Reading disagreements yet another time
Why are we back in a contentious debate about phonics again?
Here we are, back in the “reading wars” again. Golly, I wish we could abandon that metaphor. But, the Internet is awash in commentary on a recent series of publications about the appropriate role of phonics in early reading.
As noted previously, I’ll eschew the war-fight-combat language, because I disagree with the idea that in examining an educational issue it is appropriate to employ any available actions (techniques, methods, strategies) to defeat the “enemy.” We should be together in the effort to provide education that leads to better outcomes for children.
Sally Weale (2022), the education correspondent for The Guardian, reported that “a landmark study” shows that educators should reject phonics instruction. One (at least I) might wonder, what is that study?
Ms. Weale reported that a team of British authors who disparge phonics instruction (Wyse & Bradbury, 2022, but see also Bradbury, 2022; Wyse & Goswami, 2013), are hyping a paper that is garnering interest in the popular press. Ms. Weale reported that, “A landmark study has described the way primary school pupils are taught to read in England as ‘uninformed and failing children’, calling on the government to drop its narrow focus on phonics.”
Ms. Weale uses journalism’s time-honored “both-sides” method of analysis. This method captures concepts from opposing views to promote an unbiased perspective on the news (see, e.g., “two sides of a coin” from India’s National Institute of Mass Communication & Journalism statement of core principles of journalism).
One problem with the both-sides method is that it sometimes requires the journalist to cast advocates of bogus views as if those views deserve equal treatment. Essentially, the mehod elevates nonsense to the same level as sensibly reasoned analyses of scientific evidence. In Harvard University’s Nieman Lab, Shannon McGregor and Carolyn Schmitt (2022) discussed how this problem applies when journalists cover political debates.
Here’s an example of Ms. Weale’s presenting both sides:
In synthetic phonics, children begin by pronouncing individual sounds in words and are then encouraged to blend them together to make words. For example “s-t-r-ee-t”. Supporters say it has had a positive impact on literacy, and point to significant long-term benefits for disadvantaged pupils.
Critics say phonics training only helps children to do well in phonics tests – they learn how to pronounce words presented to them in a list rather than understand what they read – and does nothing to encourage a love of reading. England’s Pisa reading scores are virtually unchanged since 2006.
Ms. Weale refered specifically to a study by Wyse and Bradbury (2022). Without commenting further on the accuracy of her reporting, let’s see what’s in that study.
Hunh?
What is that study? Well, there is a video
available on the intertubes that summarizes the argument. What do you think of that video?
I, of course, will argue that it’s routinely wise to dig into the evidence a bit. Video provides a powerful way to convey ideas, but scholarly analysis requires a medium that permits calm, systematic, and thorough analsyis...like an acdemic article! So, let’s examine Dominic Wyse and Alice Bradbury’s article, “Reading Wars or Reading Reconciliation? A Critical Examination of Robust Reasearch Evidence, Curriculum Policy, and Teachers’ Practice for Teaching Phonics and Reading.”
Here is the abstract:
Teaching children to read is one of the most fundamental goals of early years and primary education worldwide, and as such has attracted a large amount of research from a range of academic disciplines. The aims of this paper are: (a) to provide a new critical examination of research evidence relevant to effective teaching of phonics and reading in the context of national curricula internationally; (b) to report new empirical findings relating to phonics teaching in England; and (c) examine some implications for policy and practice. The paper reports new empirical findings from two sources: (1) a systematic qualitative meta-synthesis of 55 experimental trials that included longitudinal designs; (2) a survey of 2205 teachers. The paper concludes that phonics and reading teaching in primary schools in England has changed significantly for the first time in modern history, and that compared to other English dominant regions England represents an outlier. The most robust research evidence, from randomised control trials with longitudinal designs, shows that the approach to phonics and reading teaching in England is not sufficiently underpinned by research evidence. It is recommended that national curriculum policy is changed and that the locus of political control over curriculum, pedagogy and assessment should be re-evaluated. [I omitted the link to the YouTube videon her because I’d referred to it previously in this post.]
In the following subsections, I’ll examine the purposes described by Wyse and Bradbury one by one.
Provide a new critical examination of research evidence relevant to effective teaching of phonics and reading in the context of national curricula internationally
The authors say that they will report new empirical findings relating to phonics teaching in England. Subsumed under this purpose, it appears, is the description of “(1) a systematic qualitative meta-synthesis of 55 experimental trials that included longitudinal designs; [and] (2) a survey of 2205 teachers.”
