OpenPsychometrics is free, fast and unusually candid that its Full Scale IQ Test is an educational demonstration rather than a real professional replacement. That honesty earns credit, but the short unsupervised form and missing public technical evidence keep it short of a professional grade assessment.
free, fast and honest, but not professionally sufficient.
0 The Short Answer
Open Psychometrics is a free, non commercial research site, and reviewing it the way you would review a paid IQ product misses what it actually is. There is no checkout page, no subscription, no upsell after the result and no report locked behind a payment wall. The Full Scale IQ Test at openpsychometrics.org gives you a number and three component scores at the end, at no cost, and then offers you the option of contributing your answers to a public dataset. Most of the commercial questions a review normally asks, what does it cost, what happens if you cancel, is the price disclosed before you invest twenty minutes, simply have no object here.
What replaces those questions is a different and more interesting set. Open Psychometrics publishes its raw response data. Anyone can download it, check the scoring, run their own factor analysis and publish the result, and several researchers have done exactly that. That is a level of openness almost no paid IQ product offers, and it deserves to be said plainly before any criticism follows: on the specific dimension of letting outsiders inspect the evidence, this free site outperforms most of the industry that charges money.
The limitations are real and they are of a different kind. The scoring is anchored to self reported scores from an unverified internet sample, not to a stratified norm group. Nobody watches you take it, so the conditions under which your number was produced are unknown even to you. The Full Scale IQ Test page has carried the same substantive update date for years. And the site itself says, in more direct language than most commercial competitors would ever use, that its result should not be treated as a replacement for a real IQ test. That warning is accurate, and it is the single most useful sentence on the page.
So the honest summary: an excellent free instrument for curiosity, teaching and research participation, backed by unusual transparency, and a weak basis for any individual decision that matters. The rest of this page works through the evidence behind each half of that sentence, and builds an evaluation checklist you can apply to any test you meet afterward.
Who wrote this and why it matters
This review comes from the team behind ACIS, which sells a competing cognitive assessment. Read it as an informed competitor's evaluation, not as a neutral third party audit. Everything here is checkable: we describe only what Open Psychometrics publishes publicly, we date what we observed, and we link to the sources so you can confirm each claim yourself. Where the free competitor beats paid products, including on data transparency, we say so.
1 What Open Psychometrics Actually Is
Open Psychometrics runs at openpsychometrics.org and brands itself the Open Source Psychometrics Project. Its about page credits the project to Eric Jorgenson and dates its operation to late 2011, describing two purposes at once: educating the public about personality and ability tests, and collecting response data to support research and the development of new measures. It is not a company, it does not sell anything on the site, and it does not present itself as a clinical provider.
The catalog is broad and skews heavily toward personality rather than cognition. Alongside the Full Scale IQ Test and the 45 item Vocabulary IQ Test, the site hosts the Multifactor General Knowledge Test, the Nature Relatedness Scale, an Artistic Preferences Scale, the Woodworth Psychoneurotic Inventory (often described as the first personality inventory, developed for the United States military during the First World War) and the statistical character quiz that made the site briefly famous on social media. That last one matters for context: a large share of the site's traffic arrives for a quiz that matches users to fictional characters, not for cognitive measurement.
Understanding this shape changes how you should read the IQ test. It was not built as a consumer product competing for paying customers. It was built as one instrument in a hobbyist and research collection, published by someone who wanted the response data more than he wanted your attention. That explains both the strengths (open data, blunt disclaimers, no funnel) and the weaknesses (no dedicated technical manual, an update cadence measured in years, no support channel for score interpretation).
One practical warning about the name before anything else. There are similarly named sites on other domains, and results in search engines mix them freely. Everything on this page refers specifically to openpsychometrics.org and to the test it publishes at /tests/FSIQ/. If you land on a page with a similar name that asks you for a credit card, you are not on Open Psychometrics.
