A Tale of Two Teachers (Part 2)

Sundari is a 5th grade teacher in a primary school in East Java, Indonesia. Like Ismail, whose professional development experience was discussed last month, Sundari also attended a computer-based “training”—but not in a training center, rather in her school.

June 18, 2014 by Mary Burns, Education Development Center
13 minutes read
Ibu Sundari (L) and her former coach, Ibu Lia (R)

This is a continuation of last month’s post on the cascade approach.

Sundari is a 5th grade teacher in a primary school in East Java, Indonesia. Like Ismail, whose professional development (PD) experience was discussed last month, Sundari also attended a computer-based “training”—but not in a training center, rather in her school.

In this workshop, the use of computers was far more modest than Ismail experienced—Sundari learned to use one computer to model for students how to analyze a short story. The instructor told the teachers that she was their “coach.” Sundari was not really sure what this meant, but over the next several months, her “coach,” Ibu Lia, began coming to Sundari’s school three times a week. “I’m here to help you use the computer and adapt this lesson to your classroom.” 

The coach helped Sundari plan a lesson in which her students used the school’s one laptop to analyze a short story. Together Sundari and her coach co-taught the lesson so Sundari could practice using technology to support learner-centered instruction. The coach then helped Sundari prepare to do the lesson alone. She videotaped the lesson and together with Sundari analyzed the instruction and discussed areas for improvement.  The coach is really helpful…

For many who read last month’s post, the outcome of Ismail’s PD experience was a familiar one. But we may be less familiar with how Sundari’s tale ends. That’s because in most teacher professional development programs in developing-country contexts, the kind of support Sundari received is virtually non-existent. Her story is so rare as to be considered fiction. 

But it is real.  Sundari’s coach left the school after six months, but because of the support and skills Sundari had received,  she is still using ICTs to promote learner-centered instruction in math, social studies and science – even four years later.  

Sundari was part of a coaching program (1) developed by the Education Development Center (EDC).  It was designed, in part, to counteract the poor implementation rates associated with the cascade approach. Launched in 2010 with 300 teachers and 25 coaches, this program helped teachers implement four one-computer learner-centered activities in their classrooms.

An ace in the role

Without a highly skilled coach, coaching can hurt, not help, teacher performance and student learning. Therefore, we carefully recruited and developed our own coaches. While these recruits had been trainers, teachers and/or supervised teachers, they had never actually coached teachers, so we provided them with three weeks of face-to-face instruction on coaching, a five-month intensive online coaching course, and weekly supervision, mentoring and support. Week-by-week, the coaches learned a particular skill online, and with a coaching partner, spent 3-4 days in a school coaching teachers. Formal certification as a coach was based on successful completion of all course requirements, a coaching portfolio, and formal evaluations from their mentor, principal and teachers.

It’s not the training…

The bottom line of any professional development is transfer of learning “from the training room to the classroom”. Ninety-eight percent of the teachers in the coaching program implemented at least one one-computer activity. How did this compare to teachers in our cascade approach? We never actually kept track (like most organizations), so I once asked a group of Indonesian principals if the implementation rate of our cascade approach might approach 20%. The principals laughed... “You are very generous,” said one.

…It’s the follow-up (3)

I don’t highlight this coaching program because it was an EDC project. Rather, I do so because it is further evidence that coaching, when done well, works.  Teachers who receive quality coaching “practice new skills and strategies more frequently and apply them more appropriately than teachers who participate in workshops with no coaching” (Showers & Joyce, 1996:14)—on average an estimated differential of 80% versus 10% (Ibid). Students whose teachers receive intensive coaching show consistent gains in learning outcomes (Allen, Pianta, Gregory, Mikami & Lun, 2011). In my own work as a school-based coach in several US school districts, I saw first-hand how struggling teachers flourished with the support of a caring and competent coach—and how their students flourished too.

Why coaching?

We know quite a bit about workshops. We know that teachers who participate most frequently in workshops or trainings often have more experience and expertise than those who don’t participate (OECD, 2008). Conversely, those who participate least in workshops are “at-risk” teachers—new teachers, unqualified teachers, teachers teaching out of their content area, and teachers teaching in fragile contexts—the very teachers professional development is supposed to help.

Workshops are effective for experienced teachers because such teachers can more easily assimilate new knowledge into their existing repertoire of knowledge. Workshops are less effective for at-risk teachers who don’t have this existing repertoire or who have gaps in their skill set.  Metaphorically, if the cascade approach is Procrustean, as I wrote last month, truncating teacher learning, coaching is Promethean. Like the mythological Greek figure who gave fire to humans, coaching, when done well, instills in teachers a greater sense of efficacy and agency to do their jobs well.

Why is this so? Research on change (Rogers, 1995) categorizes those going through any change as falling into five “types” (See Figure 3).  The smallest percentage—“innovators” (about 2.5% of any population) will immediately implement an innovation with no prodding. Early adopters (13.5%) are also eager to implement new ideas but not as quickly, effortlessly or independently as innovators. Further along the continuum, individuals become more resistant to change. Most individuals—about 68% (a normal distribution) are in the middle. About 16% are classified as “laggards” (Roger’s term) or “resistors.”

