Bad Data

Steven Dutch, Natural and Applied Sciences, Universityof Wisconsin - Green Bay
First-time Visitors: Please visit Site Map and Disclaimer. Use"Back" to return here.

The best defense against being taken in by the pseudoscientist is an awareness oflogical fallacies and a strict determination not to fall for emotional appeals. Inaddition to the logical fallacies, however, pseudoscientists also sooner or later alwaysend up using bad data, and a non-specialist may be hard put to tell good data from bad.Some of the data found in pseudoscience, and some examples, include:

Obsolete Data

The "Ghost Literature" of Pseudoscience

Many of the wonders science supposedly "cannot explain" actually containnothing to explain. The "facts" are based on obsolete data, hoaxes, coincidence,and the like. In fact, it is useful to have a broad overview of pseudoscience becausepseudoscientists commonly cite one another's theories. Erich von Daniken, for example,cited a theory about the capture of the earth's moon that was a popular cult before WorldWar II, and cited numerical coincidences involving the Pyramids that originated in a 19thcentury Pyramid cult. Creationists cite criticisms of radiometric dating that wereoriginally put forth by Velikovsky supporters. These "facts" look veryimpressive to the perplexed non-scientist.

Some items of bad data have become dogma in pseudoscience. The "mysteriousdisappearance of Flight 19", a group of Navy planes which disappeared off Florida in1945, is a case in point. Complete recordings of the radio traffic between the base andthe planes were transcribed in the official investigation into the disaster, and showclearly that the planes headed out into the Atlantic in the mistaken belief that they werewest of Florida. Recall that radar was in its infancy in those days. John Wallace Spencer,in Limbo of the Lost, actually quotes a flier as saying "Can't tell whether overAtlantic or Gulf". Yet this incident has become a stock item in the Bermuda Triangleliterature. Another stock item are the frozen mammoths of Siberia, supposedly flash-frozeninstantly with plants in their stomach that now grow far to the south. One gets thegeneral impression of bananas and papayas, but actually the vegetation in the mammoths'stomachs consisted of pine and spruce twigs and tundra flowers -- normal Siberianvegetation. The best-known specimen was buried when a riverbank caved in on him, breakinghis shoulder and hip bones. Somehow these inconvenient details never make it into thepopular literature, despite the fact that the frozen mammoths, like Flight 19, are derigeur in any pseudoscience best-seller. In effect, the literature of pseudoscienceis a sort of parallel literature to that of science.

Outdated Science

Another very common approach to data might be called the "lobster trap"approach. Lobster traps are built so that once the lobster gets in, he cannot get out.Pseudoscientists assume that once an idea is published in the scientific literature it isforevermore legitimate evidence, regardless of its original validity or any subsequentfindings -- after all, it's "science", isn't it? If the information is refutedor modified later, that just proves once again how fickle science is, so that the finaljudgment comes down to a subjective one. The experts disagree, so the individual has todecide for himself which alternative to accept. Relativism strikes again.

For example, as the theory of continental drift won widespread acceptance in the late1960's, a small number of conservative geologists wrote papers arguing against the theory.That was their right and duty as scientists, but most of the arguments they employed wereoutright balderdash at the time and unlike wine they have not improved with age. To anyoneeven remotely familiar with the data, these arguments were transparently specious andfilled with glaring fallacies. Nevertheless, scientific creationists continue to citethese papers in an attempt to support their own version of earth history with"scientific" evidence. This is a game science cannot win. If scientists revisetheir ideas, the very fact that they do is taken not as evidence of improvements inknowledge, but as evidence of how changeable and uncertain the results of science are. Ifscience remains firm in a belief, that's evidence not that the idea is correct but thatscientists are too pig-headed and dogmatic to change.

