Educational systems regularly adopt learning methods that appear effective because they produce visible improvements in student performance, engagement or course completion. A particular teaching strategy may become popular across schools, universities and digital learning platforms because students respond positively to it, educators find it convenient and institutions can implement it at scale. However, the widespread adoption of a learning method does not necessarily mean that it is beneficial in every educational context. A strategy that improves short-term performance may also create weaknesses that become visible only when students encounter unfamiliar problems, advanced subjects or real-world professional situations.
This creates an important challenge for modern education: how can an educational system determine whether a popular learning method is genuinely strengthening students’ capabilities or gradually creating weaknesses across an entire learner population? The problem is particularly difficult because systemic weaknesses do not always appear as immediate failures. Students may continue achieving good examination scores while developing limited independent reasoning, weak knowledge transfer, excessive dependence on hints or difficulty applying concepts outside familiar examples. Since the method appears successful according to conventional performance indicators, these weaknesses may remain undetected for years.
The concept of Learning Method Systemic Risk Detection offers a way to examine this problem. Instead of evaluating teaching methods only through their immediate outcomes, educational systems can investigate their long-term effects on knowledge, reasoning, practical application, independence and adaptability. The objective is not to reject popular learning strategies or assume that widely used methods are harmful. It is to understand whether a method creates unintended consequences when applied repeatedly, across different subjects or to large groups of students.
For an education ecosystem such as EasyShiksha, this idea creates an opportunity to think beyond course completion and assessment scores. Online courses, quizzes, projects, games, internships, certificates, university programmes and career guidance can provide different forms of evidence about how students are developing. When these experiences are considered together, an intelligent learning platform could potentially identify patterns suggesting that a particular instructional approach is producing strong immediate results but weaker long-term capabilities.
The future of educational intelligence may therefore depend not only on identifying which learning methods work, but also on detecting when successful methods begin creating hidden weaknesses.
Understanding Systemic Weaknesses in Education
A systemic weakness is a recurring limitation that develops across a group of learners, educational programmes or an entire learning environment. It differs from an individual knowledge gap because it may originate from the structure of the teaching process rather than from a student’s personal circumstances. When the same weakness appears repeatedly among learners exposed to a similar instructional approach, it becomes important to investigate whether the method itself contributes to the pattern.
For example, imagine that a large number of students learn programming through tutorials that demonstrate every step of building an application. Students watch the instructor write code, reproduce the same instructions and successfully complete the exercises. Their completion rates may be high, and their quiz performance may suggest that they understand the material. However, when asked to build a small application without a tutorial, many students may struggle to decide how to begin. They may know individual programming commands but lack the ability to combine them independently.
The weakness in this situation is not necessarily a lack of exposure to programming concepts. It may be an overdependence on demonstrated procedures. If the same pattern appears across multiple courses and student groups, the issue could be connected to the instructional structure.
Systemic weaknesses can emerge in many forms. Students may become skilled at recognising correct answers but less capable of producing answers independently. They may perform well when questions follow familiar patterns but struggle with unfamiliar applications. They may remember information for an examination but fail to retain it over longer periods. They may become comfortable with automated explanations while finding it difficult to evaluate evidence or construct their own arguments.
These weaknesses are particularly challenging because conventional educational measurements often focus on visible achievement. Examination results, completion percentages and certificates can show whether learners have completed specific activities, but they do not always reveal how independently students can use their knowledge.
A more comprehensive educational system must therefore examine not only what students achieve, but also the capabilities they develop while achieving it.
Why Popular Learning Methods Deserve Continuous Evaluation
Popular learning methods often become widely adopted for legitimate reasons. A strategy may improve accessibility, reduce instructional costs, make difficult subjects easier to understand or help educators manage large groups. Digital platforms may adopt interactive quizzes because they provide immediate feedback, while institutions may use recorded lectures because students can access them at convenient times.
The problem begins when popularity becomes a substitute for continuous evaluation. Once a method becomes established, its underlying assumptions may receive less scrutiny. Educators may continue using it because it is familiar, students may expect it and institutions may have invested in the technology required to support it. Its success may be measured through indicators that were originally selected to justify its adoption.
This can create an educational feedback loop. The method produces positive short-term results, the results encourage wider adoption and increased adoption makes the method appear increasingly legitimate. Meanwhile, weaknesses that are not measured remain outside the institution’s understanding.
