Contextual Factors that Foster or Inhibit Para‐Teacher Professional Development: The Case of an Indian, Non‐Governmental Organization

The appointment of para‐professionals to overcome skill shortages and/or make efficient use of expensive resources is well established in both developing and developed countries. The present research concerns para‐teachers in India. The literature on para‐teachers is dominated by training for special needs settings, largely in developed societies. Little has been published about para‐teachers working in developing countries with children without disabilities, despite this being a common occurrence. The present research investigates how contextual factors influence the design and implementation of professional support for para‐teachers. The research participants were 12 para‐teachers and five management and administrative staff. Contextual factors included characteristics of para‐teachers, classrooms and students from under‐resourced settings; and the practices and policies found in a non‐school educational setting, here a non‐governmental organization. The study indicates that each factor has potentially enhancing and hindering effects which need to be taken into consideration when designing and implementing professional development work.


Introduction
Over the last few decades, educational non-governmental organizations (NGOs) in India and other developing countries have taken on a range of roles, starting with service delivery in underserved areas, designing innovative, locally relevant curriculum and teacher education systems, providing replicable models to strengthen mainstream education and participating in policy reform (Archer, 1994;Blum, 2009;Rose, 2009). NGOs usually employ para-professional staff or volunteers; they are recruited from the community on a small honorarium, work in or close to their residential locality and are acknowledged for their local insights and rapport with communities (Desai, 2003). But their educational qualifications are usually limited, up to grade 10 or 12 in India (pre-university), and they have no formal training in education. In order to sustain the growth and credibility of their educational efforts, NGOs must pay close attention to the professional development of their staff (cf. Jagannathan, 2000).
Unlike the training of para-teachers involved in special needs settings, little has been published on development of para-teachers who work with children without disabilities. This is even more so for those in developing countries, in challenging circumstances and in non-school settings like NGOs. NGOs, as well as formal educational organizations, can benefit by a rich understanding of the factors that surround parateachers' work and their professional learning and whether or not they have a fostering or hindering influence. The authors of this paper seek to meet this need by reporting a context and needs assessment study in the largest educational NGO in India, which endeavors to establish self-sustaining community learning centers in underserved areas. A brief context of education in India and the NGO is provided before the methods, findings and recommendations of the study are put forward.

Context of the study: education in India
The quality of primary education in India is a matter of national concern among government and non-government actors. Backed by a constitutional commitment to basic education, about 182 million children between the ages of 6 and 14 years attend government schools (Kant et al., 2007). However, about 40 percent of children drop out by grade 5 and 60 percent by grade 8. About 47 percent of children going to grade 5 cannot read a text from grade 2 and about 55 percent of children cannot solve a three-by-one digit division problem (Kingdon, 2007). The government's response has been a flagship initiative called Sarva Shiksha Abhiyan, which aims to ensure relevant elementary education for all children in the 6-14 age group by 2010. A National Curriculum Framework recommends guidelines for a move towards a system-wide learner-centered approach. NGOs, with experience in implementing large-scale educational services, are viewed as important catalysts and partners in such educational improvement. One such NGO is Maitri.

Maitri (pseudonym): an NGO involved in educational improvement
Maitri has a commitment towards 'every child in school and learning well'. It reaches out to about 300,000 children in 21 Indian states through about 10,000 teachers. It provides educational support to slum children by establishing learning centers that usually run in one or two rooms, rented within the slum community. Teachers in these centers are women who usually live nearby, have bachelor's degrees, are interested in teaching and have no prior teaching experience. They are trained in instructional strategies at Maitri. Children who come to these centers are between 3 and 14 years of age and may or may not be enrolled in a school. The centers provide support in pre-school learning, literacy and basic math through learner-centered teaching strategies that include use of innovative material, joint work, activities and puzzles. Maitri's centers offer large-scale, low-cost, accelerated and child-centered (compared with the rote-based methods in government schools) ways for teaching literacy and basic math which are often replicated in government schools. However, the extent to which these activities are learner-centered is constrained by the under-resourced environment and teacher expertise.
This study took place within the community centers of Ahmedabad, a city in Western India where 40 centers have been established. The 40 centers are evenly distributed over three clusters, each run by a senior teacher called the cluster head. Each of the three cluster heads reports to a project leader who coordinates the community-centered activities. Another management member is her project assistant. For 8 years, the centers have worked with public school children up to age 14 whose attainment levels in reading and basic maths are very low. Until 2005, no fee was charged for the centers' services. In 2005, with the objective of supporting these clusters towards becoming self-sustaining learner-centered community learning institutions, three transitions were introduced: • Each cluster was legally registered as an individual NGO; thus each cluster head became an NGO head and the teachers in the center, members of the respective NGOs. • Parents were required to pay a monthly fee of about 50 Indian rupees (about US$1.00) per child to ensure financial sustainability, accountability and community ownership of the centers. (The monthly average income in the community was about 4,000 Indian rupees, which is about US$80.00). • The instructional scope of the centers' activities was extended beyond teaching literacy and basic math, to additional subjects including Gujarati (first language), English, Math, Science, History, Geography and Civics up to age 14. • Adoption of learner-centered material was mandated for teaching advanced subject matter. A central curriculum design team created materials for all centers.
Teachers were now responsible for ensuring revenue for their centers through individual home visits for collecting fees, teaching new subject matter for children up to age 14 and adopting learner-centered strategies in their instruction. In 2007, the management decided on capacity-building interventions to support the adoption of learner-centered materials. A context and needs assessment study was conducted to help determine priorities. It was shaped by the following research questions.
What contextual factors must be considered when designing learning support for these teachers?
How might contextual factors enhance or inhibit the development of learning interventions for the teachers?
Literature that framed the study to answer this question is presented below.

