Principles of Instruction
Author: Barak Rosenshine
Image credit: Oliver Caviglioli (@olivercaviglioli)
"The most effective teachers ensured that students efficiently acquired, rehearsed, and connected knowledge. Many went on to hands-on activities, but always after, not before, the basic material was learned."
An Overview
This article represents an overview of ten principles that have been regularly shown to be highly effective in the classroom. Each principle is reviewed from two different perspectives:
Research from cognitive science - this is science based research into how our brains work and acquire information
Research on classroom practices of master teachers - which the author describes as those whose students made the highest achievement gains. This is based on observations of both these teachers and those who are less effective.
Very early on the author states that "Even though these are […] very different bodies of research, there is no conflict at all between the instructional suggestions that come from each of these […] sources." This suggests that these 10 principles have both the backing of the science behind how our brains work and also the vast experiences of the teaching profession.
Below is a very brief overview of each of the ten principles along with some reflections from me. You can also see these summarised in the image above.
1. Begin a lesson with a short review of previous learning
This links to the idea that "review can help us strengthen the connections among the material we have learned". The most effective teachers "ensured that the students had a firm grasp of the skills and concepts that would be needed for the day's lesson".
I don't think this is particularly groundbreaking, and most teachers use a starter to most of their lessons. However in the past I have used puzzles and hooks to start a lesson. Now I am making more use of a series of short questions (or an exam question) to start lessons, which recall previous learning. I have been using this to make the most of the retrieval effect and spacing effect (as discussed in other blog posts). I will try to make sure these questions (or at least some of them) are directly linked to the new material for the current lesson.
2. Present new material in small steps with student practice after each step
This principle is largely linked to our limited working memories (which I have mentioned before) and in particular that "presenting too much material at once may confuse students because their working memory will be unable to process it". This involves teachers regularly checking student progress (assessment for learning under a different name) and addressing any issues before moving on to the next chunk of knowledge.
This may lead to a more teacher-led period of instruction, but "the most effective […] teachers spent about 23 minutes of a 40 minute period in lecture, demonstration, questioning, and working examples". This example was taken from a mathematics classroom, but other studies showed similar in other subjects.
I have been making a lot of use of breaking down instruction into smaller chunks lately, especially with my S3 class, and it has been working well. When we do get to the more complex tasks, this seems to help. It has also helped to implement ideas from principle 1, as I can review the previous chunks each day.
It is important to note that despite the larger proportion of teacher-led instruction, this is not completely devoted to a "lecture", but rather incorporates some of the other principles given below.
3. Ask a large number of questions and check responses of all students
"Students need to practice new material". The use of questions by the teacher is one way to ensure all students are getting this practice, as well as "allow a teacher to determine how well the material has been learned". It was also found that the "most effective teachers also ask students to explain the process they used" getting students to elaborate and hence improve their own learning, whilst also providing scaffolding for other students (see below).
I have drastically increased the number of questions I ask the class in recent months. In the past I would have set them off to work and then answered individual questions, but I have found, which is backed up by this principle, that by using more questions (and examples) in the instruction students are more able to work independently when the time comes. However, I know one area I need to work on is involving ALL students in this process. I try to ask random students (using a randomiser at times), but still find this not as effective as I would like. I have dabbled with mini-whiteboard several times in the past, and come across many issues (some administrative such as needing enough board pens, other behavioural such as drawing silly things on the boards). However, I am going to work harder at giving them another go.
4. Provide models
There are many ways for teachers to model responses for students. The important aspect is that they "allow students to focus on the specific steps […] and thus reduce the cognitive load on their working memory". Some of these approaches include: thinking aloud while demonstrating; worked examples; providing prompts for students. An example from writing an essay is "an effective teacher first modeled how to write each paragraph, then the students and teachers worked together on two or more new essays".
Doing worked examples is a large part of teaching Maths, and something that I already do a lot. I like the idea of "thinking aloud" and might try to build that into my example procedure. Recently I have started doing worked example pairs, which involves me doing an example, followed by the students doing a very similar example. I can see the thinking aloud process being very useful in this process (along with the mini-whiteboards I mentioned in the previous principle).
It is interesting that the author gives some other ideas to provide models, such as prompts, or working through writing a paragraph together, or even working through a reading comprehension together and also thinking aloud. I think many teachers think of worked examples as something to be done in Maths and nowhere else, but these ideas seem to be transferable in some ways.
5. Guide student practice
This links to the previous point, and once the models have been given it is important that students are not left to fend for themselves. To help students "store this material in their long-term memory" it is important for students to "process the new material and receive feedback, so they do not inadvertently store partial information or a misconception in long-term memory". Guided practice was found to be a common aspect of most classes, but "the most successful teachers spent more time in guided practice, more time asking questions, more time checking for understanding, more time correcting errors, and more time having students work out problems with teacher guidance".
This principle also fits in with the example pairs I have been doing with S3, since the problem they do is very much under my guidance. I have also been making use of what I have been calling "Test your understanding" which consists of up to 4 questions on the chunk of learning before moving on to more independent practice. During this stage of the lesson I am actively circling the room and checking student work carefully, giving feedback and addressing any errors.
6. Check for student understanding
The most effective teachers regularly check the understanding of their students, within each small chunk of learning (principle 2). This can take many forms such as "asking students to summarize the presentation up to that point or to repeat directions or procedures, or by asking students whether they agreed or disagreed with other students' answers" as well as to "ask students to think aloud" and "to explain or defend their position". In opposition to this, "the less effective teachers simply asked, 'Are there any questions?'".
