Did you know? An experienced emergency responder may not solve a high-pressure situation by listing ten options and scoring every one. Instead, familiar cues can make a workable action come to mind quickly; the responder mentally simulates what would happen if the action were taken, checks for obstacles, and modifies or rejects it if the scenario does not hold together. Recognition-primed decision making explains this interaction between experience-based situation recognition and the testing of a plausible action. It is neither blind impulse nor proof that experts are always right.
The practical lesson is to use pattern recognition where relevant experience genuinely applies, then test the first workable option against constraints and disconfirming cues before acting. A novice without comparable experience should not be pushed to imitate expert speed. A rare or inconsistent situation calls for deliberate analysis, an external checklist, another specialist or escalation. Time pressure changes the decision process, but it never removes the need for a sensible stop condition when people’s safety is at stake.
This article uses original fictional cases, checked time budgets, a transparent diagnostic exercise and the early naturalistic decision-making research of Gary Klein and colleagues. It distinguishes the descriptive recognition-primed decision model from a rigid procedure that anyone can safely follow. No scenario here reports a real eduKate student, real emergency call or an executed operational decision.
Contents
- Why smart people do not always compare a long menu of options
- What the original fireground research actually observed
- The first stage: a situation evokes a familiar pattern
- The second stage: run the plausible option forward in your mind
- Recognition-primed choices have more than one path
- Worked decision timeline: forty-five seconds is a constraint, not a guarantee
- How repeated experience creates useful patterns
- Why novices need support rather than demands for intuition
- RPD is not the same as a reflex or a hunch
- The disconfirming cue is often more important than the familiar cue
- Groups can coordinate recognition without surrendering dissent
- A classroom transfer: diagnose a mathematical error without guessing
- A second worked case: when the first plan does not fit the remaining time
- How to evaluate decision quality without rewarding lucky speed
- Limits of recognition-primed decision making
- Frequently asked questions
- Original research and next reading routes
Why smart people do not always compare a long menu of options
Traditional classroom decision problems often present three possible actions, a known set of consequences and enough time to compare numerical payoffs. Many real decisions are messier. Information arrives while events change, deadlines matter, goals are not completely defined and experienced practitioners have already seen related patterns. Under those conditions, generating a long menu can impose delay without improving the response.
Recognition-primed decision making, usually abbreviated RPD, is a descriptive model developed within naturalistic decision-making research. It explains how experience can bring a plausible course of action to mind and how a decision maker can check whether it will work. The person does not necessarily rank every option simultaneously. Instead, they recognise the situation, consider an action that fits it, mentally evaluate that action and proceed, adapt or look for another.
That is not a claim that rapid decisions are inherently superior. If the first recognised pattern is wrong, it can lead the decision maker toward the wrong action. Familiarity can become a trap. A good explanation has to include both the speed advantage of accumulated experience and the danger of carrying an old solution into a genuinely new situation.
Gary Klein’s own description of the RPD model explains that the original research focused on experienced fireground commanders. The central assertion is that expertise can generate an acceptable action as the first one considered, not that every experienced person chooses perfectly or that novices should dispense with analysis.
Recognising an acceptable action is different from guessing. It depends on detecting meaningful cues, understanding plausible goals, recognising what is typical or atypical, and evaluating what would probably happen next. The quality of each component determines whether the resulting speed is useful or dangerous.
What the original fireground research actually observed
Klein, Calderwood and Clinton-Cirocco investigated high-pressure decisions made by experienced fireground commanders. Their early study was published as Rapid Decision Making on the Fire Ground in 1986. A later publication with a postscript reports interviews with 26 experienced commanders, a mean of 23 years of experience, and 156 probed decision points in the studied incidents. These figures describe that research sample, not all firefighters or all emergency decisions.
The authors reported little evidence of simultaneous comparison of multiple options at many of those decision points. The 2010 article’s abstract says fewer than 12% involved evidence of concurrent comparison and more than 80% used recognition of a familiar type and a related typical action in the sampled decisions. That finding was influential because it challenged the assumption that skilled choices must routinely look like a matrix of competing alternatives.
The evidence should be read with its limits. Critical-incident interviews reconstruct decisions after the event. Experienced respondents may not remember every microscopic cognitive step or describe it identically across situations. The research highlights plausible regularities in naturalistic expertise, not an immutable law that a person will always choose the first option in every domain.
