Quick takeaways
- What it measures: Argument analysis — inference, identifying assumptions, applying formal deduction, interpreting passages, and evaluating argument strength.
- Flagship test: Watson Glaser Critical Thinking Appraisal — 5 subtests, 30-40 minutes. Used by Magic Circle law firms.
- Five subtests: Inference, recognition of assumptions, deduction, interpretation, evaluation of arguments. Each has a distinct decision rule.
- Where it appears: Law (Allen & Overy, Linklaters, Clifford Chance, Freshfields, Hogan Lovells, Slaughter and May), consulting, audit, civil service Fast Stream, journalism.
- Typical cutoff: Magic Circle firms typically target the 70th-80th percentile against a graduate norm. Some firms publish their threshold.
What is a critical thinking test?
A critical thinking test measures whether you can analyse arguments, evaluate evidence, and draw valid conclusions from information that may be incomplete, contradictory or deliberately misleading. Unlike numerical or abstract reasoning, the questions are language-heavy — you read short passages and decide what they prove, what they don't prove, and which conclusions follow.
Critical thinking is used by employers in roles where structured argument and evidence evaluation are core to the job: law, consulting, journalism, audit, intelligence, policy and senior management. The flagship instrument is the Watson Glaser Critical Thinking Appraisal, used by Magic Circle law firms and other employers that value precise reasoning over fast calculation.
The Watson Glaser five subtests
The Watson Glaser is the most widely used critical thinking test in graduate hiring, especially in law. It has five distinct question types, each measuring a different reasoning skill:
- Inference. Given a passage, evaluate whether each proposed inference is true, probably true, insufficient data, probably false, or false. The discipline is staying inside what the passage supports.
- Recognition of assumptions. Given a statement, identify which proposed assumptions are taken for granted. The trap is treating any plausible-sounding fact as an assumption — only unstated premises that must be true for the argument to work count.
- Deduction. Given a set of premises, decide whether each proposed conclusion follows necessarily. Pure formal logic — accept the premises as true even if implausible.
- Interpretation. Given a passage, decide whether each proposed conclusion follows beyond reasonable doubt from the passage. Stricter than "probably true" — the standard is "the passage proves it".
- Evaluation of arguments. Given a topic and proposed arguments for or against, judge each argument as strong or weak. Strong = relevant and important to the issue. Weak = irrelevant, trivial, or based on unsupported assumptions.
The full Watson Glaser typically runs 30-40 minutes and is used to screen graduate hires at Allen & Overy, Linklaters, Clifford Chance, Freshfields, Hogan Lovells, Slaughter and May and many other Magic Circle and US firms with London offices.
Where critical thinking appears beyond Watson Glaser
- Allen & Overy and other Magic Circle firms — Watson Glaser used as a primary screen.
- Consulting — McKinsey Solve, BCG Casey and Bain Online Case include critical-reasoning components; Watson Glaser also appears at some consulting firms.
- Journalism and policy — Critical thinking blends into editorial tests and policy-analysis exercises.
- GMAT / GRE Critical Reasoning — Conceptually the same question style as Watson Glaser, used in MBA and graduate-school admissions.
- Civil Service Fast Stream — UK civil service uses critical reasoning-style questions in the e-Tray and judgement assessments.
Critical thinking vs verbal reasoning
The two often get confused. Both involve reading passages and evaluating statements, but the questions are subtly different:
- Verbal reasoning (SHL-style true/false/cannot say) asks whether a statement is supported by the passage. It tests reading discipline.
- Critical thinking (Watson Glaser-style) asks whether a conclusion follows logically from the argument, whether an assumption is needed for an argument to work, whether an argument is strong or weak. It tests argument analysis.
A strong verbal reasoning scorer is not automatically a strong critical thinking scorer. Critical thinking requires familiarity with the specific judgement standards (necessarily true vs probably true vs insufficient data) and with what counts as an unstated assumption.
Why critical thinking is hard
Three specific traps catch most candidates:
The plausibility trap. Candidates accept conclusions that sound reasonable but are not actually supported by the passage. The rule is: even a true-sounding conclusion is wrong if the passage doesn't establish it.
The premise trap. In deduction questions, premises may be absurd ("All cats can fly"). Candidates instinctively reject them as false. The instruction is to accept the premises and check whether the conclusion follows from those premises — not whether the premises are realistic.
The strong-argument trap. In evaluation-of-arguments questions, candidates pick the argument they personally agree with. The test asks whether the argument is strong on the issue, regardless of personal opinion. An argument can be on the side you agree with and still be weak (irrelevant, trivial, ad hominem).
Worked example: an inference question
Passage: "The company's quarterly report showed profits declined for the third consecutive quarter, despite a marketing campaign launched at the start of the period that increased website traffic by 40%."
Proposed inference: "The marketing campaign was not effective."
