That Baseball Study Can’t Tell You How to Teach

The baseball study is a 1988 article frequently cited to support pedagogies based around highly-structured teacher-led classrooms.

14 min read
That Baseball Study Can’t Tell You How to Teach

To really engage with research into learning, we need to understand what studies can and cannot say.

Interact with a certain subset of educators on the internet, and it’s only a matter of time before you’re directed to something known as the “Baseball Study.”

The baseball study is a 1988 article from the Journal of Educational Psychology by Donna Recht and Lauren Leslie, with the full title “Effect of Prior Knowledge on Good and Poor Readers’ Memory of Text,” and it is frequently cited to support pedagogies based around highly-structured teacher-led classrooms.

In the study, Recht and Leslie examine how prior knowledge affects short-term memory and recall, using both verbal and nonverbal assessment. Sixty-four junior high students participated, divided into four groups based on two factors: reading ability and prior knowledge of baseball (creating groups of high-high, high-low, low-high, and low-low). Each student was first asked to silently read a narrative text that recounted a half-inning from a baseball game. Then, the students recreated the narrative, by moving wooden pieces around a miniature baseball field and verbally describing what had happened. Next, the researchers asked the participants about their school (to interfere with working memory) before finally asking the students to summarize the half-inning and sort 22 sentences, “on the basis of importance of ideas to the text.”

In all measurements, the students with high prior knowledge performed better, regardless of which reading level they had been grouped with. Recht and Leslie drew the following conclusion from their work:

Although direct strategy instruction makes a needed contribution, it is not enough to consider [reading] strategies without consideration of the subjects’ knowledge base (Chi, Glaser, & Rees, 1982). In light of the importance of adequate prior knowledge, strategy instruction and the knowledge base should be equally considered in the design of instruction.

In other words, broad one-size-fits-all reading strategies will not help students fully comprehend every text they read, and if you want them to understand a particular text, you need to consider their background knowledge.

Pictured: A webpage from Core Knowledge on the Baseball Experiment.

The study is a favorite of movements such as E.D. Hirsch’s Core Knowledge and the “science of reading” advocated by Emily Hanford and Natalie Wexler. It’s referenced in their blogs and op-eds and videos. Engage in dialogue with educators who identify with these groups, and it will come up sooner or later.

It’s easy to see why — if you could just put Recht and Leslie’s “low prior knowledge” students through a set of structured lessons that systematically introduce them to all the standard concepts of baseball (and, well, everything else), they could easily recreate the half-inning after reading the text. Their literacy, we assume, would be improved.

The study’s basic conclusion — that background knowledge in a topic improves a person’s ability to comprehend complex texts about that topic — is clear, and unlikely to meet any serious disputing. Yet, I take issue with the way the baseball study is typically applied, because there are several questions it cannot answer.

Most importantly: why should all these students learn about baseball?

A couple months ago, I inadvertently participated in what you might call a small-scale recreation of the baseball study.

About once a month, I meet with my book club (virtually, these days). The only theme connecting our reading list is that, amongst the four of us, we take turns picking the book. That’s it. We’ve read weird science fiction like Jeff Vandermeer’s Dead Astronauts, so-called “canonical” texts like John Steinbeck’s East of Eden, and dense biography like Manning Marable’s Malcolm X: a Life of Reinvention. One recent meeting was to discuss The Natural, a 1952 baseball novel by Bernard Malamud (later adapted into a 1984 film starring Robert Redford).

Pictured: A VHS cover of The Natural with Robert Redford throwing a baseball.

I do not like baseball. It’s slow, boring, and technical, and my memories of sitting through minor-league games as a kid mostly consist of waiting for them to end. As such, I’ve never taken the time to learn much about the game.

And just like the sad children from the 1988 study, I had a little difficulty understanding portions of the book. Long chapters describing play-by-play action, the importance of various statistics held by Roy Hobbes, even the general shape of the season — all these were unfamiliar to me.

Had I been been asked to recreate certain chapters of The Natural (or, for that matter, been given a multiple-choice quiz over them), I likely would have failed. Yet, these assessments would have missed other dimensions: that I read the entire book, that I participated in a book club discussion, that I understood the characters and their conflicts, and that I had an opinion about the book and could articulate it. The core of the story was intelligible to me, and I may have even picked up a little baseball knowledge along the way.

