What are “levels of evidence” and why do we need to talk about them in education?

  

Image created using ChatGPT

Around ten years ago, when my colleague, Dr Caroline Bowen AM and I were writing a book on evidenced-based support for children and young people with additional needs I found myself going down all kinds of rabbit holes to examine what were sometimes extraordinary claims about equally extraordinary practices. Examples of such practices were cross-body physical exercises to improve reading and writing skills, using coloured lenses and overlays to improve reading skills, “Earducators” to provide auditory-integration training to children diagnosed with made-up disorders, psychoastrology for children with autism spectrum disorders, Chelation therapy, and neurolinguistic programming – to name a very small sample.

In early 2026, we released a “companion” text that is aimed more fairly and squarely at a parent readerhsip, to support them in their increasingly complex journey through the noisy, invariably expensive and often-alluring intervention marketplace.

No-one should be surprised that the “average parent” struggles to access, appraise and distil research evidence but it is disappointing that many teachers and allied health professionals find this daunting also – invariably because of a paucity of preparation for this in their pre-service education.

Regardless of their lack of preparation for the task, parents, teachers and clinicians* all have vested interests, however, in being able to answer questions such as:

• What are the hallmarks of effective interventions to support struggling early readers?
• Does the timing of speech-language pathology intervention matter for young children whose speech is dysfluent?
• Are children with pre-school language disorders at higher risk of later reading and writing difficulties?
• Do coloured lenses/overlays improve the reading of children who have a diagnosis of dyslexia?
• If I read to my toddler, will that make them a strong reader when they attend school?
• Should phonemic awareness be taught in isolation in early years classrooms, or should printed letters be present?
• Can children with an intellectual disability be taught to read? 
• Will a gluten-free diet lessen a child’s autism symptoms?

*Policy-makers also.

This is where the “fun” starts.

I’m reminded of an email exchange some years back with an education academic at a major Australian university who was promoting so-called “brain-based interventions” for children with special needs. The term “brain-based” struck me then (and still does) as odd, given that all learning (declarative and/or procedural) ultimately comes back to the brain and its complex inter-connected systems. It seems a little like what my late dad would have described as “selling ice-creams to Eskimos”.  When I wrote to the academic in question and asked for published evidence to support said “brain-based interventions” their reply rather glibly informed me that “you can find a journal article to support just about anything” – as if this was a good thing.

While it is true that peer-reviewed journals publish material of wide-ranging academic merit, this is a problem – for the scientific community and for parents, teachers and clinicians (not to mention policy-makers and taxpayers) who are the consumers of said research. It is not a loop-hole that should be exploited by those who want to argue some version of:

• If there’s a peer-reviewed journal article showing an association between two factors, those factors must be linked in some important way;
• If there’s a peer-reviewed journal article showing lack of an association between two factors, those factors must not be linked;
• If there’s a peer-reviewed journal article that reports positively on an intervention, the intervention is trustworthy;
• If there’s a peer-reviewed article saying parents/teachers/clinicians liked a particular assessment or therapy, it should be adopted;
• If there’s a peer-reviewed journal article saying parents/teachers/clinicians disliked a particular assessment or therapy, it should be discarded;
• All research methodologies are equally valid for answering all research questions;
• We should not appraise individual papers or bodies of evidence to assess the weight they can add to the decision-making scales.

One paradox we need to address right here is the fact that peer-review of new research is both (a) well-intentioned and preferable and (b) imperfect.

Peer review is well-intentioned and preferable because

• It means that manuscripts are critiqued, often in a “double-blind” way, by academic peers who can identify flaws and gaps in arguments and analyses and suggest ways in which the paper can be strengthened before it is published.
• It helps to uphold academic standards regarding coverage of prior literature, application of ethical principles to recruitment and data collection, selection and justification of research methodology, data analysis, and appropriate extrapolation without over-reach.
• It sets published research apart from other written sources, such as newspaper and magazine articles, blogposts and newsletters, and product websites.

Peer review is imperfect because

• Some journals are inundated with submissions and can be highly selective, based on rigorous criteria, about what even goes out for review, let alone what is ultimately accepted for publication. Others have editors who are sitting by the phone, waiting for copy, and necessarily need to lower the bar to meet publication schedules.
• Sometimes editors can locate reviewers (who do their work without payment, often in their own time) with the necessary content knowledge and research background, and sometimes they can’t. Since the COVID-19 pandemic, this seems to have become more challenging than ever.
• Two reviewers may arrive at markedly different assessments of a paper, and the editor needs to exercise judgement on whose review takes precedence and/or how tensions are resolved. There is inevitably some cognitive bias and human error in this process.  

So - what does it mean to ask about the “level of evidence” for an assertion or a practice?

We can ask about the level of evidence behind an assertion, for example, “Vitamin D plays an important role in bone density” or about the level of evidence that can address a question about a practice, for example “Does a daily Vitamin D supplement improve bone density”? Corollaries in education might be “Knowledge of the alphabetic principle is important for beginning readers” and “How do we best establish knowledge of the alphabetic principle in beginning readers?”.

