[CNI News thanks Errol Bruce-Knapp of UFO UPDATES-Toronto for providing this article, which appeared in the British magazine Fortean Times, edition #83, April 1996. In the U.S., the March/April edition of UFO Magazine ran an abbreviated version, but the article's author, Peter Brookesmith, said UFO Magazine editor Vicky Cooper Ecker "missed the most important part of the argument." Here is Brookesmith's complete text from Fortean Times.]
Ever since the findings of the famous Roper Organization poll of 1991 became known, [Budd] Hopkins and his associates have touted them as an indication of the vast scale of alien abductions in the USA. In particular they have done so through a pamphlet titled "Unusual Personal Experiences," published by Bigelow in 1992, with an introduction by John Mack. The interpretation, by Hopkins, Jacobs and Ron Westrum (HJ&W), of the survey's results is 100% hogwash.
The presumption behind their reading of the figures was that a person had possibly been abducted if he or she answered Yes to five key questions and No to a "lie detector" question.
Out of 5947 respondents, Roper found precisely 18 people who fulfilled this criterion. This represents 0.3% of the sample. It does not represent 0.3% of the US population or (using HJ&W's figures) 555,000 people. This is because the margin of error in the poll is plus or minus 1.4%. Any number below that might, statistically, just as well be zero. To find out if this number is representative, you would need to question at least five times as many people as Roper did, and probably many more to overcome a law of diminishing returns.
All the Roper research tells us is that in this particular sample there were 18 people who had had five experiences that, according to HJ&W, indicate they have been abducted. It is even possible, statistically, that they are the *only* 18 such people in the USA.
It shouldn't, then, be hard to speculate why HJ&W decided that if a person gave positive answers to only four of the five questions it would show, in their words, "there is a strong possibility [that] that individual is a UFO abductee."
Not surprisingly, this relaxation of standards gave a rather more startling result. HJ&W concluded that 119 people in the sample, representing 2% of the population or 3.7 million people in the continental USA, were abductee candidates. Taking the margin of error into account (which they don't), this could actually mean that as 'few' as 1.11 million or as many as 6.29 million adult Americans may have been abducted.
The trouble is that no known system of logic will support any such conclusion at all. In the words of veteran [UK] market researcher James R. Adams: "What they are saying is, if abduction, then all these other symptoms. All these other symptoms (or some of them, even), therefore abduction. This does not follow; the logic has what is known as an 'undistributed middle.' If it is raining, the pavements are wet. But, the fact that the pavements are wet does not mean that it is raining."
In other words, HJ&W are offering a classic false syllogism, of the kind that amuses schoolboys, as "fact": "A dog has four legs; a table has four legs; therefore tables enjoy marrowbones, chase cats, and bark."
Whatever else the Roper results may tell us, they reveal absolutely nothing about the incidence or reality of abductions by aliens in the USA.
[Brookesmith further notes: Jimmy Adams is one of the most experienced and knowledgeable market researchers in the UK, and is very highly regarded within his profession. I worked with him on many research projects during the 1980s and myself had a reputation as a skilled interpreter of poll results. As with therapy, H&J have no such background; Westrum might have known better, as a sociologist. Roper are absolutely respectable; but they will of course ask whatever questions you pay them to ask. They do not interpret or, as I understand it, design questions for outside clients.]
Original file name: .CNI - Roper on Ropes? 3.28
This file was converted with TextToHTML - (c) Logic n.v.