Conversation with Paul Cooijmans on Community Dynamics, Heterogeneous and Homogeneous Tests, and Qualification: Administrator, Glia Society (3)
Author(s): Scott Douglas Jacobsen
Publication (Outlet/Website): In-Sight: Independent Interview-Based Journal
Publication Date (yyyy/mm/dd): 2021/01/08
Abstract
Paul Cooijmans is an Independent Psychometitor and Administrator of the Glia Society, and Administrator of the Giga Society. He discusses: the benefit of the “members-only web site and e-mail forum”; meetings; the birth years of the Glia Society membership; dynamics of the community; Gliaweb Riddled Intelligence Test; membership “in other societies… not accepted as proof of intelligence”; real issues with membership; retests not accepted; differences between heterogeneous and homogeneous tests; the requirement for a coinciding score with “one of the Verbal, Spatial, or Numerical tests”; Reason – Revision 2008 and the Daedalus Test; “Spatial tests”; “Numerical tests”; “Verbal tests”; both unsupervised and supervised tests for membership in the Glia Society; the Ultra Test, the Mega Test, and the Titan Test, or the SAT, ACT, GRE, and Army GCT, score acceptance deadlines.
Keywords: community, Glia Society, heterogeneous, homogeneous, Paul Cooijmans, qualifications, tests.
Conversation with Paul Cooijmans on Community Dynamics, Heterogeneous and Homogeneous Tests, and Qualification: Administrator, Glia Society (3)
*Please see the footnotes, bibliography, and citation style listing after the interview.*
Scott Douglas Jacobsen: To continue with the formal information as presented on the Glia Society website, let’s continue with the general information and the qualification information, what is the benefit of the “members-only web site and e-mail forum”? (Cooijmans, n.d.a)
Paul Cooijmans[1],[2]*: The members-only journal can be found on that web location, as well as information on how to join a few online fora or groups for members. On those fora, one may communicate with other members. I must say there has been a lot of activity over the past few years, but before that, there have also been periods with little to no activity.
Jacobsen:
Cooijmans: Good heavens, that is a profound question. For once, I am dumbstruck. I fear that the remainder of my life may not suffice to formulate its answer.
Jacobsen: What are the forms of the “meetings organized by small groups of members” happening “very occasionally”? (Cooijmans, n.d.a)
Cooijmans: There has been a small meeting at the society’s tenth anniversary in Brussels, where I held a lecture and played guitar, and we went to an eating place. I have also heard there has been a meeting in the United States, but I was not present so I can not say how that went. Then, I remember attending two other meetings in Belgium where Glia Society members were present, but those were not specifically Glia Society meetings. And on several occasions, members have visited me at home; one of those visits concerned two members at once.
Jacobsen: What explains the birth years of the Glia Society membership ranging from approximately the 1920s to the 2000s with the bulk between 1950 and 1990? Is there a more precise range for an unusual hunk of the membership?
Cooijmans: The explanation for this bizarre phenomenon is that people tend to be born in different years rather than all at once, hence the range of birth years. There is not so much an unusual hunk, but the mode of birth years is shifting upward and is now in the 1970s. Before very long, this will be the 1980s. Also, members remain on the member list as long as their membership has not formally ended, so there may be deceased people on the list, or people who have forgotten that they are members. Astoundingly few are so conscientious as to inform the Administrator, “Okay, I am dead now, you can take me off the list”.
It has occurred to me to deactivate the membership of people who have not been active for a number of years, but logistically, that is not as easy as it seems. When I experimented with this a few times in the past, the members in question tended to suddenly appear out of nothing and say, “Hey, why am I not on the member list any more?!” Somehow, they know. There must be underhand contact between members that you can not see as an Administrator, a “grapevine” so to speak.
Jacobsen: Reflecting on the membership demographics, and the growth rate variation of the membership noted in the first interview on the Glia Society (Jacobsen, 2020), did this change the internal conversation dynamics of the community of the Glia Society?