Astute readers will already be seeing the questions within questions in these examples.

• What do we mean by “important role”?
• What daily dosage of Vitamin D are we referring to? Which population(s) are we referring to? Children? Adolescents? Adult males? Adult females? Pre-menopausal females? Post-menopausal females?  People with particular medical conditions?
• What do we mean by the “alphabetic principle”?
• How does the alphabetic principle apply in a morphophonemic language like English? 
• What do we mean by children’s knowledge of the alphabetic principle? Which children? At what age/year level? How is this determined?

These questions within questions are often what leads academics, when asked about evidence to rightly respond by beginning with “Well it depends”. There’s not always a cut-and-dried answer for people schooled in the rigours of research critical appraisal and design. Yet we still need to respond to questions about whether current evidence supports a particular proposition or practice."It depends" is not code for "choose your own adventure". 

So, we must engage with the idea of levels of evidence.

This is a principle borrowed from health (medicine, nursing, allied health, public health), where it has gained wide acceptance in the last thirty years to guide policymakers and practitioners who must sieve through vast amounts of peer-reviewed (and increasingly, non-peer-reviewed) sources to guide their decision-making and optimise outcomes for their end-users.

Some readers may take exception right here, to me applying a framework derived from health to education. A gentle reminder is warranted then, that the idea of levels of evidence is equally applied in many other fields such as psychology, criminology and criminal justice, economics, aviation, engineering, road safety, environmental science and conservation, and forensic science – to name a few. As I have noted in a previous blog-post about education’s relationship with evidence, education is not special.

The evidence hierarchy is simply a means of organising, filtering, and appraising published evidence.

The hierarchy of evidence does not say that one research paradigm (e.g., positivism Vs autoethnography) is inherently superior or inferior to another. You will see in the infographic above, that both quantitative and qualitative data sources can be used right across the hierarchy. Choice of research methodology should align with the study aims/questions posed. Many education researchers, me included, use mixed methods to address their research questions. This might mean (for example) collecting student data on standardised measures as well as carrying out in-depth qualitative interviews with teachers to understand their experiences of a new teaching approach.

You will note too, that the hierarchy is a "stack" that tapers towards the top, because higher-level evidence is more difficult to collect. RCTs, for example, are usually expensive, so there will be fewer of them, meaning that there will be fewer to draw on for systematic analyses. This applies in all fields, not just in education. Evidence at lower levels is often where new knowledge begins, as occurred in the early 1980s, with the appearance of case reports heralding the arrival of the illness we now know as AIDS. As fields advance, though, new hurdles emerge. We would not want to rely on case-reports alone to answer vital questions about effective treatments at scale. For those questions, we need studies with higher levels of experimental rigour built in. 

Being open to the idea of levels of evidence helps to guard against the many ways in which cognitive bias can influence the thinking of researchers, policy-makers and/or practitioners, as described by the late Daniel Kahnemann, in his 2012 text Thinking Fast and Slow. The levels of evidence framework also offers pushback against the "fads and fashions" state of play that education has normalised in recent decades.

There are many myths and misconceptions about research methodologies and levels of evidence in education, and I blogged about these earlier this year: The dog ate my homework: Education’s evidence-excuses echo-chamber. The ideas expressed in that blogpost were uncomfortable for some, particularly my reminder that education is no different from any other social science that examines the behaviour, attitudes, and beliefs of humans. Ultimately, we all need to be able to apply rigour in the laboratory of the real world, if we are to make meaningful advances for our end-users.

Evolving evidence needs to be monitored, weighed probabilistically, and considered alongside available resources and the reasoned judgement of professionals. It is not a black-box but nor is it a means of arguing, as did my fellow education academic some years ago, that “anything goes as long as I can find a published paper somewhere that vaguely suggests this might be OK”.

When education academics and policy-makers attend a medical practitioner, drive across a bridge, or buckle up in an aeroplane, they are privileged beneficiaries of the respect shown for the hierarchy of evidence in those disciplines.

Education - you owe your stakeholders the same respect.

© Pamela Snow (2026)

Guest post: The “Yeah But” Brigade: Moving on from education’s uncomfortable relationship with data.

This is a guest blog post by James Benson*, a South Australian secondary teacher with nearly 20 years classroom and leadership experience across a range of education sectors and settings. 

 

Image source: ChatGPT 

 

My father always warned me to be wary of people who began sentences with two words: “Yeah, but”. He believed that whatever followed was usually going to be an excuse, a qualification, or an attempt to make an inconvenient fact a little less inconvenient.

I have been thinking about those two words a lot lately, particularly in relation to the way we use data in education.

