The critical thinking test: what Watson-Glaser and Halpern really score

0 The short answer

A critical thinking test measures how well you can judge arguments and evidence on the day you sit it: the Watson-Glaser tradition scores five operations (drawing inferences, recognizing assumptions, deducing, interpreting and evaluating arguments), and the Halpern assessment scores reasoning on everyday scenarios. Those scores overlap substantially with general reasoning ability but are not the same thing as the habit of reasoning carefully, which the research treats as partly separate. An adult usually meets such a test through an employer, a college or a program, not as a public product. ACIS is not a critical thinking test and does not measure thinking dispositions or argument evaluation: it is one option for seeing the reasoning abilities around critical thinking on adult norms, at 15, 30 or 50 dollars for about 45, 110 or 175 minutes (prices read October 29, 2026).

1 What Is a Critical Thinking Test?

A critical thinking test is a performance test of reasoning about written arguments, evidence and everyday decisions, and the published ones come from three traditions that measure different things under one name. The first is the Watson-Glaser tradition: timed, mostly multiple choice items built on passages and scenarios, scored for skills such as inference and evaluating arguments. The second is the Halpern tradition: everyday scenarios that ask for a written answer and then for a choice among alternatives. The third is the disposition inventory, a questionnaire about how a person says they approach problems, which scores attitudes and not performance. Searches for a "critical thinking test", "critical thinking test questions" or "critical thinking test examples" usually mean the first.

The conceptual question, whether critical thinking is just intelligence under another name, has its own page on how critical thinking differs from intelligence, and the closely related Cognitive Reflection Test is covered on the page about the Cognitive Reflection Test. This page is about the tests: what they contain, how they behave as instruments and who can sit them.

A short definition from the research helps. The ETS review by Liu, Frankel and Roohr (2014), which surveyed the assessments in use in higher education, notes that researchers debate critical thinking's definition, its amenability to assessment, its degree of generality or specificity, and the evidence of its practical impact. That is a candid starting point. When a test publisher says its instrument "measures critical thinking", it is describing a chosen definition, and the definitions differ. Halpern's, as the ETS review reports it, treats critical thinking as a tool for decision making and problem solving, and names verbal reasoning, argument analysis, assessing likelihood and uncertainty, making sound decisions and thinking as hypothesis testing as its key aspects. The Watson-Glaser family names inference, assumptions, deduction, interpretation and evaluation. Both are reasonable, and both are narrower than the everyday idea of thinking well.

2 Which Critical Thinking Tests Can an Adult Actually Take?

Most published critical thinking tests are sold to organizations, not to individuals, so an adult usually reaches one through an employer, a college or a program, and the test an adult can take on their own is often not the one named in the search. The table sets the real options side by side. It holds only what we could read on a publisher page or in a paper on October 29, 2026, and where the source is silent the cell says so, because a guessed time or cost would be worse than a blank.

Three points follow from the table. First, the Watson-Glaser page we read does not say that individuals can buy the test, and it directs buyers to sales; it links administration to a British Psychological Society training course. A candidate therefore meets it by invitation. The page does say a free practice test exists and states that it "is not a live test", meant to show how the real test works. Second, the Insight Assessment page frames the CCTST around admissions, advising, retention and accreditation in higher education, which is why a college reader will meet it; the page on IQ tests for college students covers that setting from the ability side. Third, the Halpern listing carries a warning that matters to anyone searching for it: the Buros Center for Testing page for the assessment, which we read on October 29, 2026, states that in September 2026 the test author advised Buros that the test is out of print. We could not confirm from the publisher whether it is still sold, so a reader who needs it should ask Schuhfried directly.

If your real question is "where can I take an IQ style test myself", the page on where to take an IQ test compares every route, and the page on cognitive ability tests explains how a normed ability test works. Neither is a critical thinking test, and ACIS makes no claim that it is.

