Another reason that Sold A Story is relevant to Special Education Today. A lot of kids get placed in special ed who would not be there if they had just had good instruction to begin with. We could be freeing up special ed case loads by providing better regular ed.
I'm currently researching the history of reading instruction going back 200 years, as part of a book I'm writing. That section of the book is a prequel to Sold A Story. Problems with reading instruction and arguments over it have a long history! I'm looking for threads of false beliefs about reading that started in the 1820's, and seeing how some of those false beliefs about literacy still play out in reading, and also just as importantly in writing instruction.
Sally, thanks for the comment and the important observation. As I hope you and other readers know, I like to think that evidence-based reading instruction in general education settings would reduce the proportion of students who have sufficiently severe reading problems that they would be said to have learning disabilities. It's an argument that I find very appealing. It plays to my biases quite strongly!
I would like to have evidence verifying the argument, but I don't know of much. Perhaps you could help me (us) by passing along studies that demonstrate lower incidence (ore prevalence) of LD, for example. Do you (or other readers) have strong evidence multi-tier systems do result in reductions in, say, LD?
In August of 2022, I published a post that presented evidence in agreement with the idea that strong instruction reduces failure rates. I posted a graph comparing the percentage of students who read one or more grade levels behind the norms on standardized tests after 2 or 3 years of instruction (Link: https://www.specialeducationtoday.com/i/135816483/what-are-the-results). The graph is based on data from Project Follow Through, which implemented greater than 10 different approaches to early education (K-2) with tens of 1000s of children, nearly all of whom had many "strikes against them" (poverty, being probably the biggest; almost all were enrolled in what some folks would call "Title I" schools these days ). The models had on average (median 45-51; mean 46.5) what I would consider pretty high failure rates. One model (you guessed it: the Direct Instruction folks) had only 19% failure rates.
I can't argue that this is slam-dunk evidence that good instruction eliminates failure. First off, there were still kids who were more than one year below the ideal of "grade level reading." Second, this study is quasi-experimental (e.g., no random assignment). Third, the models included more than just literacy instruction. Etc.
But I consider the evidence as suggesting support for the idea you advanced, Sally. It does appear likely that there would be have been fewer students identified for, say, LD in the schools that were employing the DI model. Because Follow Through was implemented in the years before US education has "The Law" or just as it was beginning, we don't have good data on how many students were being identified for special education. In addition, even if we could get data about identification rates, we would likely have questions about whether that measure (rates of identification) was reliable and valid—were all the eligibility teams applying criteria consistently?
Still, I'm interested in seeing some trustworthy studies examining the consequence of employing evidence-based teaching. I'd like to be able to argue the position you raised on something more than my faith and intuition.
So it’s not as simple as looking at schools who have implemented high impact reforms (like Steubenville) and comparing special Ed referrals before and after? Or just that you’re not aware of those studies? Yes I admit I’m going on faith and intuition. Thanks for reminding me of the need for evidence, you’re a good scientist!
Right, Sal! It's not a simple as looking at the case of Steubenville, pre-post.
Now, if we diligently searched for every LEA that (a) implemented a whole-school reform model (Success for All, in the case of Steubenville, but perhaps SFA or DI or XYZ--models that all included defined literacy curricula) or adopted an evidence-based literacy curriculum (again, defined, known models) and (b) maintained a consistent and documented referral-eligibility processes...then we could count how many of them experienced reduced referrals for reading problems, and we would have another quasi-experiment examining whether our hypothesis about evidence-based literacy instruction in early grades ==> fewer cases of LD.
Two buts: The study I just described is only one way we could investigate the hypothesis. It'd be fun to describe others. However, I doubt that educators will ever get the definitive studies that might would satisfy crusty old evidence chasers such as moi.
We Crusty-Old-Skeptics would want replicated series of randomized control trials conducted by independent researchers and using pristine methods. Little chance we'll get those studies. So, we really need to aggregate what we do know from less-than-perfect studies.
I'd *like* to be aware of studies of the effects of literacy curricula on incidence-prevalence of (say) LD. If readers know of one or more, let's throw them into a stack and sift them!
Thanks for raising this topic and sticking with it!