2 What the Full Scale IQ Test Contains
The Full Scale IQ Test presents six sections, each with its own instruction screen, and reports an overall score plus three component scores labeled short term memory, reasoning and verbal. The test's own page states that the average taker finishes in roughly 10 to 15 minutes, with the design running to about 20 minutes, and it advises against rushing: finishing in under ten minutes can depress your result, while stretching the session out buys you very little. Pacing advice of that sort matters more than most takers expect, and how to take an online IQ test properly covers the same ground: one uninterrupted sitting, no rushing, and no reference material within reach.
Two administration rules on the page deserve attention, because both are honest and both are unenforceable. The first is that you should complete it alone, without references or outside help. The second is that only your first attempt should be treated as valid, since familiarity with the items changes what a second attempt measures. Nothing in the software can check either condition. They are requests, and the site says as much rather than pretending otherwise.
Six sections is more structure than the single matrix grid that most free IQ quizzes offer, and less than a full battery. For calibration, professional adult batteries typically run ten to fifteen subtests across roughly an hour and a half of individual administration. A twenty minute self administered form has to sample far less. It can produce a defensible rough ordering of people, which is what it claims, and it cannot produce the fine grained profile a longer battery is built for, which is what it declines to claim.
The three component labels are worth reading precisely. Short term memory, reasoning and verbal are not the six broad abilities of the CHC framework that longer batteries organize around. Visual spatial processing, processing speed and quantitative reasoning are not reported as separate scores here. If you want to know how your fluid reasoning compares with your crystallized knowledge, our explainer on fluid and crystallized intelligence covers the distinction, and this particular test does not separate them for you.
3 How the Score Is Produced, and Why That Method Is Unusual
This is where Open Psychometrics gets genuinely interesting, and where most reviews of it stop reading too early. The site does not pretend to have a stratified national norm sample. It explains, on its results page, exactly what it does instead, and the explanation is candid enough to be worth stating in full.
Raw scores are rescaled so the general population would average 100 with a standard deviation of 15, the standard modern scale. But the people who take an online IQ test are not the general population. As the site puts it, people who like tests and are good at them are the ones who show up, so the distribution of takers is shifted and the raw distribution cannot be used directly. The fix Open Psychometrics applies is an anchoring step: at the end of the test, an optional research survey asks takers to report scores they have received on proctored IQ tests and college entrance exams, and those self reported figures are used to calibrate where the online distribution sits relative to the population.
Think about what that method assumes. It assumes people report their proctored scores accurately, that the subset who bother to answer the optional survey is representative of the subset who do not, and that self reported college entrance exam scores map onto the IQ scale in a stable way. Each assumption is arguable. Self reported ability scores are known to drift upward, people with disappointing results are less likely to volunteer them, and the survey responders are a self selected group inside an already self selected group. BRGHT drops the self report step altogether and scores each taker against its own finishers, which is what norming on site traffic looks like instead of on volunteered scores.
Here is the part that matters for your reading of this site, though: the method is disclosed. You can see the anchoring logic, agree or disagree with it, and check it against the published data yourself. Compare that with the standard commercial practice, where a page announces that your score was compared against a norm sample and never names the sample, its size, its composition or its date. A visible imperfect method is worth more to a careful reader than an invisible one that claims perfection. For how proper norming is supposed to work, see our page on how IQ scores are normed.
4 The Open Data Archive, the Site's Strongest Feature
Open Psychometrics publishes an archive of anonymized raw response data at openpsychometrics.org/_rawdata/, covering dozens of its instruments. Downloads typically ship with a codebook and the item text, which is the difference between a dataset you can actually analyze and a spreadsheet of unlabeled columns. Mirrors of the archive exist on GitHub and Kaggle, and some individual datasets carry an explicit open data license: the Big Five responses on Kaggle, for example, are released under the Database Contents License version 1.0.
The consent language is documented alongside the data. Takers were told at the start that their answers would be used for research, and were asked at the end to confirm that their responses were accurate and suitable for research use, with non confirming records removed from the published files. That is a cleaner consent trail than many commercial testing sites can show, and it is stated where a researcher can read it before downloading.