3 What Does the Watson-Glaser Test Measure?

The Watson-Glaser measures five operations on written material, drawing inferences, recognizing assumptions, deducing, interpreting and evaluating arguments, and Pearson's own summary frames its scoring through a three part model it calls RED. Bernard and colleagues (2008) describe it as the oldest and among the most widely used and studied critical thinking measures, constructed around those five subscales. The TalentLens product page, which we read on October 29, 2026, calls the five subtests "the building blocks of critical thinking and reasoning" and describes them in these terms: rating the probability that inferences are true given the information, identifying unstated assumptions behind statements, judging whether conclusions follow logically, weighing evidence to decide whether generalizations are warranted, and assessing the strength and relevance of arguments on an issue. It lists 40 items and a time of 30 to 40 minutes, timed.

Pearson's own efficacy summary, read on October 29, 2026, frames the same instrument more compactly. It says the assessment gauges a person's ability to analyze written information, interpret it and draw logical conclusions, and that its structure and scoring were informed by the RED model: recognize assumptions, evaluate arguments, draw conclusions. It adds that the test presents passages or scenarios, each with a number of items, that administration is online or on paper, that a 30 minute limit applies, and that an untimed version exists in US English for development or reasonable accommodations. Besides an overall score, it says, each test taker gets a development report that breaks the result down across the three areas.

Two things make the numbers above less tidy than they look. First, the sources do not agree on length, because the test exists in several versions. The ETS review (2014) tabulates a standard form of 80 items taking 40 to 60 minutes if timed, a short form of 40 items in 30 minutes, and a Watson-Glaser II of 40 items in 40 minutes, and reports that for the older forms only the total score was reported although there are five tests. The TalentLens page, the efficacy summary and the ETS table are three dated snapshots of a product that has been revised, so a candidate should treat the invitation from the employer, not a web page, as the statement of what they will sit. Second, the ETS review says each test contains neutral and controversial reading passages and scenarios met at work, in the classroom and in the media. That is a prose heavy format, so reading comprehension and acquired language knowledge plausibly load into the score. This is our reading, not a claim of the publisher: in CHC terms the operations resemble deductive and inductive reasoning (Gf) applied to text, which draws on comprehension-knowledge (Gc) as well.

Item level detail, the exact sections and how the score is reported are the job of the sibling page on the Watson-Glaser test, and this page stays at the level of what the instrument is for. We reproduce no items and link to no site that does.

4 What Does the Halpern Critical Thinking Assessment Measure?

The Halpern Critical Thinking Assessment scores five critical thinking subskills on 25 everyday scenarios, and what sets it apart from the Watson-Glaser tradition is that one form asks the person to write an answer before choosing from alternatives. The ETS review (2014) tabulates it as follows: 25 scenarios of everyday events with five per subcategory, computer based, with forced choice items (multiple choice, ranking or rating of alternatives) and open ended items. Form S1 has both kinds and takes 60 to 80 minutes, although the test is untimed; Form S2 is all forced choice and takes 20 minutes, also untimed. The five subskills are verbal reasoning, argument and analysis, thinking as hypothesis testing, using likelihood and uncertainty, and decision making and problem solving, as the review cites them from the Halpern (2010) manual published by Schuhfried.

The design choice has a reason and a cost. Open ended answers sit closer to real life, where nobody hands you four options. The ETS review notes the other side: constructed response items tend to have lower reliability than multiple choice items for the same testing time, and, as an example from a different test, it quotes a reliability of .43 for the 60 minute constructed response section of the CLA+ against .87 at the test level, which is driven by its 30 minute multiple choice section. That example is not about the Halpern assessment, and we do not read it as a verdict on it. It does show why a format that feels more authentic is not automatically a better measure.

What does the evidence say about outcomes? Butler (2012) gave the assessment to community adults (n = 50), state university students (n = 48) and community college students (n = 35) in the United States, together with a behavioral inventory of life events. Those with higher critical thinking scores reported fewer negative life events, r(131) = negative 0.38. The ETS review adds the limit that matters: the Butler study did not control for any measure of general cognitive ability, which leaves room for an alternative explanation that other aspects of general cognitive ability, rather than critical thinking, contributed. The outcome is also self-reported, a point the page on critical thinking versus intelligence already makes for the follow up study.

