The History of I.Q. Tests for the High-Range© 2001-2010 Paul Cooijmans
|
It began in 1994 with the Graduator. This unusual instrument consisted of an inventory of 300 guitar-related abilities (I am a composer and guitarist, and gave lessons at that time) to assess a guitarist's level. The Graduator was linked to a complex system that would convert any score profile out of 2300 possible to a guitar composition. This system existed on paper and its algorithms had to be worked out by hand.
In the course of several years thereafter I graded over a hundred guitarists, thus enabling some norming, but produced only one actual composition - that corresponds to the profile of my own 237 top score - because of the amount of work involved in working out the algorithms. I did perform this composition, called "For who loves truth the garrote called life is daily tightened a turn", once, and a very bad recording (bad in sound quality) remains, as do a practice recording and of course the score (sheet music).
Also in 1994, while working on the Graduator, I started designing intelligence tests which I called "A-1", "A-2", etcetera, "A" meaning Association test. My motivation was to make tests that could discriminate in the very high range, since I had been disappointed by the score reporting after my Mensa admission test in the summer of 1993 which stopped at the 99th percentile and did not specify further. Several "giftedness" experts I consulted had assured me it was not possible to measure intelligence beyond that, but I found that hard to believe given the obvious huge differences in comprehension I perceived between persons who all had scored at the 99th centile.
The first nine Association tests were written with fountain pen on paper and never administered, but the best items I later reused. The A-10, a Netherlandic Association test typed on paper, was the first to be administered to a number of Netherlandic Mensa members (about 8). This was early 1995. After the A-10 followed several other, up to A-23, in both Netherlandic and English, and these proof tests would later lead to the Association subtest of the Long Test For Genius, existing in Netherlandic, English, French and German. One A-test, the A-22 in Netherlandic, remained in use in its own right as an easy self-scoring test.
The association test principle was later improved by introducing a two-column or two-set approach, presenting the association problems in contrasting pairs to better lock in the solution (as in the Cooijmans Intelligence Test). The original form of the problem type also returned however, in the Genius Association Test. To my experience this is a sublime type of test, but when only verbal association problems are used, these should best be part of a comprehensive test also containing numerical and/or spatial problems to get a better indication of general intelligence. When the association problems themselves take on numerical and spatial as well as verbal forms (as in the Cooijmans Intelligence Test), they suffice as an indicator of general intelligence.
Others have also used association problems after me, unfortunately in some cases incorrectly calling them "associations". But an association problem contains multiple associations, rather than being an association.
In 1995 I was distributing the tests via ads in the Mensa International Journal, which drew a few hundred responses over the first few years, mostly from abroad. Since only few actually submitted answers and the scores were surprisingly low, I decided around June/July 1995 to design an easier and more conventional test to send to the many responders. This became the Short Test For Genius. Again, I got few submissions and scores were low, mostly 0 to 3 right out of over 40. There was however one very high score from a Netherlander in this period.
So in the Winter of 1995/1996 I went even lower and expanded the Short T.F.G., by combining it with the Association test and adding easier items to form the Long Test For Genius. Also I wrote out a contest for the Netherlandic version of it, awarding 2000 guilders - about $1300 then - to the highest scorer before the year 2000. The high scorer from the previous paragraph took this test as well. After that I kept corresponding with him and revealed the answers to the Short T.F.G. to discuss them with him; the only time I ever revealed answers.
Meanwhile in 1995 I had joined the One-in-A-THousand Society, led by Ronald K. Hoeflin, and a year later published the English version of the Long T.F.G. in its journal OATH, which brought in new submissions (the Mensa submissions from all over the world kept coming in too). Some materials from these first first few years can be seen on a section of my personal web location.
Then in 1997 I created the Final Test and its Netherlandic counterpart the Laatste Test, consisting of verbal analogies only. The English version drew relatively many submissions (a few dozen) rapidly via Hoeflin's journals, but for the Netherlandic version very few came in. In general, there is little interest in my Netherlandic tests. As the Final Test proved to be relatively easy, I made a harder sequel to it, the Test To End All Tests (Test Der Testen in Netherlandic). This was also more humorous and appeared in the Upperland Satires, a series of satirical articles by me published in OATH. Around this time two of Hoeflin's readers, Bill Bultas and Darryl Miyaguchi, began to publish my materials on their web sites, which produced extra test submissions and interest in my work. After the Test To End All Tests I would produce one more verbal analogies test, Analogies #1.