The practical payoff is that outsiders can and do check the work. A mirror maintained on GitHub notes that these datasets have been used in more than 25 published papers, with the caveat that plenty of lower quality work has used them too. That second half is not a knock on the site. It is what open data always produces: more scrutiny, some of it good, some of it careless, and a net gain over a closed archive nobody can question.
Set this against the paid end of the market. Most commercial online IQ products publish neither their item pool, their raw response distributions, nor their norm construction. A reviewer who wants to check whether the score you paid for means anything has nothing to work with beyond the marketing page. Open Psychometrics hands over the underlying responses and invites the audit. Whatever else is true about the Full Scale IQ Test, that decision is the more scientifically serious one, and it is worth saying so from a competitor that charges money. Openness of that kind does not by itself buy a scale, which is the lesson of the public domain ICAR item pool: its founding study reported 96,958 respondents from 199 countries, 66 percent female with a median age of 22 and 78.1 percent in the United States, an excellent basis for studying how items behave and no basis at all for turning one person's raw score into a percentile.
5 What Independent Researchers Have Found
Because the data is open, Open Psychometrics has attracted something almost no free online IQ test attracts: published external analysis. Two lines of evidence are worth knowing about, and they point in slightly different directions. Public technical evidence of that kind is one of the criteria used in a ranked comparison of the top online IQ tests, where most entrants have no external analysis to point at.
The most detailed sits on the Vocabulary IQ Test rather than the Full Scale IQ Test. Emil Kirkegaard's 2021 paper in OpenPsych analyzed responses from 9,278 native English speakers on the 45 item vocabulary instrument and reported a strong technical picture for a free web test. The site's own scoring rules, which penalize wrong answers, correlated at .92 with scores derived from item response theory, meaning the simple scoring is close to what a modern psychometric model would produce. Empirical reliability came out near .90. Most items carried substantial g loadings, averaging .59, though four items fell below .25 and would be candidates for replacement. Thirteen items showed statistically detectable sex bias, but at the level of the whole test the difference was negligible, below 0.05 standard deviations.
9,278
Native English speaking respondents in the Kirkegaard analysis of the Vocabulary IQ Test.
r = .92
Agreement between the site's own scoring rule and scores derived from item response theory.
.90
Empirical reliability estimated for the 45 item vocabulary instrument in that paper.
.59 mean g loading
Average across items, with four items below .25 flagged as weak contributors.
r = .44 and r = .40
Correlations of vocabulary scores with self reported educational attainment and with age.
r = .02
Near zero relationship between how long takers spent and what they scored, in absolute terms.
The second line concerns the Full Scale IQ Test directly. The site's results page reports that researchers administered it to 64 undergraduates alongside a Wechsler measure in a laboratory setting, finding a correlation of .57 between the two, with the Wechsler mean at 111 and the Full Scale IQ Test mean at 107. The underlying work is by Katie Logos, Neil Brewer and Robyn Young, circulated in 2021 as Convergent Validity of a Quick Online Self-Administered Measure of Verbal IQ for Psychology Researchers, and the professional comparison instrument was the Wechsler Abbreviated Scale of Intelligence, Second Edition. The authors' own framing is the honest one: the online measures work as a crude indicator suitable for screening large samples cheaply, not as a substitute for individual assessment.
A correlation of .57 in a sample of 64 undergraduates deserves careful reading rather than either dismissal or celebration. It is far above zero, so the test is measuring something real. It is well below the agreement two professional batteries show with each other, and a restricted range sample of university students tends to shrink correlations, so the true figure across a full ability range could be higher. What it clearly does not support is treating your Full Scale IQ Test number as interchangeable with a proctored result.
6 The Three Component Model and What It Licenses
The Full Scale IQ Test page attributes its short term memory, reasoning and verbal structure to work by Adam Hampshire and colleagues. The reference is Fractionating Human Intelligence, published in Neuron in 2012 by Hampshire, Highfield, Parkin and Owen, which combined a very large internet cohort completing twelve cognitive tasks with functional imaging, and argued that performance across those tasks resolves into several separable components mapping onto distinct brain networks rather than a single unitary factor.
