The Triple Nine Society asks for one percentile, many scores

0 The short answer

The Triple Nine Society is a high IQ society for adults at or above the 99.9th percentile, which is roughly one person in a thousand, and the decisive point is that it publishes a minimum score for each listed test, so the same rank is 146 on a test with a standard deviation of 15 and 173 on one with a standard deviation of 24. Its own pages say members score at or above the 99.9th percentile "on selected intelligence tests", that only listed tests with a signed score report qualify, and that a test may not be self-administered. ACIS is not accepted by any high IQ society and does not predict a pass, so the honest use is to learn where you stand before you pay for a supervised test: the Quick form is 25 dollars and about 45 minutes, Optimized is 35 dollars and about 110 minutes, and the Full Scale form is 59 dollars and about 175 minutes (prices read on November 4, 2026, and they can change again). The decisive limit is that an online, unsupervised score cannot be submitted, and the page on how the major societies compare sets out each society's own rules.

1 What is the Triple Nine Society?

The Triple Nine Society (TNS) is a social organization for adults with very high measured intelligence, founded in 1978, with a stated purpose of socializing among peers and exchanging ideas. The society's What is TNS page (read November 4, 2026) says it was founded in 1978, that it has more than 2,000 members, and that its membership spans 53 United States jurisdictions and 39 countries. It names its guiding principles as democracy, collegiality, respect for diverse philosophies, and openness to innovation and evolution. The admission rule is on the same page: scoring at or above the 99.9th percentile on selected intelligence tests.

What members actually receive is access to other members. The society's home page (read November 4, 2026) lists local gatherings, a North American meeting called ggg999, a Discourse forum, Facebook groups and discussion groups, and its journal, Vidya. The What is TNS page adds annual meetings in the United States and in Europe. The home page describes the purpose in two phrases, socializing among peers and intellectual exchange. That is a modest, honest statement of what a society of this kind offers. It does not promise higher income, a credential for employers or a certificate of genius, and a reader who joins expecting any of those will be disappointed by the society's own description.

Two things are worth keeping apart. The first is the claim about members, which is a fact about who is admitted: adults who documented a qualifying score. The second is any claim about what members are like, which would be a fact about people and which this page does not make. The society publishes its admission rule, not a study of its members, and the pages read for this article contain no member survey. When this page says that a member "scored at or above" a threshold, it is describing a test result on a document, not a personality or an achievement.

The society sits in a family. Mensa admits the top 2 percent according to Mensa International (read November 4, 2026), and other societies publish standards rarer than the Triple Nine one. The page on the Prometheus Society covers the next rung up, the Mega Society covers a rung above that, and the page comparing Mensa and five other societies lines up all of their thresholds. This page stays with Triple Nine: its percentile, its tests, its rules and its application.

2 What does the 99.9th percentile mean, and how rare is it?

The 99.9th percentile means a person scores higher than 99.9 percent of the reference group, which leaves about one person in a thousand at or above that rank, and on a normal curve that is 3.09 standard deviations above the mean. The 3.09 is our arithmetic on the normal distribution, not a figure from the society. A percentile is a rank, so it does not depend on any scale. The conversion to an IQ number does: the score is the mean of 100 plus the z score times the standard deviation of the scale. The page on why the standard deviation of 15 matters walks through that conversion in detail, and the page on how IQ scores differ from percentiles explains why the two cannot be averaged.

The first column of numbers comes from multiplying 3.09 by each standard deviation and adding 100. The last column comes from the society's Test Scores page (read November 4, 2026). The society's minimums do not match the arithmetic exactly, and the page does not explain how each one was set, so this page does not guess. What the table shows is that a listed minimum is the society's own decision about a specific test, not a universal conversion. The page on how rare a given IQ is lets a reader check any score on any of the three scales, and the page on turning a percentile into an IQ gives the full conversion table.

How rare is one in a thousand in practice? In a group of 1,000 randomly chosen adults, the expected count at or above the cutoff is one. In a stadium of 50,000 people it is about fifty, and the same arithmetic puts the Mensa threshold of the top 2 percent at about one person in fifty. That makes the Triple Nine threshold about twenty times as selective as Mensa's, which is our arithmetic on the two published percentiles. The page on the 99.9th percentile at IQ 145 is the place to see that region of the curve in plain terms.

There is a second consequence of working at one in a thousand, and it is about the instruments, not the people. A norm sample of 2,000 adults contains, on average, two people at that rank. A published norm table cannot rest on two observations, so the numbers at the top of any IQ scale depend on how the test developer shapes the tail of the distribution. A later section of this page returns to that point, because it governs how much weight a society cutoff can bear.

