Critical Thinking is the practice of examining information and claims rather than accepting them unconditionally. It involves evaluating their meanings, evidence, assumptions, reasoning, and conditions of applicability to make well-founded judgments about what to believe and how to act.
The word critical does not necessarily mean negative or oppositional. Critical Thinking involves examining both supporting and opposing reasons, assessing the available evidence, and determining how much confidence a conclusion deserves.
For example, suppose a company receives the following proposal:
“Introducing generative AI will significantly improve employee productivity.”
Accepting this claim immediately would not constitute sufficient examination. However, rejecting it simply because the supporting evidence is unclear would not necessarily represent a well-reasoned judgment either.
First, what exactly does productivity mean? Does it refer to reduced working hours, increased output per unit of time, or results that also account for the quality of completed work?
Under what conditions has the claimed improvement been observed? Can similar effects be expected in other departments or organizations? Has the time required to verify and correct AI-generated work been considered?
Furthermore, could comparable improvements be achieved through changes in business processes or better use of existing software without introducing generative AI?
Critical Thinking uses such questions to clarify what a claim means and what supports it. It then distinguishes established information from uncertainty and uses that distinction to improve judgment.
However, Critical Thinking does not require suspending every decision until all uncertainties have been eliminated. In practical situations, decisions often must be made with incomplete information.
The purpose of Critical Thinking is neither to avoid conclusions nor to find faults in other people’s arguments. It is to clarify the evidence and limitations underlying our judgments—including our own—and remain willing to revise them when necessary.
The History and Theoretical Background of Critical Thinking
Although the term Critical Thinking became widely established in modern education and philosophy, its intellectual foundations can be traced to ancient Greek philosophy, logic, and traditions of systematic questioning.
The Socratic Method used questions to examine the meanings, assumptions, and possible contradictions underlying people’s claims.
With the development of epistemology, scientific methodology, and logic, philosophers and scientists continued investigating how knowledge and judgments can be justified.
In the twentieth century, John Dewey emphasized the importance of Reflective Thinking in his 1910 book How We Think. Dewey argued for carefully examining the grounds supporting beliefs and the consequences that follow from them rather than simply accepting ideas as they occur.
Subsequently, researchers in philosophy and education examined how Critical Thinking should be defined, taught, and assessed.
Robert H. Ennis characterized Critical Thinking as reasonable, reflective thinking focused on deciding what to believe or do.
In 1990, Peter A. Facione and colleagues published the results of a Delphi study involving 46 experts. The study identified important cognitive skills and dispositions associated with Critical Thinking.
However, researchers do not share a single, completely unified definition of Critical Thinking. Different approaches emphasize somewhat different aspects of reasoning, judgment, and intellectual responsibility.
| Researcher or Tradition | Main Contribution |
|---|---|
| Socratic Method | Examining assumptions and contradictions through questioning |
| John Dewey | Reflective Thinking based on careful consideration of grounds and consequences |
| Robert H. Ennis | Reasonable and reflective judgment about what to believe or do |
| Peter A. Facione | Identifying cognitive skills such as interpretation, analysis, evaluation, inference, explanation, and self-regulation |
| Richard Paul and Linda Elder | Analyzing, evaluating, and improving one’s own thinking |
| Cognitive Psychology | Investigating cognitive biases and processes influencing judgment |
An important principle shared by these approaches is that Critical Thinking applies not only to other people’s claims but also to one’s own reasoning.
Six Core Skills of Critical Thinking
The 1990 Delphi Report identified six core cognitive skills associated with Critical Thinking: Interpretation, Analysis, Evaluation, Inference, Explanation, and Self-Regulation.
| Skill | Main Function |
|---|---|
| Interpretation | Understand the meanings of information, statements, and situations |
| Analysis | Identify relationships among claims, reasons, assumptions, and other elements |
| Evaluation | Assess the credibility of evidence and the strength of reasoning |
| Inference | Draw conclusions or formulate hypotheses from available information |
| Explanation | Clearly communicate the reasons and evidence supporting a judgment |
| Self-Regulation | Monitor one’s own thinking and revise it when necessary |
These abilities extend beyond simply understanding a text and determining whether its statements are correct.
For example, when reading a report claiming that a technology is promising, the reader must first interpret what promising means.
Next, the reasoning and evidence supporting that assessment should be analyzed.
The reliability of the research findings or market information must then be evaluated.
The reader may also consider what conclusions can reasonably be drawn, develop additional hypotheses, and explain why a particular judgment is justified.
Finally, the conclusion should be reconsidered if new evidence becomes available.
Self-Regulation is especially important.
Critical Thinking is not a process that ends once a conclusion has been reached. It includes reviewing and modifying earlier judgments in response to new evidence or objections.
Critical Thinking Skills and Intellectual Dispositions
Critical Thinking involves not only the ability to analyze and reason but also the willingness to apply those abilities.
For example, a person may possess strong logical reasoning skills but still reach poor conclusions if they consistently ignore evidence that challenges their preferred beliefs.
Similarly, someone with extensive expertise may struggle to respond to new findings if they are unwilling to consider the possibility that their existing understanding is incomplete.