Rationale: teacher learning in context
Professional development of para-teachers working in underserviced communities and NGO environments is under-researched. Para-teachers differ from professional teachers in terms of training and expertise. However, recent insights from teacher-and adult-learning literature can be valuable for shaping para-teacher learning, because at the heart of both, is the task of teaching children. This section first documents salient features characterizing teacher learning which can be relevant towards shaping learning for para-teachers in this NGO. Given that contextual factors influence what and how teachers learn, this section also identifies contextual factors that are likely to support or constrain the way teachers learn. Finally, within the rationale that teacher learning is situated in and influenced by these contextual factors, it brings these factors together into a conceptual model that underpins this study.
The idea of involving teachers actively in their own learning is central in professional development literature. It implies, as a first step, that professional support practices must cease to view change as something that is done to teachers who are passive participants of programs (Imants & van Veen, in press). The literature on teacher development and workplace learning emphasizes that lasting changes in classroom practice are gained when teacher learning: • is supported in a sustained manner (Sparks, 2002;Supovitz & Turner, 2000); • is connected to and derived from actual practice through participation in authentic tasks and problems arising in daily work (Bredeson, 2000;Darling-Hammond & McLaughlin, 1996); Factors that foster or inhibit para-teacher professional development 25 • engages teachers actively in planning and reflection to develop their own classroom practice as well as their professional support mechanisms (Ellström, 2001;Elmore & Burney, 1999); and • is grounded in interaction and sharing of knowledge with peers (Darling-Hammond & McLaughlin, 1996;Hord, 1997).
Teacher learning characterized as above is achieved by strategies and practices that invite, structure, support and guide participation (Billet, 2006), and combine effectively in and out of classroom tasks for learning (Putnam & Borko, 2000). This implies connecting learning to tasks like teaching, lesson planning, lesson modification and assessment activities (Davis & Krajcik, 2005). Such tasks may be supported through peer coaching, workshops, mentoring, performance tasks, study groups and action research (Bredeson, 2000;Darling-Hammond, 2006;Loucks-Horsley et al., 1998). Such teacher learning and strategies flourish when organizational (school) structures and culture already support or are altered to actively support staff learning and collaboration around serious problems of practice (Darling-Hammond & McLaughlin, 1996;Silins & Mulfort, 2002).
Learning through continuous, authentic, active and collaborative engagement does not take place in a vacuum, but within a particular context. Context, in simple terms, refers to the teacher, the student and the surroundings in which these function (Fordham, 1982).