Choosing that paper as a conceptual base is defensible and clearly labeled. It is also worth knowing that the paper's central claim was contested in print. The argument that the positive correlations among mental ability tasks could be produced by task mixing rather than by a general factor drew a published critique from Michael Ashton and colleagues, with a reply from the original authors in the same journal. This is a live methodological dispute in the field, not a settled foundation, and readers deserve to know they are standing on contested ground.
The larger point applies well beyond this one site. Citing a respected paper for your factor structure is not the same as validating your own instrument. Hampshire and colleagues built their components from their own twelve task battery in their own sample. A different set of six sections, with different items, different timing and a different population, may or may not reproduce that structure. Establishing that it does would require a factor analysis on this test's own responses, published where a reader can see it. The raw data to attempt this is available, which is more than most tests offer, but the analysis itself is not presented on the test page.
What this means when you read your result: treat the three component scores as rough indicators, not as a diagnosed profile. A gap between your reasoning and verbal numbers might reflect a genuine strength, or it might reflect the measurement error you would expect from a handful of items per component. Our page on what IQ tests measure explains why domain scores need multiple reliable indicators before a difference between them means anything.
7 The Self Selection Problem, Stated Precisely
Every online test that recruits its own takers faces the same structural issue, and Open Psychometrics faces it in an unusually visible form because it publishes the data that reveals it. The people who search for a free IQ test, click through six sections of timed problems and then volunteer for an optional research survey are not a random draw from the population. They skew toward the curious, the test comfortable and the internet literate, and by most evidence they skew upward in measured ability. Who the comparison group actually contains is the norms question behind an accurate IQ score, and it is the pillar that self recruited samples strain hardest, whatever the items themselves are worth.
That would be manageable if the shift were constant, but it is not. The composition of an internet sample changes with whatever drives traffic that month. A test featured on a large forum collects a different crowd than the same test collecting slow organic search traffic a year later. So a norm anchored to that flow is anchored to something that moves, which is precisely the problem that a properly constructed norm sample, stratified by age, education and region and collected in a defined window, exists to solve.
Then there is the second selection layer discussed earlier: only some takers complete the optional research survey that supplies the anchoring information, and there is no way to verify a single self reported proctored score inside it. Someone who took a professional test and was disappointed has less reason to type the number in. This is not an accusation against anyone; it is a known and well documented pattern in self reported ability data, and it pushes the anchor in a predictable direction.
The consequence for you as an individual is specific rather than vague. Your raw performance relative to other people who took this test is measured reasonably well. The translation of that rank into a number on the 100 and 15 scale carries an additional layer of uncertainty that no confidence interval on the results page displays. If you want to understand what a given number is supposed to correspond to in the general population, our IQ percentile chart lays out the rarity ladder: 130 sits near the top 2 percent, roughly 1 person in 44, while 145 is closer to 1 in 740.
8 Nobody Is Watching, and That Changes the Number
Controlled administration is the quiet foundation of professional testing, and it is invisible until it goes missing. In a proctored session, an examiner confirms who is sitting the test, keeps timing consistent, blocks reference materials, notes if you seemed tired or distracted and can invalidate a session that went badly. None of that exists in a browser tab, on any free site, including this one. Proctoring is a property of the venue rather than of the test, so if you need it, the places that actually supply an examiner are a psychologist's office, a university training clinic, a hospital neuropsychology department or a scheduled Mensa session, each carrying its own waitlist and price.
What that removes is not just cheating protection. It removes the record of the conditions under which your score was produced. A professional score report says the test was administered on a specific date under standard conditions by a named examiner. Your Open Psychometrics result records none of this, so even you cannot later reconstruct whether your number reflects a focused morning or a distracted evening with a phone buzzing.
The retest instruction runs into the same wall. The page says only your first attempt should count, which is psychometrically correct, since practice on the specific items inflates later attempts. But nothing prevents a second or fifth run, and there is no identity linkage to enforce it. In practice, people take these tests repeatedly, keep their best result, and quietly generalize from it. If you have done that, the number you are carrying is not a measurement of your ability; it is the top of a distribution of your attempts.