Several intellectual dispositions are particularly relevant.
| Intellectual Disposition | Description |
|---|---|
| Intellectual Curiosity | Willingness to investigate unclear or unexplained phenomena |
| Open-Mindedness | Willingness to consider alternative perspectives and explanations |
| Intellectual Humility | Recognition of limitations in one’s own knowledge and understanding |
| Fair-Mindedness | Applying comparable standards to supporting and opposing claims |
| Careful Judgment | Adjusting the strength of conclusions according to the available evidence |
| Willingness to Revise Beliefs | Changing one’s position when new evidence justifies doing so |
| Persistence in Inquiry | Continuing necessary investigation despite complexity or difficulty |
These are desirable dispositions, but individuals do not necessarily demonstrate them consistently.
It may be particularly difficult to evaluate objectively an idea one has developed personally or a field in which one has invested many years of professional effort.
Critical Thinking requires both expertise and awareness of the limitations of that expertise.
Clarifying What Is Actually Being Claimed
One of the first tasks in Critical Thinking is to clarify the meaning of the claim being examined.
Consider:
“Introducing AI improves R&D productivity.”
Although this appears straightforward, it can be interpreted in several ways.
Does R&D productivity refer to the number of academic papers a researcher can review? The time required to develop an experimental plan? The number of proposed research themes? Or the number of research projects that eventually achieve commercial success?
Similarly, AI adoption may involve document summarization, programming, experimental data analysis, hypothesis generation, or other applications.
The expected effects will differ depending on the activity.
A more specific question might be:
“Can researchers using LLMs for literature reviews reduce the time required to produce research reports that meet predefined quality standards?”
This formulation makes it possible to compare working time, report quality, verification effort, and related measures.
Critical Thinking requires attention to ambiguous terms rather than allowing discussions to proceed as though their meanings were already settled.
However, it is not always necessary to define every concept exhaustively at the beginning. Important distinctions may emerge during investigation and discussion.
Distinguishing Facts, Interpretations, Inferences, and Value Judgments
Written explanations and discussions often combine reported facts, interpretations, inferences, value judgments, and recommendations.
Consider the following statement:
“After AI was introduced at one company, the number of tasks processed increased by 20%. Therefore, AI is effective for improving operational efficiency, and our company should introduce it.”
Several different elements are present.
| Type | Example |
|---|---|
| Observation or Report | The number of tasks processed increased by 20% at one company |
| Interpretation | Increased task volume represents improved operational efficiency |
| Causal Inference | AI adoption caused the increase in task volume |
| Generalization | Similar improvements can be expected at other companies |
| Value Judgment | Increasing task throughput is desirable |
| Recommendation | Our company should introduce AI |
Even if the initial observation is accurate, the subsequent interpretations and inferences do not automatically follow.
The increase in task volume may have resulted from factors other than AI.
Quality may have declined even though more tasks were completed.
Furthermore, even if efficiency improvements are established, implementation costs or information security requirements may justify a different decision.
Distinguishing these elements makes it easier to identify where additional evidence is needed.
How to Evaluate the Reliability of Information Sources
Critical Thinking requires evaluating not only the content of a claim but also the source of the information supporting it.
For example, information about the performance of a new technology may appear in a manufacturer’s promotional materials or in an independent comparative study.
These sources have different characteristics.
However, information is not necessarily false because it comes from a manufacturer, nor is it necessarily correct because it comes from an independent organization.
The important question is what evidence supports the information and under which conditions it was established.
| Evaluation Criterion | Main Question |
|---|---|
| Origin | Who published the information? |
| Expertise | Does the source possess relevant expertise? |
| Primary Evidence | Can the original experiment, statistics, statement, or document be examined? |
| Methodology | How was the information obtained or analyzed? |
| Independence | Has the finding been confirmed by independent sources? |
| Conflicts of Interest | Are there interests that could influence the presentation or interpretation? |
| Scope of Applicability | Under what conditions does the claim hold? |
| Currency | Is the information still applicable to the present situation? |
For example, numerous articles repeating the same study do not necessarily constitute independent evidence.
Conversely, several studies conducted using different methods may provide stronger support for a conclusion when their results converge.
However, even apparently consistent research findings may share measurement problems or selection biases.
The quality and origins of the evidence matter more than the number of sources alone.
Evaluating Online Information Through Lateral Reading
When evaluating information found online, it is useful to investigate sources beyond the webpage being examined rather than relying exclusively on its own presentation.
This approach is known as Lateral Reading.
Research by Sam Wineburg, Sarah McGrew, and colleagues at Stanford University examined how professional fact-checkers evaluate online information.
The researchers observed that fact-checkers often leave the original website to investigate its background, reputation, and claims through other sources.
This contrasts with evaluating credibility primarily through the appearance of a webpage, its stated credentials, or the logos it displays.
Mike Caulfield’s SIFT method provides another practical approach to online information verification.
| Step | Meaning | Activity |
|---|---|---|
| S | Stop | Avoid accepting or sharing information immediately |
| I | Investigate the Source | Examine who is responsible for the information |
| F | Find Better Coverage | Look for more reliable or informative sources |
| T | Trace Claims to the Original Context | Follow claims, quotations, or data back to their original sources |
For example, suppose an article states:
“A scientific study proves that AI substantially improves productivity.”
