11.12.24On Fluency: The Hidden Barrier to Comprehension
Yesterday I shared an overview of the topics in the forthcoming book on reading I’m writing with Colleen Driggs and Erica Woolway: seven key principles of reading instruction that should inform what we do in k-12 classrooms. (Note we’ll be discussing these principles at our Nashville workshop Dec. 5 & 6)
Today, I’m going to share a bit more about the first of those principles, the idea that fluency is an chronic and overlooked barrier to reading success at all grade levels.
Summary: Fluency is the ability to read words quickly and easily as soon as they are encountered. The quickly and easily are important because they imply the lack of reliance on working memory. If the reading itself requires conscious thought, the process will crowd out other more advanced cognitive activities that are required to make meaning of text.
The simplest definition of reading fluency is ‘the ability to read at the speed of sight.’ We take this definition from Mark Seidenberg’s book of that title. It refers to the ability to absorb the meaning of written text as soon as you look at it- to decode not only individual sounds but words and phrases as soon as you perceive them.
Fluent readers rarely have to hesitate to think about what the words they are reading say or how they link together. Of course there are times–with an especially complex text or when your attention has drifted–when even fluent readers may have to re-read a passage to make sense of it, but for the most part meaning accrues as words are glimpsed. There is no discernible delay and little conscious effort.
When you can read effortlessly, without having to think about it, your working memory, that critical part of your brain for creating memories and understanding, is free to do other things like think about the meaning of the text or perceive details within it. When the first parts of reading are automatic, conscious thought and attention are freed up and can be used elsewhere. For this reason, fluency is a prerequisite to comprehension. If you cannot read at the speed of sight, your ability to understand, think about and remember what you have read will be limited.
In fact in most cases a fluent reader can’t not read a piece of text in their native language. You see a road sign and have processed that it says “No Parking” as soon as you’ve perceived the words. The reading happens so quickly, effortlessly and automatically[1] that you do not have time to decide not to read it. Making a conscious decision takes about half a second and by then you would have read the words. Only if you were highly distracted could you look at the sign and NOT read it.
If you are a fan of Duolingo say or have experience trying to learn a foreign language, particularly as an adult, you have probably experienced the sort of effortful reading that is common for dysfluent readers. You read and must pause to consciously remember words and decipher phrases. You can often figure them out within a few seconds, but only through conscious effort and by the time you’ve done that you then have to circle back and read the sentence again, possibly several times, to figure out what it means. When listening or reading, you can keep up for a sentence or two but your working memory is soon overloaded and you lose the ability to sustain meaning-making. It’s a lot of cognitive work and you tire quickly. As long as you have to extend effort to decipher the words, you are not yet able to read in the fullest sense of the word.
In fact, fluency requires three distinct things at once.
First, it requires accuracy. You must reliably read the sounds and the words correctly. While decoding is the name of this process for speech sounds, orthographic mapping is the name for the process by which each unique sequence of letters becomes “glued” in your mind as a word. You see it and recognize the word nearly simultaneously and no longer need to fully decode it.
But fluency also requires automaticity, which is accuracy at speed. Reading rate is an important precursor to reading comprehension. Faster isn’t always better but a certain level of rapidity is almost always required. Many researchers posit 110 words per minute as a baseline rate. For example, in a 20xx study, 91% of students who had oral reading fluency scores at or above 110 words per minute also scored proficient (level 3 or above) on the Florida Comprehensive Assessment Test. Students who did not score 80 words per minute correct were almost assured to do poorly on the FCAT. Students who cannot read accurately and quickly are at high levels of risk for reading failure.
Finally, fluency requires prosody, which is “appropriate expression or intonation coupled with phrasing that allows for maintenance of meaning” (Kuhn, Schwanenflugel, & Meisinger, 2010). Prosody enables us to invest the words we are reading with meaningful expression so they sound like they might if they were spoken aloud. Prosody allows us to emphasize a certain word in a sentence or to link the words in a phrase together. It is meaning made audible.
Accuracy plus automaticity plus prosody is the fluency formula, and multiple studies have demonstrated its deep connection to comprehension as well as the fact that far fewer students have mastered it than you might expect.
Generally, most studies find that about half of demonstrated reading comprehension is predicted by reading fluency. David Paige and his colleagues at the Northern Illinois University found, in a study of sixth- and seventh-grade students, that oral reading fluency explained between 50% and 62% of differences in reading comprehension (Paige 2011a). Sabatini, Wang and O’Reilly (2019)[2] studied the connection between fluency and overall scores on the 2002 NAEP reading assessment and found that “the strongest predictor of NAEP comprehension scores was reading rate.” Bloomquist (2017) found that 45% of the variation in reading comprehension levels among 4th and 5th grade students in Colorado[3] was attributable to oral reading fluency[4]. It’s likely that the predictiveness of fluency declines slightly as students age but Schatschnieder and colleagues[5] found that reading fluency accounted for about a third (32%) of reading comprehension scores among 10th graders. Even that late in a student’s career, the connection remains strong.
