Charles Manson's IQ: what the 109 and the 121 rest on

0 The short answer

No verified Full Scale IQ exists for Charles Manson: the 109 of 1951 and the 121 of 1956 are notations in his files, reported in a 2023 journal paper that says no list of the instruments was available for either, and no raw data for the first. The only cognitive scores published with their instruments come from August 1997 and are standard scores on parts of tests, 105 on each of two WAIS-R verbal subtests, 93 and 103 on a nonverbal test and 64 on a reading test, none of them an IQ. Research on intelligence and offending describes associations across large groups, and it cannot explain or predict what any one person did.

1 Where does each IQ figure for Charles Manson circulate?

The 109 and the 121 appear on popular pages with no test, date or examiner attached, and the pages read here that cite anything at all lead back to one peer reviewed paper. On October 20, 2026, this review opened four popular pages that print the figures or discuss the evaluation and read what each says about the basis of the number. Other pages that print the figures were not opened, so the table is a sample and not a census. The table sets each claim beside the evidence the page offers for it.

Three things stand out in the table. First, the 80 on the ewash.org page has nothing under it, and no other page read for this review carries it. It is a number that appears with a "some reports suggest" and no report. Second, of the three pages that print the figures, only the Briefly News page points, through a link, toward the journal paper, and its text still reports the two figures without the one fact the paper stresses about them, which is that nothing says what was used to produce either. Third, the Psychology Today post summarizes the 1997 evaluation, cites the paper and is silent about intelligence, which is a reminder that a page about the evaluation is not a page about an IQ.

Pages that repeat one another make one claim appear many times, and they are not independent findings. The 109 on the History Collection page and the 109 on the Briefly News page may well rest on the same file note, but neither page says so. Among the sources read, this review treats the 2023 paper as the only one that reports both figures with a stated basis, and the next section reads exactly what that basis is. The wider problem of famous numbers that spread without a document is the subject of the page on the highest IQ claims, and the page on celebrity IQ claims with every number sourced applies the same questions to many names.

2 What does the 2023 paper say about the 109 and the 121?

The paper reports the 109 and the 121 as results found in his files, dated 1951 and 1956, and states that no raw data and no list of the instruments used were available for either. The paper is The last psychological evaluation of Charles Manson, by Tod Roy, Joni Mihura, Alan Friedman, David Nichols and J. Reid Meloy, in the Journal of Threat Assessment and Management, volume 10, issue 3, 2023, pages 127 to 150. The authors describe themselves as presenting the first publication of psychological test data from Manson, and they explain that the 1997 evaluation and its test results became publicly available only after his death in November 2017. This review read the full text of the paper, in a publisher formatted proof dated December 10, 2022, posted on the website of its fifth author, and the printed version may differ in wording or pagination. Everything below that is attributed to the paper comes from that text.

The passage on earlier testing is short and exact. It says that previous cognitive testing in his files showed he was tested at the National Training School for Boys in 1951, that no raw data or records of the instruments used were available, and that he was found to have an IQ of 109. It says a second assessment was conducted while he served time at Terminal Island, in San Pedro, California, in 1956, that an IQ of 121 was reported, and that again no list of instruments was available. The paper does not name the examiners and does not say what scale either number was on.

Notation is this page's own label for a figure that the paper reports from cognitive testing in his files, without the instrument or, for 1951, the raw data. Each part of that description limits what the numbers can say. A file note records a conclusion. It does not record a performance. Without the instrument, a reader cannot tell whether 109 was a deviation score on a scale with a standard deviation of 15, a ratio of mental age to chronological age, or a rough classification written down by a staff member. The page on mental age explains the older ratio scale, and the page on why a standard deviation of 15 matters explains the scale that Wechsler tests use. A reader who does not know which scale was used cannot place the number on a percentile.