The teacher
The personal characteristics, professional competencies and learning needs of teachers influence the ways they will learn. Teacher conceptions about teaching and learning are often rooted in rote-based teacher-centered notions; usually on account of traditional schooling or social experiences (Darling-Hammond, 2006). Such misconceptions can pose a challenge to embracing learner centeredness.
Existing professional competencies also influence the extent to which teachers can optimize new ideas and opportunities. These include knowledge about learners, subject matter and teaching; ability to integrate these (Darling-Hammond, 2006) and to skillfully use the knowledge in real time for making instructional decisions in classrooms (Darling-Hammond, 2006;Davis & Krajcik, 2005). Further, resources that foster their learning such as planning, reflection and collaboration abilities matter (Ball & Cohen, 1999;Darling-Hammond, 2006). Teacher beliefs also influence how they receive and learn new ideas. For instance, teachers are more open to learning when they believe that the new tasks expected of them are practical and relevant (Doyle & Ponder, 1978;Fullan, 1991;Rogan & Grayson, 2003). Their learning is also enhanced when backed by the need to make a difference; a moral purpose that brings the teachers closer to the needs of children (Fullan, 1993).
As teachers are the direct clients, professional support interventions must also take into consideration their specific learning needs. Scribner (1999) found that commonly teachers needed to engage in learning activities in order to address (1) content knowledge needs, (2) pedagogical skill gaps, (3) challenges to classroom management, and (4) gaps in student-centered strategies.

Instructional setting
The instructional context refers to content, teaching and learning methods, assessment, and classroom environment (Fordham, 1982;Loucks-Horsley et al., 1998), all of which influence the way in which teachers are able to learn to adopt new practices. For instance, the content that teachers are required to teach may or may not be aligned to learning goals, or to teacher's capacities.
Teaching and learning methods pertain to the learning strategies, the material, teachers' role, the student grouping arrangements, requirement to implement the strategies and the time available. Assessment practices include criteria for assessment, strategies and tools, and the actual learning progress of the students (Loucks-Horsley, 1998;van den Akker, 2003).
Additionally, classroom size, the availability of material, the number of children, socioeconomic and cultural aspects of the students, and norms of behaviors (Fordham, 1982;Loucks-Horsley et al., 1998) contribute to the complexity of the instructional task; the complexity increases as these factors differ from tasks with which teachers are usually confronted. When the classroom environment is under-resourced, the extent to which teachers can develop alternative practices is constrained (Brodie et al., 2002).

Organizational setting
The teacher and the instructional setting exist within an organization that determines the learning environment of teachers through its resources, leadership and its climate. Resources include time, money, space, structures, expertise and strategies needed to foster learning (Fullan, 1991;Loucks-Horsley et al., 1998). Learning happens effectively when time is released for teachers to familiarize themselves with curricula; when structures and roles encourage supporting each others' learning, continuous assistance and feedback (Darling-Hammond & McLaughlin, 1996;Howe & Stubbs, 2003).
Teachers need a leader who can foster a shared vision among the members, elicit their active participation in decisions and share authority with the members (Fullan & Pomfret, 1977;Hord, 1997). Additionally, a trusting and collaborative climate (Hord, 1997;Penuel et al., 2007) that values reflection, questioning, risk, examination of goals and practice in light of each other makes a favorable environment for teacher learning (Darling-Hammond & McLaughlin, 1996;Silins & Mulfort, 2002). These professional learning characteristics are related to and derived from theories of adult learning (e.g. Knowles et al., 1998).

Policy
The policies of an organization represent the organizational purpose and its means to achieve the purpose. Policy pertains to not just what organizations proclaim through formal documents but also what they do (Knapp, 1997). Knapp discovered that policy reaches teacher practice through its targets (embodied conceptions about teaching and learning), instruments (strategies) and avenues (routes by which influence is exerted). The content of policies gets interpreted and enacted in the organizational and the classroom settings. Desimone (2002) illustrates five attributes that influence implementation of policy. These are specificity, consistency, authority, power and stability. Specificity, or explicitness (Fullan & Pomfret, 1977), pertains to the detail and extensiveness of the policy in the form of curriculum frameworks and guidelines on use of materials and pacing aspects; and strategies to educate staff (Darling-Hammond & McLaughlin, 1999). Attention to appropriate levels of specificity is important to mitigate user confusion (Fullan, 1991;Fullan & Pomfret, 1977). Consistency or the way policies cohere or contradict with each other and with circumstances of the teachers (Darling-Hammond & McLaughlin, 1999) also affects implementation. Policies that win influence through their own authority and credibility are more effective than the ones that use power of incentives and disincentives. Moreover, organizations that are in disarray cannot achieve success with their efforts (Rogan & Grayson, 2003). Stability of policies, students, teachers and leaders also influence implementation.