We certainly do not suffer from a shortage of data. We have NAPLAN, PISA, PIRLS, TIMSS, PAT, DIBELS and an ever-growing collection of school and system assessments. We have dashboards, spreadsheets, data walls, data dives, improvement cycles and meetings devoted entirely to examining results. Being “data-informed” and “evidence-informed” has become part of the language of education.

Yet our commitment to data can be remarkably conditional. When the results accord with what we already believe, they are readily incorporated into our thinking. When they do not, we can become very good at finding reasons why they are not quite as important as they first appeared.

Of course, data needs to be interrogated. That is the point of having it. But interrogation should mean trying to understand what the data is telling us, not searching for reasons to make an uncomfortable result disappear. Different assessments illuminate different parts of student learning, and interpreting them well requires knowledge and judgement. The problem starts when scrutiny becomes a convenient form of dismissal.

Anyone who has spent time around school data will recognise the pattern. A disappointing result arrives, and the explanations (aka rationalisations) are rarely far behind. This was a difficult cohort. Attendance was poor. Students were anxious. The test did not suit them. The assessment failed to capture what teachers knew students could really do. The cohort had particular needs. The list goes on.

Some of these explanations may be correct. Context matters. But context should help us make better sense of data, not automatically provide a reason to disregard it. There is a difference between explaining why a problem exists and explaining the problem away.

I suspect there is another issue here that receives considerably less attention. For all the data we collect, I am not convinced we have invested nearly as much in developing the knowledge required to interpret it.

Data collection and data literacy are not the same thing. A school can produce an impressive dashboard without necessarily understanding what the numbers on it mean. Colour-coding students according to benchmarks, calculating percentages and producing graphs can create the appearance of analytical sophistication, but none of those things guarantees an understanding of what an assessment was designed to measure, what constitutes meaningful growth or what conclusions the evidence will actually sustain.

When that knowledge is weak, analysis can become fairly shallow. Small movements are given significance they may not deserve. Percentages are discussed without considering the size of the cohort behind them. Correlation slides quietly into causation. Measures designed for different purposes are compared as though they are interchangeable, and averages are discussed as if they describe every child in the group.

It is also very easy in these circumstances to reach for explanations that fit the educational mood of the moment. Poor achievement might be attributed to engagement, motivation, wellbeing, disadvantage, test anxiety, technology, COVID, socioeconomic circumstances or flaws in the assessment itself. Any of these could be relevant and several may operate at once. But a plausible explanation is not necessarily a demonstrated explanation.

If the data tells us students are struggling, we need to examine what they were taught, whether the curriculum was coherent and sufficiently ambitious, how effectively that curriculum was taught and whether students actually acquired the knowledge and skills they were supposed to acquire. Examination of context should not absolve us from examining the things schools can change.

Reading comprehension provides a useful example. If assessment data suggests students are struggling to comprehend what they read, describing them as disengaged readers does not get us very far. We need to know whether they can decode accurately and fluently, whether they possess sufficient vocabulary and background knowledge, whether they can navigate increasingly complex syntax and, importantly, what they have actually been taught and read.

The same applies in mathematics. Student anxiety may need to be considered, but it does not remove the need to establish what mathematics students know, what they have been taught and where the gaps lie.

An assessment result is not a diagnosis. It is the beginning of an investigation.

This is the broader context in which I have been watching the response to the latest PISA results. PISA is only one source of educational data, but the discussion surrounding it provides a useful illustration of our sometimes-complicated relationship with evidence.

The latest results provide plenty to think about. Across the OECD, achievement in reading and mathematics has fallen substantially over the past decade. Australia remains above the OECD average, but that should not obscure the longer-term deterioration in our own performance. These trends deserve serious attention.

What interests me just as much as the results themselves is the way international data has been used over time.

For years, Finland's (2000) strong performance on reading was treated as evidence worth examining. Its success generated books, conferences, international study tours and extensive commentary about what other education systems might learn from it. Professor Pasi Sahlberg became an influential advocate for Finnish education, while prominent education thinkers including Professor Andy Hargreaves drew attention to Finland as an alternative to some of the reform approaches being pursued elsewhere, e.g., England.

There was nothing unreasonable about this. When a country achieves unusually strong educational outcomes, we should be curious. We should look at its curriculum, teacher preparation, policies, demographics, culture and classroom practice and consider whether there are useful lessons.

What becomes harder to defend is changing our view of the measure when the story it tells changes.

Finland's performance subsequently declined, while England's relative performance strengthened. Over the same broad period, England pursued a substantial program of education reform: a stronger emphasis on a knowledge-rich curriculum, systematic synthetic phonics, mathematics mastery and more explicit approaches to instruction, alongside significant reforms to behaviour management, discipline and school culture. These were not minor adjustments at the margins. They reflected a quite different view of curriculum, teaching and the conditions required for learning, and many of them ran directly against approaches that had been fashionable in education for decades.

England's results do not establish that these reforms caused its international performance to lift. National education systems are far too complex for a claim of that kind. They do, however, give us a reason to be interested in what England has been doing.