A useful first step is to locate the original study.
What did it actually measure? Working time, task completion, accuracy, or the quality of the final output?
Who participated in the study, and what tasks were performed?
What comparison conditions were used?
These details determine how broadly the findings can be applied.
However, reading the original paper does not guarantee that its findings can be evaluated correctly. Understanding research methodology and statistical analysis may require specialized expertise.
Examining the Relationship Between Claims and Evidence
Even when the evidence presented is reliable, the conclusion drawn from it may not be justified.
Consider:
“Company A’s revenue increased after it introduced AI. Therefore, introducing AI increases revenue.”
The observation that revenue increased is different from the claim that AI adoption caused the increase.
Furthermore, even if a causal effect were established for Company A, the same outcome might not be expected in another company.
Critical Thinking examines the relationship between premises and conclusions.
Argument Analysis is a specialized method for investigating this relationship in greater detail.
Argument Analysis identifies claims, grounds, warrants, implicit assumptions, objections, and the forms of reasoning through which conclusions are supported.
However, Critical Thinking is broader than Argument Analysis.
It also includes evaluating information sources, examining one’s own cognitive biases, managing uncertainty, and deciding how to act.
Distinguishing Correlation from Causation
One important challenge in Critical Thinking is distinguishing correlation from causation.
Suppose a company observes a positive correlation between employees’ use of AI and their productivity.
This observation alone does not establish that AI use caused higher productivity.
It is possible that more productive employees are also more likely to use AI.
Alternatively, other factors—such as training, professional skills, or the working environment—may influence both AI adoption and productivity.
| Possible Relationship | Explanation |
|---|---|
| AI Use → Productivity | AI helps employees work more efficiently |
| Productivity → AI Use | More productive employees are more likely to adopt AI |
| Third Factor → Both | Training, skills, or working conditions influence both |
| Multiple Relationships | AI use and productivity may influence one another |
Distinguishing these possibilities may require more than ordinary correlation analysis.
Depending on the problem, appropriate research designs, comparison groups, experiments, or quasi-experimental methods may be necessary.
Even when a causal relationship is established, the magnitude of the effect and the conditions under which it occurs remain important.
Critical Thinking helps identify these questions.
However, scientifically evaluating causal relationships may require specialized methods from statistics, experimental design, and causal inference.
Examining Evidence That Could Challenge a Claim
Critical Thinking requires considering not only evidence supporting a preferred claim but also evidence that might challenge it.
For example, suppose a researcher believes that a new material performs better than an established alternative.
Rather than searching only for favorable experimental results, the researcher should investigate conditions under which the new material performs poorly.
Are there environments in which the existing material is superior?
Could the observed improvement be explained by differences in experimental conditions or measurement methods?
One particularly useful question is:
“What evidence would cause me to change my mind?”
This question encourages people to treat their beliefs and hypotheses as open to examination.
However, falsifiability in the philosophy of science and Critical Thinking in general are not identical concepts.
Testing a scientific theory and comparing alternatives in a business decision involve different objectives and evidential requirements.
Nor does attempting to falsify a hypothesis automatically settle its validity.
Considering Alternative Hypotheses
Once people identify a plausible explanation for an event, they may become more attentive to information supporting that explanation.
However, the first explanation that comes to mind is not necessarily the best one.
Suppose the defect rate in a manufacturing process increases.
A rise in equipment temperature is observed, leading to the hypothesis that temperature caused the defects.
Other explanations may also be possible.
| Hypothesis | Possible Explanation |
|---|---|
| A | Higher equipment temperature increased the defect rate |
| B | A change in the material lot increased the defect rate |
| C | A third factor influenced both equipment temperature and defect rate |
| D | Changes in measurement or inspection procedures made the defect rate appear higher |
These hypotheses are not necessarily mutually exclusive.
Several factors may contribute simultaneously.
The important principle is to consider whether the observations can be explained in more than one way before committing to a particular explanation.
Different hypotheses can then be used to derive different predictions that may be compared through observation or experimentation.
This activity is closely related to Abduction and the Scientific Method.
Cognitive Biases and Critical Thinking
Human judgment can be influenced by various Cognitive Biases.
A Cognitive Bias is a systematic tendency in information processing or judgment that can produce deviations from appropriate reasoning under particular conditions.
For example, the tendency to search for or interpret information in ways that support existing beliefs is commonly associated with Confirmation Bias.
Raymond S. Nickerson examined different manifestations of Confirmation Bias in his 1998 paper “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises.”
However, merely knowing the names of cognitive biases does not guarantee that their effects can be avoided.
| Bias or Judgment Problem | Typical Situation |
|---|---|
| Confirmation Bias | Seeking information that supports a preferred hypothesis |
| Anchoring | Allowing an initial number or piece of information to exert excessive influence |
| Availability Heuristic | Overestimating the importance of easily recalled examples |
| Survivorship Bias | Studying successful cases while overlooking failures |
| Base-Rate Neglect | Focusing on individual cases without adequately considering overall frequencies |
| Motivated Reasoning | Interpreting evidence in ways that favor desired conclusions |
For example, collecting only success stories about the adoption of a technology may lead to an exaggerated assessment of its effectiveness.