Autumn 1997 I reached a pinnacle in my creative life with The Nemesis Test, my last typewriter-on-paper work and so hard it only drew two submissions in four years. Containing various types of items, it could be seen as a successor or sequel to the Test For Genius. Around the same time I founded the Glia Society (requirement: 99.9th %ile), with me and Bill Bultas as the first members. Late 1997/early 1998 I scored and normed the Chimera High Ability Riddle Test, designed by Bultas.
Late 1997 I got an old DOS computer with WordPerfect 5.1, and began to use that for my articles, the Glia Society journal Thoth and my tests. The first test I created with that, in 1998, was the Cooijmans Intelligence Test, again a wide-range test in that it combined different item types, but much easier than the Test For Genius and Nemesis Test. The C.I.T. became relatively popular, with several dozens of submissions in the next few years. It was even published in the Finnish Mensa journal, as were the T.F.G. and Nemesis.
Around this time I also designed the unusual Creative Association Test, which turned out very unsuccessful in attracting testees - only 2 in about 4 years.
Remarkable from this period is the contact I had with Robert Jonasson, who later changed his name to Robert Lato. He was so inspired by my test "Space, Time and Hyperspace" - one of the subtests of the Long Test For Genius - that he designed a test himself following the same general idea, which was actually an original idea of mine and unique to that date: Visual problems that require one to draw the solution. This test was called Logima Strictica, later named Logima Strictica 36. He showed me the test, and had no specific plans with it at the time. I proposed to publish it in a few I.Q. society journals, and he agreed. I published it in Thoth, the journal of The Glia Society (April 1999), and in OATH, journal of the One-in-A-THousand Society (June 1999).
It must be emphasized that the appearance in Thoth was the first publication of Logima Strictica, as elsewhere it has incorrectly been stated that the first publication took place in the Swedish Mensa journal in the summer of 1999. In this initial period, submissions to Logima Strictica 36 were actually sent to me, and I sent them on to Robert Lato on a regular basis. This arrangement was practical because he was abroad a lot and could therefore not respond to mail received at his own address for long periods. He also sent me test data whenever possible so I could norm the test, and this latter arrangement has never formally ended, although for unknown reasons he has ceased to send me data. Someone else at some point began to publish alternative norms to the test, more generous than mine (and such that the score claimed by that person himself as his own made him the smartest person in the world), and I out of a sense of responsibility maintained the real norms for years thereafter as according to my data and methods my norms were more accurate (I eventually stopped publishing my norms though to avoid confusion and because the alternative norms became better for most of the range; that is, became more similar to mine).
A disturbing fact was that retests were allowed on the test, and that lists of high scorers appeared on certain web pages claiming to present the world's highest I.Q.s. However, many of those had before reported their true Logima Strictica scores to me, so that I could see, but was not allowed to say for reasons of privacy, that the original scores were in some cases markedly lower than the scores they claimed online and wherewith they were bragging to be the world's smartest people. Although I find such cults of megalomania fractionally less than praiseworthy, I in my infinite kindness remained largely silent about it for a long time.
The test and genre became quite popular, which I think is mainly because of the absence of verbal problems and many people's notion that nonverbal tests are "culture free" and therefore more fair, and also because of the fact that one-sided tests are more popular than mixed-item tests. Over the years others designed similar tests, and several societies appeared that based their admission policy on this supposedly strictly logical type of test.
Since I am the originator of the genre and posses fairly much data about it, I am entitled to give my observations as follows: This type of test gives a reasonable but not very good indication of general intelligence. It catches in spatial ability, pattern recognition, and associative horizon, but not much reasoning (the reasoning required is actually quite easy once you have seen the pattern, so that the term "strict logic" is misapplied with regard to these tests) or verbal and numerical ability. It also seems to miss conscientiousness; even very high scorers on these tests are sometimes observed behaving in uncivilized, rude, megalomanic, or antisocial ways (which is a problem seen more often with one-sided tests, and very important with regard to society admission; they let through unbalanced people). I had originally intended this test type as the spatial subtest of a larger mixed-item test, and it serves well in that function, but I do not consider it suitable as a test for intelligence it its own right.
Later I would design another test of this kind, the Lieshout International Mesospheric Intelligence Test, also in use as a subtest (of the Associative LIMIT Test).