But perhaps more to the point — I got what I personally needed out of the book (in this case, connection with my friends). My book club experience was not diminished by my incomplete comprehension of The Natural, and my life is not diminished by an incomplete knowledge of baseball. There may be better goals to education than 100% content-knowledge recall.

Studies cannot tell us how we should teach our classrooms or run our schools. That simply isn’t their function, no matter how well controlled the variables are.

A scientific study is, at its simplest, an observation of what happened. Some studies operate in the real world, with all its messiness, others in a lab, sterile and controlled; some observe over long stretches of time, others examine one particular moment. All of them, though, are descriptive rather than prescriptive — they tell you what happened, not what action you should take. The study cannot tell you that, because it does not know your situation, goals, or values.

In the case of the baseball study, it is an observation of a constructed environment. The researchers created an artificial scenario with particular parameters, controlled it carefully to isolate the variable they wanted to observe, then ran their test.

Their results tell us that students who had higher knowledge of baseball were better able to recall and recreate factual details about a text related to baseball. It does not tell us whether or not the other students (those who scored poorly) connected to the text in a manner unrelated to factual recall, or how to meaningfully build background knowledge, or for that matter whether or not baseball is a worthy topic of study at all. What you do with something like the baseball study is a reflection of your values.

For educators with authoritarian values — that student behavior must always comply to adult expectations, that the best teaching tells students what to think, that the purpose of education is to assimilate people into a particular story about America — the lesson of the baseball study is simple. To teach kids to read, simply fill their heads with as much background knowledge as possible, as efficiently as possible. And the most efficient way to do that is for the teacher to decide what background knowledge they learn and when, and for the students to comply. The teacher decides what is important and worthy of study.

For progressive educators, those of us who value liberation — that kids deserve respect as fully autonomous humans, that diversity of values and goals is good in a pluralistic society, that people need not conform to anyone’s expectation to lead full lives — the baseball study yields a different conclusion. To teach kids to read, make available as much diversity in books (and other media) as possible. Stock your shelves with books of every topic and complexity and let students read whatever they’re drawn to. Encourage wide reading to expand their interest and deep reading of what they’re already interested in.

To argue, from the baseball study, for a comprehensive knowledge-building curriculum, is to believe that the goal of education is to conform kids in the image of the adult in the room. The adult knows what is important to believe and will structure a series of lessons that faithfully guide the student to the same place. But the only way to build that is to focus your curriculum on the accumulation of facts.

But Hirsch, Wexler, and the others erase their values from this narrative, simplifying the story into one where they follow “the science of reading,” where they follow the plain conclusions of the baseball study, and everyone else is foolish, casting aside the insights of modern science in favor of their own superstitions. At times, they go so far as to compare progressive educators to climate-change deniers, an analogy which both ignores the immense weight of the scientific observations supporting anthropogenic climate change (not in any way matched by the “science of reading”) and does harm to the hard-won advances of climate activists by watering down their language.

Progressive educators do not deny that background knowledge contributes to greater ease in reading comprehension — it does, every reader knows that, regardless of whether or not they’ve encountered the baseball study — we just don’t think it’s our place to micromanage our student’s background knowledge. Rather, we consider it may be better, in the long run, to let students read whatever they already have background knowledge in. This will allow them to easily read texts of increasing complexity, because they will be building off what they already know.

Pictured: The book cover of The Natural Approach, featuring an abstract yellow and blue design.

This approach taps into the students’ intrinsic motivation, because the things they like are in fact the same as the things they have background knowledge in. Give them books they’re interested in, books that answer their questions and solve their problems, and they’ll read. Stephen Krashen, an expert on second language acquisition,* wrote, in regards to selecting texts for students based on unfamiliar vocabulary (ie background knowledge) and syntax, that

It has been widely observed informally that if readers are genuinely interested in content, this interest can outweigh other factors to a large extent. In fact, interest in content may be the most important consideration in selecting appropriate texts. It may also be the most difficult requirement to satisfy, far more difficult than controlling for syntax or vocabulary . . . In addition to an interesting text at the appropriate level of difficulty, the reader also needs to have a goal; there needs to be some reason for reading, some information or message in the text that the reader is looking for.

