Teaching reading has always been a subject of debate, a debate that involves a balancing act of multiple factors. Currently, the “reading wars” are best embodied by debates that are centered on the “science of reading,” a term that has caught fire even though nobody can clearly define it.
But a new working paper does supply some data about one element of reading instruction that provides a significant boost to reading comprehension.
The paper, a product of nine co-authors (from University of Virginia, Notre Dame, and Auburn University) including David Grissmer and Daniel Willingham, starts by laying out the basics of reading debate.
One element of teaching reading is phonics, and there is little true debate about that. Despite charges to the contrary, few-to-no educators believe that reading instruction should be either 0% or 100% phonics instruction. But the ability to decode words, to sound out what the marks on the page say, is only part of the process of reading. When you click on the “listen to article” link at the top of this column, software will perfectly decipher the sounds represented by the letters on this page, but it will not “understand” a single thing it translates into audio.
In students this is termed “word calling.” The student can read aloud fluently, but understands very little of what they’ve read.
So decoding is not enough. Teaching reading means developing the ability to comprehend what has been read. The paper points out that there are two schools of thought about how to teach reading comprehension.
One approach is to address what the paper calls “procedural skills.” For the past couple of decades high-stakes testing has assumed (and therefor teachers have been pushed to teach) that skills like using context clues, spotting main ideas, and drawing conclusions can be taught as discrete skills that exist in a sort of academic vacuum, like studying waves in a lake with no water.
The high stakes tests have followed through with this by using testing strategies that attempt to level the playing field by negating the effects of any prior body of knowledge (such as a reading selection about 12th century trade in Turkey or the infamous talking pineapple questions). The assumption here is that a skill like “spotting the main idea” can be applied equally to all reading selections regardless of whether the content covers familiar or unfamiliar subjects.
If you find that notion counter-intuitive and improbable, many teachers and researchers agree.
The second approach to reading comprehension is to build a “stronger base of previously stored General Knowledge.” Students better comprehend readings about subjects about which they already know something; this does not seem like a radical notion.
Unfortunately, the paper notes, the approaches and methods to building that base are “much more complex, less understood and more difficult to measure” than “procedural skills.” It’s easier to write policy about skill building than knowledge accumulation. And while “students need to know things in order to be better readers,” may seem sensible, the implication— that young students should spend less time in “reading” class and more time in content area subjects—runs against the grain of much elementary school practice.
To test the notion that building knowledge could have a positive effect on reading comprehension, the researchers looked at charter schools that use the Core Knowledge curriculum, based on the work of E. D. Hirsch and his ideas about cultural literacy. Hirsch is embraced by some (mostly conservative) writers and criticized heavily by others.
How you feel about Hirsch likely echoes how you feel about the proposition that all citizens should learn a certain body of knowledge, but even if we all agree on the principal, we can expect the argument about what should be included (or not) in that body of knowledge to rage endlessly. As the authors point out, this kind of General Knowledge education in early grades requires both “not only scientific validity, but also political viability.”
For the Denver-area students studied, the Core Knowledge curriculum appeared to result in a marked improvement in reading scores on the big standardized test (the PARCC). The effects were larger for the low-income schools studied (an effect size of 0.445 for middle and high-income schools, and 1.299 for low-income schools). Scores raised students around 16 percentile points. That’s better than most other interventions attempted over the past few decades.
There are several questions we can ask about the results, starting with the question of just how well the PARCC measures reading comprehension. In addition, the paper doesn’t address the question of how such a program might work out in a public school.
And while the paper is being seen as supportive of E. D. Hirch’s work, we can also ask if any shared body of general knowledge would have similar effects. In other words, would schools need to use Hirsch’s particular list of Things One Should Know to get this test score effect, or could a school’s focus on any reasonably expert body of knowledge get similar effects?
All of those questions aside, this paper suggests that a solid body of knowledge is a critical part of learning reading comprehension, and that schools would be wise to broaden focus from simply teaching decoding and skills to also include a strong element of content knowledge. Practically speaking, to raise test scores, consider fewer hours spent on phonics and reading “skills,” and more time on teaching students about science, history, and other content areas.