Verbal IQ testing and the index that replaced it

0 Quick Answer

A verbal IQ test samples acquired verbal knowledge and reasoning over that knowledge, and no current Wechsler scale reports the result as a Verbal IQ. The term survives in ordinary speech and in old textbooks. What test publishers actually print is an index. Pearson's published WAIS-IV sample score report lists a Verbal Comprehension Index of 145 for its sample examinee and contains the phrase Verbal IQ nowhere. The WAIS-5 sample score report, dated 2024, lists a Verbal Comprehension Index of 124 described as Very high, and it does not contain the phrase either.

The reason for the change was structural rather than cosmetic. Wechsler sorted his subtests into a Verbal Scale and a Performance Scale and reported an IQ for each. Factor analytic work on later editions did not support that two way split as a description of what the subtests share, and publishers moved to four or five index scores that do match the factor structure. What a verbal index means, what task families feed it, and where its limits sit are the subject of the rest of this page.

ACIS builds its Verbal Comprehension Index from five subtests: Antonyms, Vocabulary, Information, Synonyms, and Similarities. The published figures from the ACIS normative model are a composite omega of .9745, a standard error of measurement of 2.40 IQ points, and a loading of .864 on the general factor. That last number is the fourth highest of the six primary domains, which places verbal knowledge firmly inside general ability without making it the center of it. The six domains and how they relate are laid out on the cognitive domains page.

1 What Verbal Ability Means in CHC Terms

In the Cattell Horn Carroll framework, verbal ability is not a separate intelligence but a broad ability called Gc, comprehension knowledge, defined as the depth and breadth of knowledge acquired from a culture and the ability to reason using that knowledge. The framework matters here because it is the taxonomy behind the score structure of every major current battery. Kevin McGrew set out its provenance in Intelligence in 2009, volume 37, pages 1 to 10, tracing the model to the merger of Raymond Cattell and John Horn's extended Gf-Gc theory with John Carroll's three stratum survey of the factor analytic literature.

Gc is not one thing. Jack Schneider and Kevin McGrew list several narrow abilities beneath it in their chapter in the fourth edition of Contemporary Intellectual Assessment, published by Guilford Press in 2018. Three of them account for almost every task that appears on a verbal scale.

• Lexical Knowledge. The extent of vocabulary that can be understood in terms of correct word meanings. This is what a definition task, a synonym task, and an antonym task all sample, from different angles.
• General Verbal Information. The range of general knowledge a person has stored and can retrieve. This is what a general information task samples, and it is knowledge rather than reasoning by design.
• Language Development. General development in the understanding of spoken language at the level of words, sentences, and paragraphs, independent of any single one of the narrower skills.

Three more sit under Gc and are rarely tested by an intelligence battery at all: Listening Ability, Communication Ability, and Grammatical Sensitivity. Their absence is worth noticing. A person can have unusual sensitivity to grammar or unusual skill at making themselves understood and receive no credit for either on any standard verbal index, because the index samples what is easy to score reliably rather than everything the construct contains.

The ACIS subtests map onto these narrow abilities explicitly rather than by resemblance. Vocabulary, Antonyms, and Synonyms are classified under Lexical Knowledge. Information is classified under General Information. Similarities is classified under Induction and Lexical Knowledge together, which is the reason it behaves differently from the other four, as the loadings later on this page show.

2 The Five Task Families and What Each One Isolates

Verbal testing looks repetitive from the outside and is not, because each format holds a different thing constant. The differences are small and consequential. A task that asks you to produce a definition measures something a task that asks you to recognize a synonym does not, and the gap between a person's scores on the two is interpretable information rather than noise.

Consider what separates the second and third rows, since they look nearly identical. To pick the synonym of a word you need its center of meaning. To pick its opposite you need its edges, because opposition is defined against a specific dimension and a word can be near a target in general sense while being wrong as its opposite. Someone with broad but shallow vocabulary tends to do better on the first than the second, and the pattern shows up in profiles often enough that the ACIS technical manual flags it as an interpretive case. You can see the two formats side by side on the same meaning subtest and the opposite meaning subtest.

The first row is different again, and it is the reason production tasks survive in professional batteries despite being slower and harder to score. A multiple choice item gives four candidate meanings and asks for a match. A definition item gives nothing and asks what the word means, so a person who has met the word without ever pinning it down produces a partial answer and receives partial credit. That graded scoring is what distinguishes depth from acquaintance, and it is described on the definition production subtest page. The tradeoff is that production tasks reward the ability to phrase an explanation, which is a real ability and not the one being targeted.