Later work expanded the model’s attention to mental simulation and the recognition of whether a scenario fits. A primary source is Rapid Decision Making on the Fire Ground: The Original Study Plus a Postscript. It connects the original field study to subsequent interpretation. A different experimental setting might reveal other strategies or different frequencies.
Use historical research to learn a mechanism, not to manufacture a universal speed claim. It would be wrong to tell students that 80% of every expert’s decisions require no thought simply because a particular research sample used recognition often. The distribution of tasks, authority, risk and expertise matters.
The first stage: a situation evokes a familiar pattern
Experience changes what a person notices. A novice reading a complex chart may see many equally important lines. An experienced analyst may notice a small combination of unusual values that changes the interpretation. A chess beginner sees pieces; a strong player recognises a structure that suggests threats and possible moves. These examples illustrate pattern recognition, although their domain-specific evidence should not be treated as identical to fireground-command research.
In RPD, a familiar pattern can include relevant cues, plausible goals, things expected to happen next and actions that have worked in similar cases. The decision maker may not consciously enumerate each component, especially when the situation feels recognisable. Expertise can compress earlier learning into an immediate sense of what kind of situation this is.
That compression is useful only if the underlying patterns remain relevant. A technician trained on one generation of equipment may recognise a fault code from the old version and miss a new safety state. A school tutor may see a familiar algebra sign error but overlook that a student’s real problem is misreading the question. Recognition can accelerate diagnosis, but it can also hide a mismatch if new evidence is forced into an old story.
A practical reader should ask which cues caused the pattern to be recognised. Were they the most diagnostic cues, or simply the most vivid? Would another competent observer notice the same combination? Which contradictory cue would make this classification doubtful? Those questions turn expertise from an unchallengeable intuition into a provisional explanation that can be checked.
A decision maker does not need to stop and write an essay whenever a familiar pattern appears. The point is to make the mechanism accountable. In repeatable low-risk tasks, an immediate response may be entirely reasonable. In consequential or novel tasks, a brief check for surprises can be the difference between reliable expertise and dangerous overgeneralisation.
The second stage: run the plausible option forward in your mind
Recognition alone does not complete the model’s richest account. An experienced practitioner may mentally simulate what will happen if the first plausible option is implemented. They ask whether the action can realistically be carried out, whether important constraints are present and whether the expected consequences make sense. If the simulation reveals a problem, the action can be modified or rejected.
Imagine a coordinator who recognises a familiar event pattern and proposes diverting participants to a second entrance. Before ordering the change, the coordinator mentally walks through the route: Is the second entrance open? Can people with mobility needs use it? Does it create a bottleneck or conflict with an emergency exit? The imagined sequence is not an exact simulation of the future, but it can reveal obvious constraints before an irreversible action.
The critical feature is the willingness to notice a mismatch. If the mental scenario only rehearses reasons why the preferred action will succeed, it becomes confirmation bias rather than a useful check. A stronger simulation actively searches for an obstacle that would make the first option unworkable. It also allows adaptation: perhaps the entrance works only after staff are placed at a junction.
People with deep domain experience may simulate with more relevant details than novices because their memories contain failures, near misses and mechanisms. Yet expertise in one environment does not guarantee valid predictions in another. A fireground decision maker’s pattern knowledge does not automatically translate into cybersecurity incident control or a new medical procedure.
The skill is to pair the speed of recognition with a deliberate question: ‘If we do this, what is the first important thing that could go wrong, and how would we notice?’ That question improves the decision’s testability without requiring an exhaustive search of all imaginable alternatives.
Recognition-primed choices have more than one path
A simple recognised situation may suggest an action that is immediately adequate. A more complicated situation may lead the expert to run a conscious mental simulation of that first action. In a still more difficult case, the simulation exposes a flaw, so the person adjusts the action or examines another workable response. These are variations within the recognitional framework, not proof that any choice made quickly is a recognition-primed decision.
Consider an experienced maintenance operator who hears a familiar vibration and sees the expected temperature pattern. A routine corrective step may come to mind without constructing five competing theories. If another sensor displays a new pattern, the operator needs to stop assuming the case is routine. The action’s plausibility depends on both matching and mismatching observations.
