The Science of Reading and the Media: Does the Media Draw on High-Quality Reading Research?

The Science of Reading and the Media:

Does the Media Draw on High-Quality Reading Research?

Maren Aukerman, University of Calgary


Recent stories about the “science of reading” in the popular press have often distorted how children learn reading. Reporting can be biased, as my last Critical Conversations piece discussed.  But there is another problem, equally grave: reporters telling these stories may have only a partial grasp of reading research. To be fair, developing deep proficiency in a research domain can be elusive and take years of intense study. 1 For example, to report well on astrophysics news, science reporters need more than interest and a grade-school understanding of the scientific method; they require meaningful grounding in astrophysics and its research. The field of reading is no less specialized, involving thousands of scholars using a variety of research methods in robust dialogue with one another. Thus, to report well on developments in reading education, reporters must understand more than just one vocal corner of this research landscape. They should know how the information they communicate connects to a large body of scholarship that frequently includes conflicting findings and scholarly disagreement. 


Unfortunately, education journalists frequently fail to meet this bar. In this post, the second of three, I use two recent articles to illuminate 4 error patterns I call errors of insufficient understanding. One, from Time, discussed “an enormous rethink of reading instruction that is sweeping the U.S.”,  the push for more phonics. 2 The other, from the New York Times (NYT), reported on updates in a widely used published curriculum called Units of Study, one that the article lambasts as unaligned with the “science of reading” because it is anchored in an approach known as balanced literacy. 3


Error of Insufficient Understanding 1: Weak Connection to Actual Research


The Time and NYT articles both made claims about reading research, but neither appeared substantively grounded in it. Links provided in both articles mostly take the reader not to peer-reviewed research, an expected standard in scholarship, but to other news articles, podcasts, and/or popular (but not always scholarly) books. Both journalists often relied on the word of “science of reading” advocates, failing to verify those statements against research.  For example, The NYT cited a Units of Study critic claiming that the program’s author was “disconnected from research” without clearly situating this as opinion. 


By drawing mostly on vociferous advocates of one approach and bolstering their claims primarily with other journalism, journalists create an echo chamber which itself is disconnected from reading research. Indeed, Time cited no individual research studies, even where those would clearly be relevant. For example, it discussed widespread adoption of LETRS training, a professional development program supposedly aligned with the “science of reading;” yet it never acknowledged a rigorous study indicating that LETRS had no positive effect on student achievement, a critical detail from a research-based perspective. 4


When it is present, coverage of reading research also often fails to pass muster. Both Time and the NYT heavily rely on a research synthesis over twenty years old, the National Reading Panel Report. 5 Indeed, Time laments the ostensible fact that “the National Reading Panel’s recommendations and all the research were so blithely ignored” (again citing a journalist as proof of the claim). The NRP Report is an extensive, important research analysis, but it is far from complete (given research available at that time), up to date (given research conducted since then), or irreproachable (given limitations in methodology and scope). It has been the subject of numerous critiques, including scathing criticism by a participating panel member as well as reanalysis of its findings on phonics that calls those into question. 6 7 8 9 10 It is fair to call it influential, but false to equate the whole of reading research with this document, to dismiss scholars who fail to fully embrace it as ignoring research, or even to assume “science of reading” advocates and the NRP Report are themselves fully aligned.


Troublingly, Time treated the NRP Report and books from the popular press as the whole research story; it never mentioned individual research studies. The NYT did discuss two studies relevant to its topic – one that observed substantial student gains with a treatment sample of 40 schools using Units of Study 11 and one showing no effects from its use at a single school. 12 The article treated the studies as though they were of equal import and thus inconclusive, with no acknowledgment that the first, far larger in scale, was more robust. Such reporting is not research-aligned.


Recommendation: Ensure that “science of reading” news meshes with up-to-date, varied, peer-reviewed research before accepting its claims. 


Error of Insufficient Understanding 2: Inaccurate, Distorted Use of Terminology


Another problem is faulty use of terminology. For example, both articles criticize something they call “cueing” or “three-cueing,” describing it respectively as “a word-guessing method” (NY Times) and “get[ing] children to ask a lot of questions about the word they’re stuck on” (Time). But “cueing” is not an instructional approach, reading technique, or guessing game – except in the minds of certain detractors. What the reporters are likely referring to is something that those who actually use it call the three-cueing system, a framework for analyzing errors to understand children’s decoding attempts. 13