A study of the Dutch version gave a more muted picture. De Bie, Wilhelm and van der Meij (2015) studied it with 240 university students in communication science and psychology and a Dutch inventory of negative life events. They report adequate reliability (alpha .75, lambda 2 .77), a factor analysis that supports using both the constructed response and forced choice formats, each with five subscales, and a total score not significantly related to the life events inventory (r = negative 0.11). The two studies used different samples, languages and instruments, so they do not contradict each other so much as show how fragile a single validity coefficient is. One more fact belongs here: the Buros listing for the assessment says that in September 2026 the author reported it out of print, so its use today is mostly in the research literature. That last clause is our inference from the listing, not a statement of the publisher.

5 How Reliable and Valid Are Critical Thinking Tests?

Critical thinking tests produce usable total scores, but their subscores are weaker than their totals, their validity evidence varies widely by instrument, and studies by the test authors and their affiliates tend to report better psychometric quality than independent ones. Begin with the publisher's own claims, labeled as such. Pearson's efficacy summary reports positive correlations of 0.16 to 0.58 between Watson-Glaser scores and a number of on the job performance indicators, correlations of 0.38 to 0.62 with final course grades of college students, retest correlations of 0.73 to 0.89 across occasions, an internal consistency of 0.83 and correlations of 0.82 to 0.88 between forms. Those are ranges across studies, and a range from 0.16 to 0.58 says that the relationship in some settings is weak and in others moderate. For the general question of how ability relates to work performance, the page on IQ and job performance gives the meta-analytic picture.

Independent reviews are more guarded. Liu, Frankel and Roohr (2014) conclude that common problems with existing critical thinking assessments include insufficient evidence of distinct dimensionality, unreliable subscores, noncomparable test forms and unclear evidence of differential validity across groups. They report, citing Leppa (1997), internal consistency of only .21 to .51 for the CCTST subscales, and, citing Ku (2009), that studies by researchers not affiliated with the authors of a test tend to report lower psychometric quality than studies by the authors and their affiliates. We did not open Leppa or Ku, and we repeat them only as the ETS review reports them. For the Watson-Glaser, the study listed in the ETS table is Ejiogu and colleagues (2006), 84 analysts in a government agency, with corrected correlations of .32 to .52 with supervisory ratings of job performance behaviors.

Two papers go to the structure of the Watson-Glaser. Bernard and colleagues (2008) meta-analyzed 13 sets of subscale intercorrelations from published studies and ran a principal components analysis on 60 sets of subscale means; each of two versions gave a one factor solution, accounting for 82.69 and 79.55 percent of the variance. They conclude that the test should be viewed as a measure of general competency and that the subscales should not be interpreted individually. Possin (2014) argues in Informal Logic that the test has serious construct validity problems stemming from ambiguous and misleading instructions that have remained unaltered for decades, and that erroneously scored items further diminish validity, so that enhanced knowledge of formal and informal logic could result in lower scores. That is one critic's argument, and the abstract we read does not name the edition he reviewed, so it should not be taken as a statement about every current version. Read together with the publisher's figures, it is a reason to ask for the technical documentation of the exact version in use.

The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014), which we read in the open access edition, speak to the subscore problem directly. Standard 1.14 says that when interpretation of subscores, score differences or profiles is suggested, the rationale and relevant evidence should be provided, and its comment says that when a test provides more than one score, the distinctiveness and reliability of the separate scores should be demonstrated. Standard 2.3 adds that reliability estimates should be reported for each total score, subscore or combination to be interpreted, and its comment warns that total scores may be acceptably consistent while subscores have unacceptably low reliability. A candidate who is shown a profile of five or three skill scores should treat the total as the more dependable number. The page on reliability and validity explains in plain terms what those words promise.