It’s important to understand that a qualitative meta-synthesis is not the same as a meta-analysis. The goal of such a synthesis is not to provide objective, quantitative evidence about outcomes, but to assess, systematically, features of studies (often qualitative studies) that are consistent with reviewers’ research questions (Finlayson & Dixon, 2008; Nye et al., 2016; Zimmer, 2006; cf. Cooper et al., 2009). Although the methods are usually employed with a corpus of qualitative research (e.g., case studies, grounded theory studies, etc.), they can be adapted to examine a body of quantitative studies, too.
To conduct their meta-synthesis, Wyse and Bradbury read some reviews about phonics and chose one by Suggate (2016): “All 55 research papers reporting the experimental trials cited in the Suggate (2016) [Systematic Review] were located and read in full.” Good. But, wait, does Suggate’s review provide an exhaustive catalog of research on the effects of phonics instruction. Of course not. The US National Panel (see Ehri et al, 2001 for a summary about phonics) reviewed original studies (not people’s takes on them) > 15 years before Suggate’s analysis! Now, in a way, that makes sense, but it also indicates that Suggate’s review was restricted.
Well, indeed, Suggate (2016) didn’t review original studies...nor did Wyse and Bradbury. They stayed out of the details and stuck to analyzing what people had said about research on reading. That is, their evidence is predicated on people’s opinions and observations about the evidence, not the actual evidence itself.
Wyse’s and Bradbury’s strongest argument is predicated on how one of their observations—that longitudinal outcomes for different forms of early reading instruction may not favor early emphasis on phonics— and it is suspect, too.
This theme about focusing on what people say rather than what happens is important, as I see it. Please remember that people can assert that just about anything is true (e.g., Earth’s Moon is made of cream cheese).
The survey
Wyse and Bradbury also reported the results of an online survey examing teachers’ reports about how they are teaching reading and using a UK-mandated assessment with second-year students. That is, they asked a sample of educators for those people’s opinions.
It’s important to examine the Wyse and Bradbury sample, because any survey’s value is based on who completes the survey. If one wants to know what people say, one needs a sample of respondents who are representative of some larger group, one in which every member of the group has an equal chance to respond to the survey.
Do you want to know what professors of special education say? Don’t just ask friends of John Lloyd! That is, “19 of 20 dentists say that brushing for 60 seconds is over rated” doesn’t work unless every dentist (special education professor) who answered had an equal chance of all dentists (or sped profs) to respond to the survey...and there’s more, but rather than belabor the probably obvious point, let’s let it go at that.
Wyse and Bradbury do not provide data about how representative their sample is. They note that they used a particular software (“Opinio”) to solicit responses from on an on-line group. Let’s imagine what might have happened: People sent an e-mail message to some people whom they thought might be interested in the survey, perhaps because they are pals who share interests in particular views of a topic (e.g., reading instruction); those folks respond to the survey and pass it along (e.g., through social media links) to others who may have a similar view and those folks respond and pass it along, too. Can we see the “echo chamber” amplifying the views in this “sample.” Similar methods are sometimes called “snowball sampling”; one can find lots of academic discussion of the pros and cons of such methods (e.g., Falter & Brunet, 2012; Hancock & Gile, 2011), but those sampling methods are generally less trustworthy than rigorous random sampling methods.
But, what was the sample for the Wyse and Bradbury (2022) study? Uhm, from reading the report of their research, I dunno. Help?
A second problem with the Wyse and Bradbury survey is that in research it is generally important to employ an instrument that has known psychometric features. That is, one would want to know about reliability and validity of any measures employed. These words—reliabilty and validity—refer to very important ideas in any assessment, so serious scientists want to know how good a survey is...is it “reliable” (consistent results?) and “valid” (do the consistent results relate to other measures?). Wyse and Bradbury provided no evidence, other than “face validity,” that I found about the reliability and validity of their measure.
In a nutshell: If you don’t know who’s answering your questions and you don’t know if the instrument is trustworthy, why would you believe the data...let alone the arguments based on those data.
Examine some implications for policy and practice.
Even though Wyse and Bradbury seem to have very little evidence on which to base their case, they explain that there are implications for policy and practice. They did well in accomplishing this goal, I think. At least, they wrote a lot about policy and practice.
However, the arguments seem to be based more on beliefs than evidence. I encourage readers to examine those arguments on their own. But, consider the analyses of the science on which Wyse and Bradbury based their case. From my perspective, I think they whiffed on (a) evidence and (b) reasoning.
People can disagree about “logic,” but in my view, when you start from faulty premises, it’s hard to reach tenable conclusions. And, then if you compound the problem of evidence by using faulty reasoning (Anyone who teaches phonics is mis-serving children; this teacher teaches phonics; therefore, she is mis-serving children. All spotted dogs are dalmations; this animal has spots; therefore, it’s a dalmation), the problem grows.