Cooijmans: Yes, in particular I have noted that the younger generation, say people in their late teens to late twenties, has become more active than in the past, and in a positive way. When I started joining I.Q. societies, the bulk of the active members were older, mostly over 30 and often middle-aged. I was 28 myself then and one of the youngest. I suspect this shift has to do with the circumstance that these people, born in the 1990s and later, have grown up with the Internet and with Internet-based communities. It is native to them, and they are also better informed regarding topics like I.Q. testing, intelligence, statistics, and psychology than older generations tended to be when first getting in contact with I.Q. societies. Although, as everyone knows, my enormous modesty forbids me to blow my own trumpet, it might just be that the free availability and hitherto unseen quality of the sublime information on my web locations is playing a role in the education of the best-informed generations ever that are emerging now and in the decades to come.
Jacobsen: For those with an interest in exploring the website, independently, on their own course of discovery about the Glia Society, they can view the links here:
General information on the Glia Society
Thoth, the Glia Society journal
Text of lecture given at the society’s tenth anniversary in 2007
Contact the Administrator (Cooijmans, n.d.b)
For the qualifications of the Glia Society, there exist a substantial number of tests. To pre-empt questions in some prospective members’ minds, you answered some of the inevitable questions, even in precise terms. To quote you, “Required for membership is the 99.9th adult population centile on any of the accepted heterogeneous tests (that is, tests with a mixture of item types) or on each of two accepted homogeneous (one-sided) tests with different item types. Only first attempts are accepted, not retests (that is, one attempt per test is allowed). Only the tests listed are accepted. Membership in other societies is not accepted as proof of intelligence” (Cooijmans, n.d.c). For a peripheral, but important, side question, why is “the Gliaweb Riddled Intelligence Test (Revision 2011 or original version)” (Ibid.) no longer accepted?
Cooijmans: That test has never been accepted, it has always been intended as an easy and less serious test, not a high-range test. You would have to score near the top of the test’s range to qualify, where tests are less accurate.
Jacobsen: Why is membership “in other societies… not accepted as proof of intelligence”? (Ibid.)
Cooijmans: Because many other societies are not strict in their admission policies and let in people who are far below the advertised level. I have explained that in earlier answers.
Jacobsen: Have real issues arisen where membership is shown, and assumed by the shower, as proof of intelligence, so as to make one qualify for membership? Any particular narrative examples of difficult personalities being highly aggressive about the matter? No need for names or a name, merely an illustration.
Cooijmans: I do not remember any such issues. My general impression is that people who try to join with proof of membership in other societies (despite the qualification page stating that this is not possible) are mostly not very intelligent. And the more memberships they show, the less intelligent they are. Speaking of unintelligent ways to “prove” intelligence, people have also tried to join with screen shots of automatically scored online tests that can be taken indefinitely and do not even display one’s name; and the most hilarious “proof” was an A.C.T. score report whereon all of the identifying information, including the name of the candidate (!) had been blacked out. A score report to bearer, so to speak.
Jacobsen: Also, from the previous question, why are retests not accepted? Famously, this happened with the Mega Society and the Mega Test with, at least, two individuals utilizing pseudonyms, or fake names, and then taking the test twice, so as to claim a higher score – so a purportedly higher intelligence score as measured by the Mega Test. While, with the practice of no retests, presumably real names only, individuals with a sincere and honest attempt and effort can acquire an accurate, i.e., real, assessment of their general intelligence, inasmuch as alternative tests compare to the mainstream intelligence test scores. Also, should real names and first attempts only become a pervasive admissions policy of high-I.Q. and higher-I.Q. societies? If so, why? If not, why not?
