Format
Short written passages with a statement to judge.
- Questions
- 40
- Time
- 30 to 40 minutes
- Languages
- US and UK English, French, Dutch, Spanish and more
The Watson-Glaser Critical Thinking Appraisal, from Pearson TalentLens, is 40 questions in 30 to 40 minutes on short written passages. Its five subtests are drawing inferences, recognizing assumptions, deducing, interpreting and evaluating arguments, and you answer from the information given, not what you know. The Profile Report gives a percentile against a norm group, overall and for three areas: Recognize Assumptions, Evaluate Arguments and Draw Conclusions. For logic under a clock, Soreno, an AI voice interview practice app for consulting and finance candidates, has a free brain teaser trainer with no sign-up for the interviews that often follow the test.
Updated
A critical thinking test that TalentLens describes as measuring the ability to analyze arguments, evaluate information and make objective decisions. Employers use it in recruitment and development for graduates, professionals, supervisors, managers and executives, and TalentLens says university selection teams use it too.
Short written passages with a statement to judge.
Online, from an item bank, with instant results for the employer.
Three reports the employer can use.
TalentLens calls these the building blocks of critical thinking, with questions of varying difficulty and format. The definition is TalentLens's, the tip after it is ours.
1
Rating the probability that an inference is true, based on the information given. Our tip: it's a scale of likelihood, so outside knowledge and a hunch about causes both pull you off.
2
Identifying unstated assumptions or presuppositions under a statement. Our tip: ask whether the statement would still make sense if the assumption were false. If not, it was assumed.
3
Deciding whether a conclusion follows logically from the information. Our tip: it must follow every time. A rule that says all X are Y says nothing about what else is Y.
4
Weighing the evidence and deciding if a generalization or conclusion from the data is warranted. Our tip: check the group and the claim. A finding about some people rarely covers everyone.
5
Judging the strength and relevance of arguments on a question. Our tip: strong means relevant and important to the question, not that you agree.
TalentLens: Watson-GlaserInferences, deducing and interpreting all ask about a conclusion, and they read alike. This is how we separate them, by the standard of proof each asks for.
Inferences
A degree of probability. The trap is reading a cause into two things that happened together.
Deducing
Strict logic. The trap is reversing a rule: all big deals are reviewed does not mean every reviewed deal is big.
Interpreting
Warranted or not by the evidence given. The trap is a claim wider than the data, a new group or a cause the study never tested.
40 questions in 30 or 40 minutes. Check your invite for which.
Per question
That includes reading the passage, so read it once with care before you look at the statement.
Checkpoints
At questions 10, 20 and 30. If you are behind, move faster on the next passage rather than rushing the last one.
Where the clock should be (our checkpoints)
| Question | 30-minute test | 40-minute test |
|---|---|---|
| 10 | 7.5 min | 10 min |
| 20 | 15 min | 20 min |
| 30 | 22.5 min | 30 min |
TalentLens's FAQ lists three reports. Your result is read against other people, not as a raw mark.
Profile
An overall percentile plus one for each of Recognize Assumptions, Evaluate Arguments and Draw Conclusions. Number Correct, T-score, STANINE and STEN sit in its technical section. Employers can choose from a wide range of norm groups, or map a company norm for a fee.
Fairness
Each item is coded by difficulty. Tests may differ slightly in difficulty overall, which TalentLens says the scoring accounts for. Passages are coded by theme, such as Business, Science and Education, and every candidate gets enough business passages.
Development
TalentLens's sample Development Report, a French-language sample, compares each area with people in your profession and rates it as a strength to build on (top of the group), an area to deepen (the middle) or a development opportunity (the bottom).
TalentLens: sample Development Report (French)Interview
The Interview Report helps an interviewer run a structured critical thinking behavioral interview, with sample questions. Expect to be asked how you test assumptions.
From TalentLens: Watson-Glaser Critical Thinking Appraisal, TalentLens: free critical thinking practice test and TalentLens: Watson-Glaser III sample Development Report (French).
TalentLens defines critical thinking as understanding a situation from several perspectives while separating facts from opinions and assumptions. The Profile Report scores it in three areas.