A remarkable event with Logima Strictica is the exceptionally high top score of 32 which occurred in the early 2000s, and of which I have seen and recognized as beyond doubt authentic the score report. The test scorer however later sent that candidate a second report with a much lower score, and claimed the earlier report had been a "joke" by him (the scorer). This manipulation struck me as improper and unethical, but again I did not make much fuss about it as I sensed that would mainly offend people and not do much good. My conscience though was never entirely at ease since the candidate involved had been disadvantaged, and in my own formal and objective approach to statistics the original score of 32 is the true score, the later weak retraction being inadequate to devaluate the earlier authentic and valid report. Had the first report been an accidental mistake and just one or two points off, a rectification would have been valid, but an afterwards claim that a report was deliberately very far off is less trustworthy than the original report itself, which should therefore be considered true.
The next major thing was the Daedalus Test, summer 1999. A four-dimensional labyrinth of logic and agony dominated by a real Minotaur. I was quite confident this was the hardest test ever created, and indeed no submissions at all came in during its first few years. But to my surprise in 2001 suddenly a perfect score appeared, qualifying that person for The Giga Society, which I had founded earlier in 1996 (requirement: 99.9999999th %ile or 1 in a billion).
Also in 1999 there were problems with test items being published on the Internet without my permission, thus making fraud very easy on one of my tests, the Numbers subtest of the Long Test For Genius (some of it also appearing in the Short T.F.G.). I removed this most popular subtest from the L.T.F.G. for that reason; it can still be taken unofficially though as there are people who greatly enjoy solving difficult number series, and this test contains some of exceptional quality. It is regrettable that the fun is spoiled if one can so easily stumble upon the solutions.
These problems arose because the then President of the Prometheus Society had, without my permission or knowledge, submitted several of my test items with solutions for publication to someone who has been publishing them ever since. A letter of confession from the President regarding this crime is in my possession, however he has never taken adequate action to correct the situation or compensate for the damage in any way. It must be stressed that this behaviour not only violates copyright law, but also breaks the professional code of the President himself, who is a neuroscientist. Professionals in fields like that ought to know that test materials must be prevented from becoming unusable through wrong or unauthorized application (for instance by making them public).
In the winter of 1999-2000 I contacted the person who was illegally publishing my work and asked him to stop. He did not respond and kept publishing the problems with solutions without my permission. Dealing with this was hard for me because at the time I did not have access to the Internet myself, and because early 2000 I got serious health problems which lasted for months.
Later I found out that the test items were not only still being published, but that other names than mine were being given as author names, apparently to thus hide the copyright violation; to make it appear as if I was not the original author; to make it appear as if I had used problems in my test that had been created by others. And of course it is thinkable that somewhat similar series have been found by others independently of me. But those series only ended up in this particular publication because they appeared in my test, as proven by said letter of confession. The other author names were added after I had informed the violator that he did not have my permission to publish my work, apparently to disguise the fact that he was publishing without my permission.
The worst part is that it now appears to the public as if I have used test items created by others. This is not only copyright violation but also character assassination. In case there was ever any doubt: All of the problems in Numbers (and in the Short T.F.G.) have been created in 1995 by me, Paul Cooijmans, and me alone. Who says otherwise is committing a crime. It is sad, but there are those who possess no talent or creativity of their own and seek fulfilment in life through the destruction of the good work of others. They are like museum visitors cutting up invaluable paintings. Our contempt for them can never be deep enough.
Apart from unsupervised take-at-home tests, I designed two supervised tests in 1997 - the Giga Test - and 1999 - the Grail Test. The latter went with founding the Grail Society, requirement 99.999999999th %ile (1 in 100 billion). Both are in Netherlandic and have been taken only by 5 (Giga) and 2 (Grail) persons. The Grail Test is unusual in that it contains items not just for mental but also for various physical abilities, as meant in my article Definition of G. Later I would combine the best sections of the two tests into the IGNIT - the Individuele Gesuperviseerde Nederlandstalige Intelligentietest.
Meanwhile only about 5 submissions had come in to the Prize For Genius 2000, that would end December 31 1999. One of those was a very high score from a South-African colleague of the Netherlander who had scored high on my tests several years earlier. The year ended and I informed the winner. He indicated he would rather not have an official ceremony, and we agreed I would simply convert the prize money of 2000 guilders to his bank account, which I did February 2000. We kept in contact via letters for a few months thereafter, and then he went back to South-Africa to work there in his father's company (he had been working in the Netherlands as a trainee).
