How pseudoscience makes its case. Part 3.

This is part of my ongoing discussion on how quacks use pseudoscience to push their myths and potions on the world.  Part 1 discussed the scientific method, which allows us to objectively analyze the natural world.  Part 2 discussed the best way for us to examine the difference between science and pseudoscience.

I just read an outstanding analysis, by Steven Novella, MD, a clinical neurologist at Yale University, of how pseudoscience (those who pretend to praise the scientific method, yet do it in a way that is not actually science) and anti-science (those who repudiate science outright, or even undermine science, with subjective analysis and untestable spirituality) to reject evidence-based medicine.

Dr. Novella clearly states how science in medicine works:

This leads us to the final continuum – the consensus of expert opinion based upon systematic reviews can either result in a solid and confident unanimous opinion, a reliable opinion with serious minority objections, a genuine controversy with no objective resolution, or simply the conclusion that we currently lack sufficient evidence and do not know the answer. It can also lead, of course, to a solid consensus of expert opinion combined with a fake controversy manufactured by a group driven by ideology or greed and not science. The tobacco industry’s campaign of doubt against the conclusion that smoking is a risk factor for lung cancer is one example. The anti-vaccine movement’s fear-mongering about vaccines and autism is another.

Basically, science evolves over time.  A conclusion that lacks sufficient evidence may eventually be supported by better analysis or groundbreaking research.  You’ll notice that anti-science and pseudoscience pushers do not allow themselves to participate in the this continuum of research–they state emphatically that they are right.

Science, by its very nature, must be falsifiable, meaning that any hypothesis or theory has the logical possibility that it can be contradicted by an observation or the outcome of a physical experiment. Just because a hypothesis or theory is “falsifiable,” we do not conclude that it is false.   To the contrary, we understand that if it is false, then some observation or experiment will provide a reproducible result that is in conflict with it.  Simply put, science assumes that it has it all wrong, and attempts to determine why a particular theory or hypothesis is wrong.  Of course, in these attempts, usually more evidence is found to support the original theory.  Just because science requires falsifiability, that does not mean that it will ever be falsified, but science is open to the possibility.  In other words, science evolves.

Pseudoscience, by its very nature, is not falsifiable.  It is mostly based on assertion rather than scientific observation, so it cannot be tested by experiment or observation.  Creationism is a perfect example.  It is based on a human text (the bible), so there is no experiment that could be designed to test the text, since it non-responsive in a natural sense.  It would be like trying to scientifically show that the muppets existed.

Two of the most misused and misunderstood terms in evaluating scientific evidence are correlation and causation, two powerful analytical tools that are critical to evidence based medicine.  Correlation is the grouping of variables that may occur together.  For example, smoking correlates with lung cancer in that those who smoke tend to develop lung cancer at a statistically significant rate.  It’s important to note that correlation does not prove causation.  However, once you have numerous well-designed studies that correlate lung cancer to smoking, along with adding in biological and physiological models that support the correlation, then we can arrive at a consensus that not only is smoking correlated with lung cancer, it causes it.

We observe correlations every day.  But they are subjective observations for which we cannot state a causal relationship without substantial research.  The anti-vaccination movement is rife with these observations which they use to “prove” a cause.  An anti-vaccine conspiracy website claims that pregnant women are miscarrying babies after getting the shot.  The fact is that there is a statistical chance that women will miscarry during any pregnancy.  This is random variability not a cause.  In fact, based on the rate of miscarriage, we could expect that thousands of women would miscarry within 24 hours of getting the H1N1 flu shot.  But it’s not correlation, unless significant studies show a causal relationship.  For example, I’m also sure that thousands of people broke a bone or had a desire to eat a burger after getting the shot, but that’s because in a large enough population of individuals, you can find literally millions of different actions after getting a shot.

So, the miscarriage rate after receiving the swine flu shot is not correlated.  It’s just a random observation.  And there is no biological cause that could be described.  Nevertheless, the “flu vaccine causes miscarriage” conspiracy has been thoroughly debunked by research, but still the internet meme continues.  Pseudoscience sometimes uses the same methodology (or lack of methodology) to make positive assertions.  Homeopaths will claim that their dilutions will cure whatever disease, yet they do not have scientific evidence supporting them, but there plenty of evidence that debunks what they practice.

As part of my analysis of medical claims of causation or “cures”, I often use this logic to test the possibility of the usefulness of any alternative medicine–is there any physical, chemical or biological mechanism that will allow the quack procedure to work?  If you cannot imagine it without violating some of the basic laws of science, then we should stand by Occam’s razor, which states often times the simplest solution is the best.  So, if there is no evidence of vaccinations being correlated, let alone causal, to autism, then that remains the simplest solution.  To explain a possible tie without any evidence would require us to suspend what we know of most biological processes.

