Pseudoscience

A few weeks before the beginning of my senior year in high school, I decided to check out James Randi's web site. For those who are unaware, Randi is a professional skeptic, so to speak; his job is to debunk paranormal myths of all sorts. He has a standing challenge against any practitioner of paranormal powers: if they can prove their powers under scientific procedures, he will give them one million dollars. Not too surprisingly, no one has won his money yet. Anyway, I was checking out his site, and I noticed that there was an IRC message board associated with his site. So, I decided to see what was happening there. I haven't left since. I really enjoy talking with the people there, especially the nuts. It's almost comic to see them try to defend ideas such as creationism, remote viewing, and the idea that HIV doesn't cause AIDS. Anyway, I've decided to put some of my best counterarguments to these fallacies here; whether you want them to use for your own arguments in a debate, or actually yourself believe these fictions and need some help understanding science, feel free to read them.

Before I go on, however, I'll start with one of the major misconceptions which overlays all of these: What is science? Science is the use of the scientific method and experimentation to make guesses about how the world works. To prevent myself from being accused of circular definitions, the scientific method is the process of formulating a hypothesis based on observations, designing and running an experiment to test the hypothesis, and modifying the hypothesis based on the results of the test. Science does not claim to be perfect. This is why scientific principles are called "theories": Because they are only what is suggested by evidence, and can never be fully proven. However, it is important for the layman to understand that, while scientific theories are not absolutes, they are very certain. Just because evolution is only a theory does not mean it is in doubt, any more than Einstein's relativity is. Science is the use of experiments to find fundamental properties of the universe.

How does one know the value of a given hypothesis? If a hypothesis predicts future results well, or explains past results better (or at least as well) than any other existing theory, then it is presumed to be the "right" theory. This does not mean it cannot be disproven; a classic example is Einstein's relativity disproving Newton's laws of gravitation. Relativity was shown to be correct because it both explained all past results to the same level of accuracy that Newton's laws of gravitation did, and made new predictions which were proven correct. This is the model of how to disprove a scientific theory: You must explain all of the current evidence, and also find new evidence.

Burden of proof comes into play, as well. A good general rule for assigning who has the burden of proof is that whoever's claim most contradicts known science. The person who has the burden of proof must supply evidence to prove their side. They cannot merely say, "You have little evidence to support your side, therefore my side is correct." You would be amazed how many people do not understand this. The person whose side correlates with known science cannot be required to do any major proof to show that the opposing claim is false; they must put forth the effort to prove their side to be true.

In these matters, it is important to understand what science is not. Science is not anecdotal evidence. Citing one example when a psychic reading appeared to work does not constitute scientific evidence. For any "hits" to count, all the "misses" must be counted. For example, imagine this experimental procedure: A remote viewer will try to guess at an object, without knowing what this object is. He will write down his guesses about what the object is. Then, the guesses are compared to the real object. If the remote viewer is able to correctly state 10 characteristics of the object being viewed, this is not alone scientific proof. For all we know, he guessed a total of 100 characteristics, and only 10 were correct. Without seeing all of the results, none of the results are valid.

However, even if the above stated experimental procedure showed that the results were statistically significant, there still is a fatal flaw. The above mentioned experiment is a single blind experiment. As such, it has little power to prove anything. One must conduct a double blind experiment. The key principle of a double blind test is that the judges are blind as to which group is the experimental group, and which is the control group.

The epitome of the double blind test is as follows: Say you have a new drug, and you want to test its effectiveness. You get a sample group, and give half the drug and half a placebo. The subjects do not know if they have drug or placebo. This is the first "blind." Then, doctors examine the sample group, and try to asses what effects the drug (or placebo) has had on the patients. These doctors *also* do not know whether their patients had the drug or the placebo. This is the second "blind." At the very end, the code is broken, and the doctor's analysis is correlated with which patients had the actual drug. This way, no one's hopes for the drugs success can affect the outcome.

A similar double blind test can be done for remote viewing. Take 10 pictures, and put them into 10 envelopes. Then, put a random 8 digit number on each envelope. Then, give the remote viewer one of the envelopes, and write down the code number somewhere where no one can see it. The remote viewer will then write down what they think the picture is. Next, all 10 envelopes and the words of the remote viewer are given to a judge of some sort. The judge does not know which envelope was chosen. (S)he will try to figure out which envelope was selected, based only on what the remote viewer wrote down. Finally, at the end, the envelope the judge selected and the envelope originally selected are compared. They should be the same 10% of the time by random chance; if the remote viewer does significantly better than that over at least 20 trials, then they have been shown to have actual powers.

Naturally, no pseudoscientist has ever done an actual double blind test on their powers (or, at least, they've done them, but not published the results...). Their arguments are composed of logical fallacies. Sometimes, they use "straw man" arguments, where they purposely misrepresent the side they are trying to oppose ("Evolution says that modern mammals evolved from modern reptiles, and there is no evidence suggesting this!"). Another common one is Appeal to Consequences of a Belief ("Evolution leads to Communism and genocides."). However, the perennial favorite is ad hominem arguments, saying that there is something wrong with the debater, so the argument must be wrong ("You're part of the conspiracy!"). For more information about logical fallacies, visit the Nizkor project.

Anyway, that's enough about science in general. If you want more information about a specific topic, then feel free to visit that topic.