3 Why is the same percentile 146 on one test and 173 on another?

Because IQ is not a unit like the meter: it is a rank re-expressed on a scale whose standard deviation the publisher chooses, and the Triple Nine Society handles that by listing a minimum for each test instead of one number. American Mensa's guidance on its qualifying scores (read November 4, 2026) lists 132 for the Stanford-Binet in two of its tables, 148 for the Cattell and a full scale IQ of 130 for the Wechsler scales, and all of those are the same top 2 percent. Mensa's own explainer on the question of genius (read November 4, 2026) says the Stanford-Binet and many school tests have a mean of 100 and a standard deviation of 16, while the Cattell IIIB and the old form of Raven's Advanced Progressive Matrices have a standard deviation of 24, and that an IQ score means nothing without the name of the test that produced it.

That is the whole mechanism, and it explains most of the confusion on the open web about the Triple Nine Society. A person who reads "146" on one site and "173" on another is not looking at a contradiction. The two numbers sit at nearly the same rank on two different rulers. The society's own text on testing outside the United States states the practical consequence for one of those rulers: many international Mensa organizations use the Cattell IIIB, but the society does not accept their usual maximum of 161, requires a test whose scale reaches at least 173, and asks for a document that shows the actual score and not merely a percentile statement. On the arithmetic above, a score of 161 on a standard deviation of 24 is 2.54 standard deviations above the mean, which is about the 99.45th percentile and cannot certify the 99.9th. The page on the Cattell IQ test shows the conversion between the 24 point scale and the Wechsler scale in full.

Percentile equivalence is also approximate, because standard deviations are not the only difference between tests. Different tests have different norm samples collected in different years, and those norms age. The society's table makes this visible in one line: its Wechsler row carries a note that the third edition is not valid for tests taken after February 15, 2015. The page returns to that date below, because it shows how a society handles the problem that norms go out of date.

4 Which tests does the Triple Nine Society accept, and at what scores?

The society accepts a closed list of intelligence tests and academic aptitude tests, each with its own minimum score and, for most, a date window, and nothing outside the list counts, including certificates from other high IQ societies. The Test Scores page (read November 4, 2026) says only the tests listed in its table can be used, that the applicant submits an official score report, and that there are no exceptions, under any circumstances, to the minimum scores and the age and date restrictions. The table runs to dozens of rows because several tests appear once per scoring era. The rows below are the ones this page can state with confidence, and a reader should check the live table before relying on any of them.

Several patterns in that table matter more than any single row. First, most rows are academic aptitude tests (ACT, GMAT, LSAT, SAT, GRE, Miller Analogies Test), which are not intelligence tests in name. The pages read do not say why they are included, and the society simply sets a minimum for each test and era. Second, the minimums on aptitude tests are tied to eras, because the tests were rescored or restructured over time, which is why the GMAT line has a start and an end date. Third, two rows close their windows entirely: the Mega Test and the Titan Test count only if scored before January 1995, and the pages on the Mega Society and the Prometheus Society explain the history of the very high range tests that sit above this one.

Fourth, and most important for a reader choosing a route, the list names specific tests and, for some, specific editions. A Wechsler report counts if it carries a full scale IQ at or above the listed minimum and the date is valid, and the society's text for Wechsler and Stanford-Binet reports requires the testing professional's letterhead and signature. The page for testing outside the United States and Canada adds that these reports must be given by a licensed clinical psychologist or another specially licensed professional, which is why they tend to be costly, and that the society has accepted Wechsler reports in many languages. A bare score sheet, or a test given by a psychology student without the supervising psychologist's signed attestation of validity, is not acceptable. The Wechsler row as read does not name a later edition, so an applicant with a recent Wechsler report should ask the Membership Officer before paying for anything.

For readers outside the United States, the society's page on testing outside the US and Canada (read November 4, 2026) lists three groups of options: tests that depend on English or mathematics (ACT, Miller Analogies Test, LSAT), culture fair tests that do not depend much on language or mathematics (the Cattell CFT 20-R and the Cattell IIIB), and psychologist-administered batteries (Wechsler and Stanford-Binet). It gives vendor price estimates of about 40 dollars for the ACT, 50 to 130 dollars for the Miller Analogies Test depending on the test center, about 175 dollars for the LSAT, and 50 to 250 dollars or more for the Cattell IIIB. The page says plainly that the society does not set those prices and does not guarantee them, and this page repeats them only as the society's report of vendor figures, not as current quotes. For the psychologist-administered batteries the page gives no price at all.