Random Events Collected Into Spurious Patterns

In1932, one of  parapsychology'sstar performers, Hubert Pearce, guessed 25 consecutive ESP cards correctly. Thechances of getting a run of 25 psychic research cards (the standard deck hasfive each of five different symbols) is one in five to the twenty-fifth power,or one in 295,023,223,876,953,125. Believers in parapsychology insist theprobability is so extreme that this incident is proof of a real phenomenon.Skeptics are struck by the inability to repeat these results under controlledconditions and suspect either that the cards were not entirely random (poorlyshuffled) or that the subject somehow figured out what the cards were by seeingthem reflected in the table top or the experimenter's glasses, or by cuing in onfaint nicks, marks, and smudges on the cards. This process need not involveconscious dishonesty. Let's say, however, that the experiment took place exactlyas described, without any information reaching the subject by ordinary means,and that the subject actually did overcome enormous odds. What have we proven?Only that an extremely rare event took place. We haven't proven that ESP existsbecause we are no closer to knowing the mechanism by which the event took placethan we ever were. The event could have been chance, or perhaps some othereffect completely unrelated to ESP was involved. If we had some theoreticalbasis to go on we might accept an extremely rare event as evidence because wewould have some solid basis for believing the event had a real cause, but in theabsence of a workable theory and withoutthe ability to reproduce the event so that we can test different hypotheses, theevent just sits in splendid isolationand tells us nothing.

Eventhe most avowedbelievers in paranormal phenomena agree that apparent runsof successful guesses come and go erratically. When a possiblyreal phenomenon is buried amid a great deal of noise, thereis a real possibility that the patterns we perceive may be nothing more than theproducts of our own imaginations. Thevery existence of the word "coincidence" is a powerful clue to howprone we can be to see spurious patterns in random data.

Real Phenomena Misinterpreted As Anomaly

Revisionist Interpretations of Well-understood Phenomena

Very often, things science "can't explain" turn out to have perfectlystraightforward explanations. Despite the claims of Erich von Daniken and many others,there is no mystery at all how the Pyramids were constructed. Museums have thestonecutting tools and rollers that Von Daniken and other Pyramid cultists claimhave never been found. Contrary to the claims ofcreationists, intermediate forms between most major groups of related organisms do exist.We do know what happened to Flight 19, how frozen mammoths got frozen, and so on.

Psychological Phenomena

Pseudoscientists tend to take some kinds of evidence with near-absolute literalness:legends of various kinds, eyewitness accounts of UFO's or mysterious creatures, personalaccounts of "after death" experiences, astral travel, clairvoyance or telepathy,or hypnotic accounts of reincarnation. The general rule governing all this sort ofevidence is that it is highly personal and subjective, impossible to evaluate or verifyrigorously, not subject to controlled experiment, and based on personal experience ratherthan objective physical evidence. There is probably a close relationship between thepreference for this sort of evidence and the widespread feeling that science andtechnology have mechanized and dehumanized humanity. Like the humanities' support forVelikovsky, there seems to be a desire to see subjective human experiences make acontribution to the physical sciences and thereby "prove" their usefulness andscientific validity.

When I first began researching pseudoscience, I was struck by how often apparently sanepeople swore they had seen something really outlandish. I wondered how often dootherwise normal people hallucinate? The answer, it turns out, is a lot more oftenthan one might think. In fact there's a whole range of waking and semi-wakinghallucinatory phenomena. Hypnogogic and hypnopompic dreams are dreamsthat occur while a person is dozing off or waking up, and therefore mix waking sensationswith dreams. A lot of the alleged instances in which people see angels standing at thebedside are probably of this sort. I once talked with an otherwise very rational personwho told of waking up one night and seeing a UFO land near his house; it's very likelythis was also an example of a waking dream. I've had a couple of these myself. Generally I'll be in bed (that's where you usually are when going to sleep or waking up) and will "remember" odd things. The "memories" have a harsh, discordant feel to them, and sometimes I'll even wonder "did I dream that or did it really happen?" Or sometimes I'll think the memory is so odd it must have been a dream.

