Watson Glaser Critical Thinking Test: The Complete 2026 Guide
The Watson Glaser is the world's most widely used critical thinking assessment — and it trips up sharp candidates who treat it like a reading comprehension test. Here is everything you need to know to approach all five sub-tests with confidence.
What is the Watson Glaser Critical Thinking Test?
The Watson Glaser Critical Thinking Appraisal (WGCTA) is a psychometric assessment that measures your ability to reason clearly, evaluate evidence objectively, and draw valid conclusions — without importing your own prior knowledge or assumptions. It was first developed by Goodwin Watson and Edward Glaser in 1925 and has been revised multiple times since; the current version is published by Pearson TalentLens and administered through TalentLens Online or a partner platform.
Unlike numerical or verbal reasoning tests that measure knowledge domains, the Watson Glaser tests thinking process — specifically, how rigorously you apply logic to structured arguments. Candidates with strong academic grades or extensive industry knowledge regularly underperform on Watson Glaser because their domain expertise actively works against them: the test requires you to evaluate arguments based solely on what is stated in the passage, not what you already know to be true.
Test Formats
The Watson Glaser is available in two main formats for graduate-level screening:
| Format | Questions | Time Limit | When Used |
|---|---|---|---|
| Watson Glaser Short Form (WG-S) | 40 questions | 30–40 minutes | Most common in graduate recruitment screening |
| Watson Glaser Full Form | 80 questions | 40–60 minutes | Less common; used for senior or specialist hires |
| Timed online version | Varies by employer | Set by employer | Administered via employer portal (e.g. EY, law firms) |
Watson Glaser passages often cover legal, business, scientific, or political topics. You do not need background knowledge in any of these areas — all the information you need is contained in the passage. The danger is that candidates with genuine expertise in a topic area answer based on what they know to be true in the real world, rather than what the passage actually states. This is one of the most common sources of incorrect answers.
How Watson Glaser is Scored
Your raw score (the number of questions answered correctly) is converted to a percentile rank against a professional or graduate-level norm group. There is no negative marking — incorrect answers receive zero, not a penalty. Most employers using Watson Glaser as a screening tool set a percentile-based cut-off, above which candidates progress to the next stage. The specific cut-off is set by each employer independently and is not typically published.
The 5 Sub-Tests Explained
The Watson Glaser test is divided into five sub-tests, each targeting a distinct aspect of critical thinking. In the Short Form, the 40 questions are distributed across all five sub-tests, typically 8 questions per section. Understanding what each sub-test is actually measuring — and the common errors specific to each — is the most efficient way to improve your score.
Every sub-test has only 2–5 answer options. The difficulty is not in choosing from a long list of plausible answers; it is in applying exactly the right logical standard to each question. The line between "Probably True" and "Insufficient Data", or between "Conclusion Follows" and "Does Not Follow", is often very fine — and is intentionally so. Speed and precision both matter. Candidates who are overly cautious (defaulting to "Insufficient Data" or "Does Not Follow" on everything) score just as poorly as those who over-infer.
| Sub-Test | Key Skill Tested | Typical Error |
|---|---|---|
| Inference | Calibrated probability judgement | Choosing "Insufficient Data" when "Probably True/False" is correct |
| Recognition of Assumptions | Identifying what is taken for granted | Confusing an assumption with a stated fact, or over-identifying assumptions |
| Deduction | Formal logical validity | Importing real-world knowledge to override the logical form |
| Interpretation | Evidence-based conclusion drawing | Treating "beyond a reasonable doubt" too loosely or too strictly |
| Evaluation of Arguments | Argument relevance and substantiveness | Judging argument strength based on whether you personally agree with it |
Inference & Recognition of Assumptions
Sub-Test 1: Inference
In the Inference sub-test, you are given a short passage of factual observations and then presented with a proposed inference. Your job is to rate how strongly the inference follows from those facts, using a five-point scale:
- True: The inference definitely follows — there is no other reasonable interpretation.
- Probably True: The inference is more likely than not to follow from the facts, though certainty is not possible.
