A TPD intervention in Ceará state of Brazil had four components:

1. Performance feedback on teacher practice (from classroom observations undertaken at the end of the previous school year)
2. Self-help materials
3. Face-to-face interaction with high-skill coaches
4. Expert coaching support provided by Skype

The treatment group of 156 schools in the randomized control trial had a 0.05 to 0.09 standard deviation higher performance on the state test and a 0.04 to 0.06 standard deviation higher performance on the national test. The expert coaching support provided through Skype kept the costs of the program at $2.40 per student and produced cost-effective impacts on learning compared to other rigorously evaluated TPD interventions that have cost data (Bruns et al., 2017).

A randomized control trial which looked at different delivery models of structured learning programs in a pilot in South Africa, found that on-site coaching is more cost-effective (0.41 standard deviation increase in test scores per US$100) than centralized training workshops (0.23 standard deviation increase in test scores per US$100) and short coaching interventions (no significant impact).

Given the challenge to scale on-site coaching, a variation of the program was tested using virtual coaching. The results after a year showed that this variation had the same effectiveness as on-site coaching in improving teacher instruction practice and the literacy outcomes of children. The cost of the virtual coaching was US$41 per learner, whereas the on-site model cost US$48 per learner (Kotze et al., 2019).

In 2015, the Foundation for Information Technology Education and Development (FIT-ED) developed and piloted the Early Language, Literacy and Numeracy (ELLN) Digital for K-3 Teachers in the Philippines as an alternative to the Department of Education’s (DepEd’s) traditional cascade model (10-day face-to-face workshop).

ELLN Digital uses a blended approach combining self-learning with classroom practice, co-learning with peers in a school-based professional learning community and offline, interactive, multimedia modules. Plan-Do-Study-Act cycles help to improve the design, impact, and sustainability of the program through communities of practice known as Learning Action Cells (LACs). The pilot, which aimed to develop a more cost-effective and sustainable approach to provide at scale in-service teacher training, included 240 primary schools and over 4,000 teachers.

An evaluation of the pilot, which focused on literacy teaching, found that there were statistically significant improvements in the pedagogical and content knowledge of participating teachers, particularly those in rural schools. Whilst the evaluation did not review cost-effectiveness, in 2019, the program started to scale nationally with a plan to reach over 250,000 teachers in three years due to the success of the pilot and its outcomes compared to the DepEd’s traditional cascade model (Oakley et al., 2018).