Instructional coaching cycle review

What the research suggests—and what it does not

Kraft, Blazar and Hogan’s 2018 meta-analysis combined 60 studies using causal research designs. It reported pooled effects of 0.49 standard deviations on teachers’ instruction and 0.18 standard deviations on student achievement. Much of the evidence concerned literacy coaching for pre-kindergarten and elementary teachers in the United States. The authors also found that average effects in larger effectiveness trials were only a fraction of those in smaller efficacy trials. These pooled findings describe programmes in the evidence base; they are not a forecast for a particular school, subject, age group or international setting. Kraft, Blazar, & Hogan (2018), The Effect of Teacher Coaching on Instruction and Achievement: A Meta-analysis of the Causal Evidence.

Set up the review

Choose a cycle with a defined start and end. Agree with participating teachers what information will be collected, who can see it and how it will be used. Where possible, use a measure already collected for a clear purpose rather than creating a new paperwork burden. Do not use coaching notes as covert appraisal evidence. Explain confidentiality and its limits before the cycle starts.

Coaching-cycle worksheet

1. Define the instructional focus

• What learner need or classroom practice is the coaching intended to address?
• What starting evidence supports that choice? What is uncertain or missing?
• What should a teacher be able to do differently after the cycle?
• Which feature is essential to the coaching approach, and what can be adapted to subject, phase or context?

2. Describe the actual coaching offer

Record what happened, including missed sessions and adaptations. A label such as “coaching” is not enough to show that two schools delivered the same approach.

3. Review teacher practice

Select one observable practice measure that fits the instructional focus: for example, a shared observation rubric, a sample of lesson materials or a review of pupil work. Agree the criteria before looking at results. Note who collected the evidence, how many lessons or samples were included and whether the measure captures the practice you care about. Where feasible, use a second source such as teacher reflection; do not treat satisfaction as proof of changed practice.

4. Review learner outcomes and experience

Choose a near-term learner outcome plausibly connected to the instructional change. Record the starting point, collection method and date. If comparing groups or periods, state what differs besides coaching. Check participation and results across relevant learner groups where the data and privacy rules permit. A small short cycle may not be long enough—or designed strongly enough—to establish a causal impact on achievement.

5. Check burden and trust

• Was time protected for coaching and classroom follow-up?
• Did the process feel developmental and safe enough for honest reflection?
• Did coaching add workload, displace another priority or create unequal access?
• What should the coach and participating teachers change before the next cycle?

6. Make a bounded decision

Choose: continue, adapt and test, pause or stop. Record one reason, the next action, an owner, the resources needed and a review date. If attendance was high but coaching was not delivered consistently, address the implementation barrier before judging the model. If teacher practice changed but learner evidence is mixed, inspect measurement quality, time horizon and context. If no useful signal emerges, do not extend the programme merely because it is familiar.

Fictional example

A fictional international-school department pilots coaching on checking for understanding. Most teachers attend the agreed sessions, but observation notes show uneven use of the shared questioning routine, and the six-week learner-work sample is too small to interpret. The team records no achievement effect. It reviews the coaching model with participants, simplifies one routine and agrees a later sample with a broader set of classes.