As I’ve said in other posts, the internet gives us so much information, we tend to value it equally, as if every website provides accurate and logical data points.  Maybe you have a friend who had a miscarriage 24 hours after receiving the swine flu vaccine.  Maybe you’ve heard that many people have.  But that’s not science, that’s just a subjective observation.  Or even confirmation bias.

Once again, Dr. Novella says it perfectly:

In conclusion, correlation is an extremely valuable type of scientific evidence in medicine. But first correlations must be confirmed as real, and then every possible causational relationship must be systematically explored. In the end correlation can be used as powerful evidence for a cause and effect relationship between a treatment and benefit, or a risk factor and a disease. But it is also one of the most abused types of evidence, because it is easy and even tempting to come to premature conclusions based upon the preliminary appearance of a correlation.

 

How pseudoscience makes its case. Part 2.

Recently, I discussed how science works. It’s not a belief. It’s not a random set of rules. It is a rational and logical process to determine cause and effect in the natural world. Pseudoscience, by its very nature, ignores the scientific process; instead, it claims to come to conclusions through science, usually by using scientific sounding words, but actually avoids the scientific process.  They tend to use logical fallacies to make their case.  Just to be clear, logical fallacy is essentially an error of reasoning. When a pseudoscientist  makes a claim, or attempts to persuade the public of this claim, and it is based on a bad piece of reasoning, they commit a fallacy.

There are several types of logical fallacies that they employ.  My favorites are Appeals to Antiquity, or old ideas are somehow better than new ideas; Appeals to Authority, or someone who should know better supports the claim even if everyone else does not; Appeals to Popularity, or everyone does it, so it must be useful; and the Genetic Fallacy, where the source is more matter than the merits of the evidence.  Logical fallacies are so prevalent in skeptical community, there are websites devoted to describing them.

The typical pseudoscientist will use logical fallacies to state very definitively that “it’s proven.” It’s the same whether it’s creationism (the belief that some magical being created the world some small number of years ago), alternative medicine (homeopathy, which is nothing but water, has magical properties to cure everything from cancer to male pattern baldness), or vaccine denialists (I’ve discussed this topic before, no need to belabor).  The worst problem is that in the world of the internet, if you google these beliefs, the number of websites and hits that seem to state that they are THE TRUTH overwhelm those that are more skeptical or critical.

So how can you tell the difference between science and pseudoscience in medicine? In medicine, we gather and analyze evidence in one of two ways.

Almost any medical product, device, drug or procedure must, by law, must studied in a Randomized Controlled Trial, which is sometimes called a clinical trial. Essentially, it is a scientific experiment, designed to test the hypothesis of whether the safety and efficacy of a particular medical product is better than a placebo. That is, does the medical product produce results better than doing nothing at all. This is the “gold standard” of investigation, and if the study does confirm the hypothesis, you can be assured it has a benefit to your life (although the degree may be subject to argument). Alternative medicine just doesn’t do this (most of their reasoning is that their beliefs just doesn’t fit into the clinical trial model), so their is no proof that their products work. A clinical trial usually has thousands of participants, and is done in a manner that the patient and the physician do not know who is and who is not receiving the treatment. The results are analyzed statistically and published in peer-reviewed journals. Furthermore, the results are reviewed and investigated by the FDA (and legal bodies in other countries), before a drug or device can be used by a physician. This is an expensive and time-consuming process, in which alternative medicine hardly ever participates.

Now it’s not ethical to test every medical hypothesis with a clinical trial. For example, we know that smoking is bad for your health. Yet, tobacco manufacturers love to insist that there has never been a clinical trial that makes this conclusion. The reason that is true is that it would be unethical to give one group of adults cigarettes for 20 years and another group nothing to see if one would die at a higher rate. So we use epidemiological studies to determine if we can see in a population whether a cause has an effect. We can review records of thousands of smokers to see what the effect will be. Once again, pushers of alternative medicine therapies have not published a study of all the patients who might have used their therapy and see the result. Epidemiology is a scientific process that is critical to preventative medicine–without it, we cannot know if some behavior or public health issue has a causal effect on health.

Remember, anecdotes (“my mother’s friend’s cousin’s daughter was cured by eating this leaf”) are not reasons to accept alternative medicine. Even anecdotes that try to sound like science (“90 out of 100 people think this leaf does work”) aren’t a reason to “believe” in a pseudoscience.

You might have heard that taking lots of Vitamin C helps prevent colds. It doesn’t. And that conclusions is supported by large clinical trials, so unless you are afflicted by scurvy, there’s no reason to take large doses of the vitamin. And that’s the difference between real science and pseudoscience.