An independent analysis published at Technotheoria argues that the Full Scale IQ Test's composite behaves inconsistently, reporting cases where the overall score fell below the average of the component scores, and claiming the displayed score changes on page refresh. We have not independently reproduced those observations and present them as one reader's published analysis rather than as established fact. If accurate, they would be scoring bugs of the kind an actively maintained product fixes quickly, which brings up the maintenance question directly.
9 Free Changes the Evaluation, and Maintenance Complicates It
Let us settle the commercial question in one paragraph, because there is very little to say. As of August 2026, the Full Scale IQ Test is free. There is no checkout, no card capture, no subscription, no trial that converts, and no report gated behind a payment. The result appears at the end of the session. This is not a limited free tier with the good part sold separately; it is the whole product, given away. Any page that quotes you a price for taking a test on openpsychometrics.org is not describing this site. Free is only the bottom rung of a very long ladder: what IQ testing actually costs spans roughly two hundred fold from the cheapest paid option up to a full neuropsychological evaluation, and nearly all of that spread buys supervision and documentation rather than better puzzles.
That removes an entire category of consumer risk. Nobody has been charged for a renewal they forgot about, nobody has hunted for a cancellation link, and nobody has invested twenty minutes only to hit a paywall at the reveal. On this dimension the site is beyond criticism, and it is fair to note that a meaningful share of paid online IQ products would score badly on exactly the terms Open Psychometrics makes irrelevant. Our roundup of free IQ tests covers where those tradeoffs usually land.
The cost of a free non commercial project appears elsewhere: in maintenance. The Full Scale IQ Test page has long carried a substantive update date of December 30, 2019. For a research instrument that is not fatal, since a stable form is easier to study than a moving target. It does mean the norms have not been visibly refreshed in years while the population of internet test takers has continued to change.
The broader maintenance question is worth stating plainly rather than through any single incident. A volunteer project has no support rota, no release schedule and no obligation to fix anything on a timetable. Instruments can sit unchanged for years, scoring quirks can persist, and there is no service commitment behind any of it. That is the honest trade for free access and open data, and it is the reason a research tool should not be the instrument you rely on when a result actually matters to you.
10 Privacy, Data and What You Are Agreeing To
Open Psychometrics is upfront that taking its tests means contributing data. The test page states that responses may be recorded and used anonymously for research or otherwise distributed, and the archive described earlier is where that distribution happens. This is the honest version of a bargain that many free tests make silently, and the fact that you can go and look at the resulting files is the proof that the disclosure is real.
Read the consequence carefully rather than fearfully. Your answers may end up in a public dataset that anyone can download. The published files are anonymized and the site removes records where takers did not confirm their responses were suitable for research. What typically travels with a record is item level responses, timing and whatever demographic fields you volunteered, which is exactly the material a researcher needs and exactly the material you should decide about consciously.
If you are comfortable with that, the arrangement is arguably better than the alternative on offer elsewhere. Commercial testing sites also retain your responses. The difference is that they do not tell you what happens next, do not publish the files, and in some cases attach the data to an email address you handed over to unlock a result. A public anonymized dataset with a stated consent process is a more legible deal than a private one with no stated process at all.
If you are not comfortable with it, the answer is simple: do not take the test, because contributing data is the point of the project rather than a side effect. That is a coherent position and the site's design does not try to talk you out of it. It states what happens and lets you decide, which is the standard any test, free or paid, should be held to.
11 Strengths and Limitations Side by Side
Rather than a numeric grade, here is the balance sheet with each item traceable to something you can check on the provider's own pages or in the published literature. Read the left column before the right one, because the strengths are more unusual than the limitations.
Dimension
What Open Psychometrics does well
Where it falls short
Cost and commercial pressure
Completely free with no payment step, no subscription and no gated report. The full result appears at the end of the session.
No funding model means no support channel, no score interpretation help and no service commitment of any kind.