However, failure cases may be difficult to identify because organizations are less likely to publish them.
Critical Thinking therefore considers not only the available evidence but also the information that may be missing from the dataset.
At the same time, not every poor decision should be attributed to a cognitive bias.
Inadequate information, measurement problems, insufficient expertise, or organizational constraints can also contribute to incorrect judgments.
Examining One’s Own Thinking — Metacognition
An important characteristic of Critical Thinking is that one’s own thinking becomes an object of examination.
This is related to Metacognition, the ability to recognize and regulate one’s cognitive activities.
Suppose a person has a negative initial reaction to an unfamiliar technology.
Why does the technology seem unpromising?
Is there a problem with its underlying principles or experimental results?
Or is the negative reaction partly influenced by the fact that the technology falls outside the person’s familiar field of expertise?
The same questions should be applied when evaluating one’s own ideas.
If a person believes that a technology they have proposed is highly valuable, what evidence supports that assessment?
Has it been compared adequately with alternatives?
Could favorable results have been overestimated because they align with personal expectations?
Treating one’s own judgments as objects of investigation distinguishes Critical Thinking from merely criticizing others.
However, people cannot necessarily observe their own thinking with complete objectivity.
Externalizing judgments through written records, structured comparisons, and independent review may therefore be useful.
Expertise and Critical Thinking
Critical Thinking is sometimes presented as a general ability that can be exercised without specialized knowledge.
However, evaluating real-world problems often requires understanding the relevant subject matter.
For example, assessing the durability of a new material requires knowledge of material properties, test procedures, and operating environments.
Evaluating statistical findings may require understanding sampling, measurement, estimation, and statistical testing.
At the same time, extensive expertise in one field does not guarantee sound judgment outside that field.
A person may understand a particular technology extremely well while lacking sufficient knowledge to evaluate its commercialization prospects, market demand, or organizational feasibility.
Critical Thinking may therefore require both general reasoning abilities and domain-specific knowledge.
It also involves recognizing the boundaries of one’s competence and determining when additional investigation or expert assistance is necessary.
Critical Thinking is not a substitute for expertise.
Rather, it helps people understand how their expertise can be applied and where its limitations begin.
The Relationship Between Critical Thinking and Creative Thinking
Critical Thinking and Creative Thinking are sometimes treated as opposing activities.
For example, some idea-generation methods recommend withholding criticism while generating possibilities and evaluating ideas later.
Brainstorming is often associated with this approach.
However, Critical Thinking and Creative Thinking are not necessarily opposed.
Generating new hypotheses may begin with recognizing that an established explanation is inadequate.
Questioning assumptions underlying conventional technologies or practices can also create opportunities for new ideas.
For example, suppose engineers assume that increasing the production speed of a process requires larger equipment.
Questioning that assumption—”Could the process be accelerated without increasing equipment size?”—may lead to new technical possibilities.
In this sense, Critical Thinking can support idea generation.
It is also useful when evaluating the feasibility and evidence supporting newly generated ideas.
| Activity | Creative Thinking | Critical Thinking |
|---|---|---|
| Problem Discovery | Explore new perspectives | Examine existing problem definitions and assumptions |
| Hypothesis Generation | Generate alternative explanations and solutions | Identify overlooked possibilities |
| Hypothesis Evaluation | Improve and develop ideas | Examine evidence, contradictions, and applicability conditions |
| Implementation and Improvement | Experiment with alternatives | Revise judgments in response to results |
Creative Thinking can expand the range of possibilities, while Critical Thinking helps make those possibilities open to examination.
In practice, the two processes are not always clearly separated.
Practical Example 1 — Evaluating Claims About AI Productivity
Suppose a company’s management team receives the following proposal:
“A report shows that companies introducing generative AI experienced a 30% improvement in productivity. Therefore, we should deploy generative AI to all employees.”
Critical Thinking can be applied by asking the following questions.
| Perspective | Questions |
|---|---|
| Meaning | What exactly does a 30% productivity improvement represent? |
| Evidence | Can the original research or report be examined? |
| Study Conditions | Which tasks, participants, and AI systems were evaluated? |
| Causation | Was the improvement actually caused by AI adoption? |
| Generalization | Can similar effects be expected in our organization? |
| Costs | Were subscriptions, training, verification, and corrections included? |
| Risks | Could errors, information leakage, or reduced quality create problems? |
| Alternatives | Could process improvements achieve similar results without AI? |
The analysis may not undermine the possibility that generative AI is beneficial.
However, it may reveal that the evidence is insufficient to justify deploying it to every employee immediately.
A more testable proposal might be:
“Select tasks for which AI is expected to provide meaningful improvements. Conduct a limited trial comparing working time, quality, verification effort, and costs before and after adoption. Use the results to determine whether deployment should be expanded.”
Critical Thinking can thus transform a simple disagreement about whether AI should be adopted into a decision supported by testable evidence.
Practical Example 2 — Reconsidering R&D Theme Exploration
Suppose an R&D department identifies the following problem:
“We examine large numbers of academic papers and patents, but we are not discovering enough promising new research themes.”
One proposed response might be:
“We need to collect even more information.”
However, this recommendation assumes that increasing the amount of information collected will increase the number of useful R&D opportunities discovered.