- Insufficient Data: The passage neither supports nor contradicts the inference; more information would be needed.
- Probably False: The inference is more likely than not to be wrong, though certainty is not possible.
- False: The inference is definitely contradicted by the facts in the passage.
The most common mistake is defaulting to "Insufficient Data" — candidates assume that unless the passage definitively confirms or denies something, they cannot make a judgement. This is wrong. When a passage gives you clear directional evidence, "Probably True" or "Probably False" is often the correct answer. Reserve "Insufficient Data" for cases where the passage genuinely provides no relevant information about the inference.
Watson Glaser Inference questions frequently use topics (economics, science, social trends) where you may have strong outside knowledge. If the passage states that "sales of product X fell by 30% last year" and the inference is "product X is unpopular", you must evaluate this purely on what the passage tells you — not on what you know about how consumer markets work. Outside knowledge that supports or contradicts the inference should be treated as if it does not exist.
Sub-Test 2: Recognition of Assumptions
An assumption is something that is taken for granted — an unstated premise that the argument depends on. In this sub-test, you are given a statement and must decide whether a proposed assumption is genuinely embedded in that statement.
The key question to ask is: "Does the statement make sense without accepting this proposed assumption?" If the statement can stand without the assumption being true, then the assumption has not been made.
Consider this example. Statement: "We should hire more graduates from Russell Group universities." Proposed assumption: "Russell Group graduates are more capable employees." This assumption has been made — without it, the recommendation has no logical basis. Now consider the proposed assumption: "Russell Group universities are located in the UK." This is a fact, but it is not an assumption that the statement depends on. It is irrelevant to the logical chain, so this assumption has not been made.
A reliable technique: negate the proposed assumption and ask whether the original statement becomes illogical or falls apart. If negating the assumption destroys the argument, the assumption was being made. If the statement still holds without the assumption, the assumption was not embedded in it. This technique works for approximately 80% of Recognition of Assumptions questions and is worth practising until it becomes instinctive.
Deduction & Interpretation
Sub-Test 3: Deduction
Deduction questions present you with two or more premises and ask whether a given conclusion necessarily follows from them — i.e., whether it is logically impossible for the premises to be true and the conclusion to be false at the same time.
This is the most formal of the five sub-tests. The standard of proof is strict: a conclusion "follows" only if it is logically entailed by the premises. A conclusion that is plausible, likely, or strongly suggested does not "follow" in the Deduction sense unless it is the only possible result given the premises.
| Scenario | Correct Answer | Why |
|---|---|---|
| Premises guarantee the conclusion with no exceptions possible | Conclusion Follows | The conclusion is logically necessary |
| Premises make the conclusion likely but leave room for exceptions | Does Not Follow | Probable is not the same as necessary in deductive logic |
| Conclusion contradicts one of the premises | Does Not Follow | Contradiction cannot be valid |
| Conclusion introduces new information not in the premises | Does Not Follow | Deduction cannot create new facts |
This is where expert candidates fail most often. Example: Premises state "All bankers are wealthy. John is a banker." Conclusion: "John owns property." This does not follow — the premises only tell you John is wealthy, not that he owns property. Candidates who know that wealthy people typically own property apply real-world knowledge, mark "Follows", and get it wrong. In deduction, you must work only within the logical system defined by the premises.
Sub-Test 4: Interpretation
Interpretation is superficially similar to Deduction, but the standard is different. You are asked whether a conclusion follows beyond a reasonable doubt based on the evidence in a short passage — not whether it follows with absolute logical necessity.
The distinction matters significantly. In Deduction, even a 99% likelihood is not enough — it must be 100% guaranteed. In Interpretation, a high degree of confidence based on the evidence is sufficient — you are not required to rule out every conceivable exception. Think of it as a civil law standard (balance of probabilities, leaning heavily toward certainty) rather than criminal law (beyond all doubt).
