Frequently asked questions

I.Q. Tests for the High Range

The questions and answers

Which tests are accepted for admission by the … Society?

See that society's pages at GliaWeb.

What are the norms or statistics for the … Test? How many have taken the … Test?

See the page with statistics and norms. It contains all available statistics and norms, so that includes all "how many have taken" information.

How is the reported "Total number of mental ability tests taken" computed?

Every test or subtest is counted that has a statistical report of its own at the Statistics page.

Can you recommend a test for me?

Yes, but it is preferred that you choose a test yourself based on the test descriptions provided online. If a test is recommended, the aim is always to select a test that will assess your ability level well, which is not the same as to select a test that you will "like", or that will let you score high as possible. No explanation will be provided as to why a particular test is recommended.

Are repeated attempts at tests allowed?

No. The reasons for not allowing retests are explained in Recommendations for conducting high-range intelligence tests , under "Retesting".


If a test has no time limit, what would be a reasonable amount of time to spend on it?

The time it takes until one can find no further solutions. The irrelevance of the absolute length of that time interval is implied in, and inherent to, the principle of untimed testing. Provided one adheres to the rules, there is no risk of arriving at an unrealistically high score, as some beginners fear. Failing to persist for the thus defined time interval may result in a lower score than one is capable of. Knowing when to stop is part of the test.

What did [some well-known person] score on your tests?

For reasons of privacy and confidentiality that is not revealed. Neither are any other personal details of candidates revealed to third persons.

Why is real-life achievement not accepted as a measure of I.Q.?

I.Q. is a score on an intelligence test, and the purpose of such tests is and has always been to predict (in the statistical sense of "correlate with") real-life achievement. The correlation between I.Q. and real-life achievement is ultimately the validity of the test. Although real-life achievement is obviously more important than test scores, someone who only reports real-life achievement and takes no tests is not contributing to the study of the measurability of intelligence, or to the development, validation, and norming of the tests.

Having said that, the Glia Society does have an assessment procedure in which real-life achievement is considered.

Why do I keep receiving scores of zero on the tests? Am I persona non grata? Do you do this to keep consistency between your tests?

A score of zero means that the number of correct answers was zero. With repeated scores of zero one should conclude that the tests one tried were too hard for one. All scores are reported objectively, without bias against particular persons and without taking priorly achieved scores into account.

When taking a test, is it allowed to use divination to arrive at answers? If so, should only answers from one's own subconscious be used, or are answers from external intelligences in the spirit realm allowed too?

It is allowed to use divination, and both answers from one's subconscious and from the spirit realm are allowed. It is forbidden to consult other psychics or mediums in the physical world though.

Why do I get different scores on different tests?

Because of any of the following causes:

Can you tell me my real I.Q. after I have taken a number of tests with different results?

We do not use the concept of "real I.Q." and do not claim one's real I.Q. can be known with perfect accuracy. However, score reports contain a "Qualified average I.Q." which one may take as the best available approximation of the "real I.Q." this question is asking for.

There are so many tests nowadays on the Internet. How do I recognize incompetent dilettante test creators and scorers?

Incompetent dilettante test designers and test scorers may be identified through the following forms of behaviour:

  1. They allow multiple attempts on one and the same test;
  2. They publish online lists of "high scores" with names of candidates (sometimes without the candidate's consent or knowledge, so watch it!);
  3. They offer one-sided tests (that is, consisting of only one problem type), which are the most popular with candidates, but are less valid with regard to general mental ability, and are the most vulnerable to score inflation;
  4. They publish item analysis data, thus revealing the exact level of difficulty of each problem, which greatly helps future candidates and creates inequality between earlier and later candidates;
  5. They are "empathic" and helpful toward the candidate in communication (for instance, they congratulate the candidate with one's "excellent" score — which is rather inappropriate because a test score is an objective datum and not an achievement — or make remarks that allow one to infer where in the test one's errors were);
  6. The quality of their communication and work betrays they are not particularly intelligent themselves;
  7. Their material contains linguistic and/or typographic errors and sloppiness;
  8. They try to impress with credentials that are in sharp discordance with their actual level of functioning, such as a "doctor" title or "PhD", or high I.Q. scores supposedly obtained by them;
  9. Persons you know have scores on their (the dilettantes') tests that are not in accordance with those persons' apparent ability levels;
  10. They have slow-loading web locations and provide documents (like reports) that are far too heavy in file size relative to the their contents;
  11. They give credit to more or less plausible alternative answers (rather than concluding that their test items are ambiguous and should be removed or revised);
  12. They have little regard for privacy and share test data, including names of candidates, with others, without the candidates' consent or knowledge.