The difference between ‘an action came to mind’ and ‘an action was accepted’ matters. In a high-risk setting, a first proposal should still be filtered through feasibility, policy and consequence. An option can be quickly generated yet quickly rejected. This is not hesitation; it is skilled adaptation to a real constraint.
A checklist can coexist with recognitional judgement. The checklist can demand verification of rare but critical exceptions that the human might miss, such as a known safety interlock. The experienced person’s quick recognition identifies the plausible path, while the checklist guards conditions where trusting a familiar pattern would be dangerous.
Do not turn the model into a mechanical mandate that a person must always choose the first option. The research explains observed strategies under some conditions. A thoughtful team can choose deliberate option comparison when the situation is unfamiliar, the consequences are severe, or sufficient time exists for a meaningful alternatives analysis.
Worked decision timeline: forty-five seconds is a constraint, not a guarantee
Here is a fictional classroom-safe example. A school-event coordinator has 45 seconds to choose a safe crowd-routing instruction after discovering that a planned hallway is temporarily unavailable. The coordinator recognises a familiar detour, spends eight seconds confirming that the alternative route is physically open, twelve seconds mentally tracing its accessibility and bottlenecks, nine seconds communicating the decision and seven seconds checking that staff received it.
The time used is 8+12+9+7=36 seconds. The remaining allowance is 45−36=9 seconds. This is only a hand-calculated timeline; actual tasks do not reliably fit fixed second counts. It shows how recognition and simulation may support an action within a deadline without proving that the instruction is safe or that the relevant route was actually inspected.
Now change one condition. The alternative route’s exit is locked. The twelve-second simulation has discovered a disqualifying obstacle, so executing the first plan is no longer appropriate. The coordinator might need a different authorised route or an immediate pause in movement. A time limit does not justify knowingly choosing an unsafe action.
Imagine instead that the coordinator skips the seven-second confirmation because the mental simulation felt convincing. They might save seven seconds while leaving a critical information gap: the staff may not have received or understood the instruction. An observed completion check is a different kind of evidence from internal confidence. A decision pipeline should record both the proposed action and whether it was executed as intended.
A critic might say this is simply a checklist, but the lesson is subtler. Recognition generates a plausible plan efficiently, mental simulation tests feasibility, and the final communication check confirms a real-world state. All three support speed, but none can be replaced by an ungrounded assertion that the coordinator ‘just knew’.
The time figures here are original invented numbers for arithmetic practice. They should never be used as response-time standards for a real emergency. Safety-critical procedures must follow the responsible institution’s actual training and protocols.
How repeated experience creates useful patterns
Useful patterns do not come from mere repetition of the same action. They come from repeated situations accompanied by relevant feedback about what cues mattered and which consequences followed. If a person sees only successes and never examines near misses, they may become fluent in a procedure without learning its boundaries.
An expert can accumulate a library of examples: familiar conditions, subtle differences, common failures, unusual exceptions and indicators that the usual action will no longer work. Over time the recognition process can become faster. But the learner still needs opportunities to compare cases, receive corrections and encounter changed contexts rather than memorising only one ideal scenario.
Feedback must be informative. Suppose a project leader makes a poor decision but the team succeeds through luck. Treating the outcome as proof of good judgement can reinforce an unreliable cue. The related outcome bias guide explains why reviewing the decision process and information available at the time matters in addition to final results.
Deliberate case comparisons help. Present two situations with similar appearances but different underlying mechanisms and ask what cue separates them. Have learners predict what would happen if the familiar action were applied to the less familiar case. Then reveal the evidence and explain the exception. This trains discrimination instead of just speed.
Expertise may also become stale when the environment changes. A veteran team can be very efficient within an old system and slow to notice that a new regulation, software dependency or customer pattern invalidates its routine. The competency traps guide explores why fluent routines sometimes crowd out exploration.
Why novices need support rather than demands for intuition
A novice has fewer relevant patterns and may not know which cues discriminate safe from unsafe cases. Telling such a learner to ‘trust your instincts’ can reward guessing or overconfidence. A first-year student solving unfamiliar algebra, an intern processing an unfamiliar incident report and a new technician approaching a hazardous device all need explicit structure and feedback before rapid pattern recognition can be dependable.
Begin with first principles and a small set of representative cases. Explain what makes a case typical and what makes it exceptional. Ask the novice to state the goal, identify critical cues and propose a possible action with an explanation of why it could work. Then reveal an overlooked constraint and have them revise their proposed path.