Spotting what a statement takes for granted.
Judging whether an argument is strong and relevant to the question.
Covers the inference, deduction and interpretation subtests: what follows, and how firmly, from the evidence.
All written by our team, with invented companies. None are TalentLens items, and the answer choices are ours. Work from the words in each prompt only.
Answer, then check the reason.
Inference. Larkmoor Bank moved its 120 loan officers to new software in Q3. Average loan turnaround fell from 14 to 11 days that quarter. Inference: the software caused the faster turnaround. True, probably true, not enough information, probably false or false?
Not enough information. The two happened in the same quarter, but nothing in the passage rules out other causes, such as staffing or the loan mix.
Assumption. "We must cap associates at 55 billed hours a week so first-year attrition stops rising." Is it assumed that working over 55 hours contributes to attrition?
Yes, it is assumed. If long hours had no link to attrition, capping them to stop attrition would make no sense.
Deduction. Every cross-border merger above $500 million is reviewed by the firm's antitrust group. Last month the group reviewed the Helios-Orion deal. Does it follow that Helios-Orion is a cross-border merger above $500 million?
No. The rule says what the group must review, not that it reviews nothing else. Reversing the rule is the classic trap.
Interpretation. A survey of 400 partners found 68% of those using contract-review software billed more hours than the firm median. Conclusion: the software makes any lawyer bill more. Warranted or not?
Not warranted. It is one survey of partners showing an association. It says nothing about cause or about associates.
Argument. Should banks require analysts in the office five days a week? "No, because commuting is tiring for people who like to sleep in." Strong or weak?
Weak. It is on topic but rests on a minor preference, not on output, training, clients or retention.
The kind of questions a critical thinking interview asks.
Tell me about a recommendation you made when a key assumption hadn't been checked.
Sample answer. Example answer written by our team, the applicant is invented. On a lending internship, a borrower's forecast assumed fuel costs would fall 6% after new trucks arrived, but the new trucks only replaced short-haul vans. I flagged it, reran the model with flat long-haul costs, and coverage dropped below our floor. We asked for route data and added a tighter covenant before the credit committee approved it.
Name the assumption, how you tested it, what changed in your answer and the result.
Tell me about a time you disagreed with your manager.
The evidence you brought, how you raised it and what was decided. Show you weighed their argument too. Practice it, see a model answer
Tell me about a decision you made with incomplete information.
What you knew, what you didn't, how you sized the risk and what you'd check next time. Practice it, see a model answer
Ours, built around what TalentLens publishes.
Day 1
In English or Dutch. It shows what each subtest looks like. TalentLens says it can't score it, but you get a report with advice on preparing.
Days 2 to 4
Do our five items below, then write two of your own per subtest from news articles. Say why each answer is right in one sentence.
Day 5
40 questions at 45 seconds each, with checkpoints at questions 10, 20 and 30.
Test day
Read each subtest's instructions. Answer from the passage alone, even when you know the industry.
After
If the employer uses the Interview Report, expect questions on how you test assumptions and weigh evidence. Practice two stories out loud.
Practice behavioral answers and cases out loud with Soreno's AI interviewer, which asks follow-up questions and scores every answer, quoting your own words.
TalentLens gives the format as 40 questions in 30 to 40 minutes. That works out to 45 to 60 seconds a question by our count.
Drawing inferences, recognizing assumptions, deducing, interpreting and evaluating arguments. The Profile Report groups them into Recognize Assumptions, Evaluate Arguments and Draw Conclusions.
As percentiles against a norm group, overall and for the three areas. Number Correct, T-score, STANINE and STEN appear in the report's technical section.
No. Questions come from an item bank. TalentLens says each item is coded by difficulty and small differences between tests are accounted for in scoring.
Yes, on TalentLens's site in English and Dutch. TalentLens says it can't score it meaningfully, but you get a report with advice on preparing for the live test.
TalentLens lists it for graduates, professionals, supervisors, managers and executives, and lists Law and Banking & Finance among the industries on its site. University selection teams use it too.