6 How Much Does a Critical Thinking Test Overlap With Intelligence?

Critical thinking test scores overlap substantially with measures of general cognitive ability, but no study we read shows that they are the same thing, and the strongest evidence for separateness concerns dispositions, not skills. The most direct numbers come from the validity table in the ETS review, which lists a study of undergraduate educational psychology students (198 students, with 155 in the SAT correlations) that it attributes to Watson and Glaser (2010). It reports moderate significant correlations of the Watson-Glaser with SAT Verbal (r = .43) and SAT Math (r = .39), with grade point average (r = .30) and with the Ennis-Weir essay test (r = .37). College admission test scores are a related measure and not an IQ, so the figures show that the test shares variance with scholastic reasoning ability, not what share of it is intelligence.

Structure points the same way. Bernard and colleagues found that the Watson-Glaser subscales behave as one general factor, which fits the idea that the test mostly samples a general competency. Ren, Tong, Peng and Wang (2020) tested the question head on in Intelligence. With 232 university adults in the first study and 158 primary school children in the second, they measured critical thinking skills and dispositions alongside general cognitive ability, including fluid intelligence, working memory and processing speed. Both critical thinking skills and dispositions made unique contributions to academic performance with general cognitive ability controlled. But it was mainly the dispositions that contributed uniquely, while the contribution of the skills largely overlapped with general cognitive ability. The adult study used a self-report measure of academic performance, which limits what it can show.

A different emphasis comes from Butler, Pentoney and Bong (2017), who gave community adults and college students (n = 244) a critical thinking assessment, an intelligence test and an inventory of life events. People with higher critical thinking scores and higher IQs reported fewer negative life events, critical thinking more strongly predicted life events than intelligence did, and it significantly added to the variance explained by IQ. Set beside Ren and colleagues, the pattern is not a contradiction: the studies measured different outcomes with different instruments (self-reported academic performance in one, a life events inventory in the other), and both point to the same conclusion at a higher level, that whatever critical thinking tests capture is not fully contained in an IQ score.

The sensible reading is a layered one. A large part of what the skills tests score is reasoning ability of the kind that broad cognitive batteries sample: fluid reasoning for novel problems, comprehension-knowledge for language. The page on what IQ measures sets out those abilities, and the page on CHC cognitive domains maps them. What the batteries do not sample is the disposition to use them, which is the next section.

7 Why Are Thinking Dispositions and Biases Partly Separate From Ability?

Thinking dispositions and many thinking biases are partly separate from cognitive ability because the research shows that a person can be able to reason well and still not do it, which is why a skills test cannot tell you how someone will actually think. The key paper is Stanovich and West (2008) in the Journal of Personality and Social Psychology. In 7 studies the authors observed that a large number of thinking biases were uncorrelated with cognitive ability, among them the conjunction effect, framing effects, anchoring effects, outcome bias, base-rate neglect, myside bias and the sunk-cost effect. In a further experiment they showed that cognitive ability does correlate with avoiding some biases, specifically denominator neglect, probability matching rather than maximizing, belief bias and matching bias on the four-card selection task, and they present a framework for predicting when ability will and will not correlate with a rational thinking tendency. The result is not that ability is irrelevant. It is that the relationship depends on the kind of bias.

A second paper bears on the tests themselves. West, Toplak and Stanovich (2008) argued in the Journal of Educational Psychology that the biases studied in the heuristics and biases literature are legitimately aspects of critical thinking. In a study of 793 students they found that the ability to avoid these biases was moderately correlated with a more traditional laboratory measure of critical thinking, the ability to reason logically when logic conflicts with prior belief, and that the correlation remained statistically significant after the variance due to cognitive ability was partialed out. Measures of thinking dispositions, actively open-minded thinking and need for cognition, predicted unique variance in both classes of critical thinking skill after general cognitive ability was controlled. That is a measurement result with a practical reading: the skills a Watson-Glaser style test samples and the tendencies a disposition inventory samples are not interchangeable, and neither is a stand-in for IQ.

