Teachers plan lessons every day, considering their curriculum, the topics they will address, and what has worked well with their students in the past. Teachers can choose from different instructional approaches, such as direct instruction, open discussion, and others, but which approach is the most effective? Does the best instructional approach depend on the subject?

An NBER working paper by Simon Burgess, Shenila Rawal, and Eric Taylor seeks to link teacher's choices of instructional practices with student achievement by observing classrooms in public secondary schools.

The experiment was straightforward: observers in 2,700 secondary classrooms in England were given a list of activities, such as "open discussion among students and teacher" or "use of white board by teacher," and noted how often teachers used them during a 15-20 minute observation. These specific activities were grouped under categories of instructional approaches (e.g., "open discussions among students and teacher" was grouped under "student-peer interaction"). Each teacher was typically observed eight times over two years and scored by three different observers, who were all also teachers participating in the study. They compared this data against General Certificate of Secondary Education (GCSE) test scores and longer-run measures like future earnings for evidence of student outcomes.

Overall, the observations showed teachers' choice of instructional activities are largely unrelated to the subject being taught or the characteristics of students being taught. In math, teachers used more practice and assessment, and in English they used more activities related to student-peer interactions, but both types of approaches were frequently used in both subjects.

While the choice of instructional approaches appeared somewhat haphazard, the instructional activity chosen by teachers does lead to better outcomes in different subjects. For instance, in math, more time spent on practice and assessment was associated with better achievement outcomes, while in English, more student-peer interactions produced stronger student outcomes. These outcomes were "educationally and economically meaningful," as they contributed significantly to a student's GCSE scores and predicted future earnings and college going.

The study also found instructional activities chosen by a teacher could predict student achievement, independent of their teaching skills, as measured on a Danielson observation rubric. Therefore, even less-skilled teachers (as measured by observations) may help their students achieve greater outcomes if they spend their time on the right activities.

The study also found that factoring in student characteristics (e.g., prior academic performance or exposure to poverty), the relationship between teachers' instructional activities and students' outcomes held, suggesting that the choice of instructional activity matters regardless of student characteristics.

While this study was completed in a secondary context, the findings suggest homing in on specific pedagogical choices can have a lasting impact on student outcomes. However, we need to be cautious when applying the study's findings to an elementary context, where explicit, systematic instruction in foundational skills may be critical. This methodology of linking instructional approaches to student outcomes is an exciting one; researchers and curriculum developers should consider how it may be applicable to specific subjects, grades, and curricula to determine the best approaches to teaching various aspects of a subject or course.