That assumption should be examined.
Alternative hypotheses include the following.
| Hypothesis | Question to Investigate |
|---|---|
| Insufficient Information | Are important papers, patents, or technical sources missing? |
| Inadequate Analysis | Is the organization failing to generate new hypotheses from the information already collected? |
| Restricted Search Scope | Is exploration limited to familiar domains or established technological classifications? |
| Evaluation Criteria | Are highly novel proposals being rejected during evaluation? |
| Organizational Constraints | Does the organization lack the resources or capabilities to pursue the themes it discovers? |
These hypotheses are not necessarily mutually exclusive.
It is also possible that insufficient information collection contributes to the problem.
The important principle is not to select an intervention based solely on the first plausible explanation.
The statement that there are “not enough promising research themes” should itself be examined.
What counts as a promising theme?
Which criteria are used to evaluate it?
Once the problem is clarified, alternatives such as expanding information collection, introducing cross-domain exploration, or revising evaluation methods can be compared.
Here, Critical Thinking is used not to reject conventional approaches automatically, but to reconsider how the problem has been defined.
Practical Example 3 — Evaluating the Conclusions of a Research Paper
Suppose a scientific paper reports that a new technology outperforms an established alternative.
Publication in a scientific journal provides relevant information when evaluating credibility.
However, it does not establish that the technology is superior under every possible condition.
Several questions should be considered.
| Perspective | What to Examine |
|---|---|
| Comparison | Which existing technology was used as the benchmark? |
| Performance Metrics | What characteristics were actually measured? |
| Test Conditions | Were temperature, loading, materials, and environmental conditions appropriate? |
| Reproducibility | Can the result be reproduced under comparable conditions? |
| Statistical Uncertainty | Were measurement variability and sample size considered? |
| Generalization | Can laboratory results be applied to practical operating conditions? |
| Trade-offs | Did improvements in one characteristic come at the expense of others? |
For example, a material may demonstrate higher strength while having disadvantages in durability, manufacturability, or cost.
Conclusions should not be extended to applications that the original study did not examine.
Critical Thinking does not require dismissing scientific findings.
It requires distinguishing what has actually been established from what remains uncertain or untested.
A Practical Procedure for Critical Thinking
There is no universally required sequence for applying Critical Thinking.
However, the following framework can help improve practical judgments.
| Stage | Activity | Main Question |
|---|---|---|
| 1. Define the Issue | Identify the claim or decision requiring examination | What are we trying to determine? |
| 2. Clarify Meaning | Define important terms and evaluation criteria | What exactly does the claim mean? |
| 3. Identify Evidence | Organize information supporting the claim | What facts and data are available? |
| 4. Evaluate Sources | Investigate the reliability of the evidence | How was the information obtained? |
| 5. Analyze Reasoning | Examine how the conclusion follows from the evidence | Are there unjustified inferential steps? |
| 6. Explore Alternatives | Consider objections and alternative explanations | What other explanations are possible? |
| 7. Assess Uncertainty | Distinguish established knowledge from unresolved issues | What can reasonably be concluded? |
| 8. Decide and Act | Make a judgment appropriate to the available evidence | What should be done now? |
| 9. Reassess | Revise the judgment when new evidence becomes available | What evidence would change our conclusion? |
This procedure is a practical framework rather than a mandatory sequence.
When evaluating a specialized research paper, examining experimental methods and measurements may be the first priority.
When making urgent decisions, it may be necessary to focus on the most consequential uncertainties and risks.
The time and cost of investigation should also be considered.
When Is There Enough Evidence to Decide?
One of the difficult questions in Critical Thinking is deciding when sufficient evidence has been obtained.
In some situations, continued investigation is appropriate.
However, eliminating all uncertainty is rarely possible.
Further research may also impose costs or cause opportunities to be missed.
For example, a small, reversible improvement to a website does not require the same level of verification as constructing a new factory involving substantial investment.
| Decision Condition | Possible Approach |
|---|---|
| Low Impact and Easily Reversible | A limited trial may provide sufficient evidence |
| High Impact and Difficult to Reverse | More extensive investigation and verification may be necessary |
| High Uncertainty | Compare multiple scenarios or implement changes gradually |
| High Information-Gathering Costs | Consider whether additional information is likely to improve the decision |
| Urgent Decision | Prioritize the most consequential uncertainties and risks |
Critical Thinking does not always aim to achieve certainty.
It also involves making reasonable decisions while acknowledging uncertainty.
In some situations, the appropriate response is neither full adoption nor complete rejection, but a limited trial followed by further evaluation.
Critical Thinking and Logical Thinking
Critical Thinking and Logical Thinking overlap substantially, but they are not identical.
Logical Thinking focuses on organizing information, structuring reasoning, and deriving and communicating conclusions coherently.
Critical Thinking also examines whether the problem itself is properly defined, which assumptions are appropriate, and which evidence should be trusted.
For example, a Logic Tree may divide a problem into categories that appear mutually exclusive and collectively exhaustive.
However, the classification itself may not be suitable for the objective.
A market can be divided according to established product categories, but those categories may fail to reveal emerging market opportunities.
