Null Hypothesis

I am currently in Victoria visiting family and friends. My father in law subscribes to Scientific America, which had this article in it. You can find the original here. I enjoyed it, and instead of quoting bits and pieces, thought I would just post the whole thing.



In a 1997 episode of The Simpsons entitled “The Springfield Files” — a parody of X-Files in which Homer has an alien encounter in the woods (after imbibing 10 bottles of Red Tick Beer) — Leonard Nimoy voices the intro as he once did for his post-Spock run on the television mystery series In Search of…: “The following tale of alien encounters is true. And by true, I mean false. It’s all lies. But they’re entertaining lies, and in the end isn’t that the real truth? The answer is no.”

No cubed. The postmodernist belief in the relativism of truth, coupled to the clicker culture of mass media where attention spans are measured in New York minutes, leaves us with a bewildering array of truth claims packaged in infotainment units. It must be true — I saw it on television, at the movies, on the Internet. The Twilight Zone, The Outer Limits, That’s Incredible, The Sixth Sense, Poltergeist, Loose Change, Zeitgeist the Movie. Mysteries, magic, myths and monsters. The occult and the supernatural. Conspiracies and cabals. The face on Mars and aliens on Earth. Bigfoot and Loch Ness. ESP and PSI. UFOs and ETIs. JFK, RFK and MLK — alphabet conspiracies. Altered states and hypnotic regression. Remote viewing and astroprojection. Ouija boards and Tarot cards. Astrology and palm reading. Acupuncture and chiropractic. Repressed memories and false memories. Talking to the dead and listening to your inner child. Such claims are an obfuscating amalgam of theory and conjecture, reality and fantasy, nonfiction and science fiction. Cue dramatic music. Darken the backdrop. Cast a shaft of light across the host’s face. The truth is out there. I want to believe.

What I want to believe based on emotions and what I should believe based on evidence does not always coincide. And after 99 monthly columns of exploring such topics (this is Opus 100), I conclude that I’m a skeptic not because I do not want to believe but because I want to know. I believe that the truth is out there. But how can we tell the difference between what we would like to be true and what is actually true? The answer is science.

Science begins with the null hypothesis, which assumes that the claim under investigation is not true until demonstrated otherwise. The statistical standards of evidence needed to reject the null hypothesis are substantial. Ideally, in a controlled experiment, we would like to be 95 to 99 percent confident that the results were not caused by chance before we offer our provisional assent that the effect may be real. Failure to reject the null hypothesis does not make the claim false, and, conversely, rejecting the null hypothesis is not a warranty on truth. Nevertheless, the scientific method is the best tool ever devised to discriminate between true and false patterns, to distinguish between reality and fantasy, and to detect baloney.

The null hypothesis means that the burden of proof is on the person asserting a positive claim, not on the skeptics to disprove it. I once appeared on Larry King Live to discuss UFOs (a perennial favorite of his), along with a table full of UFOlogists. King’s questions for other skeptics and me typically miss this central tenet of science. It is not up to the skeptics to disprove UFOs. Although we cannot run a controlled experiment that would yield a statistical probability of rejecting (or not) the null hypothesis that aliens are not visiting Earth, proof would be simple: show us an alien spacecraft or an extraterrestrial body. Until then, keep searching and get back to us when you have something. Unfortunately for UFOlogists, scientists cannot accept as definitive proof of alien visitation such evidence as blurry photographs, grainy videos and anecdotes about spooky lights in the sky. Photographs and videos can be easily doctored, and lights in the sky have many prosaic explanations (aerial flares, lighted balloons, experimental aircraft, even Venus). Nor do government documents with redacted paragraphs count as evidence for ET contact, because we know that governments keep secrets for national security reasons. Terrestrial secrets do not equate to extraterrestrial cover-ups.

So many claims of this nature are based on negative evidence. That is, if science cannot explain X, then your explanation for X is necessarily true. Not so. In science, lots of mysteries are left unexplained until further evidence arises, and problems are often left unsolved until another day. I recall a mystery in cosmology in the early 1990s whereby it appeared that there were stars older than the universe itself — the daughter was older than the mother! Thinking that I might have a hot story to write about that would reveal something deeply wrong with current cosmological models, I first queried California Institute of Technology cosmologist Kip S. Thorne, who assured me that the discrepancy was merely a problem in the current estimates of the age of the universe and that it would resolve itself in time with more data and better dating techniques. It did, as so many problems in science eventually do. In the meantime, it is okay to say, “I don’t know,” “I’m not sure” and “Let’s wait and see.”

