The Unfortunate Reputation of Scripted Instruction
by James Hiebert - December 12, 2017
In this commentary, the author argues that scripted instruction, defined appropriately, should be the goal of researchers and teachers if the educational community wishes to improve classroom teaching over the long run.
I am writing to defend the following claim: If educators want to improve classroom instruction, they should collectively work with teachers and researchers toward increasingly scripted instruction. Readers probably find this a surprising claim because scripted instruction has been vilified as an affront to teachers and a necessarily mindless form of teaching that denies students ambitious learning opportunities.
I begin my argument by noting that scripted instruction is a prototypic example of a good idea being destroyed by poor instantiations. The image most Americans have of scripted instruction is robotic teaching, prescribed by someone outside the classroom, and delivered by teachers who have had no say in its development. If this was the only possible instantiation of scripted instruction, I wouldnt be writing this commentary.
I am writing to promote a very different image of scripted instruction. It is not instruction read to students or delivered using the same words by every teacher. But it is instruction sufficiently prescribed in detailed plans that students experience essentially the same lesson, regardless who is the teacher. The image is driven by, among other inspirations, the October, 1995 testimony of Albert Shanker (former President of the American Federation of Teachers) before the House of Representatives Committee on Economic and Educational Opportunities. Shanker portrayed a future time when the best wisdom of the teaching profession becomes its standard practice, when the best lessons for helping students achieve particular learning goals become the lessons that all students receive.
I want to use Shankers image to first address the major indictments of scripted instruction: it de-professionalizes teachers, and it is insensitive to the needs of students. The critique of de-professionalizing teachers is warranted by most U.S. versions of scripted instructiona set of lesson plans handed to teachers by someone outside of their world. But, what if the lesson plans were designed by the teachers themselves? What if teachers were given the responsibility, the time, and the rewards to gradually refine the details of lesson plans that they, and other teachers, could use when teaching toward the same learning goals? What if educational researchers were responsible to assist them by providing the information they needed to do this work? What if this new effort began today with the goal of developing really effective lessons, one lesson at a time, for each grade level within 20 years? This would be the most intellectually challenging task teachers have ever attempted. Creating scripted lesson plans that work for all students treats teachers as genuine professionals charged with improving their craft.
The second critique of scripted instructionnot being sensitive to students needsassumes several things about scripted instruction that dont need to be true. One is that every teacher, in every setting, needs to use the same scripted lesson plan. Why not test the effectiveness of different plans with different classrooms of students? Perhaps teachers will find, for example, that students with different needs achieve the same learning goals more easily using different versions of the lesson. A second assumption is that teaching is inherently uncertainteachers cannot anticipate how their students will respond on any given day. But, what if teachers had the time and incentive to keep careful records of how students responded to each instructional activity, and what if they did this every year? And, what if all the same grade-level teachers in the district did the same thing and then pooled their information? Every group of teachers I know who has tried this quickly see patterns in student responses and begin to predict, with increasing accuracy, how their students will respond to particular instructional activities. Now, teaching is not so uncertain.
To describe what scripted instruction could be, I propose a second claim: Scripted instruction is effective to the degree that educators possess knowledge about how instruction affects learning. How can such knowledge be acquired? What kind of educational R&D system would yield evidence-based effective instruction that is scripted, what would the lesson plans look like that prescribed instruction, and what would such instruction look like in practice? We dont have to guess at the answers; there are countries that already have such systems in place, and there are pockets in the U.S. that are experimenting with such systems.
R&D systems that yield this kind of instruction often follow the principles of continuous improvement, a research-based process proven to yield more effective practices in many professions. The goal is steady improvement over the long run. Cycles of implementing, testing, and refining the products used by practitioners of the profession define the work enabled by the system.
For the teaching profession, the products are lesson plans. The plans usually contain at least the following elements, all specified in great detail: precise statements of the learning goals for the lesson; instructional activities aligned with the learning goals along with the time allocated for each activity; anticipated responses from students; prescribed responses by the teacher; special questions, explanations, or summary statements that teachers use verbatim; rationales for the activities and the verbatim presentations; brief histories of other activities and presentations that have been tried with explanations for why they didnt work. In short, the lesson plans contain the best wisdom of the profession, at the time, for helping students achieve the learning goals. This is exactly the knowledge currently lacking and preventing the creation of effective scripted instruction.
Teaching from scripted lessons is highly prescribed. But, instruction does not look scripted. It is not robotic. Teachers present ambitious instructional activities but they know what to expect. A lesson that is student-centered, focused on solving open-ended challenging problems with follow-up class discussions, can be completely scripted because teachers can predict, not what each student will do, but how the class as a whole will respond. Multiple trials of the same lessons in many classrooms arm teachers with the knowledge they need to plan lessons in detail, even student-centered lessons. Few surprises occur, and when they do, teachers keep track of them and alert other teachers in their lesson plan notes. In addition, data gathered on the effectiveness of lessons enable teachers to select the best version of the lesson for their students.
Better lessons appear as more teachers try the same plans for teaching toward the same learning goals, gather more data, and use the new information to tweak the plans. Teaching from the same lesson plans is critical because improvement is possible when variation can be interpreted as variation in students responses to the same instruction, not variation in instruction. If variation in instruction occurs, detailed records are kept to determine whether the variation should be tested more widely or abandoned. Even though this process yields only incremental improvement from year to year, it is steady and, when gains are made, they last. New teachers can build on what is already known; they do not need to start over.
Clearly, the United States does not have an infrastructure in place to enable teachers to engage in this work, to record what they learn each time they teach a lesson, and to share it with colleagues who face the same problems. A corollary to my opening claim is that, if educators want to be sure teachers are teaching better 20 years from now, then building this infrastructure is what educators should begin doing today.
To not build a system that uses teachers knowledge to create increasingly effective lessons used by all teachers is a huge waste of intellect and effort. Every day millions of teachers enter their classrooms and try to help students learn very similar things and face very similar problems. Capturing what they are learning in ways that can be recorded, shared, tested, and refined is not a new idea. But, it is an idea that has been dismissed for too long.
Without scripted instruction, as I have described it, there is too much variation in the learning opportunities U.S. students receive. In the Feb 15, 2009 edition of the New York Times, Nicholas Kristof said our greatest national shame was the large variation in teaching and learning across our nations schools. If scripted instruction had not acquired its unfortunate reputation, educators would have a path toward ending these inequities and, at the same time, improving the learning opportunities for all students.
Thanks to my colleagues, especially Jim Stigler and Anne Morris, for their conversations and co-authorships that generated the ideas expressed in this commentary. The particular representations of these ideas and the opinions presented, however, are solely mine.