Critical Thinking therefore examines not only the internal consistency of a logical structure but also the assumptions underlying the choice of structure.
| Perspective | Logical Thinking | Critical Thinking |
|---|---|---|
| Main Purpose | Organize, reason, and explain logically | Examine the foundations and limitations of judgments |
| Main Focus | Information structures, reasoning, problem decomposition | Claims, evidence, sources, assumptions, uncertainty |
| Representative Methods | MECE, Logic Trees, Pyramid Principle | Evidence evaluation, alternative hypotheses, counterevidence, self-correction |
| Distinctive Contribution | Constructing coherent explanations | Examining whether explanations and judgments are justified |
This does not mean that Critical Thinking operates independently of logic.
The two approaches are complementary.
Critical Thinking and Argument Analysis
Argument Analysis identifies claims, evidence, warrants, premises, objections, and other components in written or spoken arguments.
For example, if a proposal states that Action B should be taken because of Reason A, Argument Analysis examines how A is supposed to support B.
Critical Thinking can extend the examination to the proposal’s objectives, underlying values, possible information biases, and alternative courses of action.
Argument Analysis is therefore one useful method for making Critical Thinking more explicit and systematic.
However, Argument Analysis does not automatically establish the truth of the underlying evidence.
An argument can be structurally coherent while relying on false information.
Critical Thinking and the Scientific Method
Critical Thinking is closely related to the Scientific Method.
Scientific research often involves formulating hypotheses, deriving predictions, and evaluating those predictions through observation or experimentation.
This requires examining assumptions, reasoning, and possible alternative explanations.
For example, even when experimental results agree with a hypothesis, the researcher must consider measurement problems and other explanations.
However, Critical Thinking and the Scientific Method are not synonymous.
Critical Thinking can also be applied in business, education, public policy, and everyday situations where controlled experimentation may not be possible.
Ethical and value-based judgments may also involve considerations that cannot be settled through empirical evidence alone.
The Scientific Method concerns the development and evaluation of empirical knowledge, while Critical Thinking applies to a broader range of judgments and actions.
Can Critical Thinking Be Improved Through Education?
Research has investigated whether Critical Thinking can be improved through education.
Philip C. Abrami and colleagues conducted meta-analyses of instructional interventions affecting Critical Thinking in 2008 and 2015.
Their 2015 paper, “Strategies for Teaching Students to Think Critically: A Meta-Analysis,” examined educational approaches including explicit Critical Thinking objectives, dialogue, and engagement with practical problems.
These studies suggest that it is useful to provide opportunities to practice Critical Thinking rather than merely explain its principles.
However, educational effects vary according to participants, instructional methods, and assessment measures.
Improvements in one learning context do not necessarily transfer automatically to every specialized domain.
For example, becoming better at evaluating written arguments does not by itself provide the statistical or engineering knowledge required to assess technical research correctly.
Critical Thinking therefore benefits from being practiced through concrete problems as well as studied through general principles.
LLM-Assisted Critical Thinking
Large Language Models can support several activities associated with Critical Thinking.
For example, an LLM may extract claims, evidence, and implicit assumptions from a proposal.
It may also generate possible objections, alternative explanations, and questions requiring further investigation.
| Support Function | Possible AI Operation |
|---|---|
| Claim Extraction | Clarify conclusions and recommendations in text |
| Assumption Identification | Suggest premises that have not been stated explicitly |
| Evidence Organization | Associate claims with documents and supporting data |
| Counterargument Generation | Propose questions from alternative perspectives |
| Alternative Hypothesis Generation | Suggest other explanations for the same observations |
| Verification Planning | Identify issues requiring additional investigation |
| Argument Mapping | Organize relationships among claims, supporting reasons, and objections |
| Reassessment | Reconsider conclusions when new information becomes available |
However, AI-generated counterarguments are not necessarily valid.
LLMs may generate plausible explanations or supporting references that contain factual errors or even refer to nonexistent sources.
The wording of a user’s question can also influence the direction of the response.
If the user asks an LLM to justify a preferred conclusion, it may generate reasoning consistent with that conclusion.
Conversely, asking it to find problems with a claim may encourage excessive or poorly justified criticism.
Therefore, AI-assisted Critical Thinking should involve examining supporting and opposing evidence fairly and checking important claims against their original sources.
Critically Evaluating AI-Generated Answers
AI-generated answers present particular challenges.
First, fluent writing does not guarantee factual accuracy.
Second, the presence of citations does not necessarily mean that those sources support the claims being made.
Third, multiple cited sources may ultimately rely on the same original information.
Fourth, interpretations generated by an AI system must be distinguished from statements actually contained in source documents.
Fifth, a response may reflect assumptions embedded in the user’s question rather than independently examining those assumptions.
For example, consider this prompt:
“Explain why this new technology will succeed.”
The question assumes that the technology will succeed.
The resulting answer may focus on reasons supporting that assumption.
A more useful prompt might be:
“Examine the conditions under which this technology could succeed or fail. Distinguish established evidence from unverified hypotheses, and identify the most important questions requiring further investigation.”
Including existing technologies and alternative approaches as comparison targets can also improve the evaluation.
However, carefully designed prompts do not guarantee that AI-generated answers will be accurate.
Important factual claims and judgments still require verification against reliable evidence.