How pseudoscience makes its case. Part 1.

I decided to write a three-part article here that partially describes how they make their case, not necessarily why humans accept it so easily.  I’m not a psychiatrist, and I certainly don’t play one on TV.  I thought we should start with the scientific method, or how real science works.

I always get suspicious when someone makes an argument with the statement of “it’s been proven to work”, “the link is proven”, or, alternatively, they state some negative about scientifically supported therapies. Typically, I hear these kinds of statements from the pseudoscience pushing crowd. For example, real science has debunked the “there is a proven link between vaccines and autism,” a common and popular pseudoscientific belief.  Or that most alternative medicine (CAM) therapies work based on numerous logical fallacies that suspends reason, and accepts “belief” in the therapy, something that evidence-based medicine just doesn’t do.

In fact, science rarely uses the term “proven”, because the scientific method is not a system to make a definitive answer on any question–scientists always leave open the possibility of an alternative hypothesis that can be tested. If the alternate hypothesis can be supported through experimentation, then it can replace the original one. When an alternative medicine or junk science supporter states “it has been proven,” ask where is the evidence.  What is more troubling is that someone who believes in these therapies cannot imagine that they don’t work, what is called falsification, which is a hallmark of good science.  Whenever I hear that a scientist say, “we were wrong, it doesn’t work,” my retort is “excellent, good science.”

The scientific method is an unbiased systematic approach to answer questions about the natural world, including medicine. It has several basic steps:

  1. Define the question–this could be anything from “does this compound have an effect on this disease?” or “how does this disease progress?”
  2. Observations–this is the subjective part of science. Do we observe trends or anomalies? Does a physician notice that every patient from a town or neighborhood exhibit the same disease? A lot of science arises from observations of the natural world. One of the most famous stories in the early history of medicine is when Edward Jenner observed that milkmaids rarely were infected by smallpox because they were exposed to cowpox, a less virulent disease.
  3. Hypothesis–taking the observations, create a hypothesis that can be tested. In Jenner’s case, he hypothesized that exposure to cowpox immunized individuals to small pox.
  4. Experiment–simply, the scientist then tests the hypothesis with experiments and collect the data. The experiments are not designed to solely validate the hypothesis but may also attempt to contradict it.
  5. Analyze–this requires statistics to determine the significance or results.
  6. Interpret–sometimes the data leads to a revision of the hypothesis, which means the scientist has to return to steps 3-6. Or it confirms or supports the hypothesis, which means the researcher can move to Step 7.
  7. Publish results–in today’s scientific community, the results require peer-review, which subjects the data, analysis and interpretation to the scrutiny of other scientists before publication. This is a critical step that ensures that the results can stand up to criticism. It does not prove anything, but it does support the hypothesis.
  8. Retesting–Many times the research is repeated by others, or the hypothesis may be slightly revised with additional data. Science is not static, it constantly revises theories as more data is gathered. For this reason alone, science is not an absolute, it is constantly seeking new data.

Science is an evidence-based systematic analysis without inherent opinion or emotion. In other words, it is a method to cut through opinions and anecdotal observations, so that one can have some reasonable expectation that a medicine or device will work as planned. CAM fails to utilize scientific method. Supporters of CAM usually perform experiments to confirm the hypothesis, never to contradict it. It is the fundamental principle of falsifiability, that is, that if a hypothesis is false, it can be shown in experimentation that allows science to have an open mind about the world. When you speak to a believer of CAM, they almost never assume that their treatment cannot work.

It’s interesting that CAM and pseudoscience start out with observations of the real world. For example, CAM therapies sometimes work, not because of the therapies themselves, but because humans just get better from many diseases. So, these CAM therapies rely upon confirmation bias, that is, the tendency to accept information that supports your beliefs, or even post hoc ergo propter hoc, a logical fallacy which says “since that event followed this one, that event must have been caused by this one.”  Humans too often conflate correlation and causation.  Just because events follow one another, that doesn’t mean one causes the other.  I suppose that’s how superstitions arise.

Part 2 of this discussion will be out as soon as I write it. It will discuss how to tell what is “proven” or what is science.  Stay tuned.

Quality of Wikipedia

Just a brief housekeeping note.  Many of the links in this blog will point to an associated Wikipedia article.  I, in fact, read and edit those articles, so I won’t make a link to an article that doesn’t make my point.  It is supposed to be the epitome of a democratic encyclopedia, one that anyone can edit.  It is free to read, and it has become the number one hit for almost any key subject, whether popular or obscure.  Other than Obama’s own websites, the Wikipedia article about him is the first google hit.