Interpretation asks "does this follow from the evidence as a reasonable person would conclude?" Deduction asks "is this the only possible conclusion given these premises?" In practice, you will find that more conclusions "Follow" in Interpretation than in Deduction for equivalent evidence strength — because the bar is slightly lower. If you apply the same strict standard to both sub-tests, you will under-mark Interpretation questions and score below your true level.
Evaluating Arguments — The Most Misunderstood Sub-Test
The Evaluation of Arguments sub-test is frequently cited as the most surprising for first-time Watson Glaser candidates — because what makes an argument "strong" or "weak" is not what most people instinctively think.
You are presented with a question (typically a policy-style statement such as "Should the UK increase its corporate tax rate?") and then a proposed argument for or against it. Your task is to decide whether the argument is strong or weak.
What Makes an Argument Strong?
According to the Watson Glaser framework, a strong argument must satisfy all three of the following criteria:
- Directly relevant: It must directly address the question posed — not a tangentially related issue.
- Important: It must deal with a genuinely significant consequence or consideration, not a trivial or peripheral one.
- Substantial: It must be based on factual substance or clear logical reasoning, not emotional appeals, personal opinions, or circular logic.
An argument can be factually correct, personally compelling, and well-reasoned — and still be weak by Watson Glaser standards if it fails on relevance, importance, or substance. This is where opinion-based thinking leads candidates astray.
If you are asked "Should the government fund arts education?" and the argument is "Yes, because art makes life worth living," this is a weak argument — it is an emotional appeal without substantive evidence. Even if you personally believe deeply in arts education, the argument as stated does not meet the standard of a strong argument in the Watson Glaser framework. Conversely, an argument you disagree with might be strong if it is relevant, important, and evidenced. Assess the argument structure, not its conclusion.
Common Patterns of Weak Arguments
| Weak Argument Type | Example | Why It Is Weak |
|---|---|---|
| Emotional appeal | "Yes, because it would make people happier." | Vague, not substantiated |
| Appeal to tradition | "No, because we have always done it this way." | Irrelevant to the actual question |
| Slippery slope without evidence | "Yes, otherwise society will collapse." | Extreme, unsubstantiated chain of causation |
| Tangential point | "Yes, because the weather in London is unpredictable." | Completely irrelevant to the question |
| Trivially true but unimportant | "No, because some people would need to fill in more paperwork." | Not sufficiently significant |
| Circular reasoning | "Yes, because it is a good idea." | Restates the conclusion without justification |
After reading an argument, ask "So what? Why does this matter to the question?" If you cannot provide a clear, substantive answer — if the connection to the question is vague, indirect, or relies on several unstated assumptions — the argument is likely weak. Strong arguments survive the "So what?" challenge with an immediate, direct, and concrete answer. If you need to do a lot of additional reasoning to explain why the argument is relevant, that additional reasoning is not in the argument and the argument is weak as stated.
Which Employers Use the Watson Glaser Test?
The Watson Glaser test is most prevalent in sectors where analytical reasoning, written argumentation, and evidence-based judgement are core job competencies. Law, financial services, consulting, and the public sector are the most common fields.
Law Firms
Watson Glaser is the de facto standard assessment for training contract applications at UK law firms. The test is used to screen candidates early in the application process — typically before interviews — because critical thinking is fundamental to legal work: evaluating arguments, identifying assumptions in contracts, and drawing valid conclusions from complex evidence.
| Firm Tier | Examples | Typical Stage |
|---|---|---|
| Magic Circle | Freshfields, Clifford Chance, Linklaters, A&O Shearman, Slaughter and May | Online application stage, before assessment centre |
| Silver Circle | Herbert Smith Freehills, Ashurst, Macfarlanes, Travers Smith | Early screening stage |
| US firms in London | Baker McKenzie, Latham & Watkins, White & Case | Varies by firm; some use Watson Glaser, others use bespoke tests |
| National firms | DLA Piper, Eversheds Sutherland | Online screening |
Most law firms administer the Watson Glaser online with a strict time limit, typically 30 minutes for 40 questions. This equates to 45 seconds per question on average. Speed under time pressure is a genuine differentiator — candidates who are technically accurate but slow will score lower than they expect compared to untimed practice. Timed practice from the very first session is essential.