5 What are the supervision and documentation rules?

The society requires an official score report, and it refuses scores from self-administered tests, from tests given by a business associate or a family member, and from administrators who earned the score before gaining access to admissions tests without disclosing that access. These are the rules on the Test Scores page, stated in the society's words in short form. The score report must name the test and show a score at or above the listed minimum. For tests given by a psychologist, the report must carry the psychologist's name, business address, professional letterhead or official stamp, and signature. Applicants must be at least 18 years old at the time of applying. The only age and date restrictions at the time of testing are the ones in the table, and some rows set a minimum age at test date, as the Stanford-Binet rows do at 16.

Notice what these rules do not say. The pages read for this article do not use the words "supervised" or "proctored" as a general condition. What they exclude is a defined list of conditions: self-administration, an administrator who is a business associate or a family member, and an undisclosed insider. For academic aptitude tests such as the ACT or GMAT, the supervision is built into the test center, and for psychologist-administered batteries it is built into the professional's role. The effect is the same in all cases: a person who is not the applicant and not a relative has produced or certified the score under a standard procedure. An online test taken alone at home is, by the society's text, a self-administered test, and no score from it is accepted.

The rule about insiders deserves a sentence of explanation, because it is easy to misread as a restriction on psychologists. Anyone who is authorized to administer or to access the society's admissions tests must show satisfactory scores earned before that authorization, and must disclose the access on the application. The society says failure to disclose can lead to dismissal. The pages do not state the reason. A plausible one, and this is our inference, is that familiarity with an instrument can raise a score, and a society that relies on one number has to guard that number. The same logic explains why the page on the hardest IQ tests is careful about difficulty: difficulty alone does not make a score trustworthy, and the conditions of testing matter at least as much.

What is not accepted is as informative as what is. The Test Scores page lists three exclusions by name: certificates of membership in other high IQ organizations, Mensa acceptance letters unless they include the exact name of the test taken and meet the score criteria, and Mensa U.K. Cattell IIIB scores. The page does not state the reason, but the effect is clear: a letter that says "accepted" does not tell the society which test was used, and the society's table works only with a named test and a number. An applicant who has been admitted to Mensa on the strength of a Wechsler or Stanford-Binet score above the Triple Nine minimum can submit that report; the Mensa letter by itself is not enough. The pages on how to join Mensa and Mensa qualifying test scores cover the Mensa side of that route.

6 How does the application work, and what does it cost?

The application is an online form followed by a signed printout and a score report sent by email, and the pages read for this article list dues of 10 or 30 dollars but no separate application fee. The society's How to Join page (read November 4, 2026) and its membership application page (read November 4, 2026) describe the steps. The applicant completes the online form with a name, email address and password (which become the permanent login), selects the test being submitted and enters the score and test date, and receives an application number. The applicant then prints the final screen with the Print Form button, signs it, and emails a scan or photo of the signed form together with the official score report to the Membership Officer. The application page says the admission process does not begin until both documents arrive.

The review follows. The How to Join page says eligibility review takes about two weeks, with delays possible, that the applicant receives an email when admission begins, and that a Dues entry in the main menu means the application has been approved. Dues come in two forms. Print Vidya costs 30 dollars and includes a mailed hard copy and online access, and Internet Vidya costs 10 dollars for online access only. The page says dues cover 12 issue numbers of the journal, which is six double issues, and that both options carry full membership benefits. The first payment can be made by card or PayPal through the Membership Store after approval, or by check to the Financial Officer, with 2 to 3 weeks allowed for processing. A score report may be in any language.

The cost that matters is therefore the cost of the qualifying score, not the cost of joining. A reader who already has a valid Wechsler report or an old SAT, GMAT or ACT score pays nothing extra to submit it. A reader who must buy a test pays the test fee reported on the testing outside the United States page, or whatever a licensed psychologist charges, which the pages read do not give. The sensible order is to check whether any score you already hold appears in the table, check its date window, and only then consider paying for a new one. The page on where to take an IQ test lists the supervised and unsupervised routes side by side, and the page on taking an IQ test as an adult covers the adult route in general.