Scaffolds can include decision trees, short checklists, worked examples, comparison tables and supervised simulations. These are not signs of weak intelligence. They provide the external structure that experience will eventually compress into useful patterns. A checklist can remain important even for experts when rare safety gates are easy to forget.
Teachers should not demand a shorter response time before understanding becomes stable. A child who arrives at an answer slowly but explains the crucial cue may be building a better foundation than a child who gives a quick answer by matching superficial features. Time pressure can be introduced later when independent accuracy and changed-context transfer are already observable.
Even after training, novices and experts should work within the appropriate authority boundary. A student may analyse a fictional emergency scenario but should not be asked to direct real responders. A junior analyst can propose a response while the organisation retains accountable review for consequential actions.
RPD is not the same as a reflex or a hunch
A reflex is a rapid physiological response to a stimulus. RPD concerns a cognitive process in which a familiar situation evokes a plausible action and the person may mentally evaluate it. A hunch, in everyday language, could refer to anything from useful experience to an unexplained guess. Without examining the cues and imagined consequences, calling an action a hunch hides the mechanism.
A fast option is not necessarily the first thing the brain randomly produces. Experience can make the first considered action meaningful because prior cases have linked certain cues to successful or unsuccessful paths. The alternative to exhaustive simultaneous comparison is not necessarily irrationality. But the first option can still be wrong if the pattern was misclassified.
Analytical reasoning remains useful. A team designing a long-term infrastructure investment may have months, financial data and a need to compare scenarios with documented costs. Insisting on recognition-primed selection would be inappropriate. The choice of decision strategy should reflect time, reversibility, expertise and the degree of novelty.
Nor is RPD equivalent to the idea that every decision involves consciously simulating a movie of future events. Some recognised situations evoke routine actions with little explicit simulation. The research model includes variations, and people may combine recognitional, procedural and analytical methods in ways that are difficult to separate through self-report.
Useful language preserves distinctions: ‘the cue suggested a familiar response’, ‘the response was tested against a likely consequence’, and ‘the response was authorised and executed’. Collapsing all three into ‘the expert knew what to do’ loses the steps that a learner can practise and a reviewer can check.
The disconfirming cue is often more important than the familiar cue
Suppose a warehouse operator recognises a familiar shipping delay and expects that changing the pickup slot will solve it. Yet the shipment’s temperature record shows a product-storage excursion. The delay pattern is familiar, but the additional evidence creates a different and potentially more consequential problem. A good recognitional decision process must be sensitive to the cue that makes the old response inappropriate.
Train people to name an expected observation and a disconfirming observation. If the situation truly matches the pattern, what should be visible? If it is a different mechanism, what might look different? This small exercise helps prevent a known story from absorbing every new fact. It also encourages a team member to share inconvenient observations without fear of slowing an expert down.
An action’s imagined success is not a proof that it will work. Mental simulation can overlook hidden constraints, unusual combinations or physical dynamics beyond the person’s experience. External sensors, second opinions, checklists or small reversible tests can reveal problems that private imagination misses. They are especially valuable when consequences are serious.
Instead of arguing that intuition is either magical or useless, ask what quality of evidence the decision maker used. An experienced mechanic’s observation may be highly diagnostic; an inexperienced observer’s confidence may be based on superficial similarity. The truth of a claim comes from how it matches the mechanism and observations, not from the speed or certainty of its presentation.
Under uncertainty, it is appropriate to pause and seek new information when the recognised pattern breaks. A system that always insists on acting immediately can turn a strength of expertise—recognising familiar cases—into a weakness—refusing to recognise exceptions.
Groups can coordinate recognition without surrendering dissent
In a team, different members may recognise different patterns from the same event. A logistics specialist sees an inventory constraint, an accessibility coordinator sees a route problem and a finance officer sees a contractual deadline. A fast decision can benefit from a well-defined process for surfacing the few cues that would invalidate the proposed action.
The leader can ask for a concise challenge rather than convene an endless debate: ‘What observable fact would make this route unsafe?’ ‘Who has current evidence about that constraint?’ ‘Is there any critical requirement that the first proposal violates?’ These questions focus attention on disconfirming information without pretending there is enough time for a full numerical ranking of every conceivable option.