The paper reports one more earlier assessment, and it too has no figure. In 1976, the paper says, his medical file documented that a psychologist administered a battery that included the Rorschach, the House-Tree-Person test and the Graham-Kendall Memory for Designs Test. No raw data were included, and the psychologist concluded that he functioned intellectually in the above-average range. The paper reports no number for that conclusion. It is therefore a third file note, and it differs from the other two in having no figure to circulate.

The paper also says something about age. It reports that he was 63 years old when he was evaluated in 1997 and that he died in 2017 at the age of 83. By our arithmetic, he was therefore in his teens when the 1951 testing took place and in his early twenties for the 1956 testing. Whatever instrument was used, a result from a teenager at a school for boys and a result from a young adult serving time at Terminal Island were obtained under conditions the paper does not describe. A repeated figure from a file is a claim by someone who wrote it down, and this review treats the 109 and the 121 as attributed figures, documented as notations and not as measurements.

3 What did the 1997 evaluation at Pelican Bay test, and why?

The 1997 evaluation was a clinical battery given to answer a placement question, and the intelligence tests in it were a short estimate, not an assessment of intellectual ability. According to the paper, Manson was transferred to Pelican Bay State Prison in August 1997 to serve a Secure Housing Unit term for a serious rule infraction. The evaluation that followed used the Rorschach, the MMPI-2, two subtests of the WAIS-R (Similarities and Vocabulary), the Test of Non-Verbal Intelligence, second edition (TONI-2), the Reading subtest of the Wide Range Achievement Test, revised (WRAT-R) and the Psychopathy Checklist, revised. The paper says the evaluation process extended over 5 days, and that its purpose was to rule in or rule out a mental illness that would decide whether he was placed in a disciplined unit or in a treatment program.

The purpose matters for how to read the cognitive numbers. The paper says the two WAIS-R subtests and the TONI-2 were given to get an estimate of his intellectual functioning, and that the reading test was given to get an estimate of whether he could read well enough to take the MMPI-2. Those are screening uses. A clinician who needs a rough idea of verbal ability, nonverbal reasoning and reading level for a personality inventory does not need, and did not collect, the 11 subtests of a full WAIS-R. The evaluation was not an attempt to rank his ability, and the paper does not use it that way.

The conditions are described plainly in the paper. It says the setting for a psychological evaluation in a prison is not ideal, with inmate movements, loud noises and interruptions by custody officers, and that test protocols had to be modified. It states that he was tested in a steel mesh interview cage with his hands and feet shackled. It reports that the examiner found him cooperative throughout and that at the first meeting his mental functioning varied. None of this makes the scores meaningless, and none of it can be removed from them. A standardized test assumes standard conditions, and the page on reliability and validity explains why departures from them add uncertainty that a reported number does not show.

The paper is also open about its own provenance. It says the first author, Roy, conducted the 1997 evaluation, that Manson gave him permission to publish the findings at some future date, and that the California Department of Corrections gave permission to publish after his death. The MMPI-2 and the Rorschach were rescored and interpreted afresh, according to the abstract, by assessment professionals, and the paper states its aim in terms of personality, psychopathology and the threat of extremist ideology. This page does not assess those parts, makes no statement about any diagnosis, and uses the paper only for the cognitive testing it reports. The authors dedicate the study to the victims of Manson's crimes, in the hope that its insights might help prevent similar tragedies, and this page follows the same restraint in what it says about them.

4 What scores did the 1997 tests produce?

The paper reports five standard scores from 1997, from 64 to 105, and describes the intelligence test results as average and the reading result as far below average. Table 1 of the paper gives a raw score and a standard score for each test, with a mean of 100 and a standard deviation of 15, and the authors converted the WAIS-R subtest scores to that scale so that all of the results could be compared directly. The table below reproduces those values. The last two columns are our arithmetic: the distance from the mean and the percentile that each standard score would have on a normal curve with that mean and spread.