“Slow, capacity draining word recognition processes require cognitive resources that should be allocated to comprehension. Thus reading for meaning is hindered,” Keith Stanovich and Anne Cunningham summarized in their 1998 analysis What Reading Does For the Mind, but they also outlined a secondary effect of dysfluency: “Unrewarding reading experiences multiply.” Struggling to understand and reading without learning much make reading appear to have less value. Place this alongside the greater cost in terms of effort required for dysfluent readers and the value calculus tips away from reading. “Practice is avoided or merely tolerated,” it is done “without real cognitive involvement.” Dysfluent readers stop wanting to read or experience reading as not being especially meaningful and so the gaps between them and their classmates widen.
Though research into the scale of dysfluency is limited, studies suggest that it is a wide-scale problem. Analyzing the data from Sabatini’s study of more than 1700 4th graders, for example, Paige notes that “41.7% of 4th grade students—almost half—appear to have reading fluency issues,” and that such issues are “strongly associated with poor performance on the NAEP.[6]” A 1995 report by the National Center for Education Statistics found similarly that just “55 percent of fourth graders were considered to be fluent,” and that after reading a passage twice silently, only about 13 percent of the fourth-graders in a NAEP study “could read with expressive interpretation and consistent preservation of the author’s syntax.[7]”
Moreover, fluency remains a pervasive issue among older students. Paige found gaps in fluency and a strong correlation to comprehension among sixth and seventh grade students, well beyond the years where most schools remain attentive to fluency and in Schatschneider’s study of Florida students, more than half of tested 10th graders demonstrated fluency rates below proficient.
And the effects are of course not limited to testing. A study of Italian[8] students, found that “Reading fluency predicted all school marks in all literacy-based subjects [we’d argue that all subjects are literacy-based!], with reading rapidity being the most important predictor.” The authors added, “School level did not moderate the relationship between reading fluency and school outcomes, confirming the importance of effortless and automatized reading even in higher school levels.[9]”
Two findings jump out from that statement- first the simple importance of rapid reading specifically and second that fluency matters at all grade levels- though it is least likely to be assessed and therefore recognized among older students. (The students in this study were in grades 4-9).
Despite this data, “fluency has been relatively neglected beyond the elementary grades,” Paige observes. There is, he writes, “little instruction occurring…to improve reading fluency” beyond the mid elementary years and by middle and high school, teachers are more likely to “employ work-arounds so students don’t have to read text” during class.
Research suggests that fluency poses a hindrance to comprehension for close to half of students even into high school in other words. “At the end of the day, my hunch is that 40% to 50% of middle school students do not have proper reading fluency,” Paige told us. “In schools where students generally struggle with academic attainment, this percent is likely closer to 80%.”[10]
Given that the texts students are expected to read become more complex and therefore demanding from a fluency standpoint, there is little reason to suspect there are not large numbers of students in high school and even college for whom reading fluency is a massive and hidden barrier to reading comprehension. And when that is the case, were are less and less likely to know about it. When was the last time the average 9th grader’s oral reading fluency was assessed?
A colleague of ours observed that not only did she as a teacher rarely ask older students to read aloud and so know little about their fluency, but that she did the same with her own children. “I never ask them to read aloud anymore,” she said of her middle school-aged children. “I suddenly realized that it has been years since I had even an intuitive sense for their fluency.”
She’s right to be worried, especially given that almost all of the existing research data was conducted before the precipitous rise of the smart phone and social media which have dramatically reduced independent reading outside of school among American teenagers to a fraction of what it once was (see page #). The problem is almost assuredly worse now.
[1] Hecotr Ruiz Martin notes, in How Do We Learn, the physical act of “reading is procedural knowledge , and as such, it is impossible to avoid doing it when we see words (if we are expert readers).
[2] https://psycnet.apa.org/record/2018-39179-001
[3] https://digitalcommons.usu.edu/cgi/viewcontent.cgi?article=6744&context=etd
[4] 53% was attributable to the the different but less-frequently assessed skill of silent reading fluency
[5] https://files.eric.ed.gov/fulltext/ED495465.pdf
[6] Interview with David Paige by the authors 12.11.23
[7] https://nces.ed.gov/pubs95/web/95762.asp
[8] We note here that English is more orthographically complex than Italian. This is to say it’s less predictable and consistent in spellings and sounds. It’s harder to read fluently and so we might conjecture that Bigozzi’s findings would be even more stronger manifested among subjects reading in English.
[9] https://pubmed.ncbi.nlm.nih.gov/28261134/
[10] Interview with the authors, 12.11.23