Cooijmans: I explain the objections against retests on my web location, but because this is so important, and so poorly understood by many incompetent dilettante test scorers, I will repeat it here in amended form:
(1) Retests are not comparable to first attempts but somewhat higher on average (that is why some candidates want them!) so accepting retests as the candidate’s true score (as incompetent dilettantes do) implies that the first score is not the true score, and therefore means to oblige all candidates to take the test twice in order to know their true score, and to require them to destroy their first score report (or not issue it at all). After all, the first score is sometimes higher than the retest, so if you let them keep it they will use it for admission, self-promotion and the like!
(2) Considering the retest score to be the true score implies that only the retest scores can be used for statistical purposes such as norming, and the first-attempt scores are useless statistically; it means to throw away the biggest part of the work one is doing, of the data one is gathering. In practice, of course, those who allow retests do use the first-attempt scores for statistical purposes, sometimes even in combination with the retests (and third attempts if not more) to arrive at a larger sample size, thus corrupting their statistics.
(3) In practice, candidates and test scorers involved in retests use the highest of the two scores, rather than the actual retest score (which should be used in all cases even when it is lower) and therefore add to the above mentioned two problems the inflation of scores caused by having “two chances”, as well as the levelling between candidates resulting from the same (inflation and levelling, when using the highest of two scores, are the necessary result of the imperfect test-retest correlation; and this correlation is imperfect or there would be no point in retesting to begin with). In case it is not at once apparent why using the highest of two scores causes inflation and levelling, one may imagine that “the highest of two” is on average higher than “always the retest”, because the retest score is sometimes lower than the first score. “Levelling” means reduction of variance.
(4) Candidates with a perfect or near-perfect score on first attempt are excluded from knowing their true score this way as there is no or too little room for their retest score to be higher than their first score.
(5) Through retests, candidates can verify the value (score) of particular answers or answer sets (more or less like in the game “Mastermind”) which endangers the secrecy of the test’s answers. From two scored submissions, very much more information can be derived than from one.
If a retest is allowed in some rare case for a special reason, the score report must mention it concerns a retest, to prevent it from being used for admission or for statistical purposes as if it were a first attempt. If the retest report does not mention it concerns a retest, this makes it impossible to distinguish it from a first attempt, and therefore reports from test scorers who fail to identify retests on their reports can never be trusted or accepted for admission purposes.
Regarding real names, of course members should only be admitted under real names, otherwise they can never be held to account for anything, and any test scores under false names have no validity. With false names accepted, people could try tests indefinitely to figure out the intended answers until they hit a qualifying score. And that is exactly what they do if you let them. I saw this happen in the early 2000s and was shocked by people’s dishonesty.
And first attempts should indeed be the only ones accepted for membership, I think the above reasons make that clear.
Jacobsen: Can you expand on the aforementioned differences between heterogeneous and homogeneous tests, please?
Cooijmans: Heterogeneous tests contain at least two different item types, when item types are classified on face value as verbal, numerical, spatial, or (sometimes) logical. Homogeneous tests contain only one item type. My experience says that the former yield a better indication of general intelligence and are less affected by fraud. Homogeneous tests are insufficient as tests of “g” and are also the preferred target of frauds and high-score chasers. To say that a test is an insufficient indicator of “g” means, in practice, that a candidate’s score on that test may deviate unacceptably much from one’s true level in general intelligence, either in the upward or in the downward direction. It is the upward error that attracts frauds and megalomaniacs.
For further clarification, it must be understood that an I.Q. test measures a general factor (= that part of its variance that is shared by all mental tests), one or more group factors (= the variance shared by some but not all tests), and specificity (= the variance that is unique to the test in question). Homogeneous tests catch in relatively too little of the first and too much of the last two, it seems. This has proven to be hard to understand for many, probably because of the abstract nature of concepts like “(part of the) variance”. For these reasons, I do not allow candidates to take homogeneous tests in their own right (only as part of a heterogeneous test) and do not express scores on homogeneous tests in I.Q.
