Using evidence to strengthen tech-supported teacher professional development in Madagascar

Reading Time: 6 minutes

Reposted from EdTech Hub. Original authors: CAITLIN MOSS COFLAN, SAALIM KOOMAR AND HASINIAVO RASOLOHERY. Creative Commons Attribution 4.0

High school students in Soavinandriana, Itasy region, Madagascar, attending an educational movie projection.
Photo credit: Hasiniavo Rasolohery, 2019

Read in French


Madagascar faces high levels of learning poverty; 97 per cent of the country’s children at late primary age are not proficient in reading (World Bank, 2019, based on 2015 PASEC data). We know teachers are one of the most crucial influences on student learning; in 2019, just 15% of primary school Malagasy teachers were qualified. How can these teachers be supported effectively to improve student learning? And how can this support be provided at a distance, at scale, using technology in ways that can support populations with limited connectivity?  Despite reportedly high internet speeds, internet penetration is low, just 2.1% of Malagasy people can access the internet (Quartz article, Lijadu, 2019). 

Since EdTech Hub’s first foray into these issues through a Helpdesk brief on teacher professional development initiatives in low-connectivity settings, we have worked with the Ministry of Education, National Teacher Training Institute, and the World Bank under the first component of the Basic Education Support Project to unpack this set of challenges. 16,000 Grade 1 and 2 teachers will be directly involved in a teacher professional development (TPD) pilot under this project to improve student literacy and numeracy outcomes. 

Evidence-based conversations 

This blog shares insights from conversations we’ve had with colleagues in Madagascar. Such conversations seek to integrate what we know about effective tech-supported teacher professional development with factors specific to the Malagasy context. Our insights are grouped under four topics: procuring hardware, scaling iteratively, building on previous lessons, and instilling a pedagogical approach to the TPD pilot. 

Procuring hardware

Early plans for the teacher professional development pilot included large-scale procurement of 16,000 smartphones and 16,0000 tablets. However, we wondered if instead of procuring these devices, the pilot could use either smartphones or tablets. Or better yet, taking inspiration from a Mike Trucano blog, could the TPD pilot make use of technology that teachers already own and regularly use? We brought these questions to the Ministry and discussed how best to approach decisions about hardware for the pilot. 

Putting our own advice to take stock of what exists into practice, we are now working to develop a  survey to understand Malagasy teachers’ existing access, usage, and needs in relation to technology. While we have household data on ownership of certain technologies (see Figure 1), data on teachers’ access to — and use of — technology is not available. 

Figure 1. Screenshot of presentation to Minister Sahondrarimalala showing household ownership of technologies in Madagascar (translated to English)

The survey will give a better understanding of what hardware is needed and how teachers will use it, helping de-risk hardware procurement decisions and creating potential to increase the cost-effectiveness and scalability of the pilot. Though no final decisions have been made,  procuring at most one hardware device per teacher instead of two would represent a potential cost saving of around £1.6 million, freeing up resources for subsequent training and maintenance. 

Beyond informing procurement decisions that could result in significant cost savings, the survey data will inform other aspects of the pilot. For example, data about teachers’ perceived tech readiness can help determine levels of  digital literacy support teachers need. We are hopeful that this approach will help save money on hardware, bring user needs into focus, and result in an effective pilot design.  

Enumerator taking notes of students’ comment during a Focus Group
Photo credit: Hasiniavo Rasolohery, 2018

Scaling iteratively

The TPD pilot aims to improve the competencies of 16,000 teachers to teach literacy and numeracy across 22 regions of Madagascar. This is ambitious and difficult, but not impossible. To get there, we are having great conversations on how to approach implementation iteratively in order to monitor and evaluate progress, and then learn and adapt. 

While challenging to integrate into existing project timelines, baking small-scale tests into project activities helps  validate the critical assumptions underpinning project delivery.  