Figure 3: Diffusions of innovation: Change types (Rogers, 1995)

A train-the-trainers’ model or workshop with no follow-up may easily capture innovators and perhaps even early adopters. But for the remaining 84% of teachers who won’t or can’t implement an innovation independently, coaching provides varying degrees and types of support, practice and handholding.  Because it is a 1:1 intervention,  as individual teachers learn new behaviors, change their practice, and experience different “stages of concern” (Hord, Rutherford, Luling-Austin & Hall, 2006), coaching can provide differentiated, personalized support (Burns, 2011) that is so critical to teacher change:

(With the cascade approach) I did not really understand what was going on…I was thinking how will I implement these ideas? But with a coach, I could talk about my plan, revise it based on his feedback and guidance, and try it out before the class with help. Because, you know, you have to make sure that everything is well prepared before the class begins. (Indonesian teacher)

Mission over margin

Fortunately, awareness about the power of instructional coaching is beginning to receive greater attention in the education development world. Coaching and more long-term and practice-based forms of professional development certainly must be part of a new pedagogy for teachers. More important, they must reflect a new psychology about how teachers learn (to be discussed in the next post). Coaching is intensive and ongoing and developmental—it is not a text message or a phone call or a couple of classroom visits over a term. We must take care not to dilute coaching as we have so many other types of professional development. Otherwise we will further denigrate the cause of providing quality professional learning to the teachers who need it most.

Ismail and Sundari are not fictitious characters (I worked with Ismail in India in 2010 in a school-based support PD program I designed.  I have communicated with Sundari and worked very closely with her coach, Ibu Lia, from 2008-2011 in Indonesia.). Like teachers everywhere, Ismail and Sundari are well aware of which professional development models work best for them and those that don’t. The last word goes to them:

Sundari: Ordinary training is just training.  That’s it. There’s no follow up.  I will never know what I lack or need to develop further for improvements.  With coaching, I receive feedback and refinements.

Ismail: Please, Madame, do (follow-up support) for all teachers in India… this (support) is the only way our teaching will progress.


  1. I designed the program and all its content and managed the work of a very talented and hard-working group of Indonesian colleagues, mentors and coaches (like Ibu Lia) who made it all happen.
  2. Such classifications are not fixed in stone. Individuals can fall into different categories, depending on the innovation. For example, a teacher may be an early adopter of IRI, but a resistor when it comes to using computers.
  3.  The mantra of my former manager at SEDL in Austin, Vicki Dimock—a strong advocate of supporting teachers.


Allen, J., Pianta, R., Gregory, A., Mikami, A. & Lun, J. (2011). An interaction-based approach to enhancing secondary school instruction and student achievement. Science, 333, 1034-1037

Burns, M. (2011). Distance education for teacher training: Modes, models and methods. Retrieved from http://go.edc.org/07xd

Hord, S., Rutherford, W. L., Huling-Austin, L.  & Hall, G. E. (2006). Taking charge of change. Austin, TX: Southwest Educational Development Laboratory.

OECD. (2008). TALIS 2008 technical report: Teaching and learning international survey. Paris, France: Author.

Rogers, E. (1995). Diffusions of innovations (4th ed.). New York, NY: Free Press.

Showers, B. & Joyce, B. (1996, March). The evolution of peer coaching. Educational Leadership, 53 (6):12-16

Post a comment or
Quality teaching

Latest blogs


Totally agree that cascade approach without continuous engagement and follow-up with teachers doesn't work. We at TeleTaleem have experienced it first hand. Our Learnign Boost / School Garee program implemented in remote rural areas of Khyper Pakhtunkhwa in Pakistan (http://vimeo.com/88227370) follows a just-in-time teacher training and mentoring philosophy with bi-monthly training sessions, supported by in-class formative assessments, classroom teacher observation and community engagement. The program has shown phenomenal results in children's learning outcomes in early grade numeracy and primary grade mathematics, confirming that structured mentoring and continuous support to teachers is much more effective that one-off train the trainer model.

In reply to by Khurram Sultan

Hi Khurram--

I didn't recognize the project at first, but someone from your organization contacted me a few weeks ago to review and give some advice about your work. You mentioned "phenomenal results." That is a pretty powerful statement. Could you share more what those results are and how you measured and ascertained them? I am sure this would probably be useful to many readers.