Theoretical framework
On the basis of the factors presented above, we propose a theoretical framework for investigating contextual influences upon teacher learning. The framework suggests Factors that foster or inhibit para-teacher professional development 27 that the teacher, the instructional setting, the organizational setting and policy share a complex web of relationships, which individually and collectively influence teacher professional learning. This nested systemic arrangement is visualized in Figure 1. Teacher characteristics are the core of the arrangement of contextual factors. Moving outwards, the instructional setting, the organizational setting and the policy factors are represented. The concentric rings are off-center to illustrate that all factors touch one another.
The aim of this study, which is to understand the contextual factors that surround para-teachers' professional learning needs, drawing on data from an NGO in India, rests in the understanding that teacher learning is situated within an influential systemic structure discussed above. It is expected that future professional development efforts can build on factors that prove to be enhancing for professional learning. Similarly, when contextual factors may prove inhibiting for the professional development of para-teachers, appropriate contextual conditions must be met prior to or in the course of the professional development interventions.

Procedures
As noted, two questions shaped this study: What contextual factors must be considered when designing learning support for para-teachers? And, how are the contextual factors likely to enhance or inhibit the learning of para-teachers? This section describes the methods used to collect data for both the questions.

Participants
The Ahmedabad learning center project leader, the project assistant, all three cluster heads and para-teachers were participants of the study. Twelve para-teachers from  across the four clusters were purposively selected for interviews. As teacher turnover was common, para-teachers who had been with the organization for at least 2 years were selected for the interview. The project leader of the organization was a qualified teacher educator and had been head of the project for 3 years. The project assistant was a social work graduate with relatively fewer educational experience.

Instruments
The first question was answered through interviews and observations. Interviews were used to understand para-teachers' personal characteristics and professional competencies, such as relevant qualifications and experience in teaching, professional purpose, confidence, their perceptions regarding their work and their learning requirements. Semi-structured non-participant classroom observation was used to understand the para-teachers' instructional skill and the different aspects of the instructional setting. The observation tool contained instructional components like the learning objective, activity, content, materials, grouping, time, para-teachers' role and assessment (cf. van den Akker, 2003). Data from participant observation of meetings between cluster heads and para-teachers supported by interviews with cluster heads shed light on organizational factors like the resources, leadership and climate. Interviews with cluster heads, project leader and project assistant were also used to construct the policy intentions driving the project. Figure 2 illustrates the methods used to collect data on the contextual factors related to the first question. The second question aimed at characterizing the contextual factors into fostering or inhibiting influences on the learning of para-teachers. This characterization was developed and verified as respondents met to reflect upon and analyse the information gathered for the first question. The findings of the first question were presented to the Factors that foster or inhibit para-teacher professional development 29 participants and they were asked to share their views on the nature of influence. The researcher participated in meetings to facilitate the reflection. Such a member-check approach was suitable to (1) facilitate a shared understanding about why and how the different factors affected teacher learning; and (2) create a common ground for the way forward.

Findings
This section presents the results of the study based on each factor mentioned in the theoretical framework.

The para-teachers
Characteristics Data related to teacher characteristics covered their personal characteristics, professional competencies and their learning requirements. Para-teachers were experienced in working with children; however, they had limited experience with learner-centered instruction and no formal training in instruction. Para-teachers expressed deep commitment and interest in their work. They shared their aspirations for their learning centers to progress towards providing 'good quality' educational support that would be 'different from the traditional school'. They valued their role as para-teachers at such centers.
In one-to-one interactions, para-teachers discussed their lack of confidence about subject matter knowledge as well as 'presenting the subject matter in a manner that was easy to understand for the children'. Classroom observations revealed that when inadequate knowledge and lack of preparation made subject matter difficult, coping strategies included (1) diverting the child's attention to easier concepts; (2) suggesting that the child should seek clarifications from school teachers; and (3) returning later with clarifications after obtaining help from peers. Para-teachers were not in the habit of planning their instruction or reviewing their practice.
When asked to share their learning needs, para-teachers enthusiastically listed over 20 learning needs that spanned issues of teaching, data analysis, financial planning and rapport building with the community. Of those, four were perceived as urgent. These were: (1) subject matter knowledge of maths and environmental science; (2) curriculum implementation of maths and science; (3) maintaining order and discipline in the classroom; and (4) managing heterogeneous student needs. Para-teachers openly expressed the view that the new role of mobilizing fees for the center through door-to-door marketing activities was 'fatiguing and very time consuming'. They stated they were 'too worried about reaching enough fee amounts and could not focus on teaching'.

Nature of influence on para-teacher learning
Cluster heads as well as management members acknowledged that the para-teachers' commitment, despite the difficult working conditions, and enthusiasm for new learning were commendable qualities and could play a fundamental role in enhancing the effectiveness of professional development. It was also acknowledged that parateachers' rapport with children and their familiarity with the sociocultural environment of the children were important aspects of prior understanding that must be capitalized upon during the learning interventions. The factors that could inhibit learning were identified as (1) the lack of professional habits such as regular planning and review and (2) a weak attention to their teaching role. These factors could interfere with their readiness to focus on instructional concerns on a sustained basis.