In fact, educators can use it to identify when a child over-relies on guessing. After identifying the primary nature of children’s errors, teachers can guide them into attending to new information so they no longer need to guess. If a child guesses “mother” in place of “mommy,” the child might be asked to slide their finger under the word to sound it out. A different child, who pronounces “said” so it rhymes with “raid,” might be asked if it makes sense in the sentence – not to replace the child’s phonics-based strategy, but to supplement it. This instructional guidance is not usually called “cueing” by those who use it. Rather, it may be called an interactive strategies approach. 14 


Well-designed instruction based on a three-cueing system analysis of student errors emphasizes tailored, progressively more refined strategies so that, for example, a child over-relying on pictures to decode would not be encouraged to keep using pictures. If a teacher or instructional program nonetheless does that, it is inconsistent with the whole basis of the approach. Some researchers argue that a subset of children need more systematic exposure to phonics than may happen with an interactive strategies approach. 15 But even those critics generally do not characterize the approach as “guessing.” That characterization is unhelpful, polarizing, and false.


This is not the only journalistic misuse of literacy terminology. For example, “balanced literacy” is frequently caricatured as involving little phonics, in contrast with what gets called the “science of reading.” Time explained it as offering “instruction in the link between sounds and letters… sprinkled in with other methods teachers thought worked.” The article wrongly suggested that the defining feature of balanced literacy is teaching idiosyncratic, non-research-based ways of decoding outside of phonics. In reality, advocates of balanced literacy see phonics as integral but also recognize that it should not displace attention to other important things like comprehension and language development, even in the early grades – a position firmly rooted in what reading research suggests. 16 17 In fact, some educators ascribing to balanced literacy may exclusively teach decoding via explicit phonics instruction (not interactive strategies).


Recommendation: Be aware that “science of reading” news may contain distorted information and wrong use of terminology: do not take accuracy for granted.


Error of Insufficient Understanding 3: Spurious Claims that One Approach is Settled Science


The NYT article argues that one particular approach to teaching decoding, called systematic phonics, “is the most effective way to teach reading,” and the Time article conveys a similar message. One can certainly make a strong case for such instruction 18, but the research terrain is complex: some analyses of phonics-intensive approaches, including studies of the much-touted Orton-Gillingham approach and the aforementioned LETRS program, have found no positive effects. 19 21 22 Moreover, little research examines whether systematic phonics ultimately improves comprehension, which is arguably the gold standard; studies that do examine this have failed to find clear benefits. 23 In short, there is insufficient evidence to conclude that any single approach, including the particular systematic phonics approach often elided with “the science of reading,” is most effective. 24 25 


The NYT not only insists that the type of phonics pushed by many “science of reading” advocates is the “most effective;” it also maintains that interactive strategies approaches – wrongly called “three-cueing” –  don’t work. 26 The actual evidence there is also more nuanced: for example, while one study suggests that gains associated with interactive strategies fade over time 27, another found positive effects even 10 years post treatment. 28 Moreover, journalists often have a double standard, pointing out the current lack of evidence for the long-term efficacy of interactive strategies approaches but staying mum about a corresponding dearth of evidence regarding the longitudinal efficacy of more exclusively phonics-based approaches. 29 30 


Worth noting, too, is that what gets called the “science of reading” is in no way analogous to the science of climate change, where 97% of scientists agree. 31 Nearly 57% of education professors favored balanced literacy in a recent survey, with only 22% favoring systematic phonics. 32 The press should be asking why buy-in by researchers is so low rather than accepting on faith that the minority position is the research-based one; few journalists appear to be doing so.


But beyond that, a genuinely research-based science of reading must be rooted in a scientific process – ways of engaging with the known, posing questions, and reconciling complexities – more than it is about adhering to fixed knowns. Disagreement is vital. This does not mean anything goes: reading researchers examine goals, what can be extrapolated from different measures, how research should be designed, whether research is generalizable to a larger sample of children, and what implications are valid. But it is a misrepresentation of research to treat it as fixed knowledge handed down by irrefutable researchers: this is perhaps what “science of reading” advocates within and outside the press get most wrong. Indeed, someone who denies the role of research-based dialogue is not a scientist at heart, but rather an ideologue.


It could be different. Early literacy researchers investigate many things. For example, there is likely a limit to how much phonics instruction children should receive. Phonics skills instruction alone produces less benefit than splitting time between skills instruction and practice in applying those skills to real texts; 25 or 30 minutes of targeted skills daily may be the sweet spot 33 34, which is pretty close to the average of what teachers already report doing. 35 Other scholars argue against one-size-fits-all solutions: stronger initial decoders in first grade benefit from less phonics instruction even as peers who begin the year with fewer of such skills benefit from more. 36


And other things similarly remain contested: how to make phonics instruction more engaging and effective 37; whether to teach phonics beyond first grade, given diminishing returns 38; what kinds of texts should be used with emergent readers 39; how much phonics should be “pushed down” into kindergarten when earlier does not mean better in developing reading 40; and so on. Researchers do have varied opinions on whether interactive strategies instruction contributes enough to warrant teaching it. 41 42 But that disagreement is not the only or even arguably the most important debate. 