Consider an educational platform that introduces frequent multiple-choice quizzes to improve engagement. Students receive immediate scores, learn to recognise important terms and become more familiar with examination formats. These outcomes can be valuable. However, if the entire assessment environment relies heavily on recognition-based questions, students may receive fewer opportunities to explain concepts, solve unfamiliar problems or construct arguments independently.
The problem is not that multiple-choice quizzes are inherently ineffective. It is that excessive dependence on one assessment format may narrow the range of capabilities being developed and measured.
Continuous evaluation allows educational systems to distinguish between a useful learning method and an approach that has become overused or poorly matched to particular learning objectives.
The Difference Between Individual Learning Gaps and Method-Induced Weaknesses
One of the most important requirements of systemic risk detection is distinguishing individual learning difficulties from weaknesses associated with an instructional method. A student may perform poorly because of limited prerequisite knowledge, insufficient study time, language barriers, unfamiliarity with technology or personal circumstances. These factors should not automatically be interpreted as evidence that the teaching method is defective.
Method-induced weaknesses require a different level of analysis. The system must investigate whether students exposed to a particular learning experience demonstrate similar limitations more frequently than comparable learners who experienced a different approach.
For example, suppose students who learn a scientific concept through demonstrations consistently perform well on questions that reproduce the demonstrated examples. However, they struggle when asked to apply the same principle to a new situation. This pattern alone does not prove that demonstrations caused the weakness. Students may have had different prior knowledge, received different levels of support or encountered assessments with varying difficulty.
A stronger investigation would compare learning outcomes across groups with similar starting conditions and examine whether the instructional approach is associated with differences in independent application. It would also consider whether the weakness appears across multiple assessments rather than relying on one examination.
This distinction is essential because an intelligent educational system should not label a method as harmful based on a superficial correlation. It should identify patterns, investigate alternative explanations and communicate uncertainty.
For EasyShiksha, this could mean analysing learning data at multiple levels. Individual learner analytics could identify personal knowledge gaps, while aggregated course analytics could reveal whether a particular instructional design is associated with recurring difficulties across a wider population.
How Short-Term Success Can Conceal Long-Term Weaknesses
A major reason systemic weaknesses remain hidden is the difference between immediate performance and long-term capability. Many learning methods are evaluated shortly after instruction, when students still remember the examples, explanations and procedures they have recently encountered.
A learner may complete a lesson and answer related questions correctly because the information is still available in working memory. However, the same learner may struggle several weeks later when the information must be recalled without assistance. Similarly, a student may successfully reproduce a demonstrated process but fail to adapt it when the task changes.
This distinction becomes important when educational systems measure learning primarily through immediate completion tests. A method may appear effective because it helps students perform well directly after instruction, even though it does not produce equally strong retention or transfer.
Delayed assessments can help reveal this difference. Students can be asked to revisit concepts after a period of time, apply them in new contexts or explain how they would solve a related problem without referring to the original lesson. The resulting evidence can show whether knowledge remains accessible and usable beyond the immediate learning environment.
The same principle applies to professional skills. A student may complete a course on project management and answer theoretical questions correctly. However, the ability to manage changing deadlines, coordinate responsibilities and respond to unexpected problems requires a different form of capability. Practical projects and internships can help reveal whether the knowledge transfers into more complex environments.
An educational system that measures only immediate success risks confusing successful exposure with durable learning.
The Role of Learning Analytics in Detecting Systemic Risk
Learning analytics can help educational systems identify patterns that would be difficult to detect through individual assessment results. A single student’s incorrect answer may not reveal much about the quality of an instructional method. However, thousands of responses collected across similar learning experiences may expose recurring patterns.
For example, analytics could examine whether students who complete a particular course structure show unusually high performance on familiar questions but weaker results on application-based assessments. It could investigate whether learners repeatedly request hints at the same stage, whether they struggle with particular types of unfamiliar problems or whether their performance declines when automated guidance is removed.
These patterns can provide early signals of potential instructional weaknesses. However, the system must analyse them carefully. A high number of hint requests may indicate that students need better explanations, but it could also mean that the questions are appropriately challenging. A decline in performance may reflect a more demanding assessment rather than a failure of the teaching method.
The purpose of learning analytics should therefore be to identify questions that require investigation, not to make automatic judgments about educational quality.