7 How does the Triple Nine Society compare with Mensa?

The Triple Nine Society asks for roughly twenty times as rare a score as Mensa does (top 0.1 percent against top 2 percent), and gives its members a smaller, more specialized community, while Mensa accepts a wider range of standardized, normed tests. Mensa International's What is IQ page (read November 4, 2026) says Mensa admits individuals who score in the top 2 percent of the population, accepts many different tests as long as they have been standardized and normed, and gives the Stanford-Binet and Cattell equivalents as 132 and 148. The Triple Nine Society's list is shorter, its minimums are higher, and the 99.9th percentile is its stated standard. The page on the score Mensa needs covers how a reader can estimate where they sit against the Mensa line.

The two societies are not rivals. A person with a qualifying Wechsler or Stanford-Binet report at or above the Triple Nine minimum also exceeds the Mensa minimum, since the Triple Nine percentile is higher, and the same report can in principle be submitted to both. That is a statement about ranks, not about how either society processes an application. Mensa's own rules about prior evidence, fees and supervised sessions are in the sibling pages on how to join Mensa and Mensa qualifying test scores, which are built from Mensa's official pages and not from this one.

8 Which routes produce a score the society could accept?

Only supervised, listed tests produce a score the society could accept, so the real options are a psychologist-administered battery, a proctored academic aptitude test you may already hold, or a culture fair test given by a licensed professional, and a free or online score is not on the list. The table below sets those routes beside the online options that readers often try first. Costs and times are given only where a page that was opened for this article states them, and anything else is marked as not stated or as varying.

Three conclusions follow from the table. First, if you already hold a score from a proctored academic test, check the society's table before you spend money, because a GMAT or ACT from the right era may qualify with no new testing. Second, if you must buy a test, the Cattell and Wechsler routes require a licensed professional, and the society's page on testing outside the United States warns that availability depends on where you live and must be researched locally. Third, nothing in the unsupervised rows can ever be submitted, whatever its quality. The page on the best online IQ tests compares the online options against each other.

9 What does a qualifying score sample, in CHC terms?

A qualifying score is a blend of abilities, and which blend depends on the test: fluid reasoning (Gf) dominates the culture fair tests, comprehension-knowledge (Gc) and quantitative knowledge (Gq) weigh heavily in the aptitude tests, and a Wechsler full scale IQ combines several broad abilities. The Cattell-Horn-Carroll (CHC) model names those broad abilities, and the page on the CHC model lists the narrow abilities beneath each. The society's own wording supports the split: it describes the ACT, the Miller Analogies Test and the LSAT as dependent on English, with the ACT also needing mathematics, and it describes the Cattell tests as not significantly dependent on language or mathematics. In CHC terms, then, the aptitude route leans on Gc and Gq, and the culture fair route leans on Gf.

That matters for an applicant because the same person can sit at different ranks on different routes. Someone with strong reasoning and a thinner formal education may do better on a culture fair test than on a verbal aptitude test, and someone with deep reading and mathematics may show the reverse. A cutoff at the 99.9th percentile on one route does not mean the person would clear it on another, because each test samples a different mix. The page on what IQ measures sets out the g factor and its limits.

The ACIS subtests illustrate the abilities without standing in for any listed test. Matrix Reasoning samples fluid reasoning through visual rule discovery, and Logic Grid samples controlled deduction under several constraints, which its page relates to Gf. The Vocabulary page treats word knowledge as a classic indicator of crystallized knowledge, tied to years of reading and schooling. These pages describe formats, not items, and no ACIS subtest reproduces any item of any test the society accepts. They matter here because a reader deciding whether to pay for a supervised test benefits from knowing which ability the cutoff will probe hardest.

10 How should you read a score near the cutoff?

A score within a few points of the cutoff cannot settle whether the true rank clears it, because every IQ score carries measurement error, the norms may be old, and the tail of the curve is the thinnest part of any norm sample. The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014), open access at testingstandards.net (read November 4, 2026), address the first point in Standard 2.14: conditional standard errors of measurement should be reported at several score levels unless there is evidence the error is constant, and where cut scores are specified for selection or classification, the standard errors should be reported in the vicinity of each cut score. A society cutoff is a cut score in exactly that sense. The pages read for this article do not report a standard error at any listed minimum, which means the reader cannot know from them how likely a score of, say, 144 against a minimum of 146 is to reflect a true rank above the line.

Think of a score as a point inside a band. The page on reliability and validity explains the vocabulary, and the page on how IQ scores are normed explains what the reference group does. For a person whose observed score is slightly below the cutoff, the true score could be above it, and for one slightly above, it could be below. The society has chosen to treat the listed minimum as a hard line, and its text says there are no exceptions, which is a legitimate administrative choice. It is not a statement that a point of difference reflects a real difference in ability.

