The Problems with “Show Me the Research” in Teaching Reading

The “science of reading” movement rehashes the stale Flesch argument that student reading is in crisis because of a failure to teach systematic intensive phonics to all students.

Paul Thomas
6 min readAug 18, 2020

I was born in 1961 and entered first grade in 1967, already able to read independently and play sophisticated card and board games.

My mother had taped index cards with words identifying objects all around our house. She had attended one year of junior college, but had no training in how to teach.

None the less, from the first day of school, I excelled in literacy, scoring in the 99th percentile on standardized tests. My learning to read has two important elements; I was of the generation taught by the Dick and Jane basal readers (whole-word focus over discrete phonics), and my learning to read overlapped with one of the most aggressive reading crisis moments in the U.S., spurred by Rudolf Flesch’s Why Johnny Can’t Read, first published in 1955.

To whom and what should we attribute my high-achieving literacy skills? How could research tease out any causal inferences about my reading achievement? I was certainly not a Johnny who couldn’t read even as I had almost no direct phonics instruction, and I am sure the basal readers and my teachers’ instructional methods had only minimal impact on my literacy development (other than that I loved Ms. Landford, my first grade teacher, and wanted desperately to please her).

Many years later, I was a high school English teacher in the rural high school I had attended. One year, a wonderful student whose mother was also an English teacher in the school scored a perfect 800 on the verbal section of the SAT.

People throughout the school and town often congratulated me and praised my role in her perfect score. For many, that student’s success was proof I was an outstanding teacher of English.

Frankly, I had almost nothing to do with that score; but her excellent literacy development — like my own — had hundreds if not thousands of causal elements leading to her scoring that 800.

Both of these examples, I think, help highlight the problem of proximity and time in making causal claims about who and what contribute to student learning and growth.

Real world teaching and learning are incredibly messy and always cumulative; any student’s measurable achievement has a mind-numbingly complex history behind that achievement that is not fairly attributed to any singular cause.

Now let me offer another complication.

Many years ago, I began confronting the poverty workbook and claims of Ruby Payne. When scholars on poverty and race began to challenge Payne (much of that in Teachers College Record), those challenges often exposed that Payne’s claims about poverty were almost entirely stereotypes and not supported by the research base on social class and race.

In other words, scholars were essentially saying that Payne needed to “show the research.” So she did — and exposed her own misunderstanding about that request.

Payne’s workshops and program became popular when No Child Left Behind injected federal money into schools in order to close the achievement gap. Payne, then, was being hired not necessarily to provide schools with credible expertise on social class and race (which Payne lacks in her training), but to raise test scores.

Payne’s rebuttal to the criticism was providing data suggesting that her workshops did raise test scores for the vulnerable populations associated with the achievement gap (although I am not suggesting that the data she provided did prove such).

This wrinkle to “show me the research” is incredibly important since it draws attention to exactly what we are looking for in that research; Payne’s critics were raising one issue about evidence, and Payne countered with evidence of a completely different kind.

More than six decades after the Why Johnny Can’t Read crisis (which is indistinguishable from the reading crisis in the 1940s before and the whole language “plummet” in the 1990s after), the U.S. is mired in another round of the Reading War, driven by advocacy for the “science of reading.”

The “science of reading” movement rehashes the stale Flesch argument that student reading is in crisis because of a failure to teach systematic intensive phonics to all students.

This round, however, emphasizes “science” and has embraced both a simple view of reading and a narrow view of “science” (evidence).

Advocates for the “science of reading” beat the “science” drum and often demand that anyone challenging their advocacy “show me the research” — while disregarding the problems I have detailed above.

“Show me the research” in this case is limited to the so-called gold standard of research, quantitative empirical research with controls and findings that are generalizable.

In virtually all fields, especially medicine, that gold standard is sacred for good reason.

However, this view of “science” and research is deeply flawed in education, for the problems I outline above.

Direct phonics instruction is easy to isolate and research in order to fulfill the mandate that only some research counts in reading research; this is also the way the National Reading Panel greatly skewed the work of the panel, excluding decades of research that did not meet the threshold of “gold standard.”

“Show me the research” seems on the surface to be a reasonable and even essential starting point when debating how to teach students to read. Yet, it isn’t.

First, demanding research on whole language or balanced literacy effectiveness misunderstands what these two terms represent. Whole language (WL) and balanced literacy (BL) are philosophies of literacy; they are not programs or even instructional practices.

To suggest we can separate in some blunt and clear way phonics fromWL or BL is as misguided as Payne’s response to her critics. WL and BL as guides for instructional practices would include a wide variety of phonics instruction (including direct phonics instruction).

When researchers do try to make that distinction, we find that there is often little or no differences in outcomes (see Bowers, 2020), or as with Payne, the “better” approach isn’t proving what we really are seeking or any difference dissipates over time (See Krashen, 2006; Krashen, 2004).

Ultimately, the “show me the research” demand about the teaching of reading is a problem because the “science of reading” movement has embraced not just simple views of reading and research, but simplisticviews.

Reading comprehension and critical literacy are not a simple formula, and learning to read is a complex and even haphazard process that occurs over many years (if not our entire lives) and depends on a wide assortment of interworking elements such as decoding, content and life knowledge, comprehension, and critical awareness.

While gold standard research seeks to isolate instructional methods, that sort of data has limited use in the real-world, where students are not static beings (and may be outliers) and are not able to control for poverty, racism, and sexism in their lives (all factors that likely have far more influence on their achievement than reading instruction, reading programs, or reading philosophies).

The work of Lou LaBrant reveals that we have had for a century a wealth of research and evidence on what can work to teach students to read.

The problem with “show me the research” about teaching reading is not a lack of research, but a fundamental failure to understand human behavior, especially how one comes to be an eager and independent reader.

Reading is not simple; just because you can reduce it to an algebra equation doesn’t mean you should.

The balanced literacy movement sought to give some philosophical structure to the recognition that learning to read is complex and haphazard. Additionally, balanced literacy was an effort to forefront the professional autonomy of the teacher.

To be blunt, balance literacy (like whole language) has never been implemented with those goals intact. The accountability movement has dominated the teaching of reading with a formula that is somehow absent in the current debate: standards + high-stakes testing = reading programs — teacher autonomy.

To be blunt again, more teachers are being compelled to teach reading in ways that seek to raise test scores (recall the Payne problem from above) and not to foster eager and independent critical readers.

The “science of reading” movement is making and perpetuating the Payne mistake in education.

“Show me the research” in this case is a misdirection and further evidence that many people are not willing to acknowledge the complexity of reading, the complexity of human behavior, or the professional autonomy of teachers.

I became a highly literate person with no direct phonics instruction. That’s a neat little story about my life, but it doesn’t prove a damn thing about anyone else.

This round of the Reading War is weaponizing the term “science” in ways that are guaranteed to distract us yet again from the complicated and politically unpopular work of addressing inequity in the lives and schooling of children.

That evidence is clear and disappointing.



Paul Thomas

P. L. Thomas, Professor of Education Furman University, taught high school English before moving to teacher education.