3 Why Vocabulary Size Is Such a Strong Indicator

Vocabulary is the single most reliable verbal indicator in psychometrics because it accumulates from millions of small exposures rather than from anything a person can study for. Two properties make it work. Words are learned mostly incidentally rather than taught directly, and the distribution of words a person meets depends enormously on what they read.

Anne Cunningham and Keith Stanovich set out the arithmetic in American Educator in 1998, drawing on an analysis by Hayes and Ahrens of how rare words are distributed across sources of language. Counting words with a frequency rank below 10,000, roughly those outside the vocabulary of a fourth to sixth grader, the density per thousand words runs as follows: 128.0 in abstracts of scientific articles, 68.3 in newspapers, 65.7 in popular magazines, 52.7 in adult books, 30.9 in children's books, 22.7 in popular prime time adult television, and 17.3 in the recorded conversation of college graduates talking to friends and spouses. Children's books contain roughly 50 percent more rare words than adult prime time television or the speech of college graduates.

The exposure differences that follow are larger than most people expect. Using diary data from Anderson, Wilson, and Fielding's study of fifth graders published in Reading Research Quarterly in 1988, Cunningham and Stanovich reported that a child at the 90th percentile of independent reading time read 21.1 minutes a day, working out to about 1,823,000 words a year outside school. A child at the 50th percentile read 4.6 minutes a day, about 282,000 words. A child at the 10th percentile read 0.1 minutes a day, about 8,000 words. A year of out of school reading for the child at the 10th percentile is roughly two days of reading for the child at the 90th.

The obvious objection is that avid readers were already able, so reading volume is just a proxy for ability. Stanovich and Cunningham tested that directly with 268 college students, entering high school grade point average, an intelligence test, an SAT style mathematics test, and an adult reading comprehension test into the model first, then television exposure, and only then reading volume. Television added nothing. Reading volume added 37.1 percent of the variance in general knowledge, and the pattern replicated across all five knowledge measures they used. In their 1997 study in Developmental Psychology, volume 33, pages 934 to 945, Cunningham and Stanovich followed first graders to the eleventh grade and found that early reading ability predicted eleventh grade vocabulary, general knowledge, and print exposure even with cognitive ability partialed out.

Both readings of that evidence are correct and the honest version keeps both. A vocabulary score is a strong indicator of general cognitive ability, which is why it survives in every battery. It is also a record of what a person has been exposed to, which is why it is the most environmentally sensitive score on a verbal scale. The two facts are not in tension. They are the same fact seen from either end, and what follows from them for anyone trying to raise a score is discussed on the page on changing an IQ score.

4 Wechsler's Verbal and Performance Division

For about six decades the Wechsler scales reported three IQs, and two of them were Verbal and Performance. David Wechsler organized his subtests into a Verbal Scale, containing tasks conducted through language such as vocabulary, similarities, information, and arithmetic, and a Performance Scale, containing tasks conducted through manipulation and figures such as block design, picture arrangement, and digit symbol coding. Each scale produced its own IQ, and the two combined into a Full Scale IQ. That structure ran from the Wechsler Bellevue scale of 1939 through the WAIS, the WAIS-R, and into the WAIS-III.

The division did practical work, and it is worth being fair to it before describing why it went. It gave an examiner a way to assess someone whose command of the test language was limited without abandoning the assessment entirely. It gave a visible contrast that could be compared within one person, so that a large gap between the two scales prompted a question rather than passing unnoticed. And it matched an intuition that people recognize immediately, that some minds work best with words and others with things, which is a large part of why the vocabulary persists in ordinary speech thirty years after the manuals dropped it.

What the division was not is a finding. It was an organizational decision about how to group tasks, made before the factor analytic tools that could evaluate it were widely applied to the scales. The distinction between a structure a test author imposes and a structure the data support runs through the whole history of the field, and it is traced on the history of IQ testing page.

5 Why Publishers Replaced It With Index Scores

The replacement happened in two steps, and both are visible in documents anyone can download. The WAIS-III, published in 1997, added four index scores alongside the existing IQs: Verbal Comprehension, Perceptual Organization, Working Memory, and Processing Speed. Confirmatory analyses of its standardization sample of 2,450 adults and adolescents, the same normative data Hartshorne and Germine later reanalyzed in 2015, supported the four factor structure. They did not support the two scale dichotomy as a description of what the subtests share.