Some people have such vivid fantasy lives that they actually experience some of thesensations of the fantasy. This phenomenon is called confabulation. (If I knewhow to induce this state deliberately, I could become obscenely wealthy, but I don't.) Afterreading an account of the phenomenon in the journal Skeptical Enquirer, oneperson who frequently experienced confabulation wrote in to describe his enormous reliefat not being alone in having these episodes. He described that he had evolved a variety oftechniques for telling which stimuli were real and which merely imaginary. It's easy toimagine someone unaware of having the condition, or who experienced it very rarely,believing that a fantasy was the real thing.

Second-hand Data: Urban Legends and Anecdotal Data

The guru of urban legends is Jan Harold Brunvand, a folklorist at theUniversity of Utah, who has published numerous books on the subject. I can do nobetter than recommend that you read them all. Alligators in the sewer? The kidwho dries off the family cat in the microwave? The girl in the 1960's with thebeehive hairdo who found critters living in it? The petting couple who areterrified by a radio report of a hook-handed serial killer and drive off, onlyto find a hook hanging in the door when they get home? They're all there. Andnot a one is true. Read Brunvand's books! You'll cringe with embarrassment athow many of these hoary myths you've passed along yourself.

The Internet has made it possible to proliferate these myths even faster. Afew prominent recent examples include:

All of these might have happened at some time or other. One of thehallmarks of urban legends is that they could be true. But in almostevery case, the people who pass these legends along have no direct knowledge ofthem. They always come from a friend of a friend, and efforts to track the storyto its original source never seem to succeed.

Anecdotal Evidence is a single, possibly true instance used to justifya generalization. Everybody who refuses to wear a seat belt has a story of howtheir Aunt Gertrude was miraculously hurled out of her car into a pile of rosepetals just before the car went off a bridge, burst into flames, and sank inquicksand. They simply choose to ignore the far larger number of cases whereAunt Gertrude has to be picked up with a squeegee. To be legitimate evidence,anecdotal evidence has to be true (often it's not) and representative.The anecdote has to illustrate something that happens as a rule, not as anexception.

"The millionaire who pays no taxes" is another popular piece ofanecdotal data, especially with political liberals. According to the 1999 StatisticalAbstract of the United States (Table 559), the 111,000 taxpayers whoreported income of over $1,000,000 dollars in 1996 (the average was about $2.8million) paid an average of $875,000, or 31 per cent. Taxpayers with $50-75,000income paid an average of $7300 or 12 per cent. Taxpayers reporting the averageincome of $35,000 paid $3400 or 9.7 per cent. So millionaires earn on theaverage about 81 times as much as the average taxpayer but pay 257 timesas much tax. Some millionaires do indeed pay no tax because they can offsetincome against business losses or are paid from tax-exempt bonds, but theanecdote, though true, is not representative.

"Gee-Whiz" Facts are facts that look impressive but have nosubstance. A prime example is the claim "suicide is the second leadingcause of death among teen-agers." Without in any way making light of thisproblem, think for a second. What can kill a teen-ager? They've already survivedchildhood diseases and are not prone to diseases of aging yet, so only threethings can be significant causes of death among teen-agers: accident, suicide,and homicide. Anything other than that order indicates a real problem, butsuicide will always be a leading cause of death among teen-agers.

Another "gee-whiz" fact that pops up during our periodic binges ofpanic over missing children is "over a million children are reportedmissing every year." Again referring to the 1999 Statistical Abstract ofthe United States, about a third of the population, or 77 million people,are 19 or younger. If a million kids vanish every year, in 19 years, 19 millionwould vanish. One child in four would disappear before reaching adulthood. Ithink we'd notice that. Yes indeed, a million children are reportedmissing every year but the vast majority are found within a few hours.