Norms add a second source of drift. The Standards say in a comment to Standard 5.1 that a score scale point originally defined as the mean of a reference population should no longer be read as average performance if the scale is held constant and the population changes. Standard 5.11 puts on publishers the responsibility to renorm with enough frequency to permit continued accurate score interpretations, and its comment leaves the user responsible for avoiding norms that are out of date. The APA Guidelines for Psychological Assessment and Evaluation (2020), in the rationale to Guideline 8, name the agedness of norms as a consideration in test selection. The society's table shows a practical response: the WAIS-III row closes on February 15, 2015, which is a way of saying that scores from the older norms should not be treated as if they were current.

The size of the drift has a research estimate. Trahan, Stuebing, Fletcher and Hiscock's meta-analysis in Psychological Bulletin (2014) analyzed 285 studies with 14,031 participants who took two intelligence tests with different norms, and found an average gain of about 2.31 standard score points per decade, and about 2.93 points per decade for the modern Stanford-Binet and Wechsler tests. They report that sample type and test order explained some of the variation, while age and ability level did not. Applied to the society's rule, the following is our reading and not the authors': a score earned on norms that are a decade or two old is likely to read several points higher than the same performance on current norms, which is the problem the society's date windows address.

The third source of uncertainty is the thin tail. As the earlier section showed, a norm sample of a couple of thousand adults holds only a couple of people at the cutoff rank, so a published table must be smoothed or extrapolated there. The Pearson page for the WAIS-5 (read November 4, 2026) says that for examinees of high ability, the WAIS-5 should be considered because of its higher ceiling than the WISC-V at age 16, which is a publisher's acknowledgment that ceilings matter near this range. It does not state what the extreme norms rest on, and this page does not claim to know. The page on the hardest IQ test makes the related point that a harder test does not make its highest numbers more certain.

Practical conditions matter too, whichever route you take. For a supervised test, the examiner controls the room. For an unsupervised practice session, you control it: a quiet space, a full-size screen, and a settled state, with breaks used the way the instructions allow. ACIS lets you take breaks, saves progress between subtests and gives 30 days to complete a form. Three ACIS subtests present items as audio (Digit Span, Alphanumeric Sequencing and Arithmetic) and need working sound. None of these conditions turns an unsupervised result into an admissible one, but they make it a fairer picture of how you perform.

11 What does a score at this level say about a person?

It says that on a particular test, on a particular day, the person scored in the top one in a thousand of the reference group, and nothing more about character, achievement or worth, because the evidence on what rare ability predicts comes from groups followed over time, not from any single score. The strongest long-running evidence on people at a rare rank comes from a different kind of identification than the society's. Kell, Lubinski and Benbow's study, Who Rises to the Top? Early Indicators, in Psychological Science (2013), followed 320 youth identified before age 13 as having profound mathematical or verbal reasoning ability, in the top 1 in 10,000, for nearly three decades. The abstract reports that many held leadership positions in business, health care, law, academia and the STEM fields by age 38. The authors also conclude that identifying such potential requires assessing multiple cognitive abilities and atypical measurement procedures.

Two cautions come with that result. First, it is a finding about a group selected at a rarer level than the Triple Nine one and measured with above-level tests taken early in life, so it does not transfer to an adult who has a Wechsler report. Second, it describes outcomes in a cohort and says nothing about whether a particular member of any society is typical of that cohort. A group result is a description of the group. The page on gifted adults discusses how adult giftedness is assessed and what the life outcome evidence does and does not show, and it does so without turning a percentile into a personality.

For readers who are exploring whether they might be gifted as adults without any society in mind, the page on the gifted test for adults covers how to confirm where you stand, and it separates that question from admission to any society. The Triple Nine Society is one use for a high score. It is not the definition of one.

12 Where ACIS sits, and what a buyer should do

ACIS is an adult normed online battery that can show you where you stand by index before you spend money on a supervised test, and it is not accepted by any high IQ society, cannot predict a pass and does not replace a psychologist's report. The site is direct about this on its research page: ACIS "is not intended for clinical diagnosis, hiring decisions, accommodation determinations or admission to high IQ societies." That is the limit that decides this page. It is online and unsupervised, it is not a clinical or diagnostic instrument, and it is offered in English only. The technical manual notes that verbal and knowledge tasks are affected by English proficiency and educational exposure, which matters for a reader weighing a Gc heavy route against a culture fair one. Anyone who needs a score a society will file must take one of the listed tests under the listed conditions.