Intelligent, informed people can disagree about these issues. Reading researchers regularly seek to persuade other researchers without dismissing those who see things differently as not listening to science. And adjusting one’s approach as one becomes persuaded of the need, as was done when the Units of Study update incorporated more phonics, is a mark of thoughtful engagement with research, not a “major retreat” (as per the NYT article).  


Recommendation: Be skeptical of “science of reading” news that touts “settled science,” especially if such claims are used to silence disagreement.


Error of Insufficient Understanding 4: Lack of Context about Previous Phonics Implementation Attempts


The NYT notes that Units of Study is being dropped in Oakland, apparently because critics believe that its less phonics-intensive approach has not seen enough success. The article does not mention that Oakland previously used a phonics-heavy program, Open Court Reading 43, introduced there in 1998 with similar media fanfare to the current hoopla. 44 The Time article does bring in a historical perspective after a fashion, quoting an Oakland-based “science of reading” advocate who claimed that the previous Open Court implementation there had led to strong reading gains before it was abandoned. But Time failed to discuss actual research on Open Court. In point of fact, there is evidence that Open Court actually had a negative effect on many kids’ reading, particularly as kids aged up and encountered complex, harder-to-understand text. 45   


The idea that phonics can fix children’s reading ills is at least 70 years old, yet results from other large-scale phonics reforms have also yielded disappointing results, including during the Reading First era in the U.S. and as England’s recent national curriculum mandates have played out. 46 47 48 And Canada, which has relied mostly on a holistic, less phonics-intensive approach, has generally had excellent reading scores in international test comparisons. 49 Both historical and international comparisons unravel the narrative that systematic phonics is a do-all fix. 


Recommendation: Consider whether “science of reading” news acknowledges the uninspiring track record of real-world phonics-intensive initiatives.


These four kinds of errors show up in numerous “science of reading” articles in the popular press beyond the two analyzed here. Particularly in combination with media bias errors, they do real harm. In my final contribution to this Critical Conversations series, to be released in the coming weeks, I will discuss some of these consequences. 


Please cite this work: Aukerman, M. (2022) The Science of Reading and the Media: Does the Media Draw on High-Quality Reading Research? Literacy Research Association Critical Conversations. CC BY 4.0 license.


Maren Aukerman is a Werklund Research Professor at the University of Calgary who focuses on literacy education and democratic citizenship. She studies educational ethics, how youth engage with information in the media, and the preparation of students for responsible citizenship and democratic dialogue, particularly in the context of literacy education. Aukerman previously was on the faculty at Stanford and the University of Pennsylvania and is the recipient of a National Academy of Education/Spencer Postdoctoral Fellowship as well as the 2009 Albert J. Harris Award and the 2018 Dina Feitelson Research Award from the International Literacy Research Association. In her current work, she is studying how young people make sense of COVID-19 information that they encounter, with a special emphasis on what they do and do not find trustworthy as information sources.


Photo by Jessica Ruscello on Unsplash

The Science of Reading and the Media: Is Reporting Biased?

The Science of Reading and the Media: Is Reporting Biased?

Maren Aukerman, The University of Calgary


At least two questions are critical for assessing journalistic quality. First, does the writing exhibit bias that distorts the truth? And, second, does the writer understand the topic deeply enough to report on it accurately? In the field of reading education, these issues have come to the fore with a recent explosion of journalism about the “science of reading.” 1 2 3 Teachers, policymakers, and parents hunger for information about good reading instruction, yet journalism on the topic can be grossly oversimplified. In a three-part Critical Conversations series, I examine how well-intentioned journalism about the “science of reading” is frequently biased and inadequately research-based, ultimately making the case that such reporting has damaging consequences for the teaching of early reading. In this article, the first of the series, I highlight media bias.


From how much of the media tells it 4, a war rages in the field of early literacy instruction. The story is frequently some version of a conflict narrative relying on the following problematic suppositions:

  • a) science has proved that there is just one way of teaching reading effectively to all kids – using a systematic, highly structured approach to teaching phonics;
  • b) most teachers rely instead on an approach called balanced literacy, spurred on by shoddy teacher education programs;
  • c) therefore, teachers incorporate very little phonics and encourage kids to guess at words;
  • d) balanced literacy and teacher education are thus at fault for large numbers of children not learning to read well.

