The Watson-Glaser Critical Thinking Appraisal is Pearson's 40-item test of how well you separate fact from assumption, judge arguments and draw conclusions from a short passage. Here is the current US version, what each of the five question types asks, how the percentile score works, and where it fits next to the tests US firms name for summer 2028.
Key facts
The Watson Glaser test is a 40-item critical thinking test from Pearson TalentLens with a 30-minute limit. Its items fall into three scored areas, Recognize Assumptions (12 items), Evaluate Arguments (12) and Draw Conclusions (16), built from five question types: inference, assumptions, deduction, interpretation and evaluation of arguments. The current version, Watson-Glaser III, draws each sitting from an item bank and reports a percentile against a norm group the employer picks, so there is no fixed pass mark. You answer from the passage alone, not from what you know.
- Publisher: Pearson TalentLens. The current version is Watson-Glaser III, released in the US in 2018.
- Length: 40 items per sitting: 12 Recognize Assumptions, 12 Evaluate Arguments and 16 Draw Conclusions (inference, deduction and interpretation).
- Time: 30 minutes for the items. With instructions and untimed practice questions, the whole session can run up to 40 minutes.
- Delivery: Online and item-banked, so each sitting is drawn from a large pool. It can be taken unproctored; Pearson recommends a desktop or laptop.
- Reading level: Directions and items are written at or below a ninth grade reading level. The difficulty is the logic, not the vocabulary.
- Scores: An overall percentile plus percentiles for the three areas, compared with a norm group the employer chooses.
Try the Watson Glaser Test format
Each section as the app runs it on Pearson's practice player: the instructions, a worked example, then 3 scored questions at the real test's pace, with the report after it.
The five sections on Pearson's practice player, as the app draws it: each with its instructions and a worked example, then 3 scored questions at the full test's pace of 45 seconds a question. Each one opens in its own window, with a report at the end. Pearson reports the real test as a percentile against a comparison group the employer picks, and publishes no pass mark. This practice counts right answers only.
Inference
A set of facts, then a conclusion: judge it True, Probably True, Insufficient Data, Probably False or False.
1 example, 3 questions · 2:15
Recognition of Assumptions
A statement, then a suggested assumption: is the speaker relying on it or not?
1 example, 3 questions · 2:15
Deduction
Premises, then a conclusion: does it have to follow from them, whatever you know?
1 example, 3 questions · 2:15
Interpretation
A short passage, then a conclusion: does it follow beyond reasonable doubt?
1 example, 3 questions · 2:15
Evaluation of Arguments
A question and an argument for or against it: is the argument strong or weak?
1 example, 3 questions · 2:15
What each section looks like





What it is
The Watson-Glaser Critical Thinking Appraisal is published by Pearson TalentLens. Pearson first released it in the US in 1964, moved to two 40-item fixed forms (D and E) with Watson-Glaser II in 2010, and released the current Watson-Glaser III in the US in 2018. Version III is an online, item-banked test: each sitting has 40 items picked at random from a large pool, so two candidates rarely see the same set.
Pearson sells it for the recruitment and development of graduates, managers, executives and professionals, and says university selection teams use it too. It comes in US English among other languages, can be taken unproctored online, and Pearson recommends a desktop or laptop over a phone. Pearson's own write-ups name US users such as the University of South Florida, which uses it with students, and the US Air Force, which uses it in a leadership course for senior enlisted leaders.
For US law students the more common online test is a different one. Among the US law firms we cover, the test the firms name for summer 2028 is Suited: Skadden asks applicants to its US offices to complete a brief Suited assessment, and Simpson Thacher requires Suited's Cognitive Skills Assessment for its 2028 Summer Associate Program. Latham & Watkins asks every applicant to take a virtual assessment. Suited's cognitive part covers attention to detail, critical reasoning and logical reasoning, so the reading discipline the Watson Glaser trains carries over.
Intervyo is an independent preparation platform. Our practice materials are written from scratch to teach the format and are not affiliated with or endorsed by Pearson TalentLens.
The dimensions under test
- Inference: Rate how likely a stated inference is to be true, given a short set of facts. Five answers: True, Probably True, Insufficient Data, Probably False, False. General knowledge everyone shares is allowed here; specialist knowledge is not.
- Recognize Assumptions: Spot the unstated premise behind a statement. For each proposed assumption you pick Assumption Made or Assumption Not Made.
- Deduction: Take the premises as given and decide whether the conclusion follows logically from them: Conclusion Follows or Conclusion Does Not Follow, even when the conclusion sounds true in real life.
- Interpretation: Weigh a short paragraph of evidence and decide whether a generalization or conclusion based on it is warranted: Conclusion Follows or Conclusion Does Not Follow.
- Evaluate Arguments: Judge each argument on a question as Strong or Weak. An argument that is not directly relevant to the question is weak, however important it sounds in general. You are judging relevance and weight, not whether you agree.