An advanced system could combine assessment results with information about learning activities, revision patterns, project performance and delayed retention. It could compare students with similar prior knowledge and examine whether certain instructional approaches are associated with consistent differences in capability development.
For EasyShiksha, this creates a potential opportunity to connect information from courses, quizzes, projects and internships. Rather than treating each activity as an isolated event, the platform could explore how different learning experiences contribute to a student’s developing capabilities.
Detecting the Overuse of a Single Learning Method
One common source of systemic weakness is excessive dependence on a single teaching or assessment approach. A method may be effective for introducing a concept but insufficient for developing advanced understanding. Problems emerge when the same approach is used for every stage of learning.
Recorded video lessons, for example, can make educational content accessible to students who cannot attend live classes. They allow learners to pause, replay and revise explanations at their own pace. Yet video consumption alone cannot establish whether students can apply the information independently. If an entire learning programme relies primarily on watching lessons, students may receive limited opportunities to practise retrieval, solve problems or explain concepts in their own words.
Similarly, gamification can improve engagement by introducing challenges, progress indicators and rewards. However, if students become primarily motivated by points and completion targets, they may focus on maximising rewards rather than understanding the underlying subject. This does not mean that educational games should be avoided. It means that game mechanics must remain connected to meaningful learning objectives.
AI-generated explanations present another example. They can make difficult concepts easier to approach and provide personalised assistance. However, if students depend on generated explanations without learning how to verify claims, evaluate sources or construct their own reasoning, their confidence may increase faster than their independent capability.
A healthy educational system should therefore evaluate the balance between instructional methods. It should examine whether learners receive appropriate opportunities to observe, practise, retrieve, explain, question and apply knowledge.
The Method-to-Capability Gap
A useful way to investigate instructional effectiveness is to examine the relationship between the learning method and the capability it is intended to develop. This relationship can be described as the method-to-capability gap.
The gap appears when a teaching method produces the expected immediate activity but does not sufficiently develop the capability that activity is supposed to represent. A student may complete a coding tutorial without developing independent programming ability. A learner may finish a communication course without becoming more effective in real conversations. A student may complete a research methodology module without being able to evaluate the reliability of a research paper.
The method-to-capability gap becomes especially important when educational institutions use activity completion as evidence of competence. Completing a course demonstrates participation and may indicate progress, but it does not automatically prove mastery.
A stronger evaluation process begins by defining the capability that students should develop. The system then examines whether the chosen instructional method creates opportunities to demonstrate that capability under appropriate conditions.
For example, if the objective is independent problem-solving, students must eventually solve problems without being shown every step. If the objective is research literacy, students should evaluate evidence rather than simply memorise research terminology. If the objective is professional communication, students should have opportunities to communicate in realistic situations.
This approach encourages educational platforms to measure learning through demonstrated capability rather than relying exclusively on the activities used to deliver it.
AI as an Early-Warning System for Educational Weaknesses
Artificial intelligence could potentially help educational systems identify systemic weaknesses earlier by analysing patterns across large volumes of learning data. Unlike conventional reporting systems that may present course completion and average scores, AI-based analysis could examine relationships between instructional methods, assessment formats, learning behaviour and later performance.
For instance, an AI system might identify that students who complete a particular sequence of lessons perform strongly on immediate recall assessments but repeatedly struggle with unfamiliar applications. It could also detect that learners from several courses show similar difficulties after receiving highly guided instruction.
These findings could generate an early-warning signal for educators. The system might recommend reviewing the instructional sequence, introducing more independent practice or investigating whether the assessment design accurately measures the intended capability.
However, AI should not independently declare that a teaching method is harmful. Educational data contains many contextual factors, and automated models can confuse correlation with causation. If students using a particular learning method also have less prior knowledge or fewer study resources, the system may incorrectly attribute their performance to the method itself.
Human educators and researchers must therefore review important findings, examine the underlying evidence and determine whether further testing is necessary.
The most responsible role for AI is to make hidden patterns visible and help institutions ask better questions. It should support educational judgment rather than replace it.
Designing a Learning Method Risk Detection Framework
A proposed Learning Method Risk Detection Framework could help platforms evaluate instructional approaches across multiple stages of student development. The framework would begin by identifying the method being used and the educational objective it is intended to achieve. This is important because no learning method should be evaluated without considering its purpose. A technique designed to introduce basic terminology should not be judged by exactly the same standards as one intended to develop advanced problem-solving.
