The Washington PostDemocracy Dies in Darkness

Can reading comprehension be taught?

This is the second of five posts I am publishing this week by cognitive scientist Daniel Willingham, all dedicated to reading and based on his new book, “Raising Kids Who Read: What Parents and Teachers Can Do.”  The first post asked and answered the question of whether the Common Core State Standards could be boost reading comprehension.

Willingham is a professor of psychology at the University of Virginia and author of several books, including “Why Don’t Students Like School? and “When Can You Trust The Experts? How to tell good science from bad in education.” This appeared on his Science and Education blog. Willingham started teaching at U-Va., in 1992, when his research focused on the brain basis of learning and memory. But after 2000, his research centered on applying cognitive psychology to K-16 education. He has in recent years written a number of popular posts for this  blog, including, “Left/right brain theory is bunk” and “The 21st Century skill students most lack.

By Daniel Willingham

Can reading comprehension be taught? In this blog post, I’ll suggest that the most straightforward answer is “no.” Reading comprehension strategies (1) don’t boost comprehension per se; (2) do indirectly help comprehension but; (3) don’t need to be practiced. Let me elaborate on these claims.

To start, let’s think about what goes into reading comprehension. In yesterday’s piece, I pointed out that a critical feature of reading comprehension is the need to tie ideas together via inferences. Consider again the brief text I introduced yesterday:

“We’re not going on that vacation to Miami after all. My wife could get only get vacation time in July.”

To understand the meaning, it’s not enough to understand each sentence on its own; you must understand that the second statement is causally related to the first (time off only in July caused the cancellation of the Miami vacation) and the causal connection requires some prior knowledge of Miami (that it’s uncomfortably hot in summer) and of people’s preferences when they take vacation.

Research shows that readers, especially poor readers, often fail to tie together ideas across sentences (e.g., Cain et al, 2001). Readers notice if they can’t understand a sentence because the syntax is complex or because it contains unfamiliar vocabulary. But poor readers may not notice that sentences actually contradict one another.

Thus, we might think that an obvious route to improving reading comprehension would be (1) to get students to relate the ideas across sentences and (2) to get students to notice it’s a problem if the ideas in a text don’t hang together coherently. These approaches represent two categories of Reading Comprehension Strategies (RCS). Students are given a task to complete with a text—create a summary, for example, or draw a graphic organizer (idea web) of the text. Such tasks require coordinating the ideas across what they’ve read. Another type of RCS is an injunction to the reader to notice whether or not the text makes sense. A third type of strategy suggests that students “activate prior knowledge” in recognition of the fact that tying together the ideas in the text likely requires prior knowledge.

Here’s why I say that using these strategies doesn’t really make the child a better reader. We’re tempted to think that teaching RCS is like coaching. A baseball coach tells a batter to mimic the sort of things that an excellent hitter does: make the stance relaxed but ready, step into the ball, and so forth. The idea is that by doing what good hitters do often enough, these practices become second nature, and so the amateur’s hitting skill improves. Likewise, if we prompt the beginning reader to do what more successful readers do, in time these techniques will become second nature.

But we can’t actually tell the reader exactly what to do because comprehension depends on the particulars of the text. The sentences about vacation and Miami had to be related, but they relate in a particular way (summer, heat, vacation) that is more or less unique to this text. I can’t give a reader all-purpose instructions about how to connect sentences, other than to say “sentences must connect.”

So baseball coaching is a bad analogy for RCS. Here’s a more apt analogy. Suppose you bought a desk at Ikea which you were to assemble yourself. The instructions, in their entirety read: “Think about desks you’ve seen before. And every now and then step back and see if what you’ve got so far looks like it makes sense.” This is good advice, but it doesn’t tell you how to build your desk. For that, you need to know whether flap A goes into slot B or slot C. You need the specifics of the connections you are to make. Likewise, RCS instruction doesn’t give the specific connections to make among the ideas of a text. It can’t, because how to connect ideas depends on the specifics of the ideas.

It sounds like RCS instruction shouldn’t work, but a great deal of research shows, without a doubt, that children who receive instruction in RCSs are better able to understand texts than they were before the instruction (e.g., Suggate, 2010). Why?

I suggest that RCSs are better thought of as tricks than as skill-builders. They work because they make plain to readers that it’s a good idea to monitor whether you understand. Some students are not even clear that reading is meant to be communication—if your eyes pass over all the words, in their minds, then you’re reading. And for students who do perceive that reading is meant to convey a message, RCSs may raise the bar for comprehension. For example, a student may come to appreciate that she should understand a text well enough to be able to summarize it.

So maybe RCSs are best thought of a trick, not something that builds skill. So what? As long as it works, who cares?

The distinction matters because “skill builder” implies something worthy of extended practice. A “trick,” in contrast, is useful but quickly mastered. And in fact, there is ample research evidence that extended practice with RCS instruction is fruitless.

Gail Lovette and I (2014) found three quantitative reviews of RCS instruction in typically developing children and five reviews of studies of at-risk children or those with reading disabilities. All eight reviews reported that RCS instruction boosted reading comprehension, but NONE reported that practice of such instruction yielded further benefit. The outcome of 10 sessions was the same as the outcome of 50.

How much instructional time is devoted to RCSs in American schools? It’s hard to say, but research indicates that more than “just a little” is time that could be better spent on other things, especially (as noted yesterday) to building content knowledge.

Another way for students to build content knowledge is to read in their leisure time. Tomorrow I’ll take up the question of reading motivation.

You can read the first post here: Does the Common Core boost reading comprehension?


Cain, K., Oakhill, J. V., Barnes, M. A., & Bryant, P. E. (2001). Comprehension skill, inference-making ability, and their relation to knowledge. Memory & cognition29(6), 850-859.

Suggate, S. P. (2010). Why what we teach depends on when: Grade and reading intervention modality moderate effect size. Developmental Psychology,46(6), 1556.

Willingham, D. T. & Lovette, G. (2014) Teachers College Record, Date Published: September 26, 2014 ID Number: 1 7701 , Date Accessed: 1 0/4/201 4