Deliberate Fraud

Parapsychology has seen rampant fraud, with leading researchers repeatedly caughtfudging experiments and fabricating data. Evenreputable paranormal researchers can and have resorted tofakery. An embarrassing incident took place in 1974, when Walter Levy, Jr., thendirector of the Institute for Parapsychology in Durham, N.C., was found to havefaked experimental results. The experimental design Levy was working on wassuperb. A rat in a cage had an electrode planted in the pleasure center of itsbrain, so that the animal would get a pleasurable stimulus when a current wasapplied to the wire. Animals so wired will do almost anything to keep gettingsuch stimuli, even ignoring food and sex. Levy's rat got a stimulus every time adetector detected the decay of a radioactive atom. If the animal had the abilitythrough paranormal means to increase the number of stimuli the evidence forparanormal powers would be very strong. This is exactly the sort of evidence itwould take to demonstrate paranormal effects: the rat is not biased and cannotcheat, and there is no known process that can influence the decay of radioactivematerials. The results of such a discovery, to put it mildly, would befar-reaching. 

Workers notedthat Levy, a well-known researcher in parapsychology, was unusually attentive tothe apparatus, and finally sawhim disconnecting a wire to the recorder that registered the results of theexperiment. Each disconnection registered a false positive result and thereforemade the experimental results look more favorable to Levy's theory. Levy, whohad a promising career in parapsychology, resigned. 

Levy was sincerely convincedof the reality of his phenomenon (as most scientific frauds are) but desperateto get results. Uri Geller, who rose to worldwide fame in the 1970's for hisalleged psychic abilities, was an outright cynical fraud. PhysicistsHaroldPuthoff and Russell Targ of the Stanford Research Institute (not connected withStanford University) were very much impressed by the feats of Geller, who couldsupposedly bend metal through psychic power. Several other prominent physicistsinvolved in paranormal research were highly impressed by Geller. Unfortunately,Geller was seen cheating by professional magicians who were better equipped tospot his sleight-of-hand methods than the physicists, and his stunts weresuccessfully copied, so Geller has since fallen from favor. But not before he did a huge amount of damage. He sued hisexposers and lost. They countersued,charging that Geller's lawsuit was frivolous and malicious. They won, but Gellertook refuge in his native Israel and never paid up. Still doubt we need radicaltort reform?

Justhow easy it can be to con believers in parapsychology is shown by a hilariousincident during the heyday of Uri Geller. Magician John Randi, who was thenlittle known in paranormal circles, showed up at the offices of Psychic News inLondon. He introduced himself as James Zwinge (his real name) and for two hourskept the office in an uproar by bending metal, changing the times on clocks, andduplicating almost all the celebrated feats of psychics using perfectly standardstage techniques. Psychic News ran the story on the front page, complete withclaims of "constant surveillance" and assertions that metal had bentwhen "all could vouch Zwinge had not been near it."

Parapsychologistsare insulted by the emphasis on fraud in their field, but the plain fact is thefield has been so compromised that the only way to redeem it is to start overfrom scratch, junking everything done up till now, including the assumptionthat paranormal phenomena exist at all, and start over with a totallyclean slate.

Genuine New Discoveries

One of the heroes of pseudoscience is Ignatz Semmelweiss, who began usingantisepsis in the 1850's, a decade before it was generally accepted in medicine.Semmelweiss was driven out of medicine by his fellow doctors. Whatpseudoscientists don't tell you is that Semmelweiss was in some ways a crank. Hehad noticed that doctors who went directly from performing autopsies todelivering babies had an appalling record of post-partum deaths. When heinsisted that doctors scrub with a caustic solution, his ward soon had the bestsurvival rates. But Semmelweiss theorized that it was the smell of deathclinging to the doctor that caused infection, something that must have seemed athrowback to the Middle Ages. Nevertheless, here we have a real case of anapparently crackpot idea turning out to have a real basis in fact. 


Robert Gentry and Radiation Halos?

Return to Pseudoscience Index
Return to Professor Dutch's Home Page

Created 3 February 1998, Last Update 15 January, 2020

Not an official UW Green Bay site