The next stage would examine the evidence generated during and after instruction. Immediate quiz results could provide information about initial understanding, while delayed assessments could reveal retention. Application-based exercises could show whether learners can transfer knowledge to unfamiliar situations. Projects and internships could provide evidence of how academic knowledge performs in practical environments.
The system could then compare outcomes across different learner groups and instructional approaches. It would need to account for differences in prior knowledge, course difficulty, learner access and other relevant conditions. Where possible, institutions could use carefully designed comparative studies to investigate whether a particular method contributes to the observed weakness.
Once a possible risk pattern is identified, educators could review the instructional design and introduce targeted changes. These might include additional practice, alternative explanations, more open-ended questions or opportunities for independent problem-solving. The system would then monitor whether the revised approach produces stronger outcomes over time.
This framework should operate as a continuous evaluation process rather than a one-time audit. Learning methods evolve, student populations change and educational technologies introduce new forms of interaction. A method that works well in one context may require adjustment when applied to another.
The Importance of Multiple Assessment Formats
A major challenge in detecting systemic weaknesses is that assessment methods can reproduce the same limitations as the teaching methods they are supposed to evaluate. If students learn through repeated multiple-choice exercises and are assessed through similar questions, the system may measure familiarity with the format rather than the full range of subject understanding.
A more balanced assessment environment can help reveal capabilities that a single format may overlook. Multiple-choice questions can be useful for checking conceptual recognition and factual understanding. Written explanations can show whether students can organise their reasoning. Practical exercises can reveal whether they can apply knowledge. Projects can demonstrate how they handle complex tasks, while oral discussions can help educators examine how learners respond to questions and defend their conclusions.
No single assessment format provides a complete picture of capability. The objective is to select assessment methods that match the intended learning outcomes and provide different forms of evidence.
For EasyShiksha, quizzes could serve as one component of a wider assessment architecture. A student might first complete a knowledge check, then attempt a problem-solving activity, followed by a project that requires independent application. The resulting evidence could provide a more complete picture of development than a quiz score alone.
This approach would also help identify whether a particular instructional method is producing narrow success. If students perform well in one assessment format but consistently struggle in other relevant formats, educators would have a reason to investigate the difference.
Learning Method Risk in Digital Education
Digital education creates new opportunities for personalised learning, but it also makes certain instructional patterns easier to repeat at scale. A platform can deliver the same video lesson, quiz format, automated hint structure or gamified activity to thousands of students. This consistency can improve accessibility and operational efficiency, but it can also distribute the weaknesses of a poorly designed method across a large population.
A digital learning platform may introduce a feature because it increases engagement or reduces course abandonment. However, engagement does not necessarily indicate deeper learning. Students may spend more time on a platform because they enjoy its interface, while their ability to apply knowledge remains unchanged.
Similarly, an AI tutor may receive positive feedback because it provides quick explanations. Yet students may gradually rely on the assistant to complete reasoning tasks that they should eventually perform independently. If the platform measures satisfaction without measuring independent performance, the long-term effect may remain invisible.
Digital education therefore needs a broader definition of success. Engagement, accessibility and convenience are important, but they should be considered alongside retention, transfer, independent reasoning and practical capability.
The advantage of digital platforms is that they can collect evidence across multiple learning activities. When used responsibly, this information can help identify whether a popular feature is supporting meaningful development or simply improving short-term interaction metrics.
The Connection Between Systemic Weaknesses and Career Readiness
The consequences of method-induced weaknesses may extend beyond academic performance. Students eventually need to use their knowledge in situations where instructions are incomplete, problems are unfamiliar and solutions are not immediately available.
A student who has learned programming entirely through step-by-step demonstrations may find it difficult to debug an unfamiliar application. A learner who has studied business through memorised case-study answers may struggle when asked to make a decision using incomplete information. A student who has relied on AI-generated summaries may find it difficult to evaluate original research or explain the reasoning behind a conclusion.
These limitations can become visible during internships, recruitment assessments and early professional responsibilities. Employers may discover that candidates have completed relevant courses but need additional support when performing independent tasks.
This makes systemic learning risk a career-readiness issue as well as an academic concern. Educational systems should investigate whether their methods prepare students to operate beyond the structured conditions in which they learned.