Report new empirical findings
Well, Wyse and Bradbury did report new findings. Those findings indicated that (a) a selected sample of studies showed less-than great results (but they didn’t show that the results where the consequence of the teaching) and (b) an unrepresentative sample of teachers said, in their opinion, phonics was over emphasized.
Do those outcomes mean teaching phonics is bad?
Analysis
It seems to me that the argument advanced by Wyse and Bradbury boils down to a concern about advocating teaching phonics exclusively. To paraphrase, “Educators should not teach phonics to the exclusion of teaching other aspects of literacy.” I think that argument is correct, and I’ll come back to this point.
First, though, let me show that I am not alone in harboring concern about the argument advanced by Wyse and Bradbury. There are lots of misdirections on the Web about this topic. Does Ireland use “whole language?” Are advocates of phonics unconcerned about vocabulary and world knowledge? Here are some links:
J. Carrol in Times Education Supplement: https://www.tes.com/magazine/teaching-learning/eyfs/phonics-defence-systematic-approach
G. Ashman (2021 a, b) on his SubStack blog, Filling the Pail (which he remarks, “W. B. Yeats never said”) has two entries;
Devin Kearns Tweeted about this matter:
Read the "Reading Wars or Reading Reconciliation" paper bera-journals.onlinelibrary.wiley.com/doi/10.1002/re… after the dramatic Guardian story "Phonics is Failing Our Children" theguardian.com/education/2022… ... I'm struggling to match the data to the rhetoric. UK reading research colleagues, what am I missing?
Here’s an important question: Would a rigorous analysis of show that gains for children who received phonics-base insturction during their early years show that those gains not only faded out (did they?) but faded more than any gains by kids who received the balanced liteacy advocated by Wyze and Bradbury? In other words, suppose we gave Wyze and Bradbury a batch of schools (say 100 that were randomly chosen from all the schools in Britain) and the Direct Instruction people a similar sent of schools (random, etc.): On which approach would you bet for better outcomes on comprehension after, say, 3, 4, or 5 years?
Let me note that I agree that reading education should not focus narrowly on phonics. Teaching kids to sound out words (i.e., “phonics”) is terrifically important, but it is not all there is to reading education. It’s a dang important part of the game, but there are other parts that should be addressed.
The big idea about reading is to get meaning from printed text (which is what I hope readers are doing right now!). It’s very hard for readers to get meaning if they can’t decode. So decoding is a prerequisite for understanding.
Goddess forbid that decoding is the whole game! If it was, kids would get as much fun from reading the phone book as they would from reading “authentic” literature.
But, if they can decode and then they have no frickin’ idea about what the words they’re decoding mean, they are lost. They may be able to decode “dragon,” but if they have little knowledge of the difference between a medieval, fire-breathing, imaginary monster and a millions-of-years-old Tyrannosaurus Rex, that they can decode “dragon” and “dinosaur” accurately is insufficient.
Is decoding the only perquisite? No! Readers have to know about vocabulary so that words like “astrology” and “astronomy” don’t trip them (but note the importance of decoding in discriminating one of those from the other!). Other words—examples are myriad, but think about “drama,” “history,” “phylogenetic,” “fratricide,” “severe”—require understanding that does not come from decoding. One has to know the meaning of these (and way many other!) words, and you don’t get the meaning solely from phonics. Phonics alone may not guide your pronunciation of “severe”; whether the word is used as a verb or an adjective will guide a reader to know whether to emphasize the second syllable. To be sure, you benefit from knowing the relationship between “comb” and “combine,” but that understanding is about the concepts, not (much) about the decoding. You get it from much broader experience. But, importantly, you get some damn important hints by being good at decoding...think again about the spelling differences between “causal” and “casual.”
In addition, to being able to extract meaning from what one reads, one must have life experiences (e.g., what it’s like to ride a train and have a conductor ask for tickets). Readers need general knowledge to comprehend what they read, as Wexler, 2018, argued (see, also, Kintsch, 1988).
Conclusion
Correct, phonics instuction is not all there is to reading. Literacy should include lots about opportunities to read real literature and content, too. But those who condemn teaching children to decode will promote the sorry consequences of illiteracy.
