abelard was one of the key figures that introduced logic to the Roman Catholic church in the 12th century -- coinciding with things like the explanation that the priest may have transformed wine into the blood of Christ in the sacrament of the Eucharist but it still tastes like wine (which only happened because that was around the time they started getting the masses involved in the Eucharist). The mystery of the Trinity actually originates in the split of the Roman and Eastern Orthodox churches. The (almost) modern mass standardizes much of the mystery in the church in the late 16th century, right at the birth of modern science (the council of Trent was primarily concerned with Protestantism, but was generally concerned with questioning church dogma). The modern mysteries in Christianity follow from these and later scientific developments -- the seven days/6000 years didn't become a problem until geology got its start in the late 18th century.

Mysterium fidei is a protective response to rationalism by Christianity, not a fundamental aspect of religion. In fact, the concept of religion itself, as opposed to a collection of undifferentiated beliefs, may just be a modern Christian concept.

I personally tend to subscribe to that last view in some sense. There is just human thought, composed of heuristics and methods of de-conflicting those heuristics. The heuristics are things like the method of steepest descent, the Earth is 6000 years old, or What is Buddha? Three pounds of flax. De-conflicting methods range from ignoring particular data to giving up a heuristic to keeping both in unanalyzed conflict. These heuristics then signal particular group loyalties or abilities. Science and religion are not exactly separable concepts, but part of a continuum of belief systems. Science tends to put more emphasis on signalling curiosity, open-mindedness, and a particular form of group loyalty (e.g. citing papers) as well as its almost total dedication to de-conflicting all heuristics dealing with concepts that can be tested empirically. Scientists particularly feel like they are thinking when they do this. Fundamentalist Christians in the US signal group loyalty by "not believing in evolution". I don't mean that they are not believing it on purpose (implying a certain kind of rationalism), or they are somehow dumb (implying a certain kind of "irrationalism"). I mean that they feel like they are thinking as much as scientists do, but are signalling different abilities and group loyalties. Human thought is not preeminent, but subordinate to mating and resource strategies. Inasmuch as we are convincing do we succeed, so there is some tendency for an accurate representation of reality to eke out an existence.

And that includes this.

Monday, September 6, 2010

Hoisted from the archives: Enough with the mystery of the wave-particle duality.

Ok, so I was reading this, and wondering things to myself like -- is religion like the belly-button of our consciousness, i.e. a conduit from our ancient limbic system to our higher planning and social interaction noodly bits our brain, and the remainder of the pathway by which we bootstrapped to sentience -- but then came to this:

Just as physicists cannot fully understand the electron as either a particle or a wave, but only as both at once, says [Andrew] Newberg, so we need both science and a more subjective, spiritual understanding in order to grasp the full nature of reality.

Wave-particle duality? Yes we do have to describe electrons as both in particular experiments, but we can fully understand the electron (to 10 or so decimal places) as a single thing: an irreducible (Dirac spinor) representation of the Lorentz symmetry group. Sort of like how a cube is a (scalar) representation of the group of 90 degree rotations about the x, y, and z axes, and a sphere is a representation of the group of all rotations, an electron is a particular representation of the various 'rotations' of the Lorentz group (which gives us the weird stuff in Einstein's special relativity).

So what Newberg is saying is that there is some kind of unified quantum field theory of brain function out there waiting to be discovered. This is what dualities have meant in the past. The duality between electricity and magnetism gave us Maxwell's electromagnetism. It's not mysterious anymore ... 150 years later. The wave-particle duality lead to Quantum Electrodynamics only 20 or so years later. T-duality may be one of the links that eventually lead to the string theory description of everything.

You are welcome to continue to be weirded out by the quantum mechanics of 80 years ago if you'd like, but the uncertainty principle is nothing more mysterious than why you need measure something resonating for a longer time to more precisely determine its tone (or that something that resonates for a short time does not have a precise tone). Or if you wiggle a jumprope between two people, that you can figure out the wavelength if you make a standing wave, and figure out where the pulse is if you make a pulse, but not vice versa and never both.

There are things that are actually mysterious in quantum mechanics (like entanglement), but let's bring on the new weird dualities in Physics! The AdS/CFT correspondence! The holographic principle! We may all just be projections on a screen ... and I can't show you why with a jumprope.

Thursday, August 12, 2010

I Am Not an Economist. But.