Instead, some responses have concentrated on reasons commentary on England's performance should be qualified. Sampling, demographics, participation rates and the difficulties inherent in comparing education systems have all featured in the discussion. These are legitimate considerations. They are also considerations that apply, to varying degrees, across international assessment data more broadly. No international comparison takes place under perfectly controlled conditions, and no national sample is a miniature laboratory in which every contextual difference has conveniently disappeared.

The issue, then, is not whether we should scrutinise samples or examine the characteristics of the students who participated. Of course we should. The issue is whether we apply that scrutiny consistently.

If differences in sampling, demographics or context are sufficient to substantially discount England's performance, then the same standard needs to be applied when interpreting the performance of Finland, Singapore, Estonia, Canada, Australia or any other education system we happen to be discussing. We cannot treat contextual differences as background noise when the results support a preferred educational narrative and then elevate them to the central explanation when they do not.

Methodological caution cannot become something we remember only when a result is inconvenient.

If international assessment was sufficiently informative to support a global conversation about lessons from Finland, then it remains sufficiently informative to provoke curiosity when a system following a different policy direction performs strongly. PISA cannot be an illuminating source of evidence when it supports an educational philosophy we favour and suddenly become much less interesting when it produces a result we find uncomfortable.

That's the “Yeah, but”, right there.

The same habit appears with domestic and school-level data. When NAPLAN repeatedly identifies weakness in writing, saying that NAPLAN does not measure everything about a child is both true and beside the point. Nobody sensible believes that it does. When DIBELS identifies significant problems with fluency, pointing out that DIBELS does not measure every aspect of reading does not resolve the fluency problem. When PAT indicates weak growth, understanding the limitations of PAT should inform the investigation rather than end it.

No assessment gives us the whole picture. That is precisely why we use multiple sources of evidence, that must be interpreted with care.

A single result may be anomalous. When different measures, designed for different purposes and administered at different points in schooling, begin to tell a similar story, however, the collective evidence becomes increasingly difficult to explain away. The task is not to find the perfect assessment. It is to determine what the weight of the evidence is telling us.

There is also a deeper issue sitting underneath much of this discussion. Our arguments about what counts as meaningful educational data may partly reflect disagreement about what we think schools are actually for.

I should declare my bias here.

I believe the central educational purpose of school is the acquisition of domain knowledge. Schools should teach young people things they do not already know, including things they may have little opportunity to learn outside school. They should open access to mathematics, science, history, geography, literature, the arts and the wider cultural inheritance, while developing the literacy and numeracy that allow students to participate fully in those domains.

This does not mean schools should be cold or joyless places where children's wellbeing is irrelevant. Schools should be safe and inclusive. Children should be known, respected and cared for. Relationships matter. Behaviour matters. Belonging matters.

I regard these things, however, as important conditions for successful schooling rather than replacements for its educational purpose. A school may be warm, inclusive and caring, but if students leave unable to read demanding texts, write coherently or work confidently with mathematics, something fundamental has been missed.

There is nothing historically unusual about debating the purpose of schooling. Mass education has always carried several expectations. Schools have transmitted knowledge and culture, prepared young people for work and citizenship, contributed to social cohesion and played an important role in socialisation. The balance between these purposes has shifted over time.

What does seem increasingly apparent is how much more we now ask schools to do. Their remit has expanded to encompass wellbeing, resilience, identity, agency, engagement, creativity, collaboration, social and emotional development and an expanding collection of dispositions said to prepare young people for an uncertain future.

Many of these are worthwhile aims. The difficulty comes when they begin to compete with, rather than support, the core business of acquiring knowledge and skills, or when they become alternative indicators of success whenever academic outcomes disappoint.

Belonging, motivation and engagement can provide useful information about students' experiences, but they tell us different things from measures of achievement. Self-reported engagement does not establish whether a student can comprehend a difficult text. A sense of belonging does not tell us whether that student understands fractions. A measure of growth mindset does not tell us how much background knowledge a student possesses.

We should be particularly cautious when these softer measures become substitutes for learning itself.

If academic learning is regarded as one outcome among a very large collection of equally important goals, sustained declines in achievement can be accommodated relatively easily. Attention can move towards wellbeing, engagement, creativity or collaboration, followed by the familiar rhetoric that conventional assessments fail to measure everything that matters.

Of course they do.

The relevant issue is whether they measure something that matters.

If the acquisition of knowledge remains a central responsibility of schooling, sustained deterioration in reading, mathematics or science requires a serious response. Those results do not tell us everything about education, but they tell us enough that we should pay attention.

There is an equity dimension to this as well. Families with significant cultural and economic resources can often compensate for what schools do not provide. They can buy books, tutoring, experiences, travel, music lessons and academic support. They can fill gaps in knowledge and navigate the education system on behalf of their children.

Many children do not have that luxury.