Big Four and Consulting
Several large professional services firms use Watson Glaser either as a standalone assessment or as part of a broader test battery. EY's online assessment includes a Watson Glaser component alongside numerical and verbal reasoning for certain service lines. Consulting firms value the test for the same reasons as law firms — client-facing advisory roles require exactly the kind of structured reasoning the test measures.
Financial Services and Banking
Some investment banks and asset management firms incorporate Watson Glaser into their screening, particularly for roles in research, structuring, risk, or compliance where written argumentation is a significant part of the job. It is less universal in financial services than in law, but candidates applying to firms in this space should be prepared for it to appear.
Public Sector
The Civil Service Fast Stream and certain NHS Graduate Management Scheme pathways have historically used or drawn on Watson Glaser-style critical thinking assessments as part of their selection process. The critical thinking competencies tested map directly to the analytical requirements described in Civil Service competency frameworks.
While the Watson Glaser assessment is a standardised tool, individual employers can configure the number of questions, time limits, and which sub-tests are included. Always check the instructions provided by the specific employer when you receive the test invitation — do not assume that the format you practised on is identical to the live test. Most employers using Watson Glaser for graduate screening use the 40-question Short Form.
4-Week Preparation Strategy
Watson Glaser preparation works differently to numerical or verbal reasoning practice. Because the test measures a thinking process rather than subject knowledge, the most effective preparation builds new cognitive habits — specifically, the habit of evaluating evidence strictly within the information provided, without importing outside knowledge or opinions.
Four weeks is sufficient for most candidates to produce a meaningful score improvement. The key is deliberate practice with structured review, not high-volume repetition without reflection.
- Read detailed explanations of all 5 sub-test types
- Complete one untimed practice test per sub-test
- Review every answer — right and wrong — against the rules
- Identify which sub-tests feel least intuitive
- Focus 60% of practice time on your two weakest sub-tests
- Practise Deduction with formal syllogism exercises
- Practise Assumptions with the "negation test" technique
- Begin timed practice — 45 seconds per question maximum
- Complete 2–3 full timed 40-question practice tests
- Track score and time taken per sub-test separately
- Identify if slow sub-tests are costing you unattempted questions
- Work on pacing: if behind, skip and return rather than stalling
- One full timed test every 2 days — focus on consistency
- Review every error for sub-test-specific pattern (same mistake type?)
- Read quality opinion pieces and practise evaluating argument strength
- Day before: light review only — no new material
Supplementary Techniques That Work
- Read legal and policy commentary critically: Publications like the FT, The Economist, and quality legal journals contain arguments and counter-arguments. Practise identifying the assumptions embedded in each paragraph, not just reading for comprehension.
- Practise formal logic puzzles: Deduction sub-test performance specifically improves with practice of formal syllogisms (if-then statements, all/some/none logic). Even 10 minutes per day builds the mental habit of strict logical reasoning.
- Write arguments and then challenge them: Draft a short argument on any topic, then actively try to find the hidden assumptions and evaluate whether the argument is genuinely strong or weak by Watson Glaser standards. This active writing practice internalises the framework faster than passive reading.
- Time your pace explicitly: Use a visible countdown timer during practice. Knowing exactly how many seconds remain per question builds the intuitive pacing that keeps you on track during the real test.
Both tests require close, careful reading of short passages and strict adherence to the information provided. Candidates who combine SHL verbal reasoning practice with Watson Glaser work tend to see faster improvement in both. The "True / False / Cannot Say" discipline in verbal reasoning reinforces the same habit of not importing outside knowledge — which directly transfers to Watson Glaser Inference and Interpretation questions.
Frequently Asked Questions
Sharpen Your Critical Thinking Before the Real Test
Practice timed verbal, logical, and critical reasoning tests designed to build the precision and speed Watson Glaser demands. Track your percentile improvement across sessions.