The rest of the year, health problems forced me to cut down on the I.Q. test work. During the Autumn of 2000 I slowly recovered, and the next breakthrough came January 2001 when I bought a new computer and got an Internet connection. I acquainted myself with Windows, Word, e-mail, web design and the like, and started making a web site for The Glia Society and my I.Q. tests and other work (actually a Giga Society member had made a Giga site in 1999 already).
Mid-February 2001 I got a cable Internet connection and web space and put the web site online; GliaWeb was born. From the start on I had around 80 page views a day, which would later go up to 200-300 when I managed to get listed in the Yahoo! directory. Inspired by the new medium I created the <COLT> - Cooijmans OnLine Test. Several more new tests would follow, the hardest being the Isis Test, aiming to identify the most intelligent individual ever in the universe. And, inspired by the "Bonsai Kitten" web site, I designed the Bonsai Test, a miniature test of low difficulty containing various item types.
Mid-2001 a strange thing happened; The aforementioned Netherlander who had scored high on the T.F.G. and several other tests of mine was found near a railway tunnel, his head neatly separated from his body. As his death was mysterious, I sought contact with the South-African colleague to inform him and ask if he perhaps knew any backgrounds to this. I did not know an address in South-Africa, so I wrote to the old address in the Netherlands, hoping the current occupants would know his new address. Surprisingly, it turned out the address was occupied by relatives of the high-scoring Netherlander, who told me the South-African colleague had been an imaginary identity under which the Netherlander had resubmitted the Long Test For Genius to thus win the Prize For Genius. The name of the "South-African colleague" was actually the name of a nephew of him, a young boy living at this address. The Short T.F.G. answers I had revealed to him had enabled him to score about 20 I.Q. points above his usual level, when retaking the Long T.F.G. under a false name. I of course disqualified the "South-African" submission from the contest and awarded the Prize to the runner-up, who as it happens was the high-scoring Netherlander himself with his original score.
For reasons of piety I ignored the fact that he had never registered and paid the 10 guilder registration fee for the contest with his own name and score, which (registration) was a requirement to compete for the Prize, even though I nevertheless had notified him well before the deadline that he was still allowed to register if so desired; therefore I ignored the fact that he was not entitled to the Prize at all, not to mention the fact that he had committed fraud and robbed me of 2000 guilders, which would rightfully be a reason for disqualification altogether, if not for prosecution. For clarity it should be noted that the highest scorer among those who HAD registered had solved 20 problems less than the high-scoring Netherlander on the latter's first (genuine) attempt, so could not rightly be called "highest scorer" and was not entitled to the Prize either.
I now also understood a discrepancy I had noted when I received the fraudulent submission; although the Long T.F.G. score had been in the mid-180s, the hypothetical score on the Short T.F.G., which consisted of a subset of L.T.F.G. items, was around I.Q. 200 - the highest ever on any test in history. It had even occurred to me briefly then that the submission might have come from the Netherlander under a false name, but being too good for the world as I am, I didn't give that a second thought.
Another remarkable circumstance was that, looking back, the Netherlander, with whom I had been in regular correspondence for several years, had stopped writing to me at the time of the total eclipse of 1999, which he had witnessed. I only realized this after his death. Also the South-African's style of letter-writing somehow resembled his, be it that the handwriting was different. He had told me before the eclipse that it would be an important experience for him, a turning point. In hindsight, after the eclipse he has only written to me as the South-African.
There were more things: The Netherlander, who had taken all of Ronald K. Hoeflin's tests, had told me at some point he had retested on those under his sister's name, scoring higher. And later, writing to me as the South-African, he again mentioned taking Hoeflin's tests with again higher scores. I have informed Hoeflin of these retests around 2006. And in his very last letter to me as the South-African, a few weeks before my 35th birthday in April 2000, he gave me as a "parting gift" the answers to a test by another Netherlandic Mensa member who had awarded 300 guilders to the highest scorer on his test, and suggested I would use those to win the prize...
In the second half of 2001 I created the Qoymans Multiple-Choice test, that in its free period of about a month drew 60 submissions. Never before had I received so many submissions in such a short period. The Psychometric Qrosswords of late 2001 though were less popular with only 2 submissions in its first few months.
To be continued