For them, school may be the only institution capable of systematically providing access to the knowledge that others acquire partly through circumstance. This is why I find the argument that schools should be less concerned with knowledge particularly difficult to reconcile with claims about educational equity. Knowledge is not an optional extra for disadvantaged children. It is one of the most powerful things schools can distribute more fairly.

And that brings us back to data.

Sometimes a cohort genuinely is problematic. Sometimes a movement in results means very little. Sometimes attendance, disadvantage or disruption explains much of what we see.

And sometimes students simply have not learned enough of what we were charged with teaching them

An evidence-informed profession needs to be capable of entertaining that possibility without immediately reaching for an explanation that allows us to preserve what we already believe. That requires better assessment literacy, better data literacy and a willingness to follow evidence into places that may sit awkwardly with the prevailing educational zeitgeist.

Data needs interpretation, but interpretation should help us get closer to the problem rather than provide increasingly sophisticated ways of avoiding it. If evidence only counts when it can be used to confirm what we already believe, we are not really allowing it to inform us at all.

My father was probably right to be suspicious of “Yeah, but”. The phrase becomes most seductive at precisely the point when the evidence is telling us something we would rather not hear.

That is probably when we need to listen most carefully.

---

*James Benson is a pseudonym. The author is a real teacher, who wishes to preserve anonymity, for professional reasons. 

 (c) Pamela Snow & "James Benson" (2026)

Guest post: The Unfinished Work of Professionalizing Reading Education

I am pleased to publish this guest post* by G. Reid Lyon PhD and Doug Carnine, PhD:

Why a Common Knowledge Foundation and Professional Language Are Essential for Improving Reading Proficiency

G. Reid Lyon, Ph.D.

Senior Advisor, ALLIED Hub Advocacy, Leadership, Learning, Implementation, and Educational Policy, Drexel University

Doug Carnine, Ph.D.

Founder, Evidence Advocacy Center

Category: Leadership  

Reading Time: 15 min

About the authors

G. Reid Lyon and Doug Carnine bring complementary expertise and decades of experience in evidence-based fields. Lyon specialized in neuroscience and learning disorders at the University of New Mexico followed by postdoctoral training investigating how neural systems support reading development and where reading difficulties originate. Carnine was trained as an experimental psychologist—through a University of Illinois NSF undergraduate fellowship—and dedicated his career to designing and testing instructional methods for children at risk of academic failure.

Lyon trained neuropsychologists, neuroscientists, medical students and residents in neuroanatomy and neurophysiology within the constraints of a mature profession that had built rigorous requirements for training, certification, ethics, licensure, and hierarchies of evidence. In contrast, Carnine, who trained educators during his tenure at the University of Oregon, was shocked to discover that education lacked rigorous requirements for training, certification, ethics, licensure, and hierarchies of evidence.

Their shared conclusion, which comes from two different sets of experiences, is unambiguous: education too often views evidence as a question of "positionality" whereby rigor in assessing and ranking research findings are dismissed in favor of judging alignment with pre-determined ideologies.

Executive Summary

The United States does not suffer from a shortage of scientific knowledge about reading development, reading difficulties, or effective instruction. More than six decades of interdisciplinary research have produced a robust and increasingly convergent understanding of how reading proficiency develops and how it can be improved. Yet reading outcomes remain alarmingly low.

This paradox points to a clear conclusion: the primary obstacle is no longer a gap in scientific discovery. It is the failure to institutionalize that knowledge within the systems responsible for preparing and supporting educators and educational leaders.

Central Argument

The reading crisis is not only an instructional, curriculum, or policy problem. It is fundamentally a professionalization problem. Until education establishes a universally shared knowledge base, a common professional language, consistent preparation standards, and enforceable accountability mechanisms, significant variability in professional knowledge, instructional practice, and student outcomes will persist. This process must begin within university education faculties, where academic freedom should take second place to empirical evidence.

Unlike mature professions—such as medicine, engineering, nursing, aviation, and speech-language pathology—education has not built the guardrails that ensure scientific knowledge is translated into practice at scale. Professionalization is the missing link.

The Paradox of Reading Science: We Know Enough, Yet Proficiency Remains Low

Few areas of education have benefited from a more extensive scientific knowledge base than reading. Beginning in the 1960s and accelerating across subsequent decades, researchers identified the linguistic, cognitive, developmental, and neurobiological foundations of reading acquisition. This work culminated in landmark interdisciplinary efforts—including the NICHD Reading Research Program—that integrated findings across disciplines, providing the basis for developing instructional programs and assessment systems as well as the means of implementing them.

Alongside the science of reading, advances in the science of learning, the science of instruction, and implementation science have further enriched this knowledge base. Together, four interlocking bodies of evidence define what the field now knows:

The Science of Reading and Writing

Decades of research have established core principles about how reading proficiency develops:

• Reading and writing are not acquired naturally, as spoken language is; they must be explicitly taught.