Transparency of data
Raw anonymized responses published with codebooks and item text, mirrored on GitHub and Kaggle, with a documented consent process.
The published archive does not include a factor analysis or reliability table for this specific six section form.
Scoring method
The anchoring logic is disclosed in plain language, including the admission that online takers are not the general population.
The anchor depends on unverified self reported proctored scores from an optional survey completed by a self selected subset.
External validation
Unusually for a free test, outside researchers have published on it, including a laboratory comparison against a Wechsler measure.
That comparison used 64 undergraduates and reported a correlation well below what two professional batteries show with each other.
Honesty about limits
The page states directly that the result does not replace a real IQ test and should not drive decisions, in blunter language than most competitors use.
The name Full Scale IQ Test borrows a term from professional batteries whose meaning does not transfer to a twenty minute form.
Precision and coverage
Three component scores give more texture than a single number from a matrix only quiz.
No confidence intervals are published, several broad abilities are not measured, and the page's substantive update date sits in 2019.
Notice the pattern in that table. The failures are the ordinary limits of a short unproctored instrument built by one person as a research tool, which is what the site says it is. The successes are things a well funded paid competitor could copy tomorrow and mostly does not.
12 Verdict, and Where ACIS Sits
Use the Full Scale IQ Test at Open Psychometrics when you want a free, quick, reasonably structured look at your own performance, when you are teaching or learning about how cognitive tests are built, or when you want to contribute to a public dataset. On those uses it is one of the better free options on the internet, and the open archive makes it the one we would point a student toward first. For the paid end of the category, our reviews of the BMI Certified IQ Test and 123test show what commercial products add, and what they still leave undocumented.
Do not use it as evidence for anything consequential. Not for a diagnosis, not for gifted placement, not for accommodations, not for a job application, and not as a submission to any organization that sets an ability threshold, because those bodies define their own accepted instruments and an unproctored web test is essentially never among them. The site tells you this itself, and it is right. If a number is going to affect a real decision, the path runs through a qualified professional administering an accepted battery under controlled conditions.
ACIS occupies the middle of that range and we should be clear about our own limits. ACIS is a paid online self assessment, longer than the Full Scale IQ Test, structured around six CHC domains across twenty subtests, and it reports a profile with intervals rather than one bare number. It is also self administered and unproctored, so it is not clinical, not diagnostic and not a replacement for professional evaluation either. Where we think we improve on the free option is breadth of coverage, reporting that shows uncertainty instead of hiding it, and documentation you can read before deciding whether to trust the result. Where Open Psychometrics beats us outright is on publishing raw response data for anyone to reanalyze, and we would rather say that than pretend otherwise. Our methodology and score construction are laid out in our published technical documentation, and if you are comparing options across the market our guide to the best IQ tests sets out the criteria we think matter.
No. As of August 2026 there is no payment step anywhere in the flow and no part of the result is withheld pending a purchase.
Who is behind the site?
Its about page names Eric Jorgenson and dates the project to late 2011. It presents itself as an educational and research effort, not as a company or a clinical service.
How long does the Full Scale IQ Test take?
The page puts the typical session at 10 to 15 minutes and the design at around 20, and cautions that finishing in under ten minutes tends to hurt your outcome.
What are the six sections?
Each section carries its own instructions and feeds three reported components: short term memory, reasoning and verbal, alongside a single overall figure.
Where does the number 100 come from in the result?
Raw performance is rescaled to a distribution centered on 100 with a spread of 15 points, the convention modern batteries share.
How does the site handle the fact that its takers are unusual?
It asks an optional end of test survey for scores people received on proctored tests and entrance exams, then uses those figures to position the online distribution.
Is that anchoring approach sound?
It is a reasonable improvisation and it depends on unverified self reports from people who chose to answer. Disclosed and imperfect beats undisclosed, but it is not a stratified norm sample.
Can I download the response data?
Yes. An archive of anonymized datasets with codebooks sits at openpsychometrics.org/_rawdata/, and mirrors exist on GitHub and Kaggle.
Has academic work used those datasets?