Three points follow from the table. The two verbal subtests gave the same standard score, 105. The TONI-2, a nonverbal measure, gave 93 on its first administration and 103 when the examiner then used a testing of the limits procedure. That procedure goes beyond standard administration, with the examiner exploring what a person can do outside the usual rules, so the 93 is the result under standard conditions. The paper does not explain the 10 point difference, and this page does not try to. The reading result stands apart. The paper reports it at the third grade equivalent and at the first percentile, which it puts in plain words as 99 percent of adults having better reading ability.

The paper summarizes the intelligence test results as in the average range, and that is the only characterization of them that this page repeats. The numbers are on the standard scale, the same scale that the page on IQ score versus percentile explains, and a reader can look any of them up with the IQ percentile calculator. Each is a standard score for one test, and the paper does not combine them. It reports no Full Scale IQ, no Verbal IQ and no composite of any kind, which is consistent with a battery built for another purpose.

One more caution belongs here. The paper also reports standard scores on variables from the Rorschach scoring system known as R-PAS, among them values such as 123, 130 and 133 on variables that the authors describe as evidence of cognitive resources, for example mental complexity, verbal expressivity and the ability to analyze and synthesize situations. Those numbers come from a personality assessment method that scores how a person responds to inkblots. They are not IQs, and a reader who meets a 123 or a 130 in the paper should not put it beside the 109 or the 121. The authors use them in a clinical argument about his thinking and psychopathology. This page takes no position on that argument. It notes only that a figure printed in a paper is not an IQ unless it comes from an intelligence test and is reported as one.

5 Why are two subtests not a Full Scale IQ?

A Full Scale IQ is a composite of many subtests across several abilities, and a result from two verbal subtests and one nonverbal test describes those tasks and nothing wider. The WAIS-R, according to the University of Oslo test sheet, was published in 1981, was authored by David Wechsler, and consists of six verbal and five nonverbal subtests, from which a verbal, a nonverbal and a full scale IQ were obtained. The 1997 battery gave two of those eleven. By our arithmetic, that is fewer than one in five of the subtests the full scale was built from, and the paper reports no score for the other nine.

The two subtests that were given sample one area. In the terms of the CHC model, which names broad abilities with precise labels, Vocabulary and Similarities are measures of comprehension-knowledge (Gc), the breadth and depth of acquired verbal knowledge and the reasoning done with it. The page on fluid versus crystallized intelligence explains why a score in that area can differ from a score in reasoning with novel material. The Similarities and Vocabulary pages describe the formats without reproducing items. The paper calls the TONI-2 a nonverbal intellectual measure and describes its result as nonverbal reasoning, and in CHC terms the closest label for that is fluid reasoning (Gf). The WRAT-R Reading result is a measure of reading achievement, and it is not an intelligence test by any definition the page on what IQ measures uses.

A composite is not an average of whatever was given. It rests on a norm sample for the full set, on a combination rule and on a reliability estimate for the composite, which is why the page on the Full Scale IQ treats the composite as its own object and why a report prints an interval around it. A subtest standard score carries its own, wider uncertainty and covers a narrower slice of ability. Two scores of 105 and one of 93 do not add up to a figure that can be called his IQ, and the authors did not present them that way. They described the tests as a way to get an estimate of his intellectual functioning.

A short estimate can answer a quick clinical question, as the purpose of this battery shows. It cannot support a ranking. When a page converts a partial battery into a single number and sets it beside the figures of famous people, it asks the subtests to carry a meaning they were not built to carry.

6 Can the 109, the 121 and the 1997 scores be compared, and can anyone estimate an IQ today?

No: the 109 and the 121 have no named instrument or norm group, the 1997 scores are partial and rest in part on an instrument published in 1981, and a responsible estimate of a present figure for a man who died in 2017 cannot be made. Take the three sets of numbers in turn. The 1951 and 1956 results cannot be placed on any scale because no instrument is listed. The 1997 results can be placed on the standard scale, but the WAIS-R was published in 1981, which by our arithmetic was 16 years before it was used at Pelican Bay. The paper does not discuss the age of the norms.