Characteristics
Data were collected to understand the content, teaching-learning methods, assessment and classroom environment. Observations revealed that an overall curricular frame-work to define 'what, when and how' to teach was not available to the para-teachers. The content for teaching was selected separately for every child each day and was not linked. Subject matter concepts were usually selected arbitrarily based on their own comfort level with concepts or an imminent test or exam in the child's schools.
Para-teachers' classroom environment and the teaching-learning strategies in the classroom were also observed. The para-teachers' mandate was to provide supplementary academic support to children in their grade-specific subject matter. However, typically, a child in grade 5 or 6 (10-12 years) was barely (or not at all) able to read a text from grade 2 (7-8 years). The learning requirements of these children ranged from literacy and basic maths to the syllabus of grade 7. Children from the same age needed support in different subjects and at varying levels. Para-teachers adopted different instructional strategies to cope. They ranged from asking a child to memorize a section in the book and then checking how well the child reproduced it, oral explanations of difficult concepts with the help of examples, solving a specific math problem in front of a child and verbalizing the 'working of a math problem' as a way of solving a difficulty. Curricular material included textbooks, practice books and teaching aids supplied by Maitri. The teaching aids included some guidelines, tests and material for teaching different concepts in Environmental Sciences and Mathematics for grades 2, 3 and 4. In interviews, para-teachers expressed their opinions in relation to the material provided by Maitri, saying that: (1) the materials were interesting and more 'practical'; but difficult to implement due to low familiarity; (2) the instructions were unclear; (3) the material often served to teach certain concepts only partially, and they had no additional material to address the remaining parts of the concept; (4) the assessment tools that had practical questions were not easily usable as children were used to a traditional question-answer test; (5) some tests and teaching activities were weakly linked; and (6) pre-tests and post-tests were not consistent in difficulty levels. There were no records of assessments.
Observations indicated that classrooms were a small, poorly ventilated space in the community at the residence of a community member. The classroom had between 30 and 40 children from 3 to 14 years. These children usually attended formal government schools and came to the center for supplementary support. Para-teachers explained that the children were not enrolled for a fixed or uniform time period. About 50 percent of these children came on a monthly basis, that is, they may attend class for 1 month and not for another. Concurrently, new children were added at the end of each month to make up for the fall in the fees. The classrooms often became disruptive, with only a few children engaged in work. Para-teachers were found to be shouting out repeatedly to restore order in the class. However, on the whole, para-teachers were affectionate towards children.

Nature of influence on para-teacher learning
The instructional setting of the para-teachers posed several challenges towards developing instructional practice. Some conditions, such as the classroom space, were fixed conditions which limited the extent to which para-teachers could explore classroom management and instructional alternatives. Other constraints were modifiable, but still inhibited para-teacher practice substantially. For instance, the manner in which children were enrolled created a very heterogeneous and unstable student population, making it very difficult for para-teachers to give attention to all of them and to plan for all their learning needs. Also, material was inadequate and curricular decisions of 'what should be taught, to whom, how, and when' were left entirely to para-teachers. This reduced the instructional process to a set of precipitate decisions made by the teacher with few meaningful experiences that could be drawn upon to strengthen learning.

Characteristics
Within the organizational context, the focus was on understanding the resources, leadership and organizational climate. The daily activities of the para-teachers outside Factors that foster or inhibit para-teacher professional development 31 class hours and professional support activities gave insight into the types of learning resources para-teachers enjoyed. Meeting observations and interviews revealed that teacher schedules entailed two to three classes a day of 2 h each, followed by about 2 h of house-to-house visits for collect pending fees and weekly administrative work. Cluster heads explained that every 6 months the para-teachers went on a 'marketing' drive, to enroll children and collect fees during which classes were closed for a month. Different types of support focused on reinforcing their revenue generation capabilities. These included activities such as periodic budgeting, financial reviews and planning of marketing strategies. Organized support for addressing teaching concerns was missing. There were no time slots available in which para-teachers could prepare for their daily lessons. They spent late night hours familiarizing themselves with the subject matter. New teaching aids were the only support para-teachers received for teaching activities. But this material usually reached the para-teachers just before the term started, leaving little preparation time.
Cluster heads shared that their tasks mainly involved supporting and managing their teams to get better revenue. Their role in the improvement of the teaching task was limited. They seldom visited classrooms, and when they did, it was just to check if the class was functioning.
All the three cluster-in-charges shared that they felt pressured by the need to achieve financial targets. The para-teachers trusted their cluster heads and the cluster heads supported their para-teachers and empathized with their situation.