The WAIS-IV, published in 2008, took the second step and stopped reporting Verbal IQ and Performance IQ at all. The evidence is in Pearson's own sample output. Its Composite Score Summary lists exactly six rows: Verbal Comprehension at 145, Perceptual Reasoning at 123, Working Memory at 133, Processing Speed at 122, Full Scale at 139, and the optional General Ability Index at 139. The phrases Verbal IQ and Performance IQ appear nowhere in the document. The WAIS-5 sample report from 2024 keeps the same architecture, listing five primary composites and a Full Scale, and adds ancillary composites including a Verbal Expanded Crystallized index and a Verbal Reasoning index.

The gain from the change is specificity. A Verbal IQ mixed together tasks that shared a delivery channel, which meant a person could arrive at the same Verbal IQ through very different combinations of word knowledge, stored facts, and verbal abstraction. An index built from a defined set of indicators, reported alongside its own confidence interval, is a narrower claim and therefore a more useful one. The cost is that the reader now has to hold five or six numbers instead of three, which is exactly the complaint the old structure was designed to avoid. How the current editions differ from each other is compared on the WAIS-IV versus WAIS-5 page.

6 What a Verbal Index Actually Reports Now

A modern verbal score is a point estimate, an interval around it, a percentile, and a base rate for its distance from the other indexes, and all four belong in the reading. The two Pearson sample reports show this concretely. In the WAIS-IV report the Verbal Comprehension Index is 145, at the 99.9th percentile, with a 95 percent confidence interval running from 138 to 149. In the WAIS-5 report the Verbal Comprehension Index is 124, at the 95th percentile, with a 95 percent confidence interval from 115 to 130 and a standard error of measurement of 4.50 points.

Notice how wide those intervals are. The WAIS-5 verbal figure of 124 is consistent with a true score anywhere from 115 to 130, which spans two classification bands. That is not a defect of the instrument, it is the honest expression of a score with a standard error of 4.50 points. Any verbal number reported without an interval is being presented as more precise than it is. If you want to see where a point estimate sits in the population you can convert a score to a percentile, but the interval is the object that should travel with it.

The fourth element, the base rate, is the one most often skipped. The WAIS-IV sample report tabulates the difference between each pair of indexes, tests it for statistical significance, and then reports how common a difference of that size is in the standardization sample. For its sample examinee the gap between Verbal Comprehension at 145 and Perceptual Reasoning at 123 is 22 points, which exceeds the critical value of 8.31 at the .05 level and is therefore statistically reliable. The base rate for a difference that large is 5.7 percent of the overall sample. Reliable and uncommon are two separate facts, and the report prints both because they answer different questions.

One more feature of the current reports is worth naming because it changes what a verbal score compares you against. The WAIS-5 prints both an age based scaled score and a reference group scaled score for every subtest, the second computed against examinees aged 20 to 34. For the sample examinee, aged 70 years and 5 months, Vocabulary scores 15 against her own age group and 16 against the young adult reference group. On Coding the same comparison gives 11 against 9. Which comparison is the right one depends entirely on the question being asked, and a verbal score reported without saying which group it is a rank within is incomplete.

7 How to Read a Verbal Score Without Overreading It

The interpretive rule that survives across every serious source is to read the index rather than the subtest, and the subtest rather than the item. Alan Kaufman and Elizabeth Lichtenberger make this the spine of Essentials of WAIS-IV Assessment, published by Wiley in 2009, which is the standard interpretive reference for the scale. Their case rests on measurement rather than preference: a composite built from several indicators has a smaller standard error than any of its parts, so the same amount of confidence buys a narrower statement at the index level than at the subtest level.

The ACIS figures show the size of that difference directly. The Verbal Comprehension Index, built from five subtests, has a standard error of measurement of 2.40 IQ points. Its individual subtests, expressed on the scaled score metric with a mean of 10 and a standard deviation of 3, have standard errors of 0.67 for Vocabulary, 0.85 for Information, 0.95 for Antonyms, 1.04 for Synonyms, and 1.12 for Similarities. Converted to the IQ metric those subtest errors are several times the index error, which is the arithmetic behind the rule.

That said, the pattern within an index carries information the composite cannot, and the ACIS technical manual names the specific comparisons that mean something.

• Antonyms high with Similarities lower suggests strong word knowledge alongside weaker reasoning about semantic relations.
• Antonyms low with Synonyms adequate points to difficulty with opposite meaning contrasts specifically rather than a broad lexical limitation.
• Information lower than Vocabulary suggests strong word meaning with narrower factual exposure.
• Similarities higher than the other four suggests strong abstraction despite ordinary lexical breadth, which is the profile a purely vocabulary based verbal score would misdescribe.