The Cognitive Reflection Test sits in the same literature. Toplak, West and Stanovich (2011) showed in Memory & Cognition that it is a more potent predictor of performance on a wide sample of heuristics-and-biases tasks than measures of cognitive ability, thinking dispositions and executive functioning, that it has a substantial correlation with cognitive ability, and that it remained a unique predictor after the other measures were controlled. The authors conjecture that this is because neither intelligence tests nor executive function measures assess the tendency toward miserly processing in the way it does. We do not reproduce its items, and the page on the Cognitive Reflection Test explains what it is and how to read a result.

Disposition instruments exist, and it is worth knowing what they are. The ETS review lists the California Critical Thinking Disposition Inventory as a 75 item Likert scale taking 30 minutes, with seven scales: truth-seeking, open-mindedness, analyticity, systematicity, confidence in reasoning, inquisitiveness and maturity of judgment. It scores agreement with statements, which makes it a measure of how a person describes their approach, not of how they perform. A related but different topic, whether traits such as emotional intelligence add anything beyond cognitive ability, has its own page on emotional intelligence versus IQ.

The consequence for a test taker is concrete. A strong critical thinking score says that you can do the reasoning the test samples under test conditions. It does not say that you do it when the conclusion flatters you, when you are tired or when nobody is asking. A weak score does not say that you reason badly in daily life. ACIS does not measure thinking dispositions, and it does not run a bias battery, so it cannot speak to this side at all.

8 Can You Practice for a Critical Thinking Test?

Practice and retesting raise scores on cognitive ability tests in general, so familiarity with the format is a legitimate and modest advantage, but the honest sources of practice are the publisher's own non-live material and the format description, not a purchased stack of questions. The strongest evidence is a meta-analysis. Hausknecht, Halpert, Di Paolo and Moriarty Gerrard (2007) summarized 50 studies of practice effects for cognitive ability tests, covering 107 samples and 134,436 participants, and found an adjusted overall effect size of .26. Effects were larger when practice was accompanied by test coaching and when identical forms were used. The abstract notes that as many as 25 to 50 percent of applicants in organizational and educational settings are retested. This is a result about cognitive ability tests as a group, and we have not found a meta-analysis specific to critical thinking tests, so it should be read as an indication and not a figure for the Watson-Glaser.

What does a candidate actually have? The TalentLens page says a free practice test exists and states that it is not a live test, meant to show how the real test works. The page's FAQ says randomly generated tests are built to match in item count, difficulty and content themes, which, if accurate, means that two sittings of the same test are not the same set of items. We cannot verify how that works in practice, and we did not open the practice test, so we describe it and nothing more. The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014) support the underlying principle. Standard 8.2 says test takers should be provided in advance with as much information about the test, the testing process, the intended test use, the scoring criteria and the testing policy as is consistent with obtaining valid responses, and its comment says they should usually be told about content, item formats and imposed time limits, with general advice on test-taking strategies. A candidate is entitled to ask the invitation's sender what the test is, how long it takes and how the result will be used.

Commercial practice sets are a separate matter. Sellers of practice questions are not sources for this page, we link none of them, and we have not evaluated them. A set of practice items cannot be checked against the live test, so what a buyer gets is format familiarity at best. The page on pre-employment cognitive tests covers the employer side of this category, and its subject is hiring tests as a group, which is a different use again from the one this page is about.

The practical reading of the evidence is modest. Know the five operations. Practice reading short arguments slowly and asking what is stated, what is assumed and what follows. Expect a time limit. And do not read a better second score as a changed mind: a retest gain of the size the meta-analysis reports can reflect familiarity with the format.

ACIS is not preparation for any critical thinking test, and an ACIS report does not predict how anyone will do on one. We have no evidence on either point and make no claim.