Greg Mankiw considers Casey Mulligan's summer employment effect to be a challenge for Keynesian economists to explain. Now, I am not an economist; however, I think this represents a lapse in numeracy, not economic theory

Let's discuss orders of magnitude. Mulligan describes a summer employment effect -- employment among the 16-19 year old cohort rises relative to the 20-24 year old cohort from May to June -- representing an increase in the supply of labor accompanying an increase in the demand for that labor that purportedly contradicts the Keynesian view that aggregate demand in a recession is controlling unemployment. Here is the graph:

How big is this bump? Going back to the data, we get 4.1 million 16-19 year olds employed vs 12 million 20-24 year olds in May of 2010, for a ratio of 0.34. For June 2010, these numbers at 4.9, 12.6 and 0.39 respectively. So about ~ 800,000 kids become newly employed. At $10/hour for 500 hours (for the summer, both of which are probable overestimates), that represents $4 billion dollars of economic activity.

The US economy is about $15 trillion dollars. That $4 billion dollars represents about 0.03% (no, really, 0.0003) of the US economy, and about 0.5% (no, really, 0.005) of the ARRA stimulus.

There is no theory in economics that predicts anything to 1%. Maybe they can get you the number of pennies you'd get for change of a dollar. Maybe.

This is a bit like a physicist saying that the Lamb shift represents a serious challenge to theoretical physics even before the development of quantum mechanics.

I think this another aspect of a problem with economics that Matthew Yglesias describes as "doing work that superficially resembles physics and therefore counts as science-like". I wish they would pay more attention to how physicists work, not just how our interminable papers look. Yes, there appears to be an effect, but it is at most a 1% effect. Supply and demand appears to work at the order of magnitude level. Now show me Real Business Cycle theory or new Keynesian economics or whatever getting stuff right to 10%, then we can talk about this.

And though I am not an economist, I have a feeling that this represents 16-19 year olds getting jobs at the mall, Denny's and the movie theater so that they can have money to go to the mall, Denny's and the movie theater. It seems like it could occur largely independent of aggregate demand in the economy.

Monday, July 12, 2010

arXiv blog (no amateurs)

While I don't have much to say about the paper itself, it is again time to appreciate people who comment on the arXiv blog. The post in question:

The comment in question:
Wrong domain for graph
If you wanted to prove that there's a 27 Million year peak, wouldn't it make sense to present the frequency domain data instead of the time domain data?  If there were a peak every 27 Million years, then it would be clear from a frequency domain graph.
And the evidence bsdimp did not read the paper itself?

Sunday, June 27, 2010

Is It More Efficient to Replace Science with Public Opinion or Public Opinion with Science?

Recently, there was an article in the NYT about the declining belief in global warming, and/or its negative impacts in Europe.

I would like to suggest that all scientific problems be solved with polls. We simply ask a representative sample whether Yang-Mills theory exists and has a mass gap, whether the Riemann Hypothesis is true, or the best course of action regarding a quantized theory of gravity: is it a) Strings, b) Loops, c) Triangles, d) Entropy, e) I Don't Know, or f) I've Never Studied Gravity, But I'd Like to Say It's (a) Based on No Information Whatsoever. This is far easier than doing science.

[As an aside, society as a whole pays quite of bit of money to have scientists around in order to have opinions on scientific topics. I've never quite understood the propensity for people to then dilute their paid expert opinion with their own opinion. It's like going to a restaurant and insisting on cooking your meal yourself. The division of labor has given us so much!]

Of course, the problem with polling is always systematic bias, based on your wording, or in fact, in the case of the polls cited in the NYT article mentioned above, the outside temperature. Unfortunately, the belief that global warming is happening/bad/imminent is strongly correlated with the actual temperature.

In the article, first poll was taken in November 2009 after an above average temperature summer and the second poll was taken in March of 2010 after an unusually cold winter. That led to a drop in belief that global warming was happening and man-made from 41% to 26%.

And the German poll (42% feared global warming) was also taken in March and compared to a poll (62%) taken in one of the summer of one of the warmest years on record there.

It is entirely possible this NYT article may have been deliberately written to be deceptive. I found this.

Now is a partisan site, so I'll have to do more digging. But it does explicitly call out Rosenthal (the author) and anyone who cites Viscount Monckton is pretty much a nutcase (no offense to nutcases).

Here's a new global warming poll showing increasing belief in the US as we go from January to June of this year.