Each of those is a hypothesis to check against everything else known about the person, not a conclusion. A single subtest score with a standard error above one scaled score point cannot settle anything by itself, and the correct response to an interesting pattern is more information rather than a firmer statement. The general machinery of standard errors, confidence intervals, and what they permit is on the reliability and validity page, and what the resulting bands mean in ordinary language is on the score interpretation page.

8 Verbal Ability Across the Lifespan

Verbal knowledge holds up with age in a way that reasoning speed and working memory do not, and this is one of the most replicated results in cognitive aging. Timothy Salthouse examined it in Neurobiology of Aging in 2009, volume 30, pages 507 to 514, drawing on a cross sectional sample of 2,350 adults aged 18 to 60 and a longitudinal subsample of 729 with retest intervals averaging 2.5 years. His central argument was that some aspects of age related decline begin in healthy educated adults while they are in their twenties and thirties, and that longitudinal studies had been missing it because practice effects from prior testing masked the change.

The exception he stated explicitly is the one that matters here. Measures based on accumulated knowledge, such as tests of vocabulary or general information, are consistently found to increase until at least age 60. The decline he was documenting applies to reasoning, speed, and memory, not to the knowledge store.

Joshua Hartshorne and Laura Germine mapped the same contrast with far finer age resolution in Psychological Science in 2015, combining normative data from the WAIS-III and WMS-III with 48,537 web participants. Their finding was that different abilities peak on different schedules rather than together. Digit symbol coding performance peaked first, in the late teens. Working memory tasks peaked around age 30. Vocabulary peaked last of all, and later in their web sample, at around age 65, than in the WAIS-III normative data, where it peaked at around age 50. They described the result as considerable heterogeneity in when cognitive abilities peak, with some declining around high school graduation, some plateauing in early adulthood and turning down in the thirties, and others not peaking until the forties or later.

Two cautions belong with these numbers. Hartshorne and Germine reported peak ages, not rates of change, and no per year slope should be attributed to them. And a peak in a cross sectional or web sample is partly a statement about cohorts, since people born in different decades had different schooling and different reading environments. The broader contrast between abilities that hold and abilities that fade is set out on the fluid versus crystallized page, and what it implies for a score obtained at one age is on the page on IQ and age.

9 What a Verbal Score Means in a Second Language

A verbal index measures knowledge of a specific language, so for someone tested in a second language it measures exposure to that language as much as it measures Gc. This is not a caveat bolted onto the score, it is what the score is. Every item on a lexical knowledge subtest asks about the meaning of an English word, and a person who has spent five years in English does not have the same store of English word meanings as a person who has spent forty, whatever their comprehension knowledge in their first language.

The ACIS technical manual records this as a structural property of the instrument rather than a disclaimer. Its analysis basis lists the language context as English speaking, and states that this basis is used when interpreting VCI and other language sensitive scores. The adult reference frame is drawn from 3,243 English speaking records aged 16 to 90, and the technical analysis set is 2,750 complete records. A verbal index computed against that frame answers the question of how a person's English verbal knowledge compares with that frame. It does not answer a question about verbal ability in general.

Two practical consequences follow. The first is that a depressed VCI in a second language speaker should be read as evidence about English exposure before it is read as evidence about ability, and the rest of the profile is the check. If the fluid reasoning, visual spatial, and quantitative indices sit well above the verbal one, the most economical explanation is the language, not the reasoning. The second is that when a language sensitive score is likely to mislead, a reduced verbal composite is the appropriate object to read instead.

ACIS reports one for exactly this purpose. The Culture Fair Index is built from 15 subtests weighted toward reasoning, working memory, and speed rather than verbal knowledge, and the published figures are composite omega .9837, a standard error of measurement of 1.92 IQ points, and a g loading of .949. It is a reduced verbal composite, not a culture free one, and the difference is not a quibble: instructions still arrive in language, item conventions are still learned, and the phrase culture fair describes an intention rather than an achievement. What that label can and cannot promise is examined on the culture fair testing page, and the evidence on how language background relates to measured scores is on the average IQ by language page.

10 How the ACIS Verbal Comprehension Index Is Built

Five subtests feed the index, and one of them behaves differently from the other four in a way the loadings make visible. The published standardized loadings within the higher order model run from .814 for Information at the low end to .865 for Similarities at the high end, with Antonyms at .860, Vocabulary at .848, and Synonyms at .828. The index itself loads .864 on the general factor and reports composite omega of .9745 with a standard error of measurement of 2.40 IQ points.