Jacobsen: Some of the following questions may seem tedious. Please excuse if the next questions come across as such, however, a differentiation of reasoning, as well as a convergence, may help with seeing the administrative rationale behind these particular tests and abilities for admissions policies to the Glia Society, including the differentiation between unsupervised and supervised tests. The “Logic tests” section of the page states:
Required: A score corresponding to the 99.9th centile (unless otherwise indicated) on one of these AND one of the Verbal, Spatial, or Numerical tests.
- Reason – Revision 2008 (subtest of Reason Behind Multiple-Choice – Revision 2008)
- Daedalus Test (subtest of Psychometrically Activated Grids Acerbate Neuroticism and of Labyrinthine LIMIT) (Ibid.)
Why the requirement for a coinciding score with “one of the Verbal, Spatial, or Numerical tests”?
Cooijmans: Because homogeneous tests do not provide a sufficient indication of general intelligence in themselves, as explained in the previous answer. Only combined they do.
Jacobsen: Following from the previous section line of questioning, what makes Reason – Revision 2008 and the Daedalus Test reliable and valid metrics of the admissions policy to the Glia Society? (Ibid.)
Cooijmans: The fact that they possess enough reliability and validity to be accepted as homogeneous tests. But those two parameters are by no manner of means the only ones to be considered regarding admission; others include hardness, quality of norms, resolution, and robustness. These six parameters are also incorporated in an overall indicator of test quality. These parameters are computed from the test data using mathematical formulas.
To avoid appearing pedantic, arrogant, or like I am evading the question, the following clarification is kindly provided: Reliability and validity are technical terms from the science of psychometrics. In the event that the asker was not familiar with the technical use of these words and was merely intending to ask something like “Which concrete, tangible aspects of these tests make them possess the named properties (reliability and validity in this case)?” or even something as vague as “Which concrete, tangible aspects of these tests make them good admissions tests?” the following answer applies:
Reliability is the answer to the question “to what extent would this test give the same score if it were possible to take it repeatedly without a learning effect?” Reliability results positively from (1) test length and (2) item intercorrelations. Validity is the answer to the question “what does this test measure?” Validity results from the relation between the test and anything outside the test, including other tests but also real-world variables. The validity of a test can not be higher than its reliability, because a test can not correlate higher with any outside variable than it correlates with itself.
The two tests named in the question both happen to contain extremely original, novel tasks, and as such tap into the candidate’s raw mental abilities. Solving such tasks can probably not be learnt or improved to a great degree.
Jacobsen: The “Spatial tests” section of the page states:
Required: A score corresponding to the 99.9th centile (unless otherwise indicated) on one of these AND one of the Verbal, Numerical, or Logic tests.
- Spatial section of The Marathon Test
- Space, Time, and Hyperspace (Spatial section of “Test For Genius – Revision 2004/2016” and of “Long Test For Genius”)
- LIMIT – Lieshout International Mesospheric Intelligence Test (subtest of Associative LIMIT)
- Strict Logic Spatial Examination 48 (Wai) 17.5
- Eureka Test (Lygeros)
- Spatial Insight Test (no longer used but still accepted) (Ibid.)
What makes these tests, in particular, reliable and valid admissions tests to the Glia Society on the spatial intelligence indices?
Cooijmans: The first three paragraphs of the previous answer take effect here too. In addition, these tests appear to measure mental ability in the target range, so around the Glia Society’s pass level. This is determined by statistical analysis of the available data, and can also be monitored by observing the behaviour of persons with known scores on the tests.
Jacobsen: The “Numerical tests” section of the page states:
Required: A score corresponding to the 99.9th centile (unless otherwise indicated) on one of these AND one of the Verbal, Spatial, or Logic tests.
- Numerical section of The Marathon Test
- Numerical section of Test For Genius – Revision 2010 (Ibid.)
What makes these robust, or valid and reliable, admissions tests on numerical abilities to the Glia Society?
Cooijmans: The first three paragraphs of the answer before the previous answer apply here too. In addition, robustness is a statistic that shows to what extent the raw scores on a test rise or fall over time. It is based on the correlation of raw scores with a “time when taken” indicator, such that January 1995 is 1, and so on. Again, the suitability as admission test is determined through statistics, and also by observation of the behaviour of candidates and members.
