Improving the instructional leadership of heads of department in under-resourced schools: A collaborative action-learning approach

An unacceptable number of learners in under-resourced schools in South Africa are failing to perform adequately in national and international benchmark tests. Poor learner performance has been linked to poor-quality teaching, which, in turn, can be attributed in part to a lack of instructional leadership at schools. According to policy, heads of department (HODs) are best placed to offer such leadership, but in many schools this is not happening. We explain how we engaged HODs in one such school in a participatory action research process, to help them construct a framework for improving their instructional leadership. Qualitative data was generated through open-ended questionnaires, transcripts of recorded action learning set meetings, photovoice narratives, and reflective journals, and these were thematically analysed. The action learning framework developed by the participating HODs, while not being a definitive answer to improving