What ACIS does offer is a picture. It reports a Full Scale IQ with its 95 percent confidence interval, six primary indices (verbal comprehension, fluid reasoning, quantitative reasoning, visual spatial, working memory and processing speed), percentiles, and scaled scores for each subtest, all on the standard scale with a mean of 100 and a standard deviation of 15. Adult norms cover ages 16 to 90. The Quick form costs 25 dollars and takes about 45 minutes, Optimized costs 35 dollars and about 110 minutes, and Full Scale costs 59 dollars and about 175 minutes, all one time with no subscription (read November 4, 2026). The free trial needs no card, and a second trial from the same email address or browser is not allowed. The refund promise is conditional and is quoted exactly: "Full refund within 5 days if the report does not deliver the features described at checkout or a technical issue prevents you from accessing your results." The technical manual holds the documentation, and the page on high range measurement is the place for readers who want to know what online tools can and cannot say about very high abilities.

The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014) frame the right use. Standard 6.1 asks administrators to follow the standardized procedures specified by the test developer, which is the reason a society insists on the conditions in its table. Standard 2.14 asks for the precision of a score near a cut score, which is why a number just under a line should not be over-read. The APA Guidelines for Psychological Assessment and Evaluation (2020) call in Guideline 5 for psychologists to apply psychometric principles together with the effects of context, setting, purpose and population, and in Guideline 8 for attention to the appropriate normative comparison and standardized administration. Read together, they say a score is evidence for a purpose, under stated conditions, with stated error. An unsupervised ACIS score is evidence of that kind for one purpose only: it informs a decision to test further and does not itself settle membership.

What should a buyer do, in order? First, check the society's live Test Scores page for any score you already hold, such as an older ACT, GMAT, LSAT, SAT or a Wechsler report, and note its date window. Second, if you hold none, decide whether you want a psychologist's battery or a listed aptitude test, using the costs on the society's page on testing outside the United States as a guide, and see the page on where to take an IQ test for the full list of routes. Third, if you want an adult normed picture first, start the free trial or open an example report to see what a report contains before paying anything. Fourth, treat that picture as one input: it does not predict what a supervised test will return, and a profile near the line is a reason to book the supervised test, not a reason to skip it. The page on taking an IQ test as an adult describes the ACIS route in more detail.

13 Sources Behind This Page

Every society fact above comes from the society's own pages, read on November 4, 2026, and every conversion between scales is our own arithmetic on the normal curve, labeled as such. The society's pages can change, so the live Test Scores table is the authority on any minimum score. The group findings cited describe groups, never an individual.

• American Educational Research Association, American Psychological Association and National Council on Measurement in Education. Standards for Educational and Psychological Testing. Washington, DC: AERA, 2014 (Standards 2.14, 5.1, 5.11 and 6.1). Read November 4, 2026.
• American Mensa. Qualifying Test Scores for Mensa Membership. Read November 4, 2026.
• American Psychological Association, Task Force on Psychological Assessment and Evaluation Guidelines. APA Guidelines for Psychological Assessment and Evaluation. Approved by the APA Council of Representatives, March 2020 (Guidelines 5 and 8). Read November 4, 2026.
• Kell H, Lubinski D and Benbow C. Who Rises to the Top? Early Indicators. Psychological Science, 2013, volume 24, issue 5, pages 648 to 659.
• Mensa International. What is IQ? Read November 4, 2026.
• Mensa. The Question of "Genius". American Mensa resources. Read November 4, 2026.
• Pearson Assessments. Wechsler Adult Intelligence Scale, Fifth Edition (WAIS-5). Product page, publication date 2024. Read November 4, 2026.
• Trahan L, Stuebing K, Fletcher J and Hiscock M. The Flynn effect: A meta-analysis. Psychological Bulletin, 2014, volume 140, issue 5, pages 1332 to 1360.
• Triple Nine Society. The Triple Nine Society, home page. Read November 4, 2026.
• Triple Nine Society. What is TNS? Read November 4, 2026.
• Triple Nine Society. How to Join. Read November 4, 2026.
• Triple Nine Society. Test Scores. Read November 4, 2026.
• Triple Nine Society. TNS Membership Application. Read November 4, 2026.
• Triple Nine Society. Testing Outside US and Canada. Read November 4, 2026.

14 Frequently Asked Questions