The key to understanding the public's opinion of global warming would probably involve some kind of seasonal average, taking into account the average temperature in the location polled. Only then could you discover what the people really want. Better yet, bypass the polling of human beings and just measure the temperature and extrapolate public opinion.

Then we could move into other areas. Opinion of the president or the effectiveness of or trust in government is in fact strongly dependent on economic factors. We could account for this. Of course, if we did, we would probably find that every president's approval rating would likely be exactly the same.

And as the global temperature will likely continue to rise, eventually more people will believe it. The seasonal effect will also slowly disappear, because, well, one of the effects of CO2 heating the planet is that it retains heat in the winter months (as well as at night, and near the poles). So by the time global warming has started to wreak its havoc, we should have a well informed public regarding the havoc that is being wreaked by global warming.

The arXiv Blog

I love the arXiv blog comments. I kind of want to start a new blog called: arXiv blog (no amateurs). Mostly for the fact that it would actually more likely draw people such as those we see here.

And regardless of whether you think special or general relativity (or quantum mechanics) is a Platonic ideal, or that some other theory of your own devising is, you are limited by the fact that, because of experimental evidence, your theory must reduce in some limit to those theories -- just like how quantum mechanics and general relativity reduce to Newtonian classical mechanics and Newton's gravitational force law and special relativity to Galilean invariance.

Monday, May 24, 2010

Martin Gardner (1914-2010)

Martin Gardner was a great inspiration to me as a kid. Aside from being my favorite column in old issues of Scientific American, his review of Gödel, Escher, Bach (in the one from July 1979 my dad gave me because it had instructions on how to build a seismometer for a science project) came back to me while I was in college rummaging through a used book bin.

As a side note: that same issue of SA had an article on the MIT bag model that I would encounter again in grad school. Plus excellent articles on the Burgess Shale and the Oklo natural reactor that came back recently as a constraint on the possibility of a time varying fine structure constant that was allowed in string theory.

Sunday, May 23, 2010

Cognitive Dissonance

The same people who think the blue line on the left is a bad representation of the gray line on the left think the red lines on the right are a good representation of the gray line on the right.

Left: Hansen et al 1988 Scenario B temperature model (blue) and actual temperature data (HadCRUT3, gray).

Right: 2007 GAO/OMB model (red) and actual budget deficits (gray).


First, the title. It comes from a discussion of the attempt to get an injunction to stop the LHC from starting.
[Law Professor Eric] Johnson quotes the British physicist Brian Cox who is reported to have said: "Anyone who thinks the LHC will destroy the world is a twat." 
That is not an encouraging sign. 
Johnson is well aware that this case may never come to court (although he points out that one like it that raises the same issues may well come about in the future).
So the real test will be how the particle physics community responds, whether with spittle-flecked ire or reasoned argument.
My fellow physicist and friend from grad school had sent this to me, and it sort of stuck as a catch-all subject line for various transgressions against science we saw out on the web.
Second, the concept. A few months earlier, he had sent me a link to the results of a study of an AIDS vaccine in the NYT, accompanied by what is best described as spittle-flecked ire. The quote from the article:
Col. Jerome H. Kim, a physician who is manager of the army’s H.I.V. vaccine program, said half the 16,402 volunteers were given six doses of two vaccines in 2006 and half were given placebos. They then got regular tests for the AIDS virus for three years. Of those who got placebos, 74 became infected, while only 51 of those who got the vaccines did.

"... how hard is it to take sqrt(8000)?" Exactly. If we take half the sample population N = 8201 in one of the groups, sqrt(N) ~ 91 is a quick measure of the size of the expected fluctuation in a measurement on that population due to random chance alone. And 91 is bigger than the difference between the number who became infected, which means categorizing this as a successful trial is dubious.

After some spittle-flecked ire from both of us on the subject, I had mentioned my long running desire to start a blog pointing out these samples of innumeracy or flaws in the application of the scientific method on the web. Not that this is a novel idea. It's more for fun. And my friend may join up at some point.

One of my personal favorites is this graph. It is the worst graph ever made. It is a good example of one of the ways you get into trouble in an investigation: you have a hypothesis you know is right and you try to look for a way to prove it. Economics has started to work that way lately ... especially in the Freakonomics vein: they look for "natural experiments" to demonstrate a hypothesis they have. Another way to put this is to have a result and then look for data that supports it. Sounds like the opposite of science to me.

Anyway, this is the sort of thing that will pop up here. I also like to make political cartoons out of graphs. So those will appear, too. Probably next post.