Psychological safety matters because people may remain silent in the face of a confident senior expert. The already published Abilene paradox guide explores false agreement, while RPD concerns the generation and evaluation of a plausible action. A team can combine rapid recognition with an explicit right to flag a safety-critical mismatch.
Roles should be clear. A person may raise a contradictory cue, an accountable decision owner may select the response, and an independent observer may verify what happened. Good group design makes the handoff visible. It does not magically turn all group discussion into high-quality expertise.
After action, record what was recognised, which alternative was rejected if any, what evidence changed the decision and what happened. Such reviews help the team learn whether its shared pattern library is improving or simply repeating familiar narratives.
A classroom transfer: diagnose a mathematical error without guessing
Consider a Secondary 1 student who writes 3(x+2)=3x+2. A tutor may immediately recognise a distributive-law misconception. That recognition suggests a useful instructional action: build a concrete example, expand both terms and compare the two expressions. But the tutor should still check whether the mistake came from the distributive law, an incomplete copy or the student’s misunderstanding of multiplication.
Use a quick test. Let x=4. The original expression is 3(4+2)=3×6=18, while the student’s expansion gives 3×4+2=14. The values differ by four. A correct expansion is 3x+6, which at x=4 becomes 12+6=18. This arithmetic illustration supplies an independent check rather than relying on the tutor’s ability to recognise a familiar handwriting pattern.
Now change the example to 2(5−x). The student might expand to 10−2x, which is correct, or to 10−x, which repeats a related mistake. The tutor can ask the learner to verify with x=3: the original is 2(5−3)=4; the correct expansion 10−6=4; the wrong expansion 10−3=7. The changed case tests transfer.
The tutor’s first recognised explanation becomes more credible when the changed-context test supports it. If the learner can reliably expand but copied the question incorrectly, the correct remedy is different: slow down and map the expression accurately before solving. Good teaching combines expert pattern recognition with student evidence, not an assumption that every repeated-looking wrong answer has the same cause.
Nothing here suggests that the psychological RPD literature directly proves one teaching method raises examination grades. This is a clearly labelled analogy demonstrating how an expert’s first diagnosis should be checked before personalised instruction begins.
A second worked case: when the first plan does not fit the remaining time
Imagine a fictional project team with 24 minutes to restore a disrupted presentation system. The first familiar repair is estimated to require an eight-minute diagnosis, six-minute configuration change, seven-minute test and five-minute independent verification. The arithmetic total is 8+6+7+5=26 minutes, which exceeds the available window by two minutes even before unpredictable delays.
A recognition-primed thinker can generate this repair quickly, but mental simulation should expose the timing conflict. The right response is not to quietly skip the independent test and announce success. The team might switch to a pre-authorised rollback with its own tested steps or use a contingency presentation method. Whether those options are permitted depends on the actual organisation and task.
Suppose a separate pre-tested fallback has five-minute activation, four-minute confirmation and three-minute stakeholder communication, totalling 5+4+3=12 minutes. In the fictional timeline this fits within 24 minutes, leaving twelve minutes. But it is acceptable only if the fallback is authorised, known to work for the current system and meets the event’s requirements.
The numerical calculation is trivial compared with the cognitive challenge: recognising that the first familiar repair no longer fits the true constraint. A novice might blindly apply the old repair, while an experienced person should detect the mismatch early. Yet even an expert can misjudge durations or overlook dependencies, so observed runbooks and test receipts remain valuable.
A useful debrief records why the option was rejected and whether the fallback completed as expected. If the team later discovers that the fallback also failed, it should revise its experience base rather than celebrate the original fast recognition. Decision quality is visible in both the process and the eventual evidence.
How to evaluate decision quality without rewarding lucky speed
Counting the number of seconds until an answer appears is not enough. A fast wrong action can be costly; a deliberate choice may be appropriate when the situation is novel. Evaluation should include whether relevant cues were noticed, whether the action respected constraints, whether disconfirming information changed the plan and whether the actual outcome was verified.
Build a set of varied cases rather than only repeating the same successful scenario. Include familiar cases, ambiguous cases and deliberate exceptions. Ask decision makers to state the first plausible response and the cue that would invalidate it. This reveals whether they have learned discriminative features or merely memorised a script.