I would like to take a certain test but intend to wait until its ceiling becomes higher after renorming so that I can qualify for a particular society.

While this is not a question, it is pointed out that there is no need to wait, because the test's ceiling may become higher regardless of the fact whether or not you take the test now, for instance when other candidates take the test with lower scores than you but with higher scores on other tests than you. Your present score, even if it is in the ceiling, does not limit the future norm of the ceiling.

Can I take this old test of yours that I found on [some web archive site]?

Obviously, the mere fact that you found the test on an archive site and not on the current web location of I.Q. Tests for the High Range betrays that it concerns an outdated, withdrawn, non-current test that can not be taken any more. You should never respond to material on archive sites as such material is published there by someone other than the original publisher or author without the latter's permission (that is the inherent nature of web archive sites). Frankly, we suspect that at least some archive sites belong to criminal organizations that republish old material, hoping it will be compromising to some random individuals, from whom they will then attempt to extort money in order to have the offending data or footage removed.

Can you give me your personal estimation of my I.Q.?

Currently no, because (1) we doubt whether it is ethical to confront a candidate with a subjective I.Q. estimation and (2) this may, in some candidates, reduce confidence in the test scoring procedure; they might suspect the test scorer of manipulating all following scores to match the earlier given estimation (this would be mistaken, but some people are that suspicious). Both these points involve a potentially disturbed relationship between scorer and candidate.

Why do I have to report personal information like name and age with every test submission when I have already provided that information via the one-time test registration form?

Naturally, this is to identify you against the earlier submitted test registration data. To "know who you are". The registration form data resides in a database with thousands of other candidates; when you send answers to a test, it has to be possible to determine which one of those thousands is you; to connect you to the pertinent entry among thousands. The only safe and practical way is to have the candidate report a small number of personalia with each test submission, which can then be matched against the corresponding entry. The combination of name, age, sex, and e-mail address is preferred for this purpose. Each of those in itself does not suffice for identification, for they are not unique; for instance, there may be more candidates with that name, and there are sometimes more people using the same e-mail address (not to mention that a given person may be using different e-mail addresses on different occasions). None of this is pedantic or idiosyncratic; it is all practical necessity, and all organizations that deal with many clients will be facing this problem. It is explained here so emphatically because a minority of candidates is exceptionally persistent in not providing the required information when sending test answers. Of course, no one is claiming that this is so because they are intending to take tests repeatedly under different names; we would not dare make such a scandalous suggestion.

As an aside and for the reader's entertainment, this also helps to detect some cases of testing under false names; believe it or not, some are so disorganized that they start out taking tests under a particular name, and then later, when addressed by the scorer with that name, respond with amazement: "Why do you call me by my [uncle's, grandmother's, friend's, neighbour's] name?! — clearly not realizing that their words thus betray that (1) they used a false name and (2) they have forgotten which name they used previously. The requirement to report name, age et cetera with every submission is too high a hurdle for such forgetful frauds.

Are you willing to norm [a particular idiosyncratic set of tests] in combination?

No. There are billions and billions of possible combinations of tests out of the tests we deal with, and norming all of those combinations would likely take more time and energy than exist in the universe. Why norm a particular set of tests that given a individual happens to have taken, out of all those billions of sets? The solution to this is the concept of protonorms (to be understood as one's raw score on a hypothetical test with very many items, as it were a combination of all tests) combined with the reported qualified average I.Q. That average represent one's score on a combination of many tests, so it approximates what this question is asking for.

Why do some tests not give scores per section, for instance verbal, numerical, spatial?

Because in those tests (1) the sections are too short to yield a reliable profile and (2) reporting section scores for such a short test would give the candidate too much information as to where one's right and wrong answers resided, and thus would endanger the security of the test. The advantage of such a test is that it is relatively short, while a test that does give reliable section or subtest scores is necessarily longer and more laborious to take.

What are typical mistakes candidates make when submitting test answers?

What are typical mistakes candidates make when ordering tests?

What rules apply when taking tests?

See Rules of conduct. That page has some overlap with the present one, as it is sometimes arbitrary whether an issue belongs there or or here.

Are not your norms much too high? Is it not much too easy for people of average or below-average intelligence to score in the I.Q. 130+ range? Should you not let more people of average or lower intelligence take your tests to get better norms?

The answer to all of these is "no", and the clarification for that is given in this article as it is too long for the present page.

Does not self-selection make your data worthless?

No, as long as one realizes that high-range candidates are inherently and necessarily a self-selected group. One can not take random people from the streets and force them to take high-range tests. The resulting data and statistics are representative of the group of high-range candidates.