In education’s recent evidence-informed revolution, few names have been more influential than that of Barak Rosenshine.
But how many people who draw on Rosenshine’s work dive deeper into the publications that his key ideas about teaching and learning stem from?
There are two main publications we need to look at here - the most well-known being Rosenshine’s 2012 article for American Educator, the magazine of the American Federation of Teachers, titled Principles of instruction: research-based strategies that all teachers should know. The other is Roseshine’s 2010 booklet for Unesco, on the same topic.
What are Rosenshine’s principles of instruction?
The publications cover 17 principles, but both report mainly on 10 of them, with little coverage of the other seven. The 17 principles include things like “begin a lesson with a short review of previous learning” and “present new material in small steps with student practice after each step”.
Many of these principles are rather underspecified. For example, one of the 17 principles is “provide systematic feedback and corrections”. But what does this actually mean? Rosenshine doesn’t give us much to go on. “Feedback” can therefore mean everything to everyone (which might contribute to their appeal).
Regardless of their specificity, the principles might make intuitive sense to teachers. But what research are they based on?
Collectively, the 2010 and 2012 publications reference 26 unique sources, some more than once. Half of these are journal articles, three are book chapters, seven are books and three are reports.
Some sources are quite old. Only three were published after 2000, and only one of those is a journal article. The other sources are mainly from the 1970s, 1980s and 1990s, with the oldest being a well-known article from the 1950s, by George Miller, on the limitations of working memory.
Of course, the fact that a piece of research is old doesn’t mean that its findings are necessarily untrue. But it is rather striking that in 2024, we are placing so much emphasis on a 10-year-old article, mainly citing research that is more than 25 years old.
Another issue is that while normal conventions of scientific literature dictate that references are given to support each specific claim being made, Rosenshine gives only two or three references at the end of each principle. This makes it virtually impossible to trace back the exact origins of the claims underpinning the principles, particularly as some of the references are whole books.
The range of authors referenced is also limited. This is surprising, as researchers such as Benjamin Bloom and Jerome Bruner (who aren’t referenced) have written extensively on topics that the principles cover, such as instruction and scaffolding.
Read more:
There is no doubt that some of the studies referenced by Rosenshine are impressive in many ways: they are ambitious in scope and intent, and cover many of his points. The references also openly mention their limitations - such as small sample sizes, the education phases involved, or surprising differences between results for mathematics and English teachers.
These limitations need to be considered, and help to illustrate why the ages of the studies do matter. For example, it used to be the norm to run a test using four groups with 20 students in each group, whereas today, this wouldn’t be considered robust enough.
So, what does all of this mean for Rosenshine’s principles? Does it make them useless or untrue? Of course not. If teachers find them useful, then they are.
However, regardless of how useful a principle might be, it’s important to understand where it comes from. And when it comes to Rosenshine, the academic provenance of his work might not be quite as impressive as its reach.
Christian Bokhove is professor of mathematics education at the University of Southampton and a specialist in research methodologies