• Reading and writing are language-based developmental achievements.

• Reading and writing proficiency depends on multiple interacting components.

• "Risk for reading and writing difficulties can be identified early before formal schooling begins.

• Most struggling readers and writers can achieve proficiency with exposure to an adequate dose of explicit, systematic, evidence-based instruction, delivered by a knowledgeable practitioner.

• Effective instruction can shift developmental trajectories and strengthen the neural systems that support ongoing reading and writing achievement.

The core instructional components are based on these principles.

The Science of Learning

The relevant evidence base extends beyond “what works.” The science of learning examines how people acquire knowledge, retain it over time, generalize it to new contexts, and recover when learning breaks down. This science informs the conditions under which instruction is most likely to succeed.

The Science of Instruction

The science of instruction addresses both the design of content and its delivery. Preparing teachers to engage students is only as effective as the quality of the content they engage with. Instructional materials must be carefully structured, sequenced, practiced, and monitored to produce reliable learning outcomes.

Implementation Science

Implementation science is the scientific study of methods and strategies that facilitate the uptake of evidence-based practice and research into regular use by practitioners and policymakers.

The science of reading identifies the knowledge students must acquire. The science of learning explains how that knowledge is acquired, retained, and transferred. The science of instruction translates these findings into instructional practices comprising well-designed instruction—what to teach—and effective instructional delivery—how to teach—that reliably produce learning. Professional preparation and support should therefore include mastery of all three sciences. Implementation science focuses on closing the gap between what we know and what we do (the know-do gap) by identifying and removing barriers that impede the use of proven interventions and evidence-based practices.

From an implementation science foundation it is equally important to de-implement practices not aligned with the converging and replicated scientific knowledge. It is challenging for the learner to acquire and apply new information when they are exposed to valid and invalid practices particularly when implementing tiered instructional models.   For example, the practice of  using the  three-cuing systems model is not supported as an accurate model of skilled word recognition or as an evidence-based method for teaching beginning readers to identify unfamiliar words.   Likewise, the scientific evidence does not support using leveled predictable texts—particularly when paired with three-cuing prompts—as the primary means of teaching children to read words. In addition, the traditional running-record system—particularly MSV analysis, leveled-text placement, and self-correction ratios—is not well supported as a scientifically valid diagnostic or progress-monitoring system.

The Central Question

The question is no longer whether scientific knowledge exists. The question is why it has not become common professional practice. The answer lies in the professional structure of reading education—not in gaps in the evidence.

Education Has Not Yet Fully Professionalized

A profession is not defined merely by the dedication or rhetoric of its members. Mature professions develop the capacity to transmit, maintain, and apply scientific knowledge consistently across practitioners and institutions. They build infrastructure that makes optimal evidence-based  practice the norm—and makes suboptimal  practices harder to sustain.

Across mature professions—medicine, engineering, aviation, nursing, finance, and maritime operations—a consistent pattern emerges: evidence-based knowledge is institutionalized through low variance in shared standards, coherent training, competence-based licensure, and accountability systems that protect the public. These professions do not rely on persuasion alone. They rely on systems.

Education lacks comparable professional guardrails. As a result, children experience dramatically different learning opportunities depending on their location, the quality of their teachers’ preparation, local instructional traditions, and—most critically—whether leaders and systems align their policies and guidelines with the strongest available scientific evidence. This degree of variability would be unacceptable in any trusted profession. It should be unacceptable in reading education.

What Trusted Professions Provide That Reading Education Still Lacks

Mature professions typically institutionalize five pillars that collectively ensure coherence, competence, and adherence to evidence-based practice. These pillars also support enforcement mechanisms that reduce the persistence of ineffective practice.

Pillar 1: A Shared Knowledge Base

Trusted professions agree on foundational knowledge—even when they debate implementation details or interpret emerging findings differently. In reading education, professionals often enter practice with substantially different understandings of critical fundamentals, including:

• How reading develops across childhood

• Why reading difficulties occur

• The role of oral language—vocabulary, grammar, listening comprehension—in reading acquisition

• The role of encoding in word reading and automaticity

• The nature and identification of dyslexia and reading-related disorders

• How assessment should guide instructional decision-making

• What characteristics define effective reading instruction

When foundational constructs are not shared, professional disagreement becomes predictable and implementation unreliable. No profession can scale competence if its members do not share a core model of the phenomenon they are meant to address.

A Common Professional Language Is Central to the Shared Knowledge Base.  Even among well-intentioned and well-educated professionals, shared language matters—because it determines what people actually mean when they invoke evidence and instruction. In reading education, key terms are routinely defined differently across universities, publishers, advocacy organizations, state agencies, and professional development providers. Terms such as “science of reading,” “structured literacy,” “explicit instruction,” “balanced literacy,” “dyslexia,” and “evidence-based” carry inconsistent meanings in the field. Even the term “reading” lacks an agreed definition.