EasyShiksha’s combination of courses, projects, internships and career guidance could potentially support this broader perspective. Course assessments can indicate initial understanding, projects can test independent application and internships can reveal how knowledge transfers into workplace environments. Career guidance can then help students identify capabilities that require further development.
The objective is not to make every student an expert in every area. It is to ensure that educational experiences build capabilities that remain useful when the learner moves into unfamiliar environments.
The Role of Projects and Internships in Validating Learning Methods
Projects and internships can provide valuable evidence because they expose students to tasks that are less predictable than conventional lessons. A project may require a learner to combine several concepts, make decisions and revise an approach when the original plan does not work. An internship may introduce professional expectations that cannot be fully reproduced through a standardised quiz.
These experiences can reveal whether a learning method has developed transferable capability. If students who perform well in a course repeatedly struggle with practical tasks, the educational system should investigate the difference. The cause may be insufficient practice, limited exposure to complex problems or an instructional method that provides too much guidance.
Projects can also help educators identify which parts of a learning programme require improvement. If students consistently struggle with the same stage of a project, the issue may relate to a missing foundational concept or an instructional sequence that does not adequately prepare learners for independent work.
Internships provide an additional perspective because workplace environments require communication, collaboration, time management and adaptation. These capabilities may not be visible in academic assessments but are essential to professional development.
For EasyShiksha, connecting course learning with projects and internships could create a valuable feedback mechanism. Practical outcomes could help evaluate whether the methods used in online learning are preparing students for more demanding tasks.
Why Educational Systems Must Avoid Overcorrecting
Detecting a possible systemic weakness does not mean that an educational system should immediately abandon the method associated with it. Every learning approach has strengths and limitations, and a method that is insufficient on its own may work effectively when combined with other strategies.
For example, recorded lectures may be useful for introducing complex information, even if they need to be supplemented with retrieval practice and projects. Multiple-choice quizzes may efficiently assess factual understanding, even though they cannot measure every form of reasoning. AI tutoring may help students overcome initial confusion, provided that learners also receive opportunities to practise independently.
An excessive reaction could remove useful educational support and create new problems. If a platform responds to concerns about guided learning by eliminating all hints, beginners may become frustrated and lose access to assistance that helps them understand difficult concepts. If it reduces gamification without considering its role in motivation, engagement may decline without any improvement in learning quality.
The objective should be instructional balance rather than replacing one popular method with another. Educational systems should identify which outcomes a method supports, which outcomes it does not adequately develop and what complementary experiences are required.
A mature learning environment does not search for one universally perfect teaching method. It builds an appropriate combination of approaches around the capabilities students need to develop.
Data Privacy and Responsible Learning Surveillance
Systemic risk detection requires educational data, but collecting more information does not automatically produce better educational decisions. Learning platforms may analyse quiz responses, time spent on tasks, hint usage, course progress and project performance. These signals can help identify patterns, but they can also reveal sensitive information about a student’s educational behaviour.
Students should understand what information is collected, why it is analysed and how the results may influence recommendations or institutional decisions. Data should be used to improve educational experiences rather than create permanent labels about students or educators.
A system that identifies a learner as dependent on assistance, for example, should not treat that description as a fixed characteristic. The behaviour may reflect the design of a particular course, unfamiliarity with the subject or a temporary learning requirement. Similarly, educators should not be judged through simplistic performance comparisons that ignore differences in student populations and teaching conditions.
Aggregated analysis can help identify broader patterns while reducing unnecessary exposure of individual learner records. Where individual data is required, access should be limited to legitimate educational purposes and handled with appropriate security and transparency.
Responsible learning intelligence must protect both the quality of education and the rights of the people participating in it.
Measuring Whether a Learning Method Is Actually Improving
A learning method should be evaluated through outcomes that match its educational purpose. If the objective is conceptual understanding, students should be able to explain the concept accurately. If the objective is long-term retention, they should demonstrate knowledge after an appropriate delay. If the objective is independent problem-solving, they should be able to complete relevant tasks without excessive guidance.
The evaluation should also examine transfer. Students should be able to use what they have learned in new situations rather than only reproduce familiar examples. Practical performance can provide additional evidence of whether academic knowledge is becoming useful capability.