An example? Onboarding teachers across fewer but harder-to-reach regions, in the first cohort of the pilot. By starting with the regions which are both the hardest-to-reach and have the worst learning outcomes, we will have a steeper short-run learning curve (another Trucano blog from 2013 discusses this in more depth). Therefore, monitoring the progress of the first cohort will present invaluable insights relevant to larger-scale stages of the pilot, whilst ensuring that the pilot remains contextually-relevant for those hard-to-reach areas. 

Figure 2. Eight targeted pilot regions.

Adapted from:, 2021

Figure 3. Subnational Human Development Index (HDI) of Malagasy regions (darker red areas depicting lower HDI).

Source: Global Data Lab, Smits & Permayer (2020).

Building on previous learning

In designing the teacher professional development pilot alongside partners, we want to make sure that we are building on lessons from previous projects in Madagascar, while integrating new insights from global good practice where appropriate. The lessons learned from several teacher professional development projects and programmes implemented in Madagascar over the past few years offer a critical foundation from which to start. 

For example, this pilot builds off a 2016 project led by the ministry in which it assessed the competencies of 1,000 formal and non-formal educators. We have dedicated a section in the survey questionnaire to identify participants from the 2016 project. Following up with participants offers important insight to the impact of the 2016 programme and shows us the ways in which those activities continue to impact teaching practice (or not). This creates a cycle of feedback loops which should contribute to continuous improvements in current and future projects.    

Figure 3. Screenshot of survey questionnaire focused on historical ministry project. 

Pedagogical approach to the TPD pilot

The characteristics of effective teacher education (e.g., this EdTech Hub brief) provide the foundations for many of our conversations related to the most impactful TPD activities to support teacher and student learning. Building on these characteristics, the pilot incorporates the following features: 

  • Small group teacher communities linked to school locations to promote peer support and learning opportunities. Grouping teachers in small communities gives them space to share notes on their work both in and outside of the classroom and to develop a sense of togetherness. These spaces are where teachers work out how to apply learning from training sessions in practice and reflect on how this is working in the classroom. 
  • A remote teacher community of practice connecting teachers involved in the project to a wider network. This offers similar learning opportunities to the small teacher groups and can also be used to disseminate content such as lesson plans or classroom activities at scale. 
  • Video examples of teaching — both effective and less effective — in Malagasy classrooms as a stimulus for reflective practice. Reflecting on actual teaching in classrooms has been shown to promote improved pedagogical practices while also developing teachers’ skills in giving and receiving feedback when observing these videos. A teacher’s ability to give and receive feedback is crucial for their own professional development and that of their peers. It also informs how they interact with learners. 
  • Structured lesson plans and teacher guides can particularly support novice or less experienced teachers, guiding them in a scaffolded way. The pilot will use structured teacher guides from a previous FHI360 / USAID project (Mahay Mamaky Teny). 

Figure 4. Screenshot of terms of reference depicting multimodal tools that will be used in the pilot to support and track teacher learning. 

Looking ahead

Though this engagement can be traced back to June 2020, it is only in January 2021 that we’ve started to engage in depth on the design of the survey. In future phases of the work, we are excited to delve deeper into the design of activities aimed at improving teacher learning, which will be tracked through lesson observations and other monitoring tools to understand the impact on students. Although we have already learned so much, we are just at the beginning of this exciting collaboration. We can’t wait to share more about what we learn from the survey and the ongoing work in Madagascar. 

Note: this blog was written in March / April 2021, we are very aware of the situation in relation to the Covid-19 pandemic in Madagascar which has worsened in recent weeks. Thus, there is a possibility that certain aspects regarding planned activities that we have covered in this blog will need to change in order to adapt to the rapidly changing context.  

Koomar, S., Moss Coflan, C., & Rasolohery, H. (2021). Using evidence to strengthen tech-supported teacher professional development in Madagascar [Blog post]. EdTech Hub. 

Available under Creative Commons Attribution 4.0 International,

EdTech Hub
EdTech Hub