Unfortunately, these suppositions turn out to be highly misleading. The problem is not with recognizing that teaching phonics can play a facilitative role in having children learn to read; that insight is, indeed, important, if not particularly new 5. The problem is that this narrative distorts the picture to the point that readers are easily left with a highly inaccurate understanding of the so-called “science of reading.”


Here, I present four types of bias errors in “science of reading” journalism. To discuss these, I examine an article from the New York Times (NYT) by Dana Goldstein in 2022. It focuses on Lucy Calkins, a professor who authored a popular literacy curriculum called Units of Study 6. The article is intensely critical of Calkins’s work and zeroes in on her recent decision to add more explicit phonics to her curriculum as a “major retreat” tantamount to admitting serious flaws in the previous version.


I note for the record that I do not know Calkins, have never used her curriculum, and agree that her program (like any) can benefit from critique. My concern is with the nature of the reporting by Dana Goldstein, an NYT national correspondent. I chose this article because it appears in an otherwise reputable newspaper, but the errors that render it profoundly flawed are widespread in education journalism.


I present these bias errors partly as a call to action for reporters; they have the responsibility and opportunity to decrease divisive rhetoric that has increasingly come to dominate discourse about early literacy instruction 7. I also hope that, by offering recommendations on how to identify errors, this article might help readers assess journalistic bias.


Bias Error 1: Lack of Balance in Reporting

The “guru” reference in the title of the NYT article is dismissive, and the article continues in the same tone. It quotes five individuals sharply critical of Calkins, while including no positive or even nuanced views. Three are sources that Goldstein 8 already drew upon in an earlier article, raising questions about how well she represents a spectrum of perspectives. Calkins herself was interviewed, but her quotes were arguably selected in ways that contribute to a lack of balance (the longest begins with “All of us are imperfect”). A curriculum as popular as Calkins’s program does not get that way without strengths that stakeholders find meaningful; that needs acknowledgment.


Recommendation: Be wary of “science of reading” news that fails to include and fairly represent a range of perspectives. 


Bias Error 2: Sensationalistic “Straw Man” Arguments

The NYT article states: “Some children seem to turn magically into readers…. That has helped fuel a mistaken belief that reading is as natural as speaking” 9. This “mistaken belief” is attributed to no named human beings, and no evidence is presented that Calkins holds this belief.


Whether or not Goldstein knows it, the source of this idea is likely a 1976 essay by Ken and Yetta Goodman. (Even back then, there was nuance in the premise:  the Goodmans noted that “teaching children to read is not putting them into a garden of print and leaving them unmolested;” see p. 459.) But to the present point: nowadays, few reading researchers or teachers embrace anything like the “reading is as natural as speaking” premise.


In reality, there is considerable common ground among literacy educators that reading should be deliberately taught, and specifically that most students benefit from explicit instruction in decoding. Where perspectives vary, they are not characterized by absolutes such as “let students figure out reading by themselves,” or “phonics is a waste.” Goodwin, editor of the flagship research journal Reading Research Quarterly, which recently ran two full special issues about the science of reading 10 11, put it this way:

The version of the science of reading that has been presented in the media is very narrow, focusing mainly on alphabetics, phonics, and word reading. It’s also pretty directive, telling teachers that if they want to help kids learn to read, then they should do this, not that. But when we invited researchers to propose and submit articles on the science of reading, that’s not how they defined it…. I just don’t see anybody talking about a battle between science and non-science. 12


Yet, media rhetoric continues to hype the supposed dichotomy 13. In doing so, articles like Goldstein’s 14 15 16 conjure a protagonist (“the science of reading”) and an enemy (“balanced literacy”) characterized as anti-phonics and anti-science. It is important, then, to distinguish between the “science of reading” often presented in the press (“popular SOR”), and actual research-based science of reading (“research-based SOR”).


To be sure, a few academics (often folks who do not do classroom-based research) have fed the “popular SOR” narrative. For example, cognitive scientist Seidenberg 17 has blogged the incendiary claim that most teachers are never exposed to “facts about the bases of reading skill” due to bad teacher education. More recently, his blog highlighted his own contribution to the Reading Research Quarterly special issue on the science of reading 18 – while completely ignoring other research contributions to that issue, including ones counter to his views 19.


But outright dismissals of balanced literacy come perhaps most of all from members of the public (see, for example, Letter to the Lindbergh Board, 2019), including parents and teachers, whose perspectives appear to have been shaped by polarizing media accounts 20 and literacy websites with a questionable research basis 21. The dichotomous worldview of “popular SOR” in public discourse is arguably in part thanks to media reporting.