A worked example
Our own passage, written to show how one set of facts is tested five different ways. Passage: a city transit agency added 40 electric buses in 2025. In their first year, maintenance cost per mile on the electric buses was 30% lower than on the diesel buses they replaced, and on-time arrivals across the network rose from 81% to 84%.
Inference: "The new buses caused the rise in on-time arrivals."
Answer: Insufficient Data. The passage puts the two facts side by side but never links them; schedules, roadwork or ridership could explain the rise. Do not upgrade a correlation to a cause.
Assumption: "The agency should replace its remaining diesel buses with electric ones to cut costs."
Proposed assumption: electric buses will stay cheaper to maintain as they age. Answer: Assumption Made, because the recommendation only works if the first-year saving lasts. A second proposal, "riders prefer the electric buses", is Assumption Not Made: the argument is about cost.
Deduction: "All buses on Route 9 are electric. Some electric buses have bike racks."
Conclusion: some buses on Route 9 have bike racks. Answer: Conclusion Does Not Follow. The electric buses with racks could all run on other routes. Draw the sets if it helps.
Interpretation: two conclusions from the passage
"Maintenance cost per mile was lower for the electric buses than for the diesel buses they replaced" follows. "Electric buses are cheaper to own than diesel buses" does not: the passage says nothing about purchase price, charging or lifetime cost.
Evaluate Arguments: "Should the agency electrify its whole fleet?"
"Yes, because electric buses cut maintenance cost per mile by 30% in their first year" is Strong: important and directly on the question. "No, because some riders find the quieter buses unsettling" is Weak: related, but minor next to cost and service.
The takeaway: Each item has its own rule. Inference allows five shades of likelihood; deduction allows only a yes or no from the premises; arguments are judged on relevance and weight, never on whether you agree.
How it is marked
Pearson does not publish a pass mark. Your result is a percentile: the share of people in a comparison group who scored lower than you. The employer chooses the group.
- Overall percentile: The headline result, shown with separate percentiles for Recognize Assumptions, Evaluate Arguments and Draw Conclusions. Pearson advises employers to use only the overall score for selection.
- Not a simple count: On Watson-Glaser III the raw score is not the number you got right. Items are scored with item response theory, so harder items carry more information, and the result is converted to a percentile.
- US norm groups: Pearson's US norm groups cover occupations (accountant, consultant, engineer and others), job levels from hourly or entry level up to executive, and education from high school to graduate degree. The same score can sit at different percentiles against different groups.
- What a score looks like: Pearson's own illustration: 35 out of 40, compared with its manager norm group, is the 86th percentile. Treat that as an example of how norms work, not a target.
Versions you might be sent
- Watson-Glaser III (online, timed): The current version: 40 item-banked questions with a 30-minute limit, suitable for unproctored screening. This is the one an employer would send for hiring.
- Watson-Glaser III untimed (US English): Pearson offers an untimed version in the US for development use or when an accommodation is needed in recruiting. The timed version is the one meant for selection decisions.
- Watson-Glaser II, Forms D and E: The older fixed 40-item forms from 2010. They remain available on paper and must always be supervised.
- Pearson's free practice test: A practice set on the TalentLens site that shows each question type. Pearson does not score it like the live test; you get advice on improving your critical thinking instead.
The test US law firms name for summer 2028
These firms' US pages cover the online assessment they ask for, with timing and what it contains.
How to prepare
Learn the five rule sets first
Most lost points come from applying the wrong rule: using outside knowledge in deduction, or treating Probably True as True in inference. Read the instructions for each type until you can say the rule from memory.
Take Pearson's free practice test
It is the closest look at the real item style. Do it untimed first to learn the formats, then again against a 30-minute clock.
Practice inference daily
Inference has five options and is where most candidates hesitate. Ten items a day, checking why each answer is right, builds the instinct for the line between Probably True and Insufficient Data.
Pace for 45 seconds an item
40 items in 30 minutes is 45 seconds each. Pearson designs the limit to be generous, so most people finish; spend saved time on the inference and interpretation sets.
Practice the test your firm names
If your target firm sends Suited, CCAT or SHL instead, spend most of your time on that format. Use our firm-to-test matrix to check what each US firm names.
Common failure modes
The specific ways candidates lose marks on this test. None of these need talent to avoid, only awareness.
- Bringing in outside knowledge: In deduction and interpretation the passage is the whole world. A conclusion that is true in real life but not supported by the text does not follow.
- Collapsing inference to true or false: Probably True and Probably False exist for a reason. If the facts make something more likely but not certain, the answer is Probably True, not True.
- Grading arguments by agreement: An argument you dislike can still be Strong if it is important and directly on the question. Judge relevance and weight only.
- Linking assumptions together: Judge each proposed assumption on its own against the statement, not against the other options.
- Racing on a phone: Pearson recommends a desktop or laptop. Long passages on a small screen cost time and lead to misreads.
What strong candidates do
The habits that separate high scorers, most of them decided before the timer even starts.
- Read the statement before the options. Pin down exactly what the passage says, then test each option against it. Rereading the passage for every option is slower than reading it well once.