Norms age in a known direction. In a study of every Stanford-Binet and Wechsler standardization sample from 1932 to 1978, Flynn found that each sample established norms of a higher standard than its predecessor, a total rise of 13.8 points in mean IQ over 46 years, and he concluded that obsolete norms had acted as an unrecognized confounding variable in hundreds of studies. A later meta-analysis by Pietschnig and Voracek, covering 271 independent samples from 31 countries between 1909 and 2013, estimated gains of 0.28 IQ points a year for full scale IQ, 0.41 a year for fluid tests and 0.21 a year for crystallized tests, and found that gains had decreased in more recent decades. The page on the Flynn effect explains the pattern. The consequence for this page is a direction and not a number: a person scored against an older norm sample tends to look better than against a newer one. How large the effect is for two WAIS-R subtests in 1997, or for an unnamed instrument in 1951, nobody can say from these sources, and this page does not adjust any figure.

Professional guidance asks the same of a living client. The American Psychological Association's Ethical Principles of Psychologists and Code of Conduct, in Standard 9.08, says psychologists do not base assessment or intervention decisions or recommendations on data or test results that are outdated for the current purpose, or on tests and measures that are obsolete and not useful for the current purpose. A 1951 or 1956 notation, with no instrument named, is the kind of result that cannot meet that bar for any purpose beyond history, and the 1997 scores carry the limits above.

Could anyone estimate an IQ today? The same code is direct about opinions on a person's psychological characteristics. Standard 9.01(b) says psychologists provide such opinions only after they have conducted an examination adequate to support their statements, and when an examination is not practical, they clarify the probable impact of their limited information on the reliability and validity of their opinions and limit their conclusions. No examination of a man who died in 2017 can be conducted now. Any figure produced from his biography, his writing or his recorded interviews would be our guess dressed as a result, and this page does not offer one. The history of how scores were produced in the first place, and why the older ratio scale gave way to the standard scale, is in the page on the history of IQ testing.

The practical rule for a reader is the one the page on how IQ scores are normed sets out. A score belongs to a test, an edition, a date, a norm group and an examiner. Take any of them away and the number can no longer be placed. The 109 and the 121 have a year and none of the other four, and the 1997 results have most of them but cover only part of the ability that the word IQ refers to.

7 What does the reading result show, and what does it not show?

The reading result shows that his written language skills were far below his verbal and nonverbal results, and it cannot be read as a measure of intelligence in either direction. The paper reports a third grade equivalent and a standard score of 64 on the WRAT-R Reading subtest, and it describes the result as consistent with his formal education, which it gives as the third grade. In the paper's account, the result also had a practical consequence. The MMPI-2 was administered orally because of his reading comprehension difficulty and because no audio recording of the test questions was available, and the authors note that few significant differences are generally found between standard and oral administration.

The contrast is informative about what each test measures. The Vocabulary subtest, as the paper describes it, concerned his understanding of orally presented vocabulary, so the verbal score of 105 was obtained without any need to read. A reading test measures something else, which is the decoding and comprehension of print, and the CHC model files it under reading and writing ability and not under comprehension-knowledge. A person can have a stronger spoken vocabulary than written decoding, and the reverse is possible, which is why a good test avoids letting one skill stand in for another. The page on intelligence versus knowledge explains why a score reflects what a person has had the chance to learn as well as how well they learn.

What the reading result cannot show is a cause. The paper links it to the amount of schooling he had, and nothing in it says that reading level reflects intelligence, motivation or any single factor. It also cannot be set beside the other results as though all five were measures of the same thing. A standard score of 64 for reading and a standard score of 105 for vocabulary are on the same numerical scale, and they describe different skills. This is the same caution that applies to any profile with large gaps between parts. The page on index score differences explains how common gaps are in the general population and why a gap between parts does not diagnose anything.

