New Evidence on Teaching Reading at Frustration Levels

For generations, reading experts have told teachers that they had to teach students to read at their instructional levels. Teachers were admonished that if they taught children with books that were too easy, there would be nothing for the kids to learn. If they taught with books that were too hard, then the reading instruction would frustrate rather than improve.

In general, that kind of advice makes sense. Spend all the time you want teaching me my ABCs and it won’t likely improve my reading ability at my advanced level of performance.

And, the idea of teaching someone something that they find to be inordinately frustrating couldn’t possibly work either.

Beyond that, things get a bit fuzzy.

Reading authorities have told teachers that these levels — independent, instructional, and frustration — are the product of two factors: how well kids can read, and how hard the books are. They have also come up with formulas for determining how to match kids and books to avoid frustration. Historically, the scheme usually called for kids to be taught from books that they could read the words of with 95-98% accuracy, and about which they could answer 75-90% of the questions. A bit of challenge — but not too much, was the idea.

Unfortunately, this insightful plan (that many of us have used in our classrooms) was just made up. In today’s parlance, the instructional level is “fake news.” No one bothered to do studies to determine whether that kind of book matching was beneficial to kids or not!

That started to change in the 1960s, when various tests of this scheme were undertaken. Again and again, the instructional level came up short. Studies were finding no correlation between how well matched to texts the kids were and they were often identifying kids who made great learning gains despite being placed in books at frustration levels.

But it wasn’t until 2000 when anyone even bothered to examine the value of the instructional level using a randomized control trial. Then things got really interesting, since those studies found either that it made no difference — in terms of reading achievement — whether kids were matched to texts at their so-called instructional level, or the frustration level kids far outperformed the instructional level ones. In other words, it was either a waste of time to match kids to books or it was hurting kids!

I’ve written about that issue frequently here and between those writings and the educational standards in many states that require kids be taught to read books at their grade levels there has been some effect. For example, I’ve noticed that several of my colleagues, when prescribing instructional level texts, now suggest 90% accuracy (instead of 95%) as the true identifier of the books kids can learn best from. Of course, they don’t have data to support these new criteria — the new numbers are as made up of whole cloth as the criteria they are replacing — but it is probably enough of a smokescreen to convince some educators that what they are being sold is consistent with research and standards (though in neither case is this true).

Why am I bringing this all up again? Because this week a new study appeared, this one published in the estimable Journal of Educational Research, and conducted by Lisa Trottier Brown and her colleagues. This study (The effects of dyad reading and text difficulty on third-graders’ reading achievement (opens in a new window)) pursues this issue with third-graders. “Results indicate that weaker readers, using texts at two, three, and four grade levels above their instructional levels with the assistance of lead readers [other, better reading, third graders], outscored both proficient and less proficient students in the control group across multiple measures of reading achievement.”

As in past studies, the results suggest not that we just have the wrong criteria for the true instructional level (there was no best book match here), but that it is unlikely there is such a thing as an instructional level; at least in terms of matching kids with books.

The key, of course, is that while inordinate amounts of frustration should be avoided in instruction, that can easily be accomplished with grade level books and supportive teaching (like the paired reading that took place in this study). The instructional level is not a student-text match. Placing kids in easier, below grade level books reduces their opportunities to learn, but learning will only take place with accommodative and supportive instruction.

Past studies have suggested that the traditional instructional level would be a great goal to have for books at the end of the lesson rather than at the beginning. Instead of trying to avoid exposing kids to things they don’t know, we need to make sure that they learn what we expose them to.

I know some colleagues, Dick Allington, for example, is not impressed that after 70 years there still isn’t any research supporting the idea of matching kids to books (beyond grade 1), and believes that the studies that have been done are flawed because they don’t just vary the book levels, but the instructional approaches themselves. No one, however, is claiming that just placing kids in harder books leads to greater learning — clearly harder books require instructional adjustments by teachers that are an important part of the equation.

Exposing kids to grade level text will not automatically raise student learning. It just provides an opportunity for greater learning. Instructional techniques — like the dyadic reading in this study — are an example of that kind of instructional adjustment. Additional guidance with vocabulary, grammar, cohesion, structure and other aspects of text complexity should have their place too.

So now we have even greater evidence that teaching kids with what they refer to as “the stupid books” (the ones below grade level) doesn’t benefit kids. I wonder if teachers and reading supervisors will listen this time?