Six Insights on Teacher Training


Image of Yue-Yi Hwa

Yue-Yi Hwa

RISE Directorate

Blavatnik School of Government, University of Oxford

Teaching isn’t easy. In many education systems, a substantial portion of teachers do not adequately master the material they are expected to teach, nor do they master the pedagogical strategies for effectively conveying this material to their students. Teacher training and coaching aims to fill these gaps, so that teachers will be better equipped to cultivate student learning.

But teacher training isn’t easy either. It involves a wide range of moving parts and competing priorities—across the system, school, teacher, and student levels—all of which need to be coherent in order for a training programme to have its desired effects. It’s no wonder that in-service training has a poor track record, given that the typical training approach involves a cascade of one-off sessions away from classroom settings, offering neither hands-on practice and nor constructive feedback.

Insights from a RISE panel

The third session of the RISE Online Presentation Series highlighted the complexity of delivering effective teacher training, but also gave reasons for optimism. All four of the teacher training programmes that our panellists had studied are vast improvements over the typical in-service training model. The panellists were:

  • Janeli Kotze (Department of Basic Education, South Africa; RISE Fellow), who presented experimental evidence comparing virtual coaching and in-person coaching in South Africa
  • Andy de Barros (MIT), who studied a mathematics intervention that included new instructional materials, teacher training, and community-engagement events in India
  • Todd Pugatch (Oregon State University), who evaluated a teacher professional development programme that accompanied a major change in the entrepreneurship curriculum in Rwanda
  • Charlotte Jones (Education Development Trust), who shared early findings on two different models for teacher communities of practice in Rwanda and Kenya

Here are some lessons from the session, which was expertly chaired by Joe DeStefano from RTI International. (Some quotes have been edited lightly for clarity.)

1. If we want to fix the learning crisis, we need to equip teachers.

Janeli: “I was surprised at how difficult it is to change teacher instructional practice. … And I realised that essentially we’re asking the teachers to take something that they’ve been doing every day, for the last 10, 15, 20 years, and to do that differently. As with any other skill that requires coaching, what we’ve learned is that there needs to be a lot of input for a teacher to actually be brave enough to try something in front of a class—and then, once she’s tried it, to try it again if it didn’t work well.”

Todd: “If we’re talking about introducing more active instructional techniques, the vast majority of teachers in developing countries, if not worldwide, are more steeped in traditional techniques—both as teachers and, prior to that, as students. So usually you’re starting at a pretty low baseline, so pedagogical training is likely to be quite important in a wide variety of contexts.”

2. Be very clear about priorities—and about the centrality of learning.

Todd: “The context of this ambitious curriculum reform that happened in Rwanda is that, essentially overnight, teachers were asked to pivot from their traditional knowledge-based instructional approach to one that was focused on active instructional techniques and practical skills. … There is a misalignment in the Rwandan context between the focus on assessment and exams, with this new curricular approach. The exams continued to follow the more traditional knowledge accumulation model that was out of sync with this emerging focus on practical skills. And until those pieces are well-aligned, it makes it particularly challenging for teachers to change their instructional techniques.”

Andy: “The most important implication of our study for the education system is the importance of focusing on instructional quality. … We show that the debate about inputs versus no inputs in educational interventions is really a false dichotomy. It should be about interventions that improve instructional quality and, eventually, student learning versus those that don’t. If those require inputs, then so be it. If there’s a high-input intervention that doesn’t improve instructional quality, then there’s no worth in that intervention.” 

3. Remember that teachers face competing expectations and challenges on many fronts.

Andy: “We have teachers who really like the activity-based programme that was being implemented—they like it, they get it, and they’re curious about it. Now when you look at the classroom observations, and particularly the dimension of instructional quality that gets at activity-based instruction, or peer collaboration in the classroom, or student engagement, you actually see that this is the lowest dimension of instructional quality. And it’s extremely low. … It’s important to show how much agreement there is among teachers about the importance of activity-based instruction, but once you look at what’s happening in the classroom there’s actually a huge discrepancy between these two pieces.”

Janeli: “Teachers are often used to the district officials coming in and monitoring in quite a strict way. ‘Have you done this? Have you done that? Have you implemented the new curriculum?’ But that’s not often followed-up with supportive advice on how they should change. … With the on-site coaching, we’ve found that often the first couple of visits are spent just on building trust. … And it’s only once teachers reach that stage of emotional readiness that they actually start learning and actually start implementing new things.”

4. Evaluators and implementors should learn along the way too.

Charlotte: “One of the communities of practice was for English-language teachers. In the first year, the programme focused on building English-language skills in the teachers. One of the challenges we saw was that teachers were expected to run the community of practice in English, and they really struggled. So there were some workarounds, like allowing code-switching, that we fed back to the programme team in order to reduce the burden on the teachers.”

Andy: “We find improvements both in terms of teaching quality, and in terms of positive effects on math learning. … Those positive effects are fully driven by effects among girls, with close to zero effects among boys. … And this is not trivial. It’s great that the programme was able to more than close the gender gap that we saw at baseline. And it’s great that we were able to find a programme that improves learning for girls. But it’s a big question mark for me to better understand what drives these different effects across boys and girls.”

5. Expect hidden costs and trade-offs.

Janeli: “With technology, there’s often a lot of hidden costs. For example, we needed to appoint someone who could support just the technology side of things. So if the tablet breaks or falls, or the teacher deletes the app, you do need someone to support with that. Secondly, you need to think through aspects of hosting and data storage. And there are all these costs that you don’t actually think about upfront—and you learn as you go along the way. Actually, technology isn’t always much more cost-effective in the end.”

Charlotte: “In the cluster-based communities of practice in Kenya, which span across five schools, the analysis showed that the cluster-level instructional coaches were playing quite a dominant role. … They set the agenda, they set the logistics, and at the same time they were the subject specialists giving teachers professional feedback. So the teachers were a bit more subdued than in in the school-based communities of practice in Rwanda. … It’s not necessarily that one is good and one is bad. … The cluster model is inherently a bit more fragile because it’s a collaboration across five schools, and maybe it needs a more dominant role for the coaches to glue it together. Does this somehow undermine teacher ownership and leadership of their own professional development? Is there a trade-off? … It may not matter for instructional outcomes—and we’ll find this out later in the study—but does it matter for sustainability, if teacher ownership is a little bit stifled?”

6. Look at constraints and enablers elsewhere in the system.

Todd: “Another component in this curricular reform in Rwanda was Skills Labs: 80-minute-long double periods, where students were encouraged to do group work and present to their peers on entrepreneurship topics. We found that even though Skills Labs were meant to be implemented by all schools as part of this new national curriculum, only 8 percent of the control group schools actually scheduled these double periods. So even just the daily schedule was a barrier to changing practices in our context.”

Charlotte: “If you’re a policymaker and you’re investing in CPD [continuous professional development] for teachers, on the basis of this research I would really urge you also invest in the leaders and the facilitators of that CPD, whether that’s district officials, pedagogical coaches, or headteachers. We need to pay more attention to the professional skills that other members of the workforce need in order to enable teachers. Even the most brilliant CPD models can fail at scale if we don’t get hold of the hardwiring of the delivery structure. … Think of it as something that will protect, or unlock, the investment in teachers.”


A recording of the panel session is available on YouTube, along with pre-recorded video presentations from each panellist that give more detail on each paper. Working papers are available on the event webpage (under the “Event Programme” tab).

RISE blog posts and podcasts reflect the views of the authors and do not necessarily represent the views of the organisation or our funders.