I have mixed feelings about Wikipedia.  There has been significant criticism of the project, which itself reduces the confidence in its quality and its usefulness.  My concerns about these articles rest in a few key areas:

  • Quality of writing.  Some of the writing is just plain bad. Some of it is a result of poor writing skills from English-speaking and non-English speaking contributors.  But some of it is also from a waffling or trivial writing style.  Worse yet, the prose is often not very engaging.  Sometimes, it’s so difficult to read it, that one is apt to dismiss the article as being useless, even if it has useful knowledge.
  • Neutrality.  Wikipedia has a vaunted policy called the Neutral Point of View, often abbreviated as NPOV.  At its core, it states that an article should be written to represent all significant views that have been published by reliable sources.  Two major problems arise from this policy.  First, since it is very subjective, editors literally beat each other over the head with this policy.  Second, and more importantly, NPOV drives articles to their worst.  Academic articles usually state a point of view.  You cannot read a book about the Holocaust without reading the point of view of the author.  A medical article nearly always presents a point of view, albeit one supported by scientific research.  Worse yet, editors will claim that a neutral article should include fringe theories.  I edit medical and science articles anonymously, and I find myself either editing out strange, unscientific ramblings from these articles.
  • Vandalism.  Because articles are almost always available to be edited by anyone, there is a certain level of vandalism that occurs, everything from “Joe was here” to profanity to much more subtle, insidious vandalism that is difficult to identify.  More often than Wikipedia is willing to admit, this subtle vandalism remains in the article, often leading to inaccurate or incorrect information.  This edit to the article on Martin Luther King, Jr remained for over 4 hours, despite the overt racism written into the article.  There are numerous cases of these type of problems.
  • Lack of expertise.  Because most editors are anonymous (a small subset of editors use their real names), it is nearly impossible to know the quality of edits from specific editors.  I believe that smart people can write about anything.  But, lacking an education in neurophysiology, how do I know if the article on dementia is of high or low quality?  Whenever I read an article, I check the references.  A high number of references means that usually the writing is supported by academic research.  Wikipedia will need to fix this issue.
  • Anti-intellectualism.  As a corollary to the lack of expertise is the pervasive contempt amongst Wikipedia editors against expert editors.  There is a class of editors called “admins” who, more or less, act as the police, judge and jury for the project.  They have the ability to pass summary judgement upon any editor.  As a class, they are a part of the anti-intellectual core of the project.  There are a few admins with expertise on science, historical, medical and other areas, but they generally keep a low profile.

So why do I link to Wikipedia articles?  They may not be the best I can find on the internet, but these articles are, in general, fairly useful and provide an adequate background for a topic.  When I find better sources for particular information, I use it.

I leave my readers with this warning about Wikipedia:  read the articles with a critical eye as you do anything else you read.  The problem is critical reading is not a skill I see very often these days.

Welcome to my world

A Skeptical Raptor’s native environment is the jungles of the internet, where junk science, pseudoscience, myths, logical fallacies, and outright lies survive unchecked. The Raptor has evolved over several million years to hunt down these anti-science prey, scaring them away from the average reader. Remember, a Raptor is missing some table manners, so the prey may not be treated very nicely.

OK, let me set aside the metaphors.  As you can see in my about me page, my background has been in the sciences, medicine and business. But the great thing about a strong science background is it teaches you critical thinking skills and the scientific method.  The scientific method isn’t mixing oxygen and hydrogen to make water, but it is the logical progression from observation to hypothesis to data to analysis to publication to review.  But science is not static, it is self critical, constantly reviewing itself, improving, discarding, or just supporting its theories.  What you’ll find is that the anti-science thinking is not self critical, because it considers improvement some sort of weakness.

I’m going to get this out of the way upfront.  I am a supporter of Big Pharma and the medical products industry in general.  Do I think they do no wrong?  No I don’t, I think that too often decisions are made based on business realities rather than medical ones.  However, despite some of the appeals to conspiracy about which I constantly read, most individuals in the industry are devoted to making human life better.  It is their only goal.  And despite some of the claims of the anti-science crowd, Big Pharma has saved many many many more lives than it has harmed.  Vaccines would be the #1 piece of evidence of that.  Polio, pertussis, measles, rubella, and many other diseases are no longer (well, not until recently, thanks to another anti-science group) a part of our cultural memory because of Big Pharma.

But I’ll talk about these issues over time.  I like writing for humor and critique, not for tremendous scientific analysis worthy of a Nobel Prize.  There are lots of bloggers, all of whom I respect beyond anything, who write about these topics in depth.  I will link to them, in case my skin-deep analysis annoys you.

So here goes.  Let’s see if I can do this.