Answer: Probably false. The passage establishes that traffic increased 40% following the campaign — a plausible positive signal — while profits declined for reasons the passage never attributes to the campaign specifically (declining profit could stem from costs, pricing, competition, or a dozen other factors unrelated to marketing). Concluding the campaign "was not effective" overreaches what the passage actually supports; the passage doesn't establish a causal link between the campaign and the profit decline at all.
This illustrates the core discipline the inference subtest demands: resisting the pull toward a conclusion that feels intuitively plausible (traffic went up but profit went down, so maybe the campaign didn't work) when the passage itself never actually connects those two facts causally. The correct habit is asking, for every proposed inference, "does the passage establish this specific claim, or am I filling in a gap with outside reasoning?" — and defaulting toward the more cautious rating (insufficient data, probably false) whenever a genuine causal or logical gap remains.
How to prepare for critical thinking tests
- Learn the five Watson Glaser subtest rules cold. Each subtest has a specific decision rule (e.g. inference: true / probably true / insufficient data / probably false / false against the passage). Practise applying the rule rather than your gut.
- Drill one subtest at a time. Most candidates have one or two weak subtests — usually assumptions and evaluation. Spend disproportionate practice time on the weakest.
- Read like a lawyer, not a layperson. Treat the passage as the only source of truth. External knowledge, common sense, plausible inferences — all irrelevant. If the passage doesn't say it, the answer is insufficient or false.
- Slow down on deduction. Deduction questions are the most rule-bound and the most often rushed. Spending an extra ten seconds confirming "does this conclusion follow necessarily" is almost always worth it.
- Practice timed full tests. Watson Glaser is long (40+ items, 30-40 minutes). Pacing matters — many candidates run out of time on the final subtest.
How TestSolve fits
Critical thinking is an excellent category for screenshot-based practice review. The questions are language-heavy and the explanation gap is large — candidates often answer wrong but cannot articulate why. TestSolve can take a practice question, walk through which subtest rule applies, identify which option correctly applies the rule and why the distractors fail, and build a personal library of decision patterns. Use on practice and review materials only.
Recognising which subtest you're facing mid-test
Because the Watson Glaser mixes five subtests with distinct rules across one continuous session, one practical skill worth building is quickly recognising which subtest a given item belongs to before applying a decision rule — using the wrong subtest's rule on a question is a completely avoidable mistake. Inference and interpretation items both present a passage followed by proposed statements, but inference uses a five-point scale (true, probably true, insufficient data, probably false, false) while interpretation uses a stricter binary standard (does the conclusion follow beyond reasonable doubt, yes or no) — confusing the two scales is a common, fixable error. Deduction items are the most visually distinct, typically presenting explicit numbered premises followed by a proposed conclusion. Recognition-of-assumptions items ask what's taken for granted rather than what follows, and evaluation-of-arguments items always present a stated issue followed by arguments for or against it. Spending a few practice sessions purely on rapid subtest identification, before worrying about getting individual answers right, pays off across the whole test.
Related skill hubs
Provider guides for this skill
Frequently asked questions
What's the difference between Watson Glaser and a general verbal reasoning test?
Verbal reasoning (SHL-style) tests whether a statement is supported by a passage — pure reading comprehension under time pressure. Watson Glaser tests argument analysis — inference, assumptions, deduction, interpretation and evaluation of arguments. The decision rules are specific to each Watson Glaser subtest.
Which employers use Watson Glaser?
Predominantly UK Magic Circle and US-headquartered law firms — Allen & Overy, Linklaters, Clifford Chance, Freshfields, Hogan Lovells, Slaughter and May, plus most major US firms hiring in London. Some consulting firms and Civil Service Fast Stream also include critical reasoning elements.
Can critical thinking be improved with practice?
Yes, more than most people expect. The five Watson Glaser subtests each have specific decision rules that can be learnt. Two to three weeks of targeted practice produces measurable improvement.
Is critical thinking timed?
Yes. The Watson Glaser is typically 30-40 minutes for 40-80 items depending on the version. Pacing across the five subtests matters — running out of time on evaluation of arguments is a common failure mode.
What's a passing Watson Glaser score?
Most Magic Circle firms target the 70th-80th percentile against a graduate norm. Some firms publish their threshold (Allen & Overy's reported benchmark is around the 75th percentile). Always check role-specific stated benchmarks if available.
Do I need to agree with an argument to rate it as strong?
No. Evaluation-of-arguments questions ask whether an argument is relevant and important to the issue, not whether you personally agree with its conclusion. An argument on the side you agree with can still be rated weak if it's irrelevant, trivial, or unsupported.
Should I reject a deduction question's premises if they seem false?
No. Deduction questions ask you to accept the given premises as true, however implausible, and judge only whether the proposed conclusion follows necessarily from them — not whether the premises themselves are realistic.
Ready to use TestSolve on your next assessment?
See it in action first, then download when you're ready. No subscription, no signup.
TestSolve is independent and not affiliated with any test provider or employer named on this page. All product names and trademarks belong to their respective owners.