There is also a lesson in the 109 file note. One popular page reports the 109 together with illiteracy in one sentence, which invites reading both as results of one examination. The paper's account separates them: the 109 and the 121 are file notations with no instrument, and the reading result is a measured score from 1997. Where a source says that someone with a given IQ could not read, the reader is right to ask whether the reading claim and the IQ claim have the same origin.

8 Does any document support calling Charles Manson a genius?

No document read for this page supports a score that would justify that word, and a page about a person who caused grave harm should not offer readers a puzzle to admire. The record that exists does not point toward it. The 1976 psychologist's conclusion, as the paper reports it, was above average. The 1997 results the paper prints are described as average. The one IQ figure above 109, the 121 of 1956, is a file note with no instrument, and the higher standard scores printed in the paper come from an inkblot scoring system and are not IQs. The page on the origin of the genius threshold of 140 shows that the word has no single definition, and none of the IQ figures that circulate for this man is near 140.

Why, then, does the question come up at all? This review cannot say what moves searchers, and the pages it read give no reason. What they do show is that a range such as 80 to 121 on one page, with an average of 109, is produced without any source, so the number looks like a consensus when it is an echo. The page on the highest IQ claims describes how such numbers propagate, and the same source-first method is applied to other notorious names on the pages for Ted Bundy, Ted Kaczynski and Adolf Hitler, each of which separates what is documented from what merely circulates.

A more basic point concerns what an IQ is. A score describes how someone performed on defined tasks relative to others of the same age. It does not describe what a person chose, how persuasive he was, or why people followed him, and it does not explain the harm that was done. The cognitive tests in the paper were not given to explain the events for which he was convicted, and this page makes no such connection. To call a person a genius on the strength of a number would be to claim a link that the evidence does not contain.

For a reader who came for a number, the honest answer is narrow. The figures that carry a source are two file notations and five cognitive test scores from 1997. The first two have no instrument. The five scores belong to partial tests, and none of the seven is a Full Scale IQ.

9 What do meta-analyses say about intelligence and offending?

The best evidence is a modest association across large groups in which lower measured intelligence goes with more recorded offending, and the 2016 meta-analysis frames intelligence as a protective factor in risk groups, not as a cause or a predictor for any one person. The table lists the designs this page relies on, each with the sample, the measure and the result as the authors state it in their abstracts. They differ in design and cannot be added together, and the sections that follow take them in turn.

Start with the meta-analysis. Ttofi, Farrington, Piquero, Lösel, DeLisi and Murray, in Intelligence as a protective factor against offending, searched 18 databases and identified 15 longitudinal studies that examined whether above-average intelligence acts as a protective factor. Among high-risk participants, higher intelligence predicted lower offending, with an odds ratio of 2.32 and a confidence interval from 1.49 to 3.63. Among low-risk participants the odds ratio was 1.33, with an interval from 0.88 to 2.01 that includes 1, so that estimate was not statistically significant. A meta-regression found the two effects significantly different. The authors conclude that intelligence can protect against offending and that the protection varies with the level of risk.

Each part of that description limits what it can say. A prospective design measures intelligence first and follows participants afterward, which limits the explanation that offending or imprisonment lowered the score. It does not turn an association into a cause, because conditions such as family circumstances and schooling can influence both. The size of the effect depends on risk, so the pattern is not a single slope that applies to everyone. Offending in these studies is also usually recorded offending, which depends on detection, charging and conviction as well as on conduct. The page on IQ and crime explains that gap, and it is the place to read the full discussion of what a population correlation does and does not show. This page does not repeat it.

The abstract read for this review reports no result specific to rare and extreme forms of violence. That matters for the case on this page, because a meta-analysis of general offending was not designed to describe one person whose case was unusual. The correct statement is the narrow one: the research describes how measured intelligence is distributed among groups that differ in recorded offending.