Nature of influence on para-teacher learning
The organizational climate had mixed implications for teacher learning. On one hand, it was fraught with difficult working conditions and anxiety over revenue matters. On the other hand, para-teachers shared a very cohesive relationship with one another and the cluster heads. It was agreed that the current work routines left para-teachers no time for instructional work outside class hours. Related to this, the burden of daily non-instructional responsibilities was identified as a critical deterrent of teacher learning. It was discussed that para-teachers and cluster heads viewed revenue generation, thus not instructional work, as their primary role, and this would impede the development of instructional learning.

Characteristics
The policy of the organization was assessed in terms of its specificity, consistency, authority, power and stability in relation to adoption of learner-centered materials by para-teachers. Formal documents that could be treated as policy statements were not available. However, in discussing organizational intentions, the project head shared that overall three broad policies had been introduced: (1) a move towards a revenuebased model instead of free educational services; (2) inclusion of higher level subject matter up to grade 7; and (3) adoption of learner-centered materials to teach subject matter. Cluster heads shared that there were specific strategies and guidelines to improve revenue and they governed decisions regarding enrollment of children, professional support interventions and incentives. For instance, a detailed directive for enrolling children emphasized that the 'amount necessary to be generated for covering the operating cost of the centre divided by the amount of fees collected per child equals number of children to be enrolled in the center'. To optimize fees, the children were to be enrolled for either a period of 6 months if the parents were willing to pay for the entire period, or the children would be enrolled for 1 month at a time. Moreover, if a child discontinued after the end of a month, a new child had to be enrolled to compensate for the drop in the revenue. The revenue-based policy was adopted and promoted using incentives for those who achieved higher revenue. Specific curricular policies or guidelines to inform teaching activities were missing.
Such mandates were not consistent with the remaining policies. For instance, in order to obtain fees, para-teachers had to 'commit to teach whatever the parents expected them to teach their children'. Parents agreed to pay fees with the expectation that their child's attainment in his or her specific grade level would improve substantially at the learning center within a single term. Para-teachers had to heed this condition irrespective of the child's current attainment levels in order to secure the terms fees. Para-teachers were expected to teach in such heterogeneous classrooms without almost any ongoing support. Similarly, the head of the project was only accountable for improving revenue and was not accountable for learning improvement. The project assistant was accountable for instructional development; however, the final authority of approving or disapproving her recommendations lay with the head of the project.

Nature of influence on para-teacher learning
Organizational mandates and their implications for teacher learning were critically discussed by cluster heads as well as management members. Cluster heads appreciated that the organization's commitment to support autonomously functioning learning centers was an important backbone for improving instructional practice. However, it was argued that currently, the transition towards learner-centered instruction was being neglected in favor of the transition towards financial autonomy. The key criticisms were that student enrollment strategy and professional support were aimed at optimizing revenue; the roles and incentives were also determined to develop revenue generation practices. The overall policy environment reflected a greater ownership for revenue mobilization, which trickled down to the other parts of the organization. It was finally argued that the current approach of treating revenue mobilization and instructional development as mutually exclusive led to a conflict of interest in which instructional learning of the para-teachers was sacrificed. For instance, the policy of promising to teach children what the parents expected, or the policy of enrolling students based on fee criteria, undermined the confidence of the teachers, led to a curriculum load that was not achievable in terms of time and capacity of teachers, and created a classroom with a complexity in student needs that para-teachers could not address meaningfully. In this way, teaching-learning concerns were compromised. And, the need for replacing this with an approach that would balance the two was expressed.

Discussion
This study brings evidence for both endogenous factors (para-teacher competencies) as well as exogenous factors (classroom and organizational) which can determine the context of para-teachers' work and can inform, enhance or inhibit their professional development. Examples were para-teachers' inability to design and implement coherent instruction, unstable student populations, dominance of revenue generation activities and lack of available time and structures for practicing instruction. Based on the case of Maitri, this study finds that there are unique contextual opportunities as well as limitations that have implications on para-teacher professional development. In this study, these factors were para-teacher qualities, their under-resourced settings and the characteristics of an NGO, which is a non-educational setting. The points below elaborate upon some of these.