Recommendation: Be wary of “science of reading” news that demonizes an approach (e.g., balanced literacy) and/or creates dichotomies of good and bad.


Bias Error 3: A Myopic Lens Fetishizing Phonics Instruction


Myopia- that is, shortsighted vision that misses the larger picture – abounds in popular SOR. The NYT article (Goldstein, 2022) does briefly state (without citation) that “research points to a broad set of skills necessary to become a literate person – including phonics, vocabulary, and knowledge of current events, history, art, sports, and nature.” Yet the alluded-to “broad set of skills” is not broad; it represents nowhere near the range of domains that the reading research community emphasizes22 23 24. For example, students should develop dexterity in decoding, comprehension, using texts for real-life purposes, and critical reading; and they should develop literate dispositions such as reading motivation. 25 Research-based SOR examines all these areas, and robust bodies of research address equally vital literacy domains, like writing development, oral language development, the development of English as an additional language, and identity-related aspects of literacy development given that children bring who they are (including varied identities and racial/socioeconomic backgrounds) to bear on their reading.


But, in popular SOR, everything but phonics gets short shrift. There are 86 mentions of phonics instruction  in Hanford’s (2019) media report 26 claiming to explain why reading instruction in the United States is inadequate, and only a single reference to any other aspect of literacy. The NYT article (Goldstein, 2022) follows a similar pattern: aside from scattered mentions of having children read complex text, only phonics is given attention. Reporting with such a laser focus on phonics, often circulated widely 27, crowds from view many core aspects of literacy. Developing children’s ability to understand, to imagine, and to think critically about text? Ignored. Developing their capacity to discuss and make reasoned text-based arguments? Ignored. Developing their reading motivation, which is associated with better comprehension? 28 Ignored. In crowding out so much else that is vital, such reporting works counter to the goal of moving the fuller body of reading research into classroom practice.


Recommendation: Be wary of “science of reading” news that zeroes in on phonics (or any highly limited slice of reading instruction) rather than building a well-rounded picture of literacy learning.


Bias Error 4: Logical Fallacies


The NYT article (Goldstein, 2022) contains this troubling statement:

Goldberg, a Bay Area literacy coach and leader in the science of reading movement, said Professor Calkins’s [recent curricular] changes cannot repair the harm done to generations of students. Even before the pandemic widened educational inequality, only one-third of American fourth and eighth graders were reading on grade level.


One journalistic problem here is that exactly zero evidence was requested of the person interviewed to support that accusation of harm. Moreover, none was provided by Goldstein to support the implication that Calkins’s curriculum was at fault for students reading below grade level. The two ideas are juxtaposed so as to suggest a relationship, but the journalist fails to empirically establish any causal link. In point of fact, reading scores on the National Assessment of Educational Progress 29 were actually 12 points higher in 2020 than they were in 1971 (before Calkins’s program even existed). If test results indicate a crisis, it long preceded Calkins and cannot fairly be attributed to her.


I could find no evidence for this claim of harm anywhere else, either. The reality is that in classrooms without much systematic phonics, most children also learn to read 30, and proficient readers from such classrooms show no word reading or comprehension disadvantages over those in classrooms with more systematic phonics instruction. In fact, they can read more fluently than proficient readers from classrooms where the phonics was more systematic, presumably because of more practice reading real texts. 31 Even Shanahan, a reading researcher skeptical of the kind of decoding instruction that Calkins’ materials previously included, concedes that there is no evidence that such instruction causes harm. 32


Recommendation: Be wary of “science of reading” news that contains wild leaps of logic unsupported by actual evidence presented.


These four errors (lack of balance, straw man arguments, myopic lenses, and logical fallacies) might seem puzzling, particularly in well-known publications: how could conscientious journalists fall prey to them? The reality is that some journalists reporting on education may have an insufficiently robust understanding of the field of reading research themselves, which can make it harder to engage in rigorous reporting. Look for additional posts forthcoming on Critical Conversations for more in-depth discussion of this problem.


Maren Aukerman is a Werklund Research Professor at the University of Calgary who focuses on literacy education and democratic citizenship. She studies educational ethics, how youth engage with information in the media, and the preparation of students for responsible citizenship and democratic dialogue, particularly in the context of literacy education. Aukerman previously was on the faculty at Stanford and the University of Pennsylvania and is the recipient of a National Academy of Education/Spencer Postdoctoral Fellowship as well as the 2009 Albert J. Harris Award and the 2018 Dina Feitelson Research Award from the International Literacy Research Association. In her current work, she is studying how young people make sense of COVID-19 information that they encounter, with a special emphasis on what they do and do not find trustworthy as information sources.


Photo by Christian Lue on Unsplash