No disrespect to Krashen, but I disagree on one point: “interest in content” ought to be the easiest criteria to solve. Just let students pick the texts themselves. Some students will (for various reasons, most likely related to bad classroom experiences located in their prior knowledge) refuse to read books, but you don’t have to limit their options to books. Blogs, online message boards, and comics include written language too. Since no text exists in a vacuum, students will, over the course of their reading, acquire background knowledge in many different topics and expand their interests.

Incidentally, this is another way to read the baseball study. Those students with high prior knowledge of baseball? They were probably students who liked baseball (I mean, who else knows much about baseball?), and thus would have had an easier time engaging with the text. The other kids may have just been bored by it.

But perhaps more important than choice-reading’s ability to help students read more complex texts or tap into intrinsic motivation, it respects their autonomy. Because not everyone cares about baseball, and knowledge of the game has little-to-no value for those who don’t.

Bernard Malamud’s The Natural is, like much of American literature from its time, about the deadened ennui felt by many Americans — in particular, white Americans — in the years after World War II, and my general familiarity with that time and narrative (acquired through countless novels, movies, and television shows) helped me comprehend a story about baseball, something I have only the most basic knowledge of. Another reader, though, might have had the opposite experience, using their detailed understanding of baseball to read a story about an unfamiliar time and place.

Pictured: The book cover of Black Leopard Red Wolf with animal design.

As another example, consider Black Leopard, Red Wolf, Marlon James’ epic fantasy based on African history and mythology. I read this last summer, and came into the experience with very little background information. As a lifelong fan of fantasy literature, I’m well versed in words like “elf,” “goblin,” and “wraith,” as well as basically anything related to the architecture, landscapes, and technology of medieval Europe. My brain has no difficulty supplying images for these words.

But of course, Black Leopard, Red Wolf featured a completely different set of fantastical creatures, and my brain was not filled with ready-made images for the omoluzu, demons who walk upside-down on the ceiling as they stalk the heroes, or the impundulu, a creature somewhat like a lightning vampire.

Due to my extreme shortage of relevant background information, I had to read Black Leopard, Red Wolf slowly. But here’s the thing: I did read it, I understood it, and I enjoyed it. Had I been assigned the book in a class and tested over it, I may have had a less enjoyable experience. But reading of my own volition and at my own pace, it was one of my favorite books of the year.

If you value the teacher’s role as an authority and the state’s job setting curriculum standards, then there’s certainly logic to the Core Knowledge interpretation of the baseball study. And it’s hard to deny that, in our current system of education, the teacher is the authority and the state does set standards. Even for teachers who value liberation, who dream of a different way of doing school, it’s tempting to give in to the powers that be and design this kind of curriculum.

At times in my career, I’ve done this. Several years ago, I spent one week of my summer reading through the entire backlog of released STAAR (State of Texas Assessments of Academic Readiness) English tests and categorizing each reading passage. My thought was that I could identify the topics most frequently used, then design a curriculum that used Hirsch’s domain immersion structure to give students background knowledge in the most common topics. I was trying to address Wexler’s Knowledge Gap, before I had ever heard that term.

Pictured: "The test-score gap is at its heart, a knowledge gap." with corresponding book cover and #knowledgegap.

The result was… fine. I found (what I thought were) interesting texts for each of those topics and used them as the basis for open-ended Socratic discussions, rather than drilling factual recall, and it wasn’t the worst teacher-led curriculum I’ve ever put together. At the end of the year, though, test scores hadn’t changed significantly from the previous year, and I’d wasted time that could have been spent on more meaningful learning experiences. I had compromised my pedagogical values for nothing.

The next year, I changed course, and began to look for ways to increase student choice and autonomy. But district and state policy made this difficult. The STAAR test wasn’t going anywhere, and my school’s admin was (justifiably) concerned about what would happen if our test scores slipped. Our job was to teach the TEKS (Texas Essential Knowledge and Skills), which were based primarily around close reading, writing specific kinds of texts, and understanding various time periods in literature. It’s hard to successfully advocate for student choice — which appears far less “rigorous” to the modern education system — in that environment.

For the upcoming 2020–21 school year, despite the enormous uncertainty about what school will look like with COVID-19 on the rise, I have something to look forward to: the Texas Essential Knowledge and Skills (TEKS) have been revised, and independent, self-selected reading has been added to the state standards. Under the strictest interpretation, choice reading is now not only allowed, but actually required by Texas law.