9 What Does ACIS Measure Around Critical Thinking, and What Does It Not?

ACIS does not measure critical thinking dispositions or argument evaluation, but it does measure the reasoning abilities that those tests draw on, through Logic Grid, Similarities, Complex Relations and Matrix Reasoning, reported as scaled scores and index scores on adult norms. Four subtests are relevant, and each is described here only as its page describes it.

• Logic Grid measures controlled deduction under multiple constraints: the person keeps several rules active, eliminates impossible combinations and preserves the implications of each clue. Its page places it in fluid reasoning (Gf). It is in the Optimized and Full Scale forms, not in Quick. The technical manual says a low Logic Grid score with weak working memory may reflect difficulty maintaining rules, which is a reminder that deduction under constraint also leans on holding information in mind. The page on logic IQ tests treats deduction as one of three fluid reasoning abilities.
• Similarities measures verbal concept formation: identifying the shared concept, class or principle behind two things that differ on the surface. Its page places it in comprehension-knowledge (Gc). It is in all three forms. The page on the verbal IQ test sets out what a verbal score measures.
• Complex Relations measures relational integration: how several elements correspond, change or combine inside one problem. Its page places it in fluid reasoning (Gf). It is in the Full Scale form only.
• Matrix Reasoning measures rule discovery in unfamiliar visual patterns: inspecting relationships, inferring transformations and selecting the option that completes the structure. Its page places it in fluid reasoning (Gf). It is in every form.

The technical manual describes the verbal reasoning composite, VRI, as one that "focuses on verbal reasoning through Similarities and Complex Relations", which it says identifies relations among concepts and abstracts shared meaning. Composites are available when their subtests are available, so that view needs the Full Scale form. The sibling pages on deductive reasoning tests, verbal reasoning tests and problem solving tests take up the same ability families from the side of the tests that use those names.

What ACIS does not do is just as important. It does not present arguments and ask whether an assumption is hidden or an inference is warranted. It does not score the habit of checking a conclusion you like, it does not run a bias battery and it does not include a disposition inventory. A person who needs a critical thinking score for a course, a program or a job needs that program's test. What ACIS offers this reader is self-knowledge about the abilities around it: whether fluid reasoning, verbal reasoning and working memory are even, or whether one stands out. The page on differences between index scores explains why gaps between indices are common and how to read them.

10 How Do You Read a Critical Thinking Score?

A critical thinking score is a comparison with a stated group on one version of one test, so before you read the number, ask four questions: which norm group, which version, total or subscore, and under what conditions. The scales differ by publisher, and none of them is an IQ scale. The table lists what the sources we opened state, and where they are silent the cell says so.

The first question is the group. A percentile or a band says where you stand relative to whoever was in the comparison sample, and the Watson-Glaser page itself says that many norm groups are available, which means two organizations can read the same raw performance against different groups. The page on how IQ scores are normed explains in general terms how a norm group turns raw performance into a standing. The second question is the version. As shown above, the Watson-Glaser exists in forms of different length and different reporting, and a number from one version is not automatically comparable with another.

The third is whether you are reading a total or a subscore. The section on reliability above covered the Standards on this point, and the practical rule is to put more weight on the total. The fourth is the margin. Every score is an estimate of performance on a day, and the Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014) say in Standard 2.14 that where cut scores are specified for selection or classification, the standard errors of measurement should be reported in the vicinity of each cut score. For a candidate that is the sentence to remember: a score just under a cut is not a precise finding. Standards 1.1 and 1.4 add that validity belongs to a particular interpretation and use, and that a user who adopts an interpretation that has not been validated must justify it. A critical thinking score validated for screening applicants is not thereby a statement about your character, your intelligence or how you will think next year.

A percentile and a standard score are different kinds of number. A percentile tells position, a standard score tells distance in standard deviation units, and the two do not stretch alike, as the page on IQ score versus percentile explains. That is why a critical thinking band or percentile should never be translated into an IQ. No source we read provides such a translation, and we do not offer one.