Jacobsen: The “Verbal tests” section of the page states:
Required: A score corresponding to the 99.9th centile (unless otherwise indicated) on one of these AND on one of the Numerical, Spatial, or Logic tests.
Unsupervised
Verbal section of Test For Genius – Revision 2004 or 2016
Verbal section of The Marathon Test
Genius Association Test (subtest of Associative LIMIT Test)
Psychometric Qrosswords
The Final Test – Revision 2013 (subtest of The Hammer Of Test-Hungry and of Dicing with death)
The Test To End All Tests (subtest of Narcissus’ last stand)
The LAW – Letters and Words
Qoymans Multiple-Choice #5 (subtest of Reason Behind Multiple-Choice – Revision 2008)
De Roskam
Supervised
Miller Analogies Test (only from period before scaled scores were given; raw score) 94 (Ibid.)
What makes these tests, in particular, important indicators of verbal intelligence compared to others? Verbal intelligence correlates highly with general intelligence. Why?
Cooijmans: I would not use the term “verbal intelligence” but rather “verbal ability”. “Intelligence” is a term I reserve for the general factor in mental abilities. Also, it is not claimed that these tests are indicators of verbal ability; they contain verbal problems, but what a test measures can only be known through statistics, not on face value. The division of items into categories like verbal, numerical et cetera is a topographical, face value division. It is unrelated to what the items actually measure. So, a verbal test is not necessarily a test of verbal ability, a numerical test is not necessarily a test of numerical ability, and so on. This, too, is so abstract that few people understand it, hence the eternal confusion and the use of terms like “verbal intelligence”.
Again, statistics and observation help to know if a test is suitable. On the causal level, the quality of the items may have to do with the eventual functional quality of the test, or one would hope so at least.
Verbal ability (not “intelligence”) correlates highly with general intelligence, or technically speaking, in a hierarchical factor analysis of a variety of mental tests, the verbal factor is fairly close to “g”. Why? On one causal level, I interpret this as a reflection of selection pressures over the past centuries and millennia; there has been strong selection for verbal ability, probably since the advent of cities, which necessitated skills like writing and reckoning for purposes of administration. The hierarchy of mental ability factors is, as it were, a fossil record of (recent) evolution. Were it possible to study mental abilities of ice-age hunter-gatherers from thirty thousand years ago, one might find the spatial factor to be closer to “g” than possible verbal and numerical factors.
On another causal level, the correlation between verbal ability and general intelligence is caused by their both being dependent on the same underlying physical structure and its already mentioned properties (to remind, properties like the number of cortical neurons, neural conduction velocity, the quality of the insulation material around the axons, and the energy-efficiency of the brain). To correct a common misconception, the correlation between verbal ability and general intelligence is NOT caused by the fact that intelligence tests often contain verbal problems; verbal ability also correlates highly with a general factor extracted from non-verbal tests. The general factor expresses itself through a wide variety of item types (“indifference of the indicator”, Charles Spearman called this).
Jacobsen: Why incorporate both unsupervised and supervised tests for membership in the Glia Society here?
Cooijmans: Because any test known to be suitable is accepted, and that happens to include some supervised tests. The only criterion is known suitability for admission at this level.
Jacobsen: The “Tests with a mixture of item types” section of the page states:
Required: 99.9th centile (unless otherwise indicated).