Baseball is an interesting subject matter for Recht and Leslie to have selected. It’s an incredibly technical sport, and for those who enjoy it, that’s part of the appeal. A player’s statistics (of varying dimensions) throughout the season is more important than what happens in one particular game, and especially one half-inning. The minutia is the point.

It’s not surprising, then, that when it comes to baseball, someone without prior knowledge would have difficulty with a written narrative of the game. The example seems almost tailor-made to support teacher-led systemized curriculum.

It’s common in education to speak of scaffolding our students’ learning. The metaphor is knowledge as a tower, stretching into the sky, and learning as climbing that tower. But the walls are vertical and smooth, and the student cannot scale it on their own. The teacher, then, builds a scaffold, wooden crossbeams that encircle the tower, providing footholds. The scaffold is barely more complex than a ladder, and the teacher knows exactly which beam is needed where, in order for the student to successfully learn. This image suggests a structured experience, moving upwards until learning has been accomplished.

I would suggest a new image: that of a vast landscape, ready to be explored. In the center of the land stands a tall mountain, with multiple peaks that reach above the tree line. Here it is bare rock, with occasional moss and lichens. Moving down the mountain’s many marked paths (and infinite unmarked ones) you pass through alpine meadows and come into a dense conifer forest. Beneath that, the forests become more diverse and broken up by streams and rock faces. Many kinds of birds, mammals, and reptiles live here. Leave the mountain and you come out into rolling hills and meadows.

Pictured: A landscape with rolling hills and distant mountaintop.

Students begin spread out in different spots across this wilderness, and they are free to explore as they see fit. They may hike up to one of the peaks, but it’s unclear what path they’ll take to get there, and that’s only the goal if the student is interested in the view. Students will explore some on their own, some together, sometimes guiding one another. The teacher has spent years walking the land by themselves, and can serve as a guide, showing students to favorite spots, warning about loose stones in the stream, and pointing out different kinds of birds as they walk along. But the teacher is always aware that the amount they know is only a tiny fraction of what is out there.

It’s tempting to think that, if we could just understand how the brain works, we would know how to teach. But relying on a “baseball study” understanding gives us a simplistic view of education, and the truth is that learning is never linear. People learn everywhere, all the time, and no two students have the same background knowledge — or, for that matter, are missing the same background knowledge.

When we force students into one set curriculum, our focus shifts to the deficits: what are they missing, and how can we scaffold their thinking to bridge that gap? But it doesn’t take much work to shift our perspective — the truth is, each student that comes into our classroom has their own unique set of prior knowledge and interests. And that’s a powerful asset, if you’re willing to let them use it.


*The discussions around the so-called “science of reading” rarely, in my experience, directly call out second-language acquisition as the issue at hand. But it’s an important point of discussion. Advocates of Core Knowledge curriculum and systematic phonics argue that these approaches are necessary for fixing the “opportunity gap” in education. This connection is always made explicit, and is referenced in the title of Natalie Wexler’s The Knowledge Gap. This gap exists largely around racial lines — white kids perform better on “standardized” reading tests than BIPOC kids, and these tests are typically written by white people. If you take a broad view of what makes something a language to include regional and racial dialects as different languages, then this entire work is not about “reading” at all, but about second language acquisition. This is likely even true for white kids, as written English has different rules around its syntax and vocabulary than spoken English, regardless of dialect. That’s why I chose to cite Krashen, an authority in second-language acquisition, in this essay.

Become a Patron!

Unpacking “Neoliberal” Schooling, Part 2: Teachers Pay Teachers
Previous article

Unpacking “Neoliberal” Schooling, Part 2: Teachers Pay Teachers

While Teach Like a Champion, the bestselling secondary education book, trains teachers to control students through behavioral compliance, Teachers Pay Teachers — generally — trains teachers to control students through “fun” activities, worksheets, and slideshows.

Next article


Today, Human Restoration Project launches its course platform: (RH.) RH is part microcredentialing program, course platform, and asynchronous conference tool aimed at instilling progressive educative systems in classrooms everywhere.


🎉 You've successfully subscribed to Human Restoration Project!