Non-professionals and professionals
Para-teachers, such as the participants in this study, are often unable to perform basic tasks such as planning for daily lessons, keeping a timetable or assessing children's progress, which would normally be taken for granted in the case of professionally trained teachers. Moreover, unlike schools, in an NGO, the immediate colleagues and supervisors may also be non-professionals. Hence, opportunities for learning through peer modeling are limited or absent.

Hierarchical relationships
Depending upon many factors including cultural perspectives, people view learning and decision making as either passive or active, and as either being owned by the Factors that foster or inhibit para-teacher professional development 33 learners or being owned by an expert (Desai, 2003). The para-teachers in the study belong to a cultural background in which challenging or debating decisions taken by authority is uncommon and is often even frowned upon. This might explain the fact that those in charge of clusters and para-teachers continued to work in self-defeating circumstances, such as those evidenced in the study, for nearly 2 years, without questioning or resisting ongoing strategies. Participating as active learners can be often difficult for such people, especially at the onset.

Multiple accountabilities
An uncompromised commitment to instructional improvement is fundamental to effective professional development but data from this study reveal factors that hinder the establishment of such a commitment in this NGO. Although it is an educational NGO, the policies and practices of Maitri reflect a weak instructional focus to begin with, let alone an environment characterized by commitment to instructional improvement. Such a contradiction is not uncommon with NGOs, for they often seem to vacillate between three sources of accountability, their donors, their own vision and their clients or communities (Najam, 1996). NGOs often interpret their own visions and rationalize their activities under the influence of donor expectations and constraints, in part due to their dependency on the donors. For instance, the urgency of the financial autonomy of the clusters on account of an imminent termination of donor funds took priority over the need to make the clusters instructionally effective. Because parents are willing to pay the fees as long as their expectations are met, the prioritization of focusing effort on financial autonomy met little resistance, and in fact provided a pretext of meeting community expectations. The study indicates that if NGOs allow such inconsistent accountabilities to prevail, it may affect the way para-teachers experience and implement their responsibilities. In this study, this was evidenced in the form of feelings of pressure, overload of non-teaching tasks, and ineffective teaching and learning.

Inadequacy of resources for learner-centered teaching
The literature recommends that teachers be exposed to new theory, research and practices about teaching and learning. However, learner-centered teaching is not yet commonly found in India, and while some schools may be modeled on such teaching, published research on this approach in India is scarce, especially in local contexts and languages.
On the basis of the contextual factors found here, some implications for professional development practice are now considered.

Implications for practice
Instructional and subject matter knowledge and skills Para-teachers could be supported through induction programs and/or through in-service support in basic instructional knowledge and skills that are important to qualify for teaching roles and bring some order to their instructional practice. Subject matter support and learner-centered teaching practices were important needs expressed by the respondents in this study, but other specific gaps that also emerged included lack of lesson planning, assessment, classifying and grouping children for instruction, keeping a timetable and defining learning objectives.

The learning tasks of para-teachers
It is important to define the instructional tasks of para-teachers realistically and authentically. In this study, para-teachers were ineffective partly because their instructional tasks involved teaching a wide range of children across a wide range of 'perceived' rather than actual curricular needs.
Thinking tools Para-teachers can gain from thinking tools, in the form of templates, for different activities like planning and reviewing instruction or grouping children. Such templates must explicate and help them 'think through' the different dimensions and steps involved in these activities, through transparent, simple, detailed prompts. At the same time they should be flexible enough to invite new ideas and serve as brainstorming tools. In developing contexts, these tools would be in the form of handouts with ample writing space and the facility to store them for later reference. Such tools, if designed appropriately, can contribute to para-teachers taking charge of their own learning and becoming self-directed. Similarly, para-teachers can also be supported in using such tools within groups for discussion and brainstorming, to help structure their efforts in learning from one another.

Workshops
Besides the use of tools, para-teachers must be supported through workshops in addressing difficult concepts, process skills or 'ways of thinking'. While demonstrating and modeling teaching practices, it could be useful to integrate the lesson planning and review tools so as to expose them to the cycle of planning, enactment and review. Group sessions during such workshops must also be used as opportunities to provide practice and feedback on 'group work' abilities with which para-teachers from such contexts are not usually familiar.