Suppose a fictional trainee completes five such cases. Three decisions meet the stated safety and explanation criteria, while two fail. The observed success share is 3/5=60% in that tiny invented assessment. It is not an estimate of real-world emergency competence. The two failures may be more informative than the percentage if they expose a specific high-risk blind spot that requires more training.
Use independent observers and consistent rubrics where feasible, and distinguish a valid process from a lucky outcome. A decision made on poor evidence can succeed by chance; a well-reasoned decision can produce an adverse outcome because of an unforeseeable event. A high-quality review should preserve the information actually available when the choice was made.
Remember that a review can itself create incentives to narrate a polished story. Ask for original timestamps, communications and observations when important. Do not manufacture detailed mental processes after the fact. The most trustworthy learning is often a modest explanation of what is known, uncertain or missed.
Limits of recognition-primed decision making
The RPD model is useful in environments with experienced decision makers, meaningful patterns and time pressure. It does not guarantee correct decisions, nor should it be forced onto domains with unreliable feedback or constantly changing rules. Strong expertise can sometimes produce confident errors when unfamiliar conditions resemble familiar ones.
Mental simulation is constrained by knowledge and imagination. A person may be unable to foresee interactions they have never encountered. Formal analysis, simulation software, controlled trials and checklists can be preferable when the stakes or complexity demand them. Choosing RPD over an analytical process should itself be a reasoned decision.
Studies based on naturalistic interviews or particular occupations do not establish a universal percentage of recognitional choices across every profession and learner. Treat empirical frequencies as observations within specific samples, with methods and selection limits. Educational transfer advice is practical inference unless directly tested in a matching study.
Rapid action can also be unsafe when the operator lacks the required authority. A student recognises a plausible cybersecurity fix but cannot deploy it to a shared system without permission. An employee recognises a financial anomaly but should follow the organisation’s review procedure before changing accounts. Expertise does not erase governance.
This article is not emergency-response guidance. Real safety-critical decisions must be made by trained responsible professionals using official procedures and current situational evidence. The fictional examples exist to teach discrimination, mental simulation and responsible escalation.
Frequently asked questions
What is recognition-primed decision making? It is a model of how experienced people may recognise a situation, generate a plausible action and mentally test it rather than compare many options at once.
Does the first idea always win? No. A mental simulation may expose an obstacle, leading to modification, rejection or consideration of another action.
Is this the same as intuition? Not exactly. The model describes a possible mechanism connecting experience, relevant cues, plausible action and testing of consequences. Everyday intuition is a much broader term.
Can a beginner use RPD safely? Novices should rely on explicit instruction, supervised examples, checklists and evidence. They can learn the component questions, but should not imitate expert speed without relevant experience or authority.
Does a fast choice mean no reasoning happened? No. Recognition can draw on considerable learned knowledge, and mental simulation may occur rapidly.
When should a team use deliberate comparison instead? It is especially useful when the task is unfamiliar, the consequences are irreversible, the choice can be paused or the team lacks dependable recognitional expertise.
What is the most important safety check? Ask which observable cue would make the first plausible action wrong, then verify critical conditions before executing.
How can a tutor apply the lesson? Treat an immediately recognised student error as a hypothesis and test it with a changed-context problem before choosing the repair.
Can a successful result prove the original decision was good? Not on its own. Evaluate the process and information available at the time, while also learning from the actual outcome.
Original research and next reading routes
The original empirical article is Klein, Calderwood and Clinton-Cirocco (1986), Rapid Decision Making on the Fire Ground. The research was revisited in Rapid Decision Making on the Fire Ground: The Original Study Plus a Postscript (2010). The author’s Recognition-Primed Decision model overview provides a concise account of its intended mechanism.
For practical reflection on how knowledge is elicited from experts, the Critical Decision Method for Eliciting Knowledge examines the study of expertise in naturalistic environments. Use its methods as research context, not as a permit to reproduce emergency decisions without training.
Students can practise the mathematics example by distinguishing a pattern-based diagnosis from the changed-number evidence. Teachers can collect cases that differ on one critical cue. Teams can keep short records of proposed actions, rejected paths, disconfirming information and verified outcomes. These activities are practical learning suggestions, not claims of measured effectiveness from the original firefighter study.
For wider connections, visit How Intelligence Works, Competency Traps and The Gold Standard of Education. The research sources establish the descriptive model; the examples and numerical checks in this article are invented to help readers transfer its ideas responsibly.