Can Critical Thinking Be Delegated to AI?
LLMs can identify possible inconsistencies in reasoning and generate counterarguments.
However, this does not necessarily establish that AI systems perform Critical Thinking in the same sense as humans.
Critical Thinking involves judgments about which problems matter, which evidence deserves priority, what uncertainty is acceptable, and how to act.
These judgments depend on objectives, values, expertise, and understanding of the real-world situation.
AI can organize relevant information and present alternative possibilities.
However, it cannot automatically assume responsibility for human decisions or guarantee that the evidence has been adequately verified.
It is particularly important not to confuse the quantity of generated objections with the quality of Critical Thinking.
Producing a long list of criticisms is less useful than correctly identifying the few issues that materially affect a decision.
Recording Critical Thinking in a Structured Form
When Critical Thinking is used repeatedly, recording its results in a structured format may be valuable.
For example, a worksheet or database could include the following fields.
| Field | Description |
|---|---|
| Claim Under Examination | What is being evaluated? |
| Definition of the Claim | Important terms and evaluation criteria |
| Supporting Evidence | Facts and data supporting the claim |
| Opposing Evidence | Information challenging the claim |
| Implicit Assumptions | Premises necessary for the reasoning |
| Alternative Hypotheses | Other explanations or options |
| Uncertainty | Unresolved questions and factors affecting outcomes |
| Current Judgment | Conclusion based on available information |
| Conditions of Validity | Circumstances under which the conclusion holds |
| Reassessment Conditions | New evidence that would justify changing the judgment |
Such records make it possible to review why a decision was made at a particular time.
They also make it easier to identify which elements require revision when new evidence becomes available.
For example, suppose a technology is judged to have low feasibility because of limitations in available materials or manufacturing methods.
If new materials or processes are developed, the previous judgment can be reconsidered by examining whether the original constraints still apply.
This approach may support not only individual Critical Thinking but also continuous learning within organizations.
Limitation 1 — Doubt Alone Does Not Produce Progress
Critical Thinking is sometimes confused with a tendency to question or oppose everything.
However, endlessly searching for possible errors does not necessarily improve decisions.
For example, a new business idea may involve technological uncertainty, market uncertainty, competitive risks, and numerous other potential problems.
Listing these difficulties does not by itself reveal the conditions under which the business could become viable.
Critical Thinking should therefore examine not only what might be wrong but also how important uncertainties could be investigated or addressed.
In some cases, conducting a limited trial and examining the results is more productive than continuing abstract criticism.
Limitation 2 — Different Decisions Require Different Standards of Evidence
The evidence required to justify a judgment depends on the nature of the decision.
For example, improving a webpage that can easily be changed does not require the same level of verification as making a major capital investment.
In early-stage R&D, it may be worthwhile to investigate hypotheses that currently lack substantial supporting evidence.
If only hypotheses already known to be correct were considered, exploring genuinely new possibilities would become difficult.
By contrast, introducing a commercial product may require much more rigorous performance and safety verification.
Critical Thinking should distinguish between generating hypotheses and deciding whether to adopt or implement them.
Limitation 3 — General Reasoning Skills Cannot Replace Specialized Knowledge
An argument may be logically coherent while relying on incorrect specialized information.
For example, evaluating statistical research may require understanding sampling, measurement, and estimation.
Assessing a material or machine may require knowledge of the relevant physical or chemical mechanisms.
Learning general principles of Critical Thinking does not make a person capable of accurately evaluating every specialized field.
Additional investigation and expert knowledge may be necessary.
Limitation 4 — Critical Thinking Cannot Automatically Resolve Value Judgments
Critical Thinking can clarify the evidence supporting a decision.
However, it cannot always determine which outcomes should be considered most desirable.
For example, a company deciding whether to introduce new production equipment may need to consider productivity, cost, safety, employee workload, and environmental impact.
Stakeholders may assign different priorities to these criteria.
Critical Thinking can clarify the assumptions and values underlying their positions.
However, logic alone cannot always determine which values should take precedence.
Critical Thinking in ThinkNavi
ThinkNavi positions Critical Thinking as a fundamental capability relevant to information interpretation, problem formulation, hypothesis development, and decision-making.
Logical Thinking helps organize information and structure reasoning.
Argument Analysis makes relationships among claims, evidence, assumptions, and counterarguments explicit.
The Scientific Method provides approaches for evaluating hypotheses through observation and experimentation.
Critical Thinking operates across these activities by examining the grounds for judgment and considering whether underlying assumptions need to be revised.
For ThinkNavi Free, one possible implementation would provide structured worksheets in which users manually organize a claim, supporting evidence, opposing evidence, alternative hypotheses, uncertainties, and conditions for reaching a decision.
Functions for comparing the arguments supporting different choices could also be considered.
Potential LLM-assisted capabilities include extracting claims and assumptions from documents, proposing alternative hypotheses or counterarguments, and identifying questions requiring further investigation.
However, AI-generated analyses would need to be compared against original sources, with evidence clearly distinguished from speculation.
Critical Thinking could also be applied to the evaluation and verification of hypotheses generated by cross-domain technology exploration systems such as ConceptMiner R&D.