Establishing responsibilities and routines
It could be useful to establish simple responsibilities and routines, for instance, daily lesson planning and review, in a way that such activities constitute an important part of their own role expectations as well as the organization's expectation of their roles. It is important for para-teachers to see these activities as much as a part of their 'job' as teaching itself. Supporting them in maintaining their own 'work plan' that helps balance time for both teaching and preparation for teaching is essential. This makes them actively involved in the design and implementation of curricular products and gives them a strong feeling of ownership. Active involvement and ownership as well are effective characteristics of adult learning and professional development.

Need for focus on para-teachers' immediate tasks
The content of the interventions must focus on the immediate tasks that para-teachers must perform. Because para-teachers are untrained in their task, learning tends to be more effective when they have the opportunity to implement it soon afterwards. This need for immediate use is often reported to be effective in adult learning and professional development. In the same vein, it is also important that if workshops are organized, they are not too long. Lectures or theoretical inputs should be given in small doses and blended well with practical experiences.

Reduction of non-teaching tasks
It is important to reduce non-teaching tasks and release as much time and space for instructional preparation as is possible. While para-teachers in NGOs may need to get involved in community processes that indirectly benefit their educational outcomes, it is important for them to see this link explicitly. In other words, the majority of their activities should arise out of and contribute to their core identity as an educator.
Relations between para-teachers and those responsible for their professional development All these are more achievable when professional support planners are familiar with the learners (para-teachers), and with their zones of proximal development. Moreover, para-teachers can be supported in being more active learners only when their facilitators are trusted by them; not perceived as authority figures; capable of fielding disagreements and dealing with their inhibitions; and effective in mediating between NGO managers and the grassroots staff.
Factors that foster or inhibit para-teacher professional development 35

Implementing the recommendations in practice
Based on the recommendations above, a professional development program was designed and implemented in Maitri. The following vignettes demonstrate how each of the recommendations was used to shape activities with para-teachers.
• Using the first recommendation of focusing on development of basic teaching skills, workshops were organized at the start of a term, which focused on supporting para-teachers in developing basic skills for teaching by modeling the strategies. • Using the second recommendation, para-teachers and cluster heads in Maitri were supported in defining the teaching role more realistically. This was done by: (1) helping teachers to identify the grade level up to which they were comfortable in relation to their content knowledge; (2) supporting them in collecting current knowledge levels of their students to determine the actual learning requirements of their respective classes; and (3) asking them to calculate the actual contact time with students in terms of days and hours. Then, based on these data, they revised (1) the enrollment criteria which earlier involved enrolling any child that could pay to now being (enrolling only those students who needed support up to grade 5 which was the grade level they were comfortable with); and (2) the learning targets for each student they would enroll (and thereby also revising the promise made to the parents). Previously, para-teachers agreed to teach what the parents expected; now, para-teachers determined what the students needed based on their actual learning levels and what could be taught within the amount of contact time available. • Based on the third and fourth recommendations, the workshops also addressed ways to help para-teachers with procedural skills like planning for their lessons. This was done by combining performance tasks in the workshops along with modeling sessions. Lesson planning templates were used to help para-teachers think through different components of a systematic lesson plan; so that parateachers could use the templates to practice designing lesson plans in groups; and get feedback from peers and experts and refine their lesson plans. • Combining the fifth, sixth and seventh recommendations, daily routines were introduced for para-teachers to get time to prepare for and reflect on everyday enactment so that everyday immediate instructional learning needs could be met. This included helping them streamline various responsibilities through a task planner so that they could reorganize their work in such a way that everyday instruction preparation time was released. • Across each of these changes, attention was given to the eighth recommendation.
For example, the design and implementation of the professional development support in Maitri was undertaken by a management member. She combined two roles, one of a researcher bringing theoretical insights and rigor into the design and implementation, and another of a facilitator from within the organization who was familiar with the context, was trusted and who had demonstrated the ability to work with para-teachers without creating a feeling of power distance despite representing the management.

Implications for research
This paper contributes to a relatively limited body of literature on para-teachers from developing countries by underlining the professional learning requirements and contextual challenges that surround their work. It also contributes a theoretical framework for investigating para-teacher learning in a systemic context. This framework is valuable to researchers and practitioners interested in identifying the systemic influences or links to effective para-teacher development. The practice of engaging non-trained grassroots staff is common among NGOs and often leads to a wide expertise gap between higher management and first-level workers. This study presents an example of using objective, transparent and participative data collection methods which helped demystify ground realities for management and enabled grassroots staff to enter into an open discussion with their peers. This study is a call for further research to expand the understanding about parateacher development needs across diverse organizational and national contexts. Additionally, there is need for more research to identify models for effective interventions for this population, including attention to enabling organizational factors.