Monday, July 11, 2011

More Evidence Supporting Hard Text

The past couple of blogs have dealt with the challenging text demands required by the new common core standards. Teachers who have been used to moving students to easier texts are in for a rude awakening since the new standards push to have students taught at particular Lexile levels that match grade levels rather than "reading levels."

Last week, I explained the evidence about the importance of text difficulty that was provided by the ACT. This week, I want to expand upon that explanation to show some of the other evidence that the authors of the common core depended upon, evidence that has been persuasively described and summarized by Marilyn Jager Adams in an article published in the American Educator (2010-2011).

Adams synthesized the information from various studies of textbook difficulty and learning, to demonstrate that textbook readabilities for Grades 4–12 have significantly and steadily grown easier since 1919; the difficulty of what adults are expected to read increased during that same time; and there is a relationship between the easing of text difficulty and students’ lower performance on the SAT. Obviously, if these things are true, one would want to ratchet (as the common core does) the difficulty of textbooks back up so that students would be better prepared for the actual reading demands beyond school.

Chall and her colleagues (Chall, Conard, & Harris, 1991) found that despite the fact that SAT passages had been getting easier, scores were declining anyway. Nevertheless, they found that textbooks were getting easier even faster than the SAT, and that reading these easier books appeared to provide poor preparation for dealing with the SAT. Even more convincing was a much larger study (Hayes, Wolfer, & Wolfe, 1996) that examined the readabilities of 800 elementary, middle school, and high school textbooks published between 1919–1991. Hayes and his team correlated the trends in text simplification with student performance on the SAT and found a good fit, concluding that “Long-term exposure to simpler texts may induce a cumulating deficit in the breadth and depth of domain-specific knowledge, lowering reading comprehension and verbal achievement.” Also, the texts used in high school have been found to be significantly easier than the texts students confront after they leave high school; in fact, young people make bigger reading gains during the years following high school than during it (Kirsch & Jungeblut, 1991).

Thus, these correlational data suggest that students will learn more from working with challenging texts than from the so-called “low readability, high interest” books that have become an educational staple. This approach is similar to that taken by athletes: To get stronger, you need to use more physical resistance than your muscles are used to; the more you do, the more you will be capable of doing, so it is essential to increase the workload.

The counter-argument to this heavier-books approach is the widespread belief that there is an optimum difficulty level for texts used to teach students to read. According to instructional level theory, if a text is written at a level that is too difficult for students, then they will become frustrated and discouraged and will not learn. Instructional level theory not only doesn't agree with the idea that learning comes from working with hard books, but claims that little or no learning would accrue if the books are too hard relative to student performance levels.

The evidence that supports the challenging-text approach obviously has some research support, but this is correlational in nature. Students seem to do better when they get a steady diet of more challenging text, but I would feel much better about this evidence if it were experimental and if there wasn't such a long-cherished counterargument. Given that, the next installment will weigh the evidence that supports the idea of there being an optimum level of text difficulty that fosters learning.

4 comments:

EC said...

The studies you describe seem to ignore the key fact that many, many more people take the SAT every year. All of the effects you describe would be explained by expanding the population of SAT test-takers (and college grads) to a more representative and less elite group.

Tim Shanahan said...

EC-
Actually none of these effects are explainable by the expanding population who takes the SAT. During the years that were studied that population was pretty stable. That there is a correlation between the drop in SAT scores and the decline in text levels in schools is not definitive proof of causation. However, no statistical analysis of the SAT decline has ever been able to successfully blame the decline on anything like population changes or problems with the assessment.

Stephanie said...

As I view our videos and read your blog on text complexity I am curious to know what your thoughts are on reading intervention. Would you support that a child that is two or three levels below reading continue to read texts that are on grade level? When you refer to complex text level are you focusing on what students in Tier I should work with or do you suggest that I as a reading teacher use grade level material to work with my tier II and III students?

Tim Shanahan said...

Stephanie--
This is an important question. The issue of raising text levels is for every one (it is a common core issue). Your question really raises the issue of how much can you raise the levels. If we are teaching grade level kids with harder text than in the past, won't this represent big increases for the lowest readers--such as in intervention programs?

No one has any research that would show us how much of a mismatch is tolerable. If the teacher doesn't know how to scaffold the difference it is probably not much harder than we have usually striven for in our programs (the instructional level is not that bad). However, if the teacher is supporting the students then even the kids in intervention can handle relatively harder texts than in the past.

I always felt that you could pull along kids effectively if they were two grades below level. You might be able to justify even greater stretches IF you are in and advantaged situation--an intervention setting, etc.

tim