11 What Does an Unsupervised Online Session Need?

An unsupervised online session hands to the test taker every condition that a proctor normally controls, so the reader has to supply them: a quiet room, the time the form needs, working sound where a subtest uses audio, and honest, unaided effort. The TalentLens page says the Watson-Glaser is suitable for unproctored screening and for online completion in a supervised environment, so unproctored critical thinking tests are a normal part of recruitment. That does not make their conditions equivalent to a supervised room. The Standards for Educational and Psychological Testing say that unproctored administrations, where standardization cannot be ensured, could provide an advantage to some test takers over others and that in such situations results should be interpreted with caution, and a comment elsewhere in the Standards says the interpreter of results should be informed if a test was unproctored.

For ACIS the practical conditions are concrete. Three subtests present their items as audio and need working sound: Digit Span, Alphanumeric Sequencing and Arithmetic. Layer Rotation, according to the technical manual, carries no words at all, although the test as a whole is English only and the manual says verbal and knowledge tasks are affected by English proficiency and educational exposure. The three forms take about 45, 110 and 175 minutes. Breaks are allowed, progress is saved between subtests and a form can be completed over 30 days, so the long forms do not have to be taken in one sitting. Working alone matters: a search engine, a calculator or a friend reading along changes what is measured, and the report then describes the helped performance. The free trial, which needs no card and is limited to one per person, shows the format before anything is spent.

12 Where ACIS Sits, and What a Buyer Should Do

ACIS is a normed online battery for adults who want to see the reasoning abilities around critical thinking on adult norms, and it is not a critical thinking test: it does not measure thinking dispositions or argument evaluation, and it is not a tool for hiring, selection or any employer test. The strengths are concrete. It offers a Full Scale IQ and six primary indices, with percentiles and a 95 percent confidence interval, across six CHC domains in its Full Scale form, with adult norms covering ages 16 to 90. Its documentation is in the technical manual, and this page states nothing about it beyond what the reports and the manual describe.

The limits are as concrete. ACIS is online and unsupervised, it is not a clinical or diagnostic instrument and it is English only. On its research page the site says ACIS "is not intended for clinical diagnosis, hiring decisions, accommodation determinations or admission to high IQ societies." ACIS does not prepare anyone for a critical thinking test, does not predict a result on one, does not convert to one and is not accepted by any employer or program as a substitute. A reader who is invited to take the Watson-Glaser or the CCTST takes that test, as invited, and nothing here changes that.

The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014), published jointly by the American Educational Research Association, the American Psychological Association and the National Council on Measurement in Education, are the yardstick this page uses and not an endorsement of ACIS. Standard 1.1 asks that the intended interpretation of scores and the construct be set forth clearly, Standard 1.4 puts the burden on a user who adopts an unvalidated interpretation, Standard 1.14 asks for evidence before subscores or profiles are interpreted, and Standard 8.2 asks that test takers be given as much advance information as is consistent with valid responses. Read that way, a score from ACIS answers one question, where your reasoning abilities stand against adults, and does not answer the others on this page, such as how you handle an argument that flatters you. For anything that needs documentation, such as an accommodation or a clinical question, a licensed psychologist is the route.

A buyer should do this, in order:

1. Decide the question. If it is a critical thinking score for a job or a program, wait for the invitation and ask what version, how long and how it will be used. If it is curiosity about your own reasoning abilities, continue.
2. Pick the form that fits. Quick is 15 dollars for 6 subtests in 3 domains, about 45 minutes, and includes Similarities and Matrix Reasoning. Optimized is 30 dollars for 13 subtests in 5 domains, about 110 minutes, and adds Logic Grid among others. Full Scale is 50 dollars for all 20 subtests in six domains, about 175 minutes, and adds Complex Relations among others. Prices were read on October 29, 2026 and can change.
3. Start with the free trial. It needs no card, and the page on free IQ tests explains what a free route includes and leaves out. The refund policy, quoted literally, is: "Full refund within 5 days if the report does not deliver the features described at checkout or a technical issue prevents you from accessing your results." That is a conditional guarantee, not an unconditional money back one.
4. Set up the room. Quiet, working sound, a rested start and a plan for breaks.
5. Read the report as one source. Keep the confidence interval in view, look at the index profile and not only the total, and read the page on IQ tests for adults if you want the wider category explained.