I do ask questions, both in whole class mode and of individual students, but I like some of these other ways of checking understanding. Building on the idea of me 'thinking aloud' whilst doing examples, I am going to try getting students to do the same when answering questions on the board. I do not get students up to answer at the board often enough, and this is something I could build into the worked example pairs.
7. Obtain a high success rate
This principle states that students should experience success in the classroom most of the time. The figures stated are that for the most successful teachers, students were correct 82% of the time and that a "success rate of 80 percent shows that students are learning the material, and it also shows that the students are challenged". This is both a motivational tool and a tool for improving learning in its own right from a cognitive point of view. In a simplistic way of thinking about it, "practice, we are told, makes perfect, but practice can be a disaster if students are practicing errors". In other words, if the success rate is not high, students actually risk embedding errors in their long term memory and "once errors have been learned, they are very difficult to overcome".
Another aspect of this principle links to the previous one. When we talk about success rate, to which aspect of the lesson are we referring? Well, a "high success rate during guided practice also leads to a higher success rate when students are working on problems on their own", and so we should try to ensure students can be successful throughout the instructional period, so that they can be more successful later.
This is something I have been considering a lot lately. Both from the point of view of engaging students in the class (you enjoy what you can do) and in improving their learning. I believe there is a cycle of success->interest->engagement/hard work->success, and that our job as teachers is to get students into this cycle at some point. Previously I have tried to use interest as the entry point to this cycle, but now I think that is the wrong approach, and am now trying to increase success in the classroom as the entry point. By breaking the topics into smaller chunks (principle 2) this has become a natural aspect, as student can easily access each individual chunk.
8. Provide scaffolds for difficult tasks
Scaffolds are a "temporary support that is used to assist a learner" which importantly are "gradually withdrawn as learners become more competent". This can include the use of worked examples, providing prompts , checklists and partially completed tasks for students to complete. The idea of thinking aloud is again suggested, which helps "novice learners with a way to observe 'expert thinking'". It is also made clear that effective teachers are able to "anticipate students' errors and warn them about possible errors some of them are likely to make" as well as using this to provide extra scaffolding where necessary.
The idea of scaffolding is something I have come across many times before, but I have always thought of it as providing partly worked solutions for those that need it, with different amounts of structure. The idea that examples or prompts also work as scaffolding makes sense as these provide students with a model to follow for the simple problems, and which they can utilise in more complex problems. However, this is definitely an area I need to think about in more depth.
9. Require and monitor independent practice
This principle relies on the underlying research that says "students need extensive, successful, independent practice in order for skills and knowledge to become automatic". For this reason students should undertake a significant period of independent practice after the teacher-led instruction phase of the lesson. This is "necessary because a good deal of practice (overlearning) is needed in order to become fluent and automatic in a skill. When material is overlearned, it can be recalled automatically and doesn't take up any space in working memory". So by embedding the skill through overlearning the skill will become much easier to transfer to other situations and problems.
From the research of master teachers, it is also clear that this independent practice "should involve the same material as the guided practice", and that teachers should be moving around the room and checking on students. However, these contacts should be minimised to "30 seconds or less", and if more time than this is required at this stage, then the teacher should realise the "guided practice was not sufficient for students to engage productively in independent practice". Independent practice can incorporate elements of cooperative learning, where students can help each other in their study. In fact these situations tend to support higher learning, presumably because students are "having to explain the material to someone else and/or having someone else explain the material to the student".
Most Maths classrooms will incorporate an amount of independent practice as students need to consolidate the skills and knowledge they have learnt in the lesson. This research has highlighted the importance of the questions being accessible (and hence allowing for high success rates as in principle 7). I do not tend to enforce a silent classroom during these phases of the lesson, preferring to allow students to ask each other questions when they are stuck, however this talk does need to be limited to talking about the current learning (Maths talk only).
10. Engage students in weekly and monthly review
This principle is about ensuring new knowledge is stored in long term memory so it can be easily accessed at a later date, and the "more one rehearses and reviews information, the stronger these interconnections become". It also frees space in working memory so students can used this extra space for "reflecting on new information and for problem solving". It has been found that "classes that had weekly quizzes scored better on final exams than did classes with only one or two quizzes during the term". As well as providing students with extra practice, these make extensive use of the Retrieval Effect and Spacing Effect that I have mentioned in previous blog posts.
I have started using weekly quizzes with my S3 class to make use of the retrieval and spacing effects, and recently I have moved from topic review quizzes to mixed topic review quizzes. I think these two effects are a huge part of cognitive science that all teachers should be aware of, and ones that I am trying to plan into my teaching more and more.
There is an excellent reflection on how one teacher has applied these principles to his teaching which can be found here: https://heathfieldteachshare.wordpress.com/2018/04/23/putting-theory-into-practice/
My Reflections
These 10 principles are all backed by research both in cognitive science and looking directly at the most effective teachers and distilling what they do to be so effective. Since starting this journey looking through the research I have already started to incorporate many of these aspects into my teaching, particularly with my S3 class. I have found that they have been proving successful, though I still have some things to improve within this classroom style. In particular I need to focus on asking questions of ALL students, keep ensuring that students can experience as much success as possible in class and to make greater use of the weekly and daily reviews.
I think that these 10 principles are a great starting point for any teacher. It is not necessary to build them all into every lesson, as with anything in teaching, but they can form a part of the daily life of a teacher. The research suggests this will benefit the learning of our students. And is this not what we should be striving to do every single day?