For example, when a cross-domain application hypothesis is generated from technical elements, the system could distinguish components already supported by existing research from those resulting from new inferences.
It could then organize applicability conditions, counterevidence, technological constraints, alternatives, and experiments required for verification.
If the evidence available when the hypothesis was generated is linked to later verification results, it may become possible to trace how the assessment of the hypothesis changes over time.
However, creating numerous evaluation criteria does not automatically determine a hypothesis’s value.
During exploratory research, a hypothesis may deserve investigation even when current evidence is limited.
By contrast, decisions involving substantial R&D investment or commercialization may require stronger verification.
The important distinction is between generating a hypothesis and deciding whether to adopt it, with the required evidence examined progressively as the project develops.
These are potential directions for future development and do not imply that all such features are currently available.
Critical Thinking does not require rejecting established knowledge or expert judgment without justification.
Nor does it guarantee that every problem will yield a certain answer.
One of its most important functions is to clarify what we know, what we do not know, and which assumptions underlie our judgments.
Examining existing knowledge and beliefs is valuable not only for preventing mistakes but also for discovering problems and possibilities that have previously gone unnoticed.
Critical Thinking is not a technique for rejecting other people’s claims. It is a way to improve our understanding and judgment, including our own reasoning. Its central practice is to consider both supporting evidence and alternative possibilities, and to revise conclusions when new information justifies doing so.
Related Topics
- Argument Analysis — Examining Claims, Evidence, and the Validity of Reasoning
- Logical Thinking — Building Sound Arguments, Communicating Reasoning, and Achieving Mutual Understanding
- Induction, Deduction, and Abduction — Discovering Patterns, Drawing Conclusions, and Generating New Hypothese
- The Scientific Method and Falsifiability — Generating Hypotheses and Testing Them Critically
- Epistemology | What Can We Know, and How?
- Cognitive Biases
- Problem Structuring — Breaking Down Problems, Understanding Relationships, and Finding Paths to Solutions
- Logic Trees and Issue Trees — Breaking Down Problems and Identifying the Questions That Need Answers
- MECE — Logical Thinking Through Mutually Exclusive and Collectively Exhaustive Classification, and Its Limitations
- Root Cause Analysis (RCA) — Understanding Why Problems Occur and Preventing Recurrence
- Five Whys | Exploring Root Causes by Repeatedly Asking “Why?”
- Analogy — Connecting Knowledge Across Domains to Generate New Ideas and Hypotheses
- Structure-Mapping Theory | Understanding Relational Correspondences and Generating New Inferences Across Domains
- Systems Thinking — Understanding Interdependence, Feedback, and Change Over Time
- Design of Experiments (DOE) — Investigating Multiple Factors and Their Interactions Through Systematic Experimentation
- What Is Statistics? — Discovering Patterns in Data and Understanding an Uncertain World
- Qualitative and Quantitative Thinking — Two Perspectives on Understanding the World and Their Integration
- Decision Matrices
- Design Thinking — From Human-Centered Problem Discovery to Ideation, Prototyping, and Testing
- Concept Exploration
References and Further Reading
Theories of Critical Thinking
- Stanford Encyclopedia of Philosophy — Critical Thinking
- John Dewey, How We Think, D. C. Heath & Co., 1910.
- Robert H. Ennis, “A Concept of Critical Thinking,” Harvard Educational Review, 32(1), 1962, pp. 81–111.
- Robert H. Ennis, “A Taxonomy of Critical Thinking Dispositions and Abilities,” in Teaching Thinking Skills: Theory and Practice, edited by Joan Boykoff Baron and Robert J. Sternberg, W. H. Freeman, 1987.
- Robert H. Ennis, “A Long Definition of Critical Thinking,” CriticalThinking.NET.
- Peter A. Facione, Critical Thinking: A Statement of Expert Consensus for Purposes of Educational Assessment and Instruction, American Philosophical Association Delphi Report, 1990.
- Richard Paul and Linda Elder, The Miniature Guide to Critical Thinking: Concepts and Tools, Foundation for Critical Thinking.
Cognition and Information Evaluation
- Raymond S. Nickerson, “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises,” Review of General Psychology, 2(2), 1998, pp. 175–220.
- Daniel Kahneman, Thinking, Fast and Slow, Farrar, Straus and Giroux, 2011.
- Sam Wineburg and Sarah McGrew, “Lateral Reading: Reading Less and Learning More When Evaluating Digital Information,” Teachers College Record, 121, 2019.
- University of Chicago Library — The SIFT Method
- Stanford Graduate School of Education — How Experts Evaluate Online Information
Education and Practice
- Philip C. Abrami et al., “Instructional Interventions Affecting Critical Thinking Skills and Dispositions: A Stage 1 Meta-Analysis,” Review of Educational Research, 78(4), 2008, pp. 1102–1134.
- Philip C. Abrami et al., “Strategies for Teaching Students to Think Critically: A Meta-Analysis,” Review of Educational Research, 85(2), 2015, pp. 275–314.
- Diane F. Halpern, Thought and Knowledge: An Introduction to Critical Thinking, 5th ed., Psychology Press, 2014.
- Alec Fisher, Critical Thinking: An Introduction, Cambridge University Press, 2011.
- The University of Hong Kong — Critical Thinking Web