Hi Mary,
Great post! So many good points in here - especially that coaching is not a couple of visits over a term. Thought you might be interested in the work of the Burma Education Partnership (www.burmaeducationpartnership.org) as an example of a teacher training program based on coaching and mentoring. They do excellent work here on the Thailand-Myanmar border coaching and mentoring migrant and refugee teachers. Worth checking out!
Best wishes,

Thank you for an interesting and informative post. It is great to see that an approach that we are using on a much smaller scale is working so well on a larger one. Village Education Project Kilimanjaro has been following up early years maths seminars by working with the teachers in their schools. Lack of man-power (people with the skill and expertise to coach) means that it is impossible to follow up all of them as intensively as we would like to and they need, but working with a small group of neighbouring schools has been effective, until, of course, the teachers leave. Even then, having worked with the first three years of the primary education system together so that teachers know how to teach year groups other than their own, has meant that new teachers do receive some coaching from their colleagues.

Hi all,

Thanks so much for your posts and for sending on more examples of coaching and long-term and sustained models of professional development which I will be sure to check out. Folks are always asking me about worthwhile programs--especially in fragile contexts--it would be great to get more detail on your programs. Fee free to email me at my EDC address: mburns@edc.org or share more details here. Sounds like momentum is building.



Another great post, Mary. The Aga Khan Schools have had great success with the sort of coaching you are describing, most notably in Pakistan where there is a proliferation of schools in remote, poor areas of the country. This experience happened within the context of a funded project, and, as you suggest, the impacts on teaching and learning in the schools have endured many years past the project's end with enduring, visibly strong instruction. The challenge, which you might be planning already to address in "part 3," is that such coaching is very costly and, therefore, often unsustainable and un-scalable. In many countries - we are doing this, for example, in Tajikistan and Kenya - a less costly alternative option is to identify key resource teachers from among a faculty. In others (too few), the head teacher plays this role. These folks still often rely on cascade training. As indicated in past comments to your posts, I have been working increasingly on the use of school and cluster communities of reflective practice to help organize and motivate teachers to address key issues they face in an action-research mode. I find that this allows them to take what they have learned in a cascade modality and to "churn" it to adapt and adopt it to positive and lasting pedagogic effect. In this way, cascade delivery can serve a useful purpose, igniting a process of change rather than pretending to be the end of the process. I look forward to "part 3."

Hi again Joshua,

Sorry for the delayed response. I used to get notifications when someone commented. No more—now when I have time I check back manually.

Part 3 actually does NOT address what you wisely suggested. Sorry about that.

A few points re: your comments.

1. So…okay, coaching is said to be expensive…but is it really and if so does the differential warrant adherence to a capacity-building method that is a lovely construct but a failure in terms of implementation? I have yet to see a good study that examines coaching versus training with implementation as an outcome. Further, if we have 5-10% implementation rates with the cascade (versus 80-90% with coaching), can we really say coaching is too expensive? Lots of US districts have moved away from trainings to coaching because they see the latter as more cost effective (Of course this brings up the whole question of system wide capacity).

2. That said, though I love working with teachers, I wonder we spend so much money on sending people to work with teachers. In fact, efforts should be focused on building the capacity of subject supervisors, district officials, block group leaders, pengawas, university educators etc. so they can work with their own teachers. I think if we did more of that, we’d be farther along in our quest to scale, since, part of the difficulty with scaling initiatives is that there is often not local capacity to continue efforts once the project ends.

3. I love the idea of teacher communities and clusters but there has to be a certain amount of expertise among teachers, or at least really good facilitators, otherwise many times they rehash bad practice. I know I’m preaching to the choir on this, but I also see a rush toward “communities of learning”—which often times means teachers are brought together and left to figure it out for themselves. Like any good innovation, this can become diluted and distorted if not done well. If there is one thing I’ve learned working in international education, if there is a shortcut to be taken, we will take it!

4. Finally, though I did not address, technology play a pivotal role in providing coaching. In Indonesia, we set up teachers with Bluetooth headpieces (females wore it under their hijabs) and used Skype for offsite live coaching (that has promise but was troublesome because of bandwidth issues. We also used two-way Skype to have coaches co-teach difficult lessons with teachers. That worked better, though teachers still preferred in-person coaching (and we have data showing that that leads to better teacher outcomes). We are certainly seeing a proliferation of mobile coaching (via voice and SMS)--though I have concerns that this may propagate a reductionist view of coaching--that a few SMS and voice calls suffice (Frankly, I already see this happening). I think human coaching is the ideal, especially if working with at-risk teachers who may have so many issues and need such rigorous support, but tech-based coaching can certainly be a viable and quality coaching alternative.

You may be the only person who reads what I write--so thank you (and I may have to start sending checks your way! :)) Again, it sounds like the work you do at Aga Khan is highly impressive and of high quality. I certainly encountered Aga Khan teachers (while I was working with teachers) in Tajikistan and in Pakistan and I was really, really impressed by them. Congratulations to you and to your colleagues and all the best to you.

Leave a comment

Your email address will not be published. All fields are required.

The content of this field is kept private and will not be shown publicly.

Plain text

  • Global and entity tokens are replaced with their values. Browse available tokens.
  • No HTML tags allowed.
  • Lines and paragraphs break automatically.
  • Web page addresses and email addresses turn into links automatically.