To be fair, not all claims are subject to laboratory experiments and statistical tests. Many historical and inferential sciences require nuanced analyses of data and a convergence of evidence from multiple lines of inquiry that point to an unmistakable conclusion. Just as detectives employ the convergence of evidence technique to deduce who most likely committed a crime, scientists employ the method to determine the likeliest explanation for a particular phenomenon. Cosmologists reconstruct the history of the universe by integrating data from cosmology, astronomy, astrophysics, spectroscopy, general relativity and quantum mechanics. Geologists reconstruct the history of Earth through a convergence of evidence from geology, geophysics and geochemistry. Archaeologists piece together the history of a civilization from pollen grains, kitchen middens, potshards, tools, works of art, written sources and other site-specific artifacts. Climate scientists prove anthropogenic global warming from the environmental sciences, planetary geology, geophysics, glaciology, meteorology, chemistry, biology, ecology, among other disciplines. Evolutionary biologists uncover the history of life on Earth from geology, paleontology, botany, zoology, biogeography, comparative anatomy and physiology, genetics, and so on.

Once an inferential or historical science is well established through the accumulation of positive evidence, however, it is just as sound as a laboratory or experimental science. For creationists to disprove evolution, for example, they need to unravel all these independent lines of evidence as well as construct a rival theory that can explain them better than the theory of evolution. They have not, instead employing only negative evidence in the form of “if evolutionary biologists cannot present a natural explanation of X, then a supernatural explanation of X must be true.”

The principle of positive evidence applies to all claims. Skeptics are from Missouri, the Show-Me state. Show me a Sasquatch body. Show me the archaeological artifacts from Atlantis. Show me a Ouija board that spells words with securely blindfolded participants. Show me a Nostradamus quatrain that predicted World War II or 9/11 before (not after) the fact (postdictions don’t count in science). Show me the evidence that alternative medicines work better than placebos. Show me an ET or take me to the Mothership. Show me the Intelligent Designer. Show me God. Show me, and I’ll believe.

Most people (scientists included) treat the God question separate from all these other claims. They are right to do so as long as the particular claim in question cannot — even in principle — be examined by science. But what might that include? Most religious claims are testable, such as prayer positively influencing healing. In this case, controlled experiments to date show no difference between prayed-for and not-prayed-for patients. And beyond such controlled research, why does God only seem to heal illnesses that often go away on their own? What would compel me to believe would be something unequivocal, such as if an amputee grew a new limb. Amphibians can do it. Surely an omnipotent deity could do it. Many Iraqi War vets eagerly await divine action.

There is one mystery I will concede that science may not be able to answer, and that is the question of what existed before our universe began. One answer is the multiverse. According to the theory, multiple universes each had their own genesis, and some of these universes gave birth (perhaps through collapsing black holes) to baby universes, one of which was ours. There is no positive evidence for this conjecture, but neither is there positive evidence for the traditional answer to the question — God. And in both cases, we are left with the reductio ad absurdum question of what came before the multiverse or God. If God is defined as that which does not need to be created, then why can’t the universe (or multiverse) be defined as that which does not need to be created?

In both cases, we have only negative evidence along the lines of “I can’t think of any other explanation,” which is no evidence at all. If there is one thing that the history of science has taught us, it is that it is arrogant to think we now know enough to know that we cannot know. So for the time being, it comes down to cognitive or emotional preference: an answer with only negative evidence or no answer at all. God, multiverse or Unknown. Which one you choose depends on your tolerance for ambiguity and how much you want to believe. For me, I remain in sublime awe of the great Unknown.


  1. This comment has been removed by the author.

  2. Here's some food for thought, scott. Is Shermer's statement, "but how do we tell the difference between what we would like to be true and what is actually true? The answer is science" a scientific claim?

  3. I'll comment on that. The answer would be "yes", as science is not only the methodology of determining the truth, but since it's based on empirical evidence it can also verify what other methods word, regardless of whether or not they're science.

    Suppose Ouija boards actually worked. Regardless of how they worked, we can still verify the claims - did it predict/provide answers in double-blind settings?

    'Other ways of knowing' are theoretically possible, and would have a real, verifiable, consequences. No other ways have, however, ever proved to work beyond that expected by chance or psychology.

    So, when we state "The answer is science", that answer is a scientific claim, and is itself based on empirical evidence.



  4. Correction:

    I'll comment on that. The answer would be "yes", as science is not only the methodology of determining the truth, but since it's based on empirical evidence it can also verify what other methods word, regardless of whether or not they're science.

    is actually:

    I'll comment on that. The answer would be "yes", as science is not only the methodology of determining the truth, but since it's based on empirical evidence it can also verify what other methods worK, regardless of whether or not they're science.