10 What do registers, the psychopathy meta-analysis and a high-IQ survey add?

Large registers show that the inverse association survives comparisons between brothers in a smaller form, the psychopathy literature shows correlations near zero to small, and the one survey of high-IQ adults shows how little is known about the upper end. Frisell, Pawitan and Långström, in Is the association between general cognitive ability and violent crime caused by family-level confounders?, linked Swedish registers covering 700,514 men born from 1961 to 1975. General cognitive ability at age 18 came from the conscript register, and violent convictions came from the crime register for 1973 to 2009. In probit regression, the coefficient was -0.19 (95 percent confidence interval -0.19 to -0.18) in the whole population and -0.18 after adjusting for childhood socioeconomic factors. It was -0.16 within half-brothers raised apart, -0.13 within half-brothers raised together and -0.10 within full brothers. The authors conclude that the association is confounded partly by factors shared by brothers, and that most of it remains after adjustment.

The sibling comparison is the useful part of the design. Brothers share more of their family environment and genes than unrelated men do, so a smaller coefficient between them means that part of the population association reflects what they share. The study covers men, uses conviction as the outcome and measures ability once, at age 18, so its limits are the limits of a register.

On psychopathy, Sánchez de Ribera, Kavish, Katz and Boutwell, in Untangling intelligence, psychopathy, antisocial personality disorder, and conduct problems, report a small negative correlation between intelligence and total psychopathy (r = -.07), and negative correlations with antisocial personality disorder (r = -.13) and conduct disorder (r = -.13), against a small positive one with oppositional defiant disorder (r = .06). They found both positive associations, for example for the interpersonal facet, and negative ones, for example for the affective facet, and significant heterogeneity across studies for most effects. A correlation of -.07 is close to nothing: knowing a person's intelligence tells almost nothing about where that person stands on a psychopathy measure. The page on psychopathy, narcissism and IQ reviews this literature in full and concludes that the evil genius idea is not supported. DeLisi, Vaughn, Beaver and Wright, in The Hannibal Lecter Myth, used 840 cases from the MacArthur Violence Risk Assessment Study and report that verbal intelligence was inversely related to 8 of the 12 psychopathy items. They call the analysis exploratory and note that the sample is unusual, so the results may not generalize.

The survey by Oleson and Chappell, Self-reported violent offending among subjects with genius-level IQ scores, looks at the other end. Its abstract notes that earlier research links offending with IQ scores about 8 points below the population average of 100, and that little is known about offending among people with above-average IQ. It compared 465 high-IQ subjects with 756 controls with average IQ on eight self-reported violent acts. The high-IQ group reported higher prevalence, incidence and arrest, and lower conviction. The abstract read does not describe how either group was recruited, so it cannot say how far either represents anyone.

These studies ask different questions of different samples. One compares hundreds of thousands of conscripts using official convictions, and another compares self-reports from two groups of unequal selection. Neither shows that a high score leads to violence, and neither shows that a low score does. They show that the shape of the relationship at the extremes is not well established. For the question on this page the consequence is modest: even if a verified Full Scale IQ of 109 or 121 existed, the research would not use it to explain anything about what he did.

11 Why do group findings not describe any one person?

A difference in rates between groups describes the groups and not the individuals inside them, and the research on intelligence and offending cannot explain, predict or excuse what any one person did. An odds ratio of 2.32 compares the rate of offending in two groups. It says nothing about which member of either group offended, and it does not say that intelligence caused the difference. A comment to Standard 10.6 in the Standards for Educational and Psychological Testing, written about choosing tests for differential diagnosis, is close in spirit: differences between group means and their statistical significance provide inadequate information regarding validity for individual diagnostic purposes. A pattern across thousands of lives cannot be handed back to one life as a verdict.