Another reason that Sold A Story is relevant to Special Education Today. A lot of kids get placed in special ed who would not be there if they had just had good instruction to begin with. We could be freeing up special ed case loads by providing better regular ed.
I'm currently researching the history of reading instruction going back 200 years, as part of a book I'm writing. That section of the book is a prequel to Sold A Story. Problems with reading instruction and arguments over it have a long history! I'm looking for threads of false beliefs about reading that started in the 1820's, and seeing how some of those false beliefs about literacy still play out in reading, and also just as importantly in writing instruction.
Sally, thanks for the comment and the important observation. As I hope you and other readers know, I like to think that evidence-based reading instruction in general education settings would reduce the proportion of students who have sufficiently severe reading problems that they would be said to have learning disabilities. It's an argument that I find very appealing. It plays to my biases quite strongly!
I would like to have evidence verifying the argument, but I don't know of much. Perhaps you could help me (us) by passing along studies that demonstrate lower incidence (ore prevalence) of LD, for example. Do you (or other readers) have strong evidence multi-tier systems do result in reductions in, say, LD?
In August of 2022, I published a post that presented evidence in agreement with the idea that strong instruction reduces failure rates. I posted a graph comparing the percentage of students who read one or more grade levels behind the norms on standardized tests after 2 or 3 years of instruction (Link: https://www.specialeducationtoday.com/i/135816483/what-are-the-results). The graph is based on data from Project Follow Through, which implemented greater than 10 different approaches to early education (K-2) with tens of 1000s of children, nearly all of whom had many "strikes against them" (poverty, being probably the biggest; almost all were enrolled in what some folks would call "Title I" schools these days ). The models had on average (median 45-51; mean 46.5) what I would consider pretty high failure rates. One model (you guessed it: the Direct Instruction folks) had only 19% failure rates.
I can't argue that this is slam-dunk evidence that good instruction eliminates failure. First off, there were still kids who were more than one year below the ideal of "grade level reading." Second, this study is quasi-experimental (e.g., no random assignment). Third, the models included more than just literacy instruction. Etc.
But I consider the evidence as suggesting support for the idea you advanced, Sally. It does appear likely that there would be have been fewer students identified for, say, LD in the schools that were employing the DI model. Because Follow Through was implemented in the years before US education has "The Law" or just as it was beginning, we don't have good data on how many students were being identified for special education. In addition, even if we could get data about identification rates, we would likely have questions about whether that measure (rates of identification) was reliable and valid—were all the eligibility teams applying criteria consistently?
Still, I'm interested in seeing some trustworthy studies examining the consequence of employing evidence-based teaching. I'd like to be able to argue the position you raised on something more than my faith and intuition.
So it’s not as simple as looking at schools who have implemented high impact reforms (like Steubenville) and comparing special Ed referrals before and after? Or just that you’re not aware of those studies? Yes I admit I’m going on faith and intuition. Thanks for reminding me of the need for evidence, you’re a good scientist!
Right, Sal! It's not a simple as looking at the case of Steubenville, pre-post.
Now, if we diligently searched for every LEA that (a) implemented a whole-school reform model (Success for All, in the case of Steubenville, but perhaps SFA or DI or XYZ--models that all included defined literacy curricula) or adopted an evidence-based literacy curriculum (again, defined, known models) and (b) maintained a consistent and documented referral-eligibility processes...then we could count how many of them experienced reduced referrals for reading problems, and we would have another quasi-experiment examining whether our hypothesis about evidence-based literacy instruction in early grades ==> fewer cases of LD.
Two buts: The study I just described is only one way we could investigate the hypothesis. It'd be fun to describe others. However, I doubt that educators will ever get the definitive studies that might would satisfy crusty old evidence chasers such as moi.
We Crusty-Old-Skeptics would want replicated series of randomized control trials conducted by independent researchers and using pristine methods. Little chance we'll get those studies. So, we really need to aggregate what we do know from less-than-perfect studies.
I'd *like* to be aware of studies of the effects of literacy curricula on incidence-prevalence of (say) LD. If readers know of one or more, let's throw them into a stack and sift them!
Thanks for raising this topic and sticking with it!