Improvement should be measured over time and across relevant learner groups. A method may work well for beginners but become insufficient at advanced levels. It may be suitable for one subject but less effective for another. It may improve engagement while requiring additional support to strengthen retention.
This means educational evaluation should avoid reducing effectiveness to a single score. A method can produce positive results in one dimension and limitations in another. A more complete evaluation considers the combination of outcomes and identifies where instructional adjustments are needed.
For EasyShiksha, this could support a shift from measuring only course completion toward understanding what learners can remember, explain, apply and demonstrate after completing a learning experience.
EasyShiksha and the Opportunity for Learning Method Intelligence
EasyShiksha’s broader educational ecosystem provides a useful context for exploring Learning Method Risk Detection. Online courses introduce concepts, quizzes provide assessment, games create interactive challenges, projects encourage application, internships provide practical exposure, certificates document learning milestones and career guidance helps learners connect their development with future opportunities.
These experiences can be considered as different stages of capability development. If they are connected through a consistent learning record, the platform could potentially investigate how instructional methods influence outcomes across the student journey.
For example, a course may use video demonstrations to teach a technical skill. A quiz could check initial understanding, while a practical project could test whether students can apply the skill independently. If learners consistently perform well in the quiz but struggle with the project, the system could flag a possible gap between recognition and application.
Educators could then examine the course structure and determine whether additional independent exercises, problem-solving activities or practical examples are needed. Future cohorts could be monitored to evaluate whether the changes improve outcomes.
A similar approach could be applied to communication courses, business programmes, digital marketing, finance, cybersecurity and other learning areas. The important principle is that course design should be evaluated through the capabilities it develops, not only through the activities students complete.
This could help EasyShiksha move toward a more evidence-informed learning ecosystem in which instructional methods are continuously examined and improved.
The Future of Educational Quality Assurance
Traditional quality assurance often evaluates educational programmes through curriculum coverage, teaching resources, examination results and institutional processes. These remain important, but the growing complexity of digital learning creates a need for more continuous evaluation.
Future educational systems could use learning analytics and AI to monitor whether instructional approaches produce consistent outcomes across different learner groups. They could identify possible gaps between immediate performance and delayed retention, compare assessment results with practical performance and detect recurring difficulties that appear across courses.
This would create a more dynamic form of educational quality assurance. Instead of waiting for a programme review or an examination cycle to reveal problems, institutions could receive earlier signals that a method may need investigation.
However, technology should not become the sole authority on educational quality. Student feedback, educator experience, curriculum expertise and independent evaluation remain essential. Quantitative data can show patterns, but understanding why those patterns exist often requires human interpretation.
The strongest future model is likely to combine automated analysis with educator judgment. AI can help identify where to look, while teachers and researchers determine what the evidence means and how learning environments should change.
Conclusion: Educational Success Requires the Ability to Detect Its Own Weaknesses
A popular learning method can produce genuine educational benefits while also creating limitations when it is used too extensively or without appropriate complementary experiences. High completion rates, strong immediate quiz scores and positive student feedback do not always demonstrate that learners have developed durable knowledge, independent reasoning or practical capability. Some weaknesses become visible only when students encounter unfamiliar problems, delayed assessments or real-world responsibilities.
The concept of Learning Method Systemic Risk Detection offers a way to investigate these hidden consequences. By analysing learning patterns, comparing assessment formats, examining delayed retention and connecting academic outcomes with projects and internships, educational systems could identify whether a particular approach is producing unintended weaknesses across groups of learners.
Artificial intelligence can support this process by detecting recurring patterns and generating early warnings. However, responsible evaluation requires human oversight, careful comparisons, contextual understanding and transparent data practices. The goal is not to eliminate popular learning methods but to understand where they work, where they become insufficient and how they can be improved.
For EasyShiksha, this approach could connect online courses, quizzes, games, projects, internships, certificates, university programmes and career guidance through a broader understanding of student capability. Learning experiences could be evaluated not only by whether students complete them, but also by whether students can retain, explain, transfer and independently apply what they have learned.
The most advanced educational system will not necessarily be the one that adopts every popular teaching method first. It may be the one that continuously evaluates its own methods, recognises unintended consequences and responds before small weaknesses become widespread limitations.
True educational intelligence is not simply the ability to teach successfully. It is also the ability to recognise when the way we teach needs to change.