The studies also differ from the case this page is about. Apart from the self-report survey described above, whose recruitment is not stated, none of the abstracts read for this review reports results for specific rare acts, which general studies of offending are not designed to analyze. Applying a coefficient from a register of conscripts, or from a meta-analysis of offending in general, to a notorious individual extends a finding beyond what it measured. The Standards say, in the comment to Standard 10.5, that when no normative or validity studies are available for a relevant population, interpretations should be qualified and presented as hypotheses rather than conclusions. For one man, in one extreme case, this review found no such studies.

Two misreadings are common and should be named. The first is that high intelligence makes a person more dangerous, which the registers and meta-analyses read here do not show, and which the one high-IQ survey, a self-report study with unstated recruitment, cannot establish. The second is that a low score makes a person dangerous, which misuses the same inverse association in the other direction. The page on IQ and crime states that a below-average score on any cognitive test is not evidence of risk or criminality, and this page agrees. A fair summary covers both: intelligence tests describe performance on cognitive tasks, and they cannot sort people into categories of harm.

For a reader wondering what the 1997 results say about his actions, the research gives no explanation in either direction, and none should be sought in a number. The results describe how he performed on a few tasks in one evaluation. They do not explain why a person did what he did, and the paper does not claim that they do.

12 What does this mean for reading any IQ score, including yours?

A score is only as meaningful as the documentation behind it, and professional standards ask the same of every score, whether it belongs to a public figure or to you. The Standards for Educational and Psychological Testing, published by AERA, APA and NCME in 2014, ask that each intended interpretation of a score be clearly articulated, with appropriate validity evidence for it (Standard 1.0). They ask that test users receive clear explanations of what a scale score means and what its limits are (Standard 5.1). They ask that released interpretations describe in simple language what the test covers, what scores represent and how precise they are (Standard 6.10). And in the chapter on psychological testing and assessment they say that professionals are responsible for guarding against reliance on test scores that are outdated, in which case retesting is appropriate. A figure that comes with no instrument or norm group meets none of these, which describes the 109 and the 121 as the paper reports them. The same chapter notes that in some cases, such as death penalty evaluations, retrospective analysis is required and multiple sources of information help provide the most comprehensive assessment possible. That describes a professional assembling records for a defined purpose, not a web page converting a file note into a ranking.

The American Psychological Association's code points the same way. Standard 9.06 asks psychologists to take into account the purpose of the assessment and the various test factors, test-taking abilities and characteristics of the person assessed, such as situational and linguistic differences, and to indicate any significant limitations of their interpretations. Those words fit the 1997 evaluation. The purpose was a placement question, the setting was a shackled interview in a cage, the reading skills were far below the verbal results, and only part of one scale was given. Read with those limits, the numbers are informative about a few tasks in one evaluation, and they are not an IQ.

A reader can apply five questions to any famous number, or to their own result: which test and edition, when and at what age, who administered it, which norms were used, and what interval surrounds the score. If a page cannot answer them, the number is circulating and has not been measured. The norm group and the date decide what a score means, as the section on norms above explains.

ACIS sells a paid assessment, so treat this paragraph as a disclosure. The report gives a Full Scale IQ and six primary indices on the standard scale with a mean of 100 and a standard deviation of 15, with percentiles and a 95 percent confidence interval, and adult norms cover ages 16 to 90. It is online and unsupervised. It is not a clinical or diagnostic instrument, and it is not for hiring, school accommodations or admission to high IQ societies. It is available in English only. The technical manual holds the documentation, and no ACIS statistic is quoted on this page. The forms are a one time payment with no subscription, breaks are allowed, and there is a free trial with no card. Prices read on October 20, 2026 can change. The Quick form costs 15 dollars, with 6 subtests in 3 domains and about 45 minutes. The Optimized form costs 30 dollars, with 13 subtests in 5 domains and about 110 minutes. The Full Scale form costs 50 dollars, with all 20 subtests in six domains and about 175 minutes. A 5 day quality guarantee applies, and you have 30 days to complete a form.

