Classroom Practice: Evidence-Based Strategies for Every Teacher

Rosenshine (2012) synthesised research from cognitive science, classroom observation studies, and cognitive support studies into ten principles. The most central are: begin lessons with a short review of prior learning; present new material in small steps with practice after each step; ask a large number of questions and check all learner responses; provide models and worked examples; guide learner practice; check for learner understanding; obtain a high success rate; provide scaffolds for difficult tasks; require independent practice; and engage learners in weekly and monthly review. Together they describe teaching that manages cognitive load, builds strong schemas, and provides frequent feedback.

Formative assessment is any activity during teaching that generates information the teacher uses to adjust instruction. It is not grading: it is diagnostic. A hinge question that reveals half the class misunderstood a concept tells you to reteach before moving on, rather than discovering the gap at the end of term. Wiliam (2011) identifies five key strategies: clarifying learning intentions and sharing success criteria, engineering effective classroom discussions, providing evidence of learning, activating learners as instructional resources for one another, and activating learners as owners of their own learning. The EEF rates feedback and formative assessment at +8 months of progress, the highest single intervention in their Toolkit.

These terms are often used interchangeably but have distinct meanings. Direct Instruction (capitalised) refers to a specific, scripted, highly structured programme developed by Engelmann from the 1960s onwards, with a strong track record in early reading and mathematics. Explicit teaching (lower case) refers to the broader practice of clearly stating what learners are expected to learn, modelling thinking, providing guided practice, and checking understanding before releasing learners to work independently. Both contrast with minimally guided discovery approaches, and both are supported by strong evidence. In practice, most teachers use explicit teaching rather than the formalised Direct Instruction programme.

Three changes have the greatest impact. First, use cold calling rather than hands up: selecting any learner (not just volunteers) ensures you get information about the whole class, not just the confident front row. Second, use wait time: research by Rowe (1986) showed that increasing pause time after a question from one second to three seconds significantly improves the quality of responses. Third, use hinge questions at key conceptual junctures: design a question whose answer tells you whether the class is ready to move on. If 30% give a wrong answer, reteach before continuing. Avoid "any questions?" as your primary comprehension check: silence does not mean understanding.

Hattie and Timperley (2007) found that the most effective feedback addresses three questions: Where am I going? (the goal), How am I going? (progress), and Where to next? (the next step). Feedback is most effective when it is specific, timely, and focused on the task rather than the person. Feedback that says "good work" has near-zero effect. Feedback that says "you have correctly identified the main cause but your explanation does not include supporting evidence from the text: add one quote and explain its relevance" has high effect. Importantly, feedback only works when learners act on it: tasks that require a response to feedback, such as DIRT (Dedicated Improvement and Reflection Time), are more effective than feedback that learners simply read and move on from.

Classroom climate refers to the social, emotional, and intellectual environment of a classroom. Learners who feel psychologically safe are more willing to attempt challenging tasks, ask questions when confused, and share incorrect answers for class discussion. Cornelius-White (2007) found a 0.52 effect size for the teacher-learner relationship on achievement. A poor climate, characterised by ridicule of mistakes, competitive comparison between learners, or unpredictable responses to behaviour, increases cognitive load and reduces the emotional safety needed for risk-taking. Practical ways to build climate include: normalising mistakes as learning data, using consistent routines to reduce unpredictability, and giving learners deliberate opportunities to succeed in front of their peers.

Differentiation means adapting teaching to meet the varying needs of learners in a class. The evidence on its effectiveness is mixed and depends heavily on what is meant by it. Producing multiple versions of every worksheet is not only unsustainable but is not well supported by evidence. Adaptive teaching, which is more commonly recommended by the EEF, means using assessment information to adjust your responses during a lesson: reteaching when needed, offering more support to those who are stuck while extending those who are ready. Rosenshine's emphasis on achieving a high success rate for all learners before moving on is essentially a form of adaptive practice. Ability grouping and setting have weak evidence overall and can be harmful to lower-attaining learners.