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 29, 2026, prices can change, and any figure we derived ourselves is labeled in the text.

• TalentLens (Pearson). Watson-Glaser Critical Thinking Appraisal. Product page: five subtests, 40 items, timing, norm groups, administration and practice test. Read October 29, 2026.
• Pearson. Watson-Glaser Critical Thinking Appraisal, Efficacy Report Summary. One page PDF: RED model, time limit, reported validity and reliability ranges. Read October 29, 2026.
• Insight Assessment. California Critical Thinking Skills Test. Product page: eight skills, 40 questions, 55 minutes, scoring and purchase. Read October 29, 2026.
• Buros Center for Testing. Halpern Critical Thinking Assessment, test listing. Publisher, year and the out of print notice. Read October 29, 2026.
• Liu O, Frankel L, Roohr K. Assessing Critical Thinking in Higher Education: Current State and Directions for Next-Generation Assessment. ETS Research Report Series, 2014, volume 2014, issue 1, pages 1 to 23.
• Bernard R, Zhang D, Abrami P, Sicoly F, Borokhovski E, Surkes M. Exploring the structure of the Watson-Glaser Critical Thinking Appraisal: One scale or many subscales? Thinking Skills and Creativity, 2008, volume 3, issue 1, pages 15 to 22.
• Possin K. Critique of the Watson-Glaser Critical Thinking Appraisal Test: The More You Know, the Lower Your Score. Informal Logic, 2014, volume 34, issue 4.
• Butler H. Halpern Critical Thinking Assessment Predicts Real-World Outcomes of Critical Thinking. Applied Cognitive Psychology, 2012, volume 26, issue 5, pages 721 to 729.
• de Bie H, Wilhelm P, van der Meij H. The Halpern Critical Thinking Assessment: Toward a Dutch appraisal of critical thinking. Thinking Skills and Creativity, 2015, volume 17, pages 33 to 44.
• Butler H, Pentoney C, Bong M. Predicting real-world outcomes: Critical thinking ability is a better predictor of life decisions than intelligence. Thinking Skills and Creativity, 2017, volume 25, pages 38 to 46.
• Ren X, Tong Y, Peng P, Wang T. Critical thinking predicts academic performance beyond general cognitive ability: Evidence from adults and children. Intelligence, 2020, volume 82, article 101487.
• Stanovich K, West R. On the relative independence of thinking biases and cognitive ability. Journal of Personality and Social Psychology, 2008, volume 94, issue 4, pages 672 to 695.
• West R, Toplak M, Stanovich K. Heuristics and biases as measures of critical thinking: Associations with cognitive ability and thinking dispositions. Journal of Educational Psychology, 2008, volume 100, issue 4, pages 930 to 941.
• Toplak M, West R, Stanovich K. The Cognitive Reflection Test as a predictor of performance on heuristics-and-biases tasks. Memory and Cognition, 2011, volume 39, issue 7, pages 1275 to 1289.
• Hausknecht J, Halpert J, Di Paolo N, Moriarty Gerrard M. Retesting in selection: A meta-analysis of coaching and practice effects for tests of cognitive ability. Journal of Applied Psychology, 2007, volume 92, issue 2, pages 373 to 385.
• American Educational Research Association, American Psychological Association and National Council on Measurement in Education. Standards for Educational and Psychological Testing, 2014. Standards 1.1, 1.4, 1.14, 2.3, 2.14 and 8.2 and the comment on unproctored administration, read in the open access edition on October 29, 2026.

14 Frequently Asked Questions