Unsupervised
- Assessment
- Cooijmans Intelligence Test – any form or version
- The Marathon Test
- Test For Genius – Revision 2004, 2010, or 2016
- Associative LIMIT
- A Paranoiac’s Torture: Intelligence Test Using Diabolic Exactitude
- Test of the Beheaded Man
- Dicing with death
- Problems In Gentle Slopes of the first degree
- The Sargasso Test
- Narcissus’ last stand
- Cartoons of Shock
- Problems In Gentle Slopes of the second degree
- The Piper’s Test
- Psychometrically Activated Grids Acerbate Neuroticism
- The Nemesis Test
- Combined Numerical and Spatial sections of The Marathon Test
- Combined Numerical and Spatial sections of Test For Genius – Revision 2010 or 2016
- Problems In Gentle Slopes of the third degree
- De Laatste Test – Herziening 2019
- De Golfstroomtest – Herziening 2019
- Labyrinthine LIMIT
- The Hammer Of Test-Hungry
- Reason Behind Multiple-Choice – Revision 2008
- Reflections In Peroxide (subtest of Narcissus’ last stand)
- Laaglandse Aanlegtest – Herziening 2016
- Bonsai Test – Revision 2016
- <Cooijmans On-Line Test> – Two-barrelled version
- Isis Test
- Divine Psychometry (Scillitani)
- The Alchemist Test (Husseini)
- Magma Test (Vanhove) 17
- Ultra Test (Hoeflin; taken before 2003)
- Mega or Titan Test (Hoeflin; taken before 1998)
- Sigma Test (Melão; taken before December 2003)
- Test For Genius (old versions; no longer used but still accepted)
- Test for extrasensory perception (Cooijmans) 2
Supervised
- SAT (before April 1995) 1470
- ACT (before October 1989) 33
- GRE (before October 2002, verbal + quantitative) 1490
- Army GCT (before 1976) 156 (Ibid.)
In the cases of the Ultra Test, the Mega Test, and the Titan Test, or the SAT, ACT, GRE, and Army GCT, why only accepting the scores if taken before those particular dates for the unsupervised and the supervised tests?
Cooijmans: For the first three, the answer leakage had apparently become such by those dates that they stopped being suitable. For the next three, things changed about those tests that made them no longer possess validity in the intended range. Another problem with these educational tests, as mentioned earlier, is that when people take them purposely (and sometimes repeatedly) to qualify, they stop being good indicators of “g”. They are not robust against deliberate attempts to score high, perhaps because they rely too much on learnt skills. For the last, something changed with the test or score reporting about that date that made it no longer possible to use it as an indicator of intelligence in this range.
Of course, much of this was before my time and in a faraway country, and I learnt of it through advice from persons in the United States and communication within I.Q. societies, and also by studying old statistical data related to those tests, sent to me by someone in the U.S.
References
Cooijmans, P. (n.d.c). Qualification: The Glia Society. Retrieved from http://gliasociety.org/qualification.html.
Cooijmans, P. (n.d.a). The Glia Society: General Information. Retrieved from http://gliasociety.org/general_information.html.
Cooijmans, P. (n.d.b). The Glia Society: The World-wide Hyperbrain. Retrieved from http://gliasociety.org/.
Jacobsen, S.D. (2020, September 1). Conversation with Paul Cooijmans on Introduction to the Glia Society: Administrator, Glia Society (1). Retrieved from http://www.in-sightjournal.com/cooijmans-1.
Appendix I: Footnotes
[1] Administrator, Glia Society.
[2] Individual Publication Date: January 8, 2021: http://www.in-sightjournal.com/cooijmans-3; Full Issue Publication Date: May 1, 2021: https://in-sightjournal.com/insight-issues/.
*High range testing (HRT) should be taken with honest skepticism grounded in the limited empirical development of the field at present, even in spite of honest and sincere efforts. If a higher general intelligence score, then the greater the variability in, and margin of error in, the general intelligence scores because of the greater rarity in the population.
License
In-Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Based on a work at www.in-sightpublishing.com.
Copyright
© Scott Douglas Jacobsen and In-Sight Publishing 2012-Present. Unauthorized use and/or duplication of this material without express and written permission from this site’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Scott Douglas Jacobsen and In-Sight Publishing with appropriate and specific direction to the original content. All interviewees and authors co-copyright their material and may disseminate for their independent purposes.
In-Sight Publishing 