A verbal reasoning solver needs the passage and the statement, not just the sentence you are being asked to judge. TestSolve reads them off your screen and returns a worked answer to your phone. You can keep the original text visible while checking the reason for true, false or cannot say.
This is a format where a free chat assistant can also be useful. If the complete passage is easy to paste, ChatGPT may give you the explanation you need. The TestSolve verbal reasoning workflow is designed to remove that copying step when the question is already on your computer.
The hardest part is often deciding what the passage does not establish. A solver must distinguish a statement that is contradicted from one that is simply unsupported. Confusing those two changes the answer even when every word has been read correctly.
The three answers describe the relationship between a statement and the supplied text. True means the statement follows from the information. False means the information contradicts it, while cannot say means the text does not settle the issue.
| Answer | What the passage establishes | What the explanation should show |
|---|---|---|
| True | The statement is supported | The relevant evidence and valid inference |
| False | The statement conflicts with the evidence | The specific contradiction |
| Cannot say | The evidence leaves the statement unresolved | The missing fact or relationship |
The task is not asking whether the statement sounds realistic. It is asking what follows from the particular passage. An AI that fills in the gaps using everyday knowledge may produce a sensible general observation and the wrong test answer.
Cannot say requires the solver to stop at the boundary of the evidence. That is easy to overlook when a statement sounds likely. A narrative about expanding a business, for example, may encourage an assumption about profits even if the passage never mentions costs.
There are two opposite mistakes. One is treating something plausible as true; the other is treating something unproven as false. Both replace a judgement about evidence with a stronger conclusion than the text supports.
A solver that handles explicit facts well can still fail when the evidence is incomplete. The useful check is whether its explanation identifies the exact information needed to settle the statement. If it cannot show support or contradiction, it should leave the conclusion unresolved.
False needs a contradiction. If a passage says an office opens on Monday and a statement says it remains closed all Monday, the two claims cannot both hold under the stated conditions. The answer rests on the conflict.
Cannot say needs unresolved information. If a passage says some engineers work remotely, it does not establish whether all engineers work remotely. “Some” does not automatically mean “some but not all”, so the stronger statement is not necessarily false.
That distinction is a useful way to judge an AI explanation. If it says “false because this is not mentioned”, ask what part of the passage actually contradicts the statement. Absence of support alone may instead justify cannot say.
AI can attempt verbal reasoning because the relevant material is language. It can identify sentences, compare claims and explain why a conclusion follows. There is often less visual complexity than in an abstract pattern or a dense chart.
The difficulty is staying within the passage. An assistant may import background knowledge, infer a cause from a sequence of events or strengthen a cautious statement. The resulting explanation can sound persuasive while exceeding the evidence.
Accuracy therefore depends on more than fluent reading. A good response must preserve qualifications such as “may”, “usually”, “only” and “unless”. Those words can determine whether the answer is true, false or unresolved.
Yes, it can be a sensible free option when you can provide the complete passage and statement. The chat format also makes follow-up questions easy. You can ask which sentence supplies the evidence or which fact is missing.
Custom GPTs named “Aptitude Solver” and wrapper pages on aichatonline.org or yeschat.ai may offer a focused version of that conversation. Their branding does not establish a separate verbal reasoning score. Check whether the explanation actually respects the passage.
TestSolve is useful when copying the text is the inconvenience you want removed. Its screen reading verbal solver reads the visible question and sends the working to your phone. That lets you compare the proposed inference with the original wording.
Do not assume that a publisher's name identifies one universal format. Aon's published scales verbal guidance explicitly describes statements judged true, false or cannot say. Its candidate information lists 49 tasks in 12 minutes for that named module.
For an invitation labelled Kenexa, identify the actual verbal assessment and its instructions. The brand covers different assessment products, so a general description should not substitute for the module you have received. The relevant solver input is the wording and choices on your screen.
Saville also offers different verbal assessments, and its published Verbal Analysis example includes a question with several substantive answer choices. It would therefore be inaccurate to describe every Saville verbal question as a three-option task. Use the true/false/cannot say method when those are the options actually presented.
A useful answer identifies the relevant part of the passage and explains the relationship to the statement. For true, that means showing support. For false, it means showing the contradiction, while cannot say should identify the fact the passage leaves open.
The explanation should not merely repeat the statement using different words. It should connect the conclusion to evidence. If that connection depends on a new assumption, the answer has not yet been established.
Consider a passage saying that sales increased after a new director joined. That timing alone does not prove the director caused the increase. A solver that turns “after” into “because of” has introduced a causal claim the passage may not support.