This creates avoidable and costly problems:

• The same term can describe entirely different constructs.

• Different terms can describe the same construct.

• Collaboration slows as professionals talk past each other.

• Lack of clarity supports a “choose your own adventure” status quo, whereby no approach is seen as preferable over another.

• Implementation fidelity erodes when apparent agreement masks real disagreement.

A common professional language, which is part of the shared knowledge base, is not a stylistic preference. It is the infrastructure that makes knowledge transferable—and enforceable.

Pillar 2: Research-Aligned Preparation

Trusted professions align preparation to a shared knowledge base and require trainees to understand not only discipline-specific findings, but also the broader sciences of learning, instruction design, and delivery. Preparation in reading education must similarly include the content in Pillar 1.

In medicine, trainees learn evidence-based practice through structured, competence-centered programs. This means they learn about the process of knowledge generation via the use of rigorous quantitative and/or qualitative research methods; they also learn how to critique new research findings and to apply the principle of levels of evidence to practice decisions.  In reading education, preparation remains highly variable and discretionary. Not all preparation programs ensure mastery of essential reading knowledge and instructional competencies, including the ability to design adaptations of reading instructional programs and deliver that instruction with fidelity. This variability helps explain why evidence-based instruction in classrooms often reflects inconsistencies in  local training and support ecosystems than on a consistent and trusted professional standard.

Pillar 3: Licensure and Professional Development Rooted in Competence

In trusted professions, licensure signals demonstrated competence and the responsibility to protect  the public. Professional development is not optional enrichment—it is structured, ongoing learning that sustains competence and aligns practice with advancing evidence, especially the operational decisions required to translate research into consistent instruction and effective error correction.

Education has not consistently built licensure and professional development systems around competence in evidence-based reading instruction. Licensure too often reflects completion of coursework rather than demonstrated mastery of relevant knowledge and skills.

Pillar 4: Accreditation with Teeth

Today, a large percentage of teacher-preparation programs that fail to prepare candidates effectively still receive accreditation. No other serious profession tolerates such laxity. This is a serious breech of community trust.

Pillar 5: Accountability for Professional Knowledge and Practice Quality

Trusted professions protect the public through accountability systems that address competence, defined as adherence to evidence-based standards. Accountability does not merely track outcomes—it monitors whether core professional practices are delivered competently and consistently, including fidelity to evidence-based instructional routines and delivery requirements grounded in the sciences of learning and instruction.

In education, accountability is too often inconsistent, symbolic, or diffuse. Systems frequently emphasize reporting student results without ensuring the implementation capacity, fidelity, or professional competence required to achieve those results.

The Pendulum Problem

Education is notoriously susceptible to pendulum swings. When evidence is treated as optional, virtually any practice can be judged as “in scope”, instruction regresses, and student data reflect this. Carnine’s Substack about “stopping the pendulum” reflects a broader pattern: reading education has repeatedly moved away from evidence-supported practices only to rediscover them after costly cycles of discrediting and re-adopting.  In many cases faulty  implementation is the culprit.  When the predicted lack of progress is noted, rather than assessing implementation fidelity, the conclusion is frequently “we tried that and it didn't work” and then  stumbling on to the next bright shiny thing.

Phonics instruction has been adopted (at least partially), abandoned, distorted, rediscovered, and marginalized again. Whole-language instruction displaced phonics; phonics later regained prominence; then it was pushed aside once more. These cycles are not evidence of a scientifically-based field. They reveal that professional systems have failed to institutionalize evidence-based practice as the default.  Pendulum swings also emanate from superficial binary conceptualizations of reading development and instruction. Consider the decades of debate centered on “phonics versus whole language” and “phonics versus balanced literacy”.  Reading proficiency is nested within oral language development and  requires the acquisition of multiple componential reading skills and experiences, integrated through direct and systematic instruction.  Joe Torgesen depicted the path to reading comprehension in the figure below.

Reducing reading development and instruction to “either-or” dichotomies ignore the complexity of the multiple factors essential for skilled reading.

Understanding the Pendulum

The pendulum is best understood as a symptom of incomplete professionalization. Trusted professions do not cycle between evidence-based and discredited practices, because professional guardrails restrict individual discretion and enforce standards of practice. Education  lacks those guardrails—so local preference, ideology, and “reform churn” can override evidence repeatedly.

What Education Can Learn from Trusted Professions

Evidence alone does not protect the public. Trusted professions become trusted because they build systems that ensure evidence becomes practice—even amid disagreement and lapses among practitioners. Five lessons emerge from their example.

Lesson 1: Trusted Professions Reduce Preventable Harm

Trusted professions protect the public by institutionalizing evidence through guardrails and external accountability. They do not wait for voluntary compliance. History shows that during periods when evidence existed but was not uniformly applied, preventable harm occurred. The “period of preventable harm” can be defined as the gap between when evidence is established and when it becomes uniformly applied in practice.