- Name the rule as you answer. Say to yourself which section you are in and what its rule is. It stops you carrying inference logic into deduction.
- Watch quantifiers. All, some, most and only change everything in deduction. Underline them mentally and draw quick sets when two appear together.
- Keep moving. No single item is worth a minute. Make your best call, flag it mentally, and keep your pace at about 45 seconds.
Practice with Intervyo
Intervyo recreates Watson Glaser Test in its real format, timed and scored, with instant feedback so the structure is familiar before it counts. Start free, no card required.
Free
Try it and track any company
$0no card
Sign up free- An online assessment in the real format
- 2 HireVue-style questions, scored
- A resume review and a cover letter
- The tracker, with alerts for 10 companies
Firm Pack
Every stage at the firm you pick
$89one payment
Choose your firm- 80 online assessment runs and 50 technical questions
- 50 video interview questions for your firm, scored
- 12 live AI mock interviews
- 15 resume reviews and 45 cover letters
- Your firm's process map and 10 research briefs
Yours for the whole application season.
Premium
The pack, plus the superday
$149one payment
Choose your firm- Everything in the Firm Pack
- 5 superday simulations
- 20 live AI mock interviews in all
- 15 research briefs and a deeper write-up on the firm
Yours for the whole application season.
Season Pass
Save 63%Every company you apply to
$16.58a month
Get the Season Pass- 150 online assessment runs a month
- 130 HireVue-style questions a month, every company
- 20 live AI mock interviews a month
- 5 superday simulations a month
- 30 resume reviews and 90 cover letters a month
$199 a year, or $44.99 monthly. Cancel any time.
Common questions
How many questions are on the Watson Glaser test, and how long is it?
Watson-Glaser III has 40 items with a 30-minute limit. With instructions and untimed practice questions, the session can take up to 40 minutes. The 40 items split into 12 Recognize Assumptions, 12 Evaluate Arguments and 16 Draw Conclusions.
What is a passing score on the Watson Glaser?
There is no published pass mark. Pearson reports a percentile against a norm group the employer chooses, so each employer sets its own bar. Pearson's own example puts 35 out of 40 at the 86th percentile against its manager group.
Do US law firms use the Watson Glaser?
The US law firms we cover name Suited for summer 2028: Skadden for its US offices and Simpson Thacher, which requires Suited's Cognitive Skills Assessment. Latham & Watkins asks every applicant to take a virtual assessment. Practice the format your firm names; the Watson Glaser's critical reasoning skills carry over to Suited's cognitive section.
Can I use outside knowledge on the Watson Glaser?
Only in a narrow way. In the inference section you may use common knowledge that practically everyone shares. Everywhere else, and especially in deduction and interpretation, you treat the passage as true and answer from it alone.
Is there a free Watson Glaser practice test?
Yes. Pearson TalentLens has a free practice test that shows each question type. It is not scored like the live test; you get advice on improving your critical thinking instead.
Is the Watson Glaser timed?
The version used for hiring is timed at 30 minutes. Pearson also offers an untimed version in US English for development use or when an accommodation is needed. If you need an accommodation, ask the employer before you start.
Is the Watson Glaser hard?
The reading level is modest, at or below ninth grade, and Pearson sets a generous time limit. What makes it hard is the rules: each section asks a different question about the same kind of passage, and carrying the wrong rule across costs points.
Keep going
Sources
Firm pages first, then dated reports. Last checked Nov 5, 2026.
- 1Pearson TalentLens: Watson-Glaser Critical Thinking Appraisal
- 2Pearson TalentLens, via the Wayback Machine: Watson-Glaser III (W-G III) Frequently Asked Questions (archived Jun 2024)
- 3Pearson TalentLens, via the Wayback Machine: Watson-Glaser III Critical Thinking Appraisal: User's Guide and Technical Manual (archived Sep 2024)
- 4Pearson: Watson-Glaser Critical Thinking Appraisal III: Assessment Efficacy Report (Apr 2020)
- 5Pearson: Watson-Glaser Critical Thinking Appraisal: Efficacy Report Summary
- 6Pearson TalentLens: Practice Tests: Take a Free Critical Thinking Test
- 7JobTestPrep: Guide To The Watson Glaser Inference Section (2026) · prep site
- 8JobTestPrep: Watson Glaser Deduction Section Prep Guide (2026) · prep site
- 9JobTestPrep: Guide To Watson Glaser Arguments: Strategies & Examples (2026) · prep site
- 10JobTestPrep: Guide To The Recognition of Assumptions Watson Glaser (2026) · prep site
- 11Skadden: Summer Associate Application Information (United States)
- 12Suited: For law students
- 13Simpson Thacher: How to Apply (Summer Program)
- 14Chambers Associate: Latham & Watkins firm profile: Get Hired · press
Intervyo is an independent preparation platform. Our practice tests simulate the style and format of Watson Glaser assessments and are not affiliated with, or endorsed by, the test publisher.