Similarities carries the highest loading of the five because it is the only one that requires reasoning as well as knowledge. Naming what two concepts have in common means retrieving both meanings and then finding a category that covers them, and the scoring separates the two steps: a superordinate category earns full credit while a concrete or functional resemblance earns partial. That is the function it serves in the index, which is to stop the verbal score from collapsing into a vocabulary and general knowledge measure. You can see the task on the verbal concept formation subtest page.

Information sits at the other end for the opposite reason. It samples what has been stored rather than what can be worked out, items advance automatically after a short fixed interval so that reasoning time cannot substitute for knowledge, and its content depends on exposure, education, curiosity, reading, and cultural context. That dependence is what makes it informative about the knowledge store and what makes it the most context bound of the five, as the general knowledge subtest page sets out.

The index does not sit in isolation. VCI correlates .801 with the Fluid Reasoning Index in the published intercorrelation table, which is roughly 64 percent shared variance and about 36 percent that is specific to the verbal domain. Against the Processing Speed Index the correlation is .577. Those two numbers together describe the interpretive situation: the verbal index is substantially a measure of general ability, and it is not only that. Full derivations, model fit, and the reliability tables are in the technical manual.

11 What a Verbal Score Does Not Tell You

The most common error with a verbal index is treating it as a rating of how well someone uses language, which is a much larger thing than the index measures. Four specific limits are worth stating, because each one gets crossed regularly.

It is not writing ability. Every task on a verbal index is short, closed, and scored against a key or a rubric. Composing an argument over several pages draws on planning, structure, and revision, none of which any verbal subtest samples. A person can hold an unusually large vocabulary and write badly, and the reverse happens too.

It is not conversational fluency. Listening Ability and Communication Ability are narrow abilities under Gc that standard batteries do not test. Someone can score at the top of a verbal index and be a poor listener, because the index never asked.

It is not reading comprehension. Vocabulary is a strong predictor of reading comprehension and is not the same construct. Comprehension involves integration across sentences and inference over a text, which single word tasks do not require. A high verbal index makes strong comprehension likely rather than certain.

It is not a diagnosis. A low verbal index is compatible with a language difference, limited schooling, a hearing history, anxiety on the day, or a specific language disorder, and the score cannot distinguish among them. ACIS is a self-administered online assessment and is not a clinical instrument, so no result from it should be used for diagnosis, hiring decisions, accommodation requests, or admission to anything.

There is a fifth point that is less a limit than a correction. People often expect a strong verbal score to imply a strong overall score, and treat a gap between the two as a problem. The .801 correlation between VCI and fluid reasoning means the two travel together most of the time and separate regularly, and a profile with high verbal knowledge and mid range fluid reasoning is a normal shape rather than an anomaly. The composite that strips out speed and working memory to compare reasoning and knowledge directly is described on the general ability index page, and the contrasting domain, novel reasoning with no knowledge to draw on, is covered on the fluid reasoning page.

12 Choosing a Form If You Want a Verbal Index

A verbal index is only worth reading next to the other indices, so the form to take is the one that produces the profile rather than the one that produces the verbal number fastest. That is the whole routing argument. A verbal score on its own cannot be checked, because the interpretive work in every section above consists of comparing the verbal figure against fluid reasoning, against visual spatial performance, and against the profile's own dispersion.

ACIS is a self-administered online adult assessment covering 20 subtests across six CHC domains for ages 16 to 90, sold as a one time payment rather than a subscription. Two of its three forms return a full profile.

• Optimized, 30 dollars, about 110 minutes. Thirteen subtests returning all five primary indices plus the Culture Fair Index and the General Ability Index. This is the shortest form that lets a verbal score be read against a complete set of domain comparisons.
• Full Scale, 50 dollars, about 175 minutes. All 20 subtests, returning a Full Scale IQ, all six primary indices, and every composite. It is the only form that administers all five verbal subtests, so it is the one that produces the five indicator Verbal Comprehension Index described on this page.

The Quick form at 15 dollars runs six subtests in about 45 minutes and includes two of the five verbal subtests, so its verbal figure rests on fewer indicators than the one whose reliability is quoted above. It is a reasonable entry point and it is not the right purchase if the verbal index is the reason you came. The free trial covers five subtests weighted toward working memory and processing speed, shows scaled scores and a provisional Full Scale estimate, and locks the confirmed Full Scale IQ, the six domain profile, and the percentiles. Paid access lasts 30 days, the trial lasts seven, and there is a five day guarantee. How the forms compare on time and coverage is set out on the adult testing page and the Full Scale form page.