Small changes in scope can have large effects. “All”, “some”, “none” and “most” describe different claims about a group. “Can”, “may” and “will” also express different levels of certainty.
| Wording change | Why it matters | What the solver should check |
|---|---|---|
| Some becomes all | The statement expands the group covered | Whether the passage establishes the wider claim |
| May becomes will | Possibility becomes certainty | Whether the outcome is guaranteed by the text |
| After becomes because | Timing becomes causation | Whether a causal relationship is actually stated or entailed |
| One region becomes the whole company | The statement changes the population | Whether evidence covers the larger group |
| A past result becomes a future result | The statement changes the time period | Whether the future claim is supported |
| A condition is reversed | A one-way rule becomes a two-way rule | Whether the reverse implication follows |
None of these wording changes automatically makes a statement false. Some will produce cannot say because the stronger claim remains possible but unproven. A solver must evaluate the actual evidence instead of applying a shortcut such as “stronger wording always means false”.
A cropped passage can change the apparent answer. The visible paragraph may contain a general statement while the hidden final sentence gives an exception. An assistant cannot account for that exception if it never receives it.
The same applies when information is spread across tabs or sheets. The relevant text needs to be available to the tool. Screen reading should not be interpreted as automatic access to every hidden page in the assessment.
There is also a distinction between an incomplete input and a complete passage with insufficient evidence. The former means the solver needs more material; the latter can justify cannot say. A dependable explanation should not blur those two situations.
It depends on the work you want it to remove. If you can paste a short passage into a free assistant and get a clear, supported answer, that may be sufficient. A paid workflow becomes more attractive when copying and managing the material repeatedly is the main obstacle.
TestSolve supports Windows, macOS and browser screen sharing. It reads the question from your computer and returns the worked answer to your phone. The first question is free, followed by $14.99 for 30 questions or $19.99 for 50, with no subscription.
TestSolve reports 92% on a 50-question benchmark measured on 18 September 2026. This is an overall figure, not a verbal-only score or a result for Aon, Kenexa or Saville. Use the first question to assess the explanation on the kind of passage you actually need help with.
Personality questionnaires and strengths inventories have no universal right answers. Even when their statements look similar to verbal test sentences, the task is different: you are describing yourself rather than judging a claim against a supplied passage. A reasoning solver cannot calculate a single correct personality profile.
Game-based assessments score how you play, so a written answer to a screenshot does not replace the required behaviour. These limits apply independently of whether a tool can read the words. External assistance should also be used only where the assessment instructions allow it.
For a complete passage that is easy to paste, ChatGPT or another free assistant may be enough. TestSolve is designed for reading the question directly from your screen and delivering a worked answer to your phone. Judge either option by the evidence it gives.
Choose cannot say when the supplied passage neither establishes nor contradicts the statement. Identify what additional fact would be needed to decide. Do not assume that a plausible statement is true or that an unsupported statement is automatically false.
No. Something can be absent from a passage and still be possible. False requires a contradiction with the information provided, whereas cannot say applies when the issue remains unresolved. Check the test's instructions, then assess the evidence for the whole statement.
It can attempt them if you provide the complete passage and statement. Its explanation should rely only on that material. Watch for added background knowledge, reversed conditions and claims that something is false simply because it was not mentioned.
Yes, TestSolve reads the visible question and returns a worked answer to your phone. It is available for Windows, macOS and browser screen sharing. Make sure the relevant passage and statement are included so the tool receives the evidence it needs.
Free chat assistants can help with text questions, and TestSolve offers the first question free without an account. That lets you inspect the reading and explanation before paying. Further question packs are available without a subscription.
The useful answer is the one that shows what the passage establishes and what it leaves open. Try TestSolve on a verbal reasoning question and compare the working with the text on your screen. First question free, no account, no subscription.
<!-- Editorial source notes, not part of the article body: Aon three-option format: https://assets.aon.com/-/media/files/aon/capabilities/talent-and-rewards/practice-tasks---verbal-reasoning.pdf Aon timing: https://www.aon.com/en/capabilities/talent-and-rewards/prepare-for-your-online-assessment Kenexa product listing: https://www.shl.com/c/global/ibm-kenexa-catalog/view/33227-cat-verbal-reasoning-test Saville verbal range: https://www.savilleassessment.com/practice-tests/ Saville four-choice Verbal Analysis example, printed page 7/PDF page 9: https://www.savilleassessment.com/wp-content/uploads/2024/07/IAC-Ability-Workbook-2024-1.pdf Product facts and benchmark supplied by TestSolve. All illustrative sentences are original. No claim is made that cannot-say errors account for most lost marks. -->
First question free. No account, no subscription.