Reading education appears to remain in such a period. Evidence-aligned practice is not universally required, consistently supported, celebrated, or effectively monitored.

Lesson 2: Members Adhere to a Shared Knowledge Base

In trusted professions, members are not free to disregard core standards of care. Practitioner judgement exists, but it operates within shared, enforceable boundaries. In education, the empirically-derived scientific knowledge base is not consistently treated as a governing standard. Practice therefore depends too heavily on local interpretation, leadership preferences, and curriculum adoption cycles.

Lesson 3: Noncompliance Is Made Harder

Trusted professions do not eliminate imperfect practice—but they reduce its invisibility and its tolerance. Education often allows ineffective practice to persist precisely because professional systems do not impose meaningful consequences for departures from evidence-based standards.

Lesson 4: Local Control Does Not Override the Knowledge Base

Local control and evidence-aligned professionalism can coexist. The problem is not local decision-making per se; it is local decisions made without enforceable, knowledge-based constraints. Trusted professions permit variation in context, style, and application while maintaining non-negotiable core standards. Education too often treats scientific evidence as negotiable—especially when school boards, district leaders, curriculum vendors, or preparation programs prefer alternatives – but not always for reasons that can be empirically supported

Lesson 5: Accountability Is the Key Mechanism

Accountability in trusted professions defines who is responsible for ensuring evidence-based practice—and establishes consequences when it does not occur. In reading education, accountability must be multi-level and enforceable. It should require:

• Classroom-level competence in evidence-based instructional routines and delivery

• Leadership responsibility for coaching, monitoring, and implementation quality

• District responsibility for fidelity to evidence-based practice across schools

• State monitoring that is both supportive and corrective when achievement levels are low and evidence-aligned practice is absent

• Clear “line of sight” between instruction as an input and student data as an output.

A Policy Agenda for Professionalizing Reading Education

Improving reading proficiency requires more than adopting new curricula or passing legislation. It requires building professional infrastructure that reliably translates scientific knowledge into daily practice. Five policy priorities are essential.

Priority 1: Establish a Common Knowledge Foundation

States should define the essential professional knowledge that all educators and educational leaders must possess regarding reading development, reading difficulties, assessment, and evidence-based instruction—grounded in the science of reading, the science of learning, and the science of instruction.

Priority 2: Establish a Common Professional Language

Preparation programs, licensure systems, professional organizations, and state agencies should adopt shared terminology and operational definitions grounded in cumulative scientific evidence. Ambiguous or inconsistently defined terms undermine implementation and accountability. Academic freedom should not extend to the teaching of concepts and practices that are not well-supported by robust empirical evidence. 

Priority 3: Strengthen Evidence Standards

States should clearly distinguish between claims that are merely plausible and claims that are genuinely evidence-based. They should require stronger evidence standards for curriculum adoption, intervention selection, and professional development investments—and should make those standards transparent and publicly accountable. There should be a prioritizing of “levels of evidence” when appraising different instructional approaches.

Priority 4: Align Preparation and Licensure

Teacher and leadership preparation programs should be evaluated based on graduates’ demonstrated mastery of scientifically validated reading knowledge and instructional competencies, including the ability to design and deliver instruction consistent with the science of learning and the science of instruction. Licensure must reflect detailed assessment of course materials (e.g., assigned readings and assessments) and verified graduate competence, not mere exposure to coursework.

Priority 5: Build Accountability for Professional Knowledge and Implementation

Accountability systems should measure both student outcomes and the quality of professional preparation, professional learning, and implementation capacity across the system. This must include monitoring fidelity to evidence-based instructional practice and requiring corrective action when competence or implementation quality is absent.

Conclusion

The future of reading proficiency depends less on discovering new scientific principles than on institutionalizing existing knowledge within the professional systems that prepare and govern educators.

The unfinished work of professionalizing reading education is the construction of guardrails that ensure evidence is consistently taught, consistently understood, communicated through shared language, and consistently applied in practice.

This task is not primarily scientific. It is professional and it also requires re-calibration of the notion of academic freedom within university education faculties, so that accountability for stronger student outcomes begins as far upstream as possible. The evidence is available; the professional infrastructure remains incomplete. If education is to produce stable, sustained progress rather than perpetual and harmful pendulum swings, it must develop the shared knowledge base, professional language, competence-based preparation and licensure, and enforceable accountability systems that characterize every trusted profession.

The reading crisis is, at its core, a professionalization crisis. Addressing it requires treating professional infrastructure—not only pedagogy or policy—as the urgent priority it is.

*A version of this paper appeared in a substack article on Carnine’s substack page.

(C) Lyon & Carnine (2026) 

About the ALLIED Hub The ALLIED Hub @ Drexel brings together research, practice, and community engagement to equip leaders, educators, families, and advocates with the knowledge and tools to advance literacy for all. We champion transformational change through advocacy, leadership, learning, implementation, education policy, and data.