One boundary belongs here rather than in a footer. An unsupervised session, however carefully scored, is not a substitute for a proctored administration by a licensed psychologist when an institution has to accept the result. If a school, an employer, a clinician, or a court needs the number, the route is a supervised evaluation with a current battery, and nothing taken over the internet stands in for it.

13 Sources Behind This Page

Every figure on this page came from one of the following, and the published documents were read rather than summarized from elsewhere. Where a number sits in a specific table or report section, that location is named so it can be checked.

• NCS Pearson. WAIS-IV Score Report, sample output. Source of the Composite Score Summary showing VCI 145, PRI 123, WMI 133, PSI 122, FSIQ 139 and GAI 139, of the 95 percent confidence interval of 138 to 149, and of the index discrepancy table giving the 22 point VCI to PRI gap with a critical value of 8.31 and a base rate of 5.7 percent. The document contains no Verbal IQ or Performance IQ.
• NCS Pearson. WAIS-5 Score Report, sample output, 2024. Source of VCI 124 described as Very high with a standard error of 4.50 and an interval of 115 to 130, of the FSIQ of 111 described as Above average, of the ancillary Verbal Expanded Crystallized and Verbal Reasoning composites, and of the reference group scaled scores computed against ages 20 to 34.
• McGrew, K. S. (2009). CHC theory and the human cognitive abilities project. Intelligence, 37(1), 1 to 10. Source of the provenance of the Cattell Horn Carroll framework used throughout.
• Schneider, W. J., and McGrew, K. S. (2018). The Cattell Horn Carroll theory of cognitive abilities, in Flanagan and McDonough, editors, Contemporary Intellectual Assessment, fourth edition, Guilford Press. Source of the narrow abilities under Gc named in section one.
• Kaufman, A. S., and Lichtenberger, E. O. (2009). Essentials of WAIS-IV Assessment. Wiley. Source of the interpretive principle that index level scores are the appropriate unit of interpretation.
• Cunningham, A. E., and Stanovich, K. E. (1998). What Reading Does for the Mind. American Educator. Source of the rare word density table adapted from Hayes and Ahrens, of the reading volume table adapted from Anderson, Wilson, and Fielding, and of the 37.1 percent figure from Stanovich and Cunningham's 1993 study of 268 college students.
• Cunningham, A. E., and Stanovich, K. E. (1997). Early reading acquisition and its relation to reading experience and ability 10 years later. Developmental Psychology, 33(6), 934 to 945. Source of the finding that first grade reading ability predicted eleventh grade vocabulary, general knowledge, and print exposure with cognitive ability partialed out.
• Salthouse, T. A. (2009). When does age related cognitive decline begin? Neurobiology of Aging, 30(4), 507 to 514. Source of the sample sizes of 2,350 and 729 and of the statement that knowledge based measures increase until at least age 60.
• Hartshorne, J. K., and Germine, L. T. (2015). When Does Cognitive Functioning Peak? Psychological Science, 26(4), 433 to 443. Source of the peak ages for digit symbol coding, working memory, and vocabulary, of the WAIS-III and web sample vocabulary comparison, and of the 48,537 participant figure. No rate of change is attributed to this paper, because none was reported for the abilities discussed here.
• ACIS technical manual, current edition. Source of every ACIS figure used above, including the VCI omega of .9745, the standard error of 2.40, the g loading of .864, the subtest loadings from .814 to .865, the subtest formats and lengths, the Culture Fair Index figures, the index intercorrelations, and the reference frame of 3,243 English speaking records with a technical analysis set of 2,750.

The professional framework governing every claim on this page is explicit about what a score report owes its reader. The Standards for Educational and Psychological Testing, published jointly in 2014 by the American Educational Research Association, the American Psychological Association and the National Council on Measurement in Education, require that score interpretations be supported by validity evidence for the specific use proposed, that reliability and standard error accompany any score used in a decision, that the composition and limits of the norm sample be disclosed, and that linguistic and cultural characteristics of the test taker be taken into account when a language loaded score is interpreted. The APA standards on test use place the same duty on whoever reports the result. Applied to a verbal index, that last requirement is not a formality. It is the difference between a score that describes a person's knowledge of English and a score presented as a description of their mind.

14 Frequently Asked Questions