A GitHub mirror of the archive states they have appeared in more than 25 published papers, and notes that weaker work has drawn on them too.
What did Kirkegaard find about the vocabulary instrument?
His 2021 OpenPsych analysis of 9,278 native English speakers put empirical reliability near .90 and found the site's scoring rule tracked item response theory scores at .92.
Has anyone compared the IQ test to a professional battery?
Work by Katie Logos, Neil Brewer and Robyn Young in 2021 ran it against the Wechsler Abbreviated Scale of Intelligence, Second Edition with undergraduate participants.
What agreement did that comparison show?
The results page cites a figure of .57 across 64 students, with the Wechsler average at 111 against 107 online. The authors described the web measures as crude screening tools.
Why does a correlation like that matter?
It confirms the test tracks something genuine while ruling out treating the two results as interchangeable. University samples also compress correlations by narrowing the ability range.
Whose research model do the three components come from?
The page points to Hampshire, Highfield, Parkin and Owen, whose 2012 Neuron paper argued that performance across twelve tasks separates into distinct components tied to different brain networks.
Is that model universally accepted?
No. Ashton and colleagues published a methodological objection, and the original authors replied in the same journal. The debate over whether those results argue against a general factor stayed open.
Can I trust a gap between my component scores?
Treat it as a hypothesis. With few items behind each component and no published error estimate, an apparent difference can easily be noise rather than a stable pattern.
Does the result come with a margin of error?
No band or interval accompanies the figure, so a single integer carries an air of exactness that a short unsupervised session cannot back up.
Should I retake it?
The site says only your opening attempt counts, and that guidance is correct. Later runs benefit from familiarity with the specific problems, so keeping your highest number misleads you.
Will a school or employer accept this?
Assume not. Institutions specify which instruments they recognize and require supervised administration, conditions an open web form cannot satisfy.
What happens to my answers?
They may be logged and released anonymously through the public archive. Takers are told this at the outset and asked afterward to confirm their answers were genuine.
Was there anything odd about the site when this page was written?
Treat it as a volunteer project with no support commitment. Pages and instruments can go long stretches without maintenance, which is a normal trade for free access rather than a red flag about the research itself.
14 Sources Behind This Page
A review is only as good as the yardstick behind it. These sources document the measurement standards this review applies and the consumer practices it checks for.
American Psychological Association. The Standards for Educational and Psychological Testing, the joint AERA, APA and NCME framework that legitimate tests are built and evaluated against.
Buros Center for Testing. The independent center whose Mental Measurements Yearbook has reviewed commercial tests since 1938. The reference point for claims about test quality.
American Psychological Association. Understanding psychological testing and assessment. What professionally administered testing involves and how it differs from informal quizzes.
Federal Trade Commission (2022). Bringing Dark Patterns to Light. The FTC staff report on design practices that obscure subscriptions and charges, worth checking any paid online test against.
Pearson Clinical Assessment Scientific Council (2023). Standardized Clinical Assessment for Practitioners: A Primer. How standard scores, percentile ranks and the standard error of measurement are meant to be read together.
Pearson (2008). WAIS-IV Score Report sample. What a real report contains: every composite paired with a percentile rank, a 95% confidence interval and a qualitative description, never a bare number.
Crawford, J.R., Garthwaite, P.H. & Slick, D.J. (2009). On percentile norms in neuropsychology. The Clinical Neuropsychologist, 23(7), 1173-1195. Three definitions of a percentile coexist in practice and can return different results for the same score.
Voncken, L., Albers, C.J. & Timmerman, M.E. (2019). Improving confidence intervals for normed test scores. Behavior Research Methods. Open access. Documents the mean 100, SD 15 metric and the uncertainty that norming from samples adds to any score.
Pearson (2024). WAIS-5, Wechsler Adult Intelligence Scale, Fifth Edition. The current adult battery, covering ages 16:0 to 90:11 across five cognitive domains.
Take the assessment
You get a profile, not a number
ACIS measures six CHC domains across 20 subtests and reports each one with its own normed score and confidence interval, so you can see where you are strong and where you are not.