To choose among options, the guides to the best online IQ tests and to where to take an IQ test set out the routes, and the comparison of professional and online IQ tests explains what each can and cannot support. If you need a result for a formal purpose such as a diagnosis, accommodations or a hiring decision, ask the receiving body what evidence it accepts and consult a licensed professional. A number you can understand and defend is worth more than a famous one you cannot trace.

13 Sources Behind This Page

Every figure above is traceable to one of the following, and each is linked at the point where it is used. Web pages were read on October 20, 2026. Figures marked as our arithmetic are calculations on a normal distribution with a mean of 100 and a standard deviation of 15, and they are not attributed to any author.

• Roy T A, Mihura J L, Friedman A F, Nichols D S and Meloy J R. The last psychological evaluation of Charles Manson: Implications for personality, psychopathology, and ideology. Journal of Threat Assessment and Management, volume 10, issue 3, 2023, pages 127 to 150. Read in full in a publisher formatted proof dated December 10, 2022, on the website of the fifth author.
• American Educational Research Association, American Psychological Association and National Council on Measurement in Education. Standards for Educational and Psychological Testing. AERA, 2014. Standards 1.0, 5.1 and 6.10, the comments to 10.5 and 10.6, and the passages on retrospective analysis and outdated scores in the chapter on psychological testing and assessment, were read in the open access PDF.
• American Psychological Association. Ethical Principles of Psychologists and Code of Conduct. Adopted August 21, 2002, with amendments effective January 1, 2017. Standards 9.01, 9.06 and 9.08.
• Ttofi M M, Farrington D P, Piquero A R, Lösel F, DeLisi M and Murray J. Intelligence as a protective factor against offending: A meta-analytic review of prospective longitudinal studies. Journal of Criminal Justice, volume 45, 2016, pages 4 to 18. Abstract read.
• Frisell T, Pawitan Y and Långström N. Is the association between general cognitive ability and violent crime caused by family-level confounders? PLoS ONE, volume 7, issue 7, 2012, article e41783.
• Sánchez de Ribera O, Kavish N, Katz I M and Boutwell B B. Untangling intelligence, psychopathy, antisocial personality disorder, and conduct problems: A meta-analytic review. European Journal of Personality, volume 33, issue 5, 2019, pages 529 to 564. Figures are from the published abstract.
• DeLisi M, Vaughn M G, Beaver K M and Wright J P. The Hannibal Lecter myth: Psychopathy and verbal intelligence in the MacArthur Violence Risk Assessment Study. Journal of Psychopathology and Behavioral Assessment, volume 32, issue 2, 2010, pages 169 to 177. Abstract read through secondary listings.
• Oleson J C and Chappell R. Self-reported violent offending among subjects with genius-level IQ scores. Journal of Family Violence, volume 27, issue 8, 2012, pages 715 to 730. Abstract read through the National Criminal Justice Reference Service listing.
• Flynn J R. The mean IQ of Americans: Massive gains 1932 to 1978. Psychological Bulletin, volume 95, issue 1, 1984, pages 29 to 51. Abstract read.
• Pietschnig J and Voracek M. One century of global IQ gains. Perspectives on Psychological Science, volume 10, issue 3, 2015, pages 282 to 306. Abstract read.
• University of Oslo. WAIS-R (Wechsler Adult Intelligence Scale-Revised). Test sheet of the special education and learning lab, file dated February 19, 2019.
• Flight T. 40 facts about Charles Manson, the notorious cult leader. History Collection, September 1, 2019.
• Ntombela T. Charles Luther Manson's life story. Briefly News, December 7, 2023.
• Young L. What was Charles Manson's IQ? ewash.org, December 27, 2024.
• Vitelli R. Exploring the mind of Charles Manson. Psychology Today, Media Spotlight, December 8, 2023.

14 Frequently Asked Questions