Sources
Ashman, G. (2022, 19 January). Has synthetic phonics been demolished? I cannot believe we are all doing this again. Filling the Pail. https://fillingthepail.substack.com/p/has-synthetic-phonics-been-demolished
Ashman, G. (2022, 21 January). Decoding the ‘phonics only’ claim. Filling the Pail. https://fillingthepail.substack.com/p/decoding-the-phonics-only-claim
Baltar, F., & Brunet, I. (2012). Social research 2.0: virtual snowball sampling method using Facebook. Internet research. https://www.emerald.com/insight/content/doi/10.1108/10662241211199960/full/html
Bradbury, A. (2022, 19 January). Why are ministers obsessed with teaching children to read using phonics? The Guardian, https://www.theguardian.com/commentisfree/2022/jan/19/ministers-obsessed-teaching-children-phonics-nonsense-words
Cooper, H., Hedges, L. V., & Valentine, J. C. (Eds.). (2009). The handbook of research synthesis (2nd ed.). New York: Russell Sage Foundation.
Finlayson, K. W., & Dixon, A. (2008). Qualitative meta-synthesis: a guide for the novice. Nurse researcher, 15(2), 59-71. https://doi.org/10.7748/nr2008.01.15.2.59.c6330
Handcock, M. S., & Gile, K. J. (2011). Comment: On the concept of snowball sampling. Sociological Methodology, 41(1), 367-371.
Kintsch, W. (1988). The role of knowledge in discourse comprehension: A construction-integration model. Psychological review, 95(2), 163.
McGregor, S., & Schmitt, C. (2022). Both sides when there’s only one. NiemanLab. https://www.niemanlab.org/2021/12/both-sides-when-theres-only-one/
Ehri, L. C., Nunes, S. R., Stahl, S. A., & Willows, D. M. (2001). Systematic phonics instruction helps students learn to read: Evidence from the National Reading Panel’s meta-analysis. Review of Educational Research, 71(3), 393-447. https://doi.org/10.3102%2F00346543071003393
Nye, E., Melendez‐Torres, G. J., & Bonell, C. (2016). Origins, methods and advances in qualitative meta‐synthesis. Review of Education, 4(1), 57-79. https://doi.org/10.1002/rev3.3065
Wexler, N. (2018, April 13). Why American students haven’t gotten better at reading in 20 years. The Atlantic. https://www.theatlantic.com/education/archive/2018/04/-american-students-reading/557915/
Suggate, S. P. (2016). A meta-analysis of the long-term effects of phonemic awareness, phonics, fluency, and reading comprehension interventions. Journal of Learning Disabilities, 49(1), 77-96.
Weale, S. (2022, 19 January). Focus on phonics to teach reading is ‘failing children’, says landmark study. The Guardian. https://www.theguardian.com/education/2022/jan/19/focus-on-phonics-to-teach-reading-is-failing-children-says-landmark-study
Wyse, D., & Goswami, U. (2013). Synthetic phonics and the teaching of reading. British Educational Research Journal, 34(6), 691-710. https://bera-journals.onlinelibrary.wiley.com/doi/full/10.1080/01411920802268912
Wyse, D., & Bradbury, A. (2022) Reading wars or reading reconciliation? A critical examination of robust research evidence, curriculum policy and teachers’ practices for teaching phonics and reading. British Review of Education, 10(1), e3314. https://doi.org/10.1002/rev3.3314 [Yay for open access!]
Zimmer, L. (2006). Qualitative meta‐synthesis: A question of dialoguing with texts. Journal of Advanced Nursing, 53(3), 311-318.
If only reporters knew how to deconstruct bogus/faulty claims! Or at least asked someone who does…
Although MK is right, I confess I do depend at times on journalistic summaries of new, and seemingly meaningful, research. I suspect many people, even those with substantial academic credentials do the same. So much to read, so little time.
Of course, if you’re serious about assessing the veracity of claims, you must examine those claims with some knowledge of what the relevant background literature actually says. Otherwise, you’re left with the readily available but notoriously unreliable “that sounds right” or “I agree with this” or “maybe” tests, not unlike 15 minute home antigen tests for Covid19.
Perhaps there was a time when mainstream journalism tried for readers/listeners to collect facts, evaluate evidence, elevate probable truth, but now it seems it doesn’t want to, or can’t. So instead, busy us. We now need scholar mediators like Prof. Lloyd, to help us evaluate the evaluations offered by journalistic accounts of science, a function, by the way, he has performed masterfully.
But in the broader view of professional or merely well-informed debates — could be reading instruction, could be vaccination, could be climate , could be, it seems, most anything — how exhausting! Translating science into public understanding seems monumentally difficult before the debate even begins.
There was a time, I think, when discriminating provocative new knowledge from utter nonsense was easier. Not anymore. Now some writers knowingly and willfully try to deceive. Political, not knowledge, “victories” are being chased. We’re all worse off for it.