The in-service teacher training model in the public schooling system has often been criticized on grounds of both inefficiency and ineffectiveness (NUEPA, World Bank & CLPR, 2015; Subramaniam & Ramadas, 2011). This article presents an alternative developed at the Ravi J. Matthai Centre for Educational Innovation, IIMA, that builds on two principles: a case-based curriculum built around problem-solving practices of teachers to address the ineffectiveness problem, and online delivery methods to address the inefficiency problem. The model was developed over three rounds: a program on school leadership for 2000 principals (2017-18), and a program in science and math for 19,300 middle school teachers and an expanded version that added language and social studies, for 140,000 teachers (2018-20). All three programs were for government school teachers in Gujarat.

Program Design

The specific design principles were derived from the global literature on what makes a teacher development program effective. However, in line with the IIMA spirit, the first aspect of the design was an adaptation of the case method. About 400 cases of good problem-solving practice of government teachers were developed - each had text, videos and photographs, and in the third program, was also in audio format to address the needs of the more than 300 visually challenged participants. Each case dealt with a real-life problem identified by a teacher, the process of developing and implementing a response, and the evaluation of the outcomes. Second, appropriate theoretical content (about two to three pages, prepared by expert teams drawn from the government) and reflective questions that the teachers had to compulsorily answer, were added to the cases. Third, a mobile phone-based discussion forum and a Facebook discussion page provided the backdrop for peer discussion and learning. Webinars and expert lectures (Facebook Live and a dedicated YouTube channel) completed the set-up. Depending on the program registered for, the number of case studies a teacher studied ranged from 20 to 30; these were classified into 'modules', such as teaching science, classroom management, using ICT, and so on. The case study teachers had consented to help their colleagues, and so their contact details were available to the participants - some became very popular!

A second aspect of the design was ongoing assessment; at end of each module, the participant had to answer a set of questions based on actual classroom vignettes, and some questions that tested the application of knowledge. Each vignette presented four answer options, one of which could be considered the 'best' response. As soon as a participant answered, automated feedback on the performance along with an analysis of the four answer options would be available. This aspect of the learning was especially appreciated by the teachers.

The third aspect of the design was the 'project' - a problem-solving exercise carried out by the participants in their own settings towards the end of the training, for about two to three weeks. Each project, on submission, was automatically sent to five other participants who had completed their projects, to be reviewed using a framework drawn from educational literature.

Each of the three programs ran for about 10 weeks. The completion rates ranged from over 95 percent for the case studies and over 90 percent for the projects. A typical participant spent about 17 to 19 hours online over the 10 weeks, and about twice this amount off-line. About 75 percent of the participants completed their programs through mobile phones, and about 23 percent through laptops; the rest used tablets.

Program Evaluation

It was important to demonstrate to the government that assessment had to be an integral component of the training it carried out, and so assessment of not just the knowledge, but also the improvement in self-efficacy beliefs or behaviour of the participants, was built into the program. For example, the leadership development program focused primarily on change-oriented behaviour and secondarily on task-oriented behaviour. Hence, apart from self-reported assessments, subordinates' assessments of principal behaviour were made, both before the programme and six months after the end of the programme. The assessment of self-efficacy beliefs relied on standard internationally accepted instruments. The assessment design also demonstrated the application of a randomized controlled trial with an experimental group and a control. The impact, as assessed through what is technically known as effect sizes, ranged from moderate to high. These assessment practices exposed the government to new ways of looking at training. A second surprise was the peer discussion that happened on the discussion forums-government representatives were also enrolled on these, and had not expected the high levels of sharing that occurred.

Key Implications for Practice

The first implication is that training must counter the dominant pattern described by Subramaniam and Ramadas (2011): "in-service training is reduced to being the mode for mechanically dispersing disparate and often inadequately worked-out reform initiatives into the system" (p.2), and the government has to move away from the so-called cascade model of training, a top-down expert-driven "transmission" model of training that is "neither efficient nor effective" (p.3). Drawing on the problem-solving experiences within the system is likely to more useful and effective for the huge numbers of teachers who face similar problems.

The second lesson is the need to harness the potential of technology while understanding its limitations. An online programme can be used as an initial outreach to cover certain non-negotiables or content that all teachers need to be familiarized with; examples of such content would include changes in the syllabus or textbooks, revised learning outcomes, or the so-called 'hard spots'. Teacher performance in the online intervention can be used to identify those who would benefit from intensive and costlier face-to-face training. This will address the problem of improper or inefficient identification of trainees that has been reported in many studies. Second, the engagement patterns of the participants will provide insights for refinement of training interventions.

An example follows. The performance data of 8399 teachers in the second program was used to generate two scores: (i) an academic score that included the participants' scores on the topic-end questions, case study and module-end responses, and the scores on the projects (given by five peers for each project); (ii) a management outcomes score that comprised scores on case responses and module-end responses from the classroom management, school comprehensive evaluation and use of ICT modules (which were also part of the training). Assuming two levels of performance (high and low), and using an analytical procedure called k-means algorithm, a four-cluster solution was extracted. A cluster of 28% participants (2352 participants) with low scores in both academic and management was identified as requiring further onsite training. About 16% do well on both and so do not constitute a priority group. Another group of 31% of the participants may need specific attention in academic subjects, and about 24% in school/ classroom management. The simple point is that a general, universal online programme can be combined with specific targeted programmes to ensure better targeting of scarce financial and training resources. The point about cost-effectiveness should be obvious. Our experiences shows that online programs, conducted at scale and combined judiciously with targeted onsite programs, can bring down overall resource outlays by about two-thirds.

Another possibility is to examine the levels of engagement and participation as revealed by online logs and off-platform activities to identify motivational issues that could be addressed through additional training. The details are not given here, but using mixture modelling five latent profiles of online engagement and seven latent classes based on off-platform activities were identified. A more detailed analysis showed a number of patterns (for instance, those who spent more time online were significantly older; groups with greater page-viewing time consisted of more female and graduate degree holders). What this analysis makes possible is once again better targeting, but this time for non-content oriented training, and generating ideas for refining future training designs.

In sum, state training establishments will have to develop their capabilities in program design and delivery, and in leveraging technology for change. However, this is easier said that done. Till such time that the state develops its capabilities, it would do well to partner with a range of institutions, academic as well as non-academic, that have the necessary expertise.

References:

NUEPA, World Bank and CLPR (2015): Draft Report of Teachers in the Indian Education System: Synthesis of a Nine-State Study, New Delhi: NUEPA, World Bank and CLPR.

Subramaniam, K. and J. Ramadas (2011): Working Paper on In-service Teacher Professional Development for Elementary Education Presented to Government of India. Mumbai: Homi Bhabha Centre for Science Education.

Note:
The evaluation results reported above formed part of the work of two doctoral students of the author, Samvet Kuril and Ketan S. Deshmukh. For further details, please see:

Kuril, S. (2019). Change in Leadership Behaviour through Online Professional Development Program: Contextualizing "Community" based on Identity, Cohesion, and Intentionality. PhD Dissertation. Ahmedabad: Indian Institute of Management Ahmedabad.

Deshmukh, K. S. (forthcoming). Determining Effects of a Web-Based Teachers' Professional Development Programme On Teaching Self-Efficacy Beliefs and Classroom Practice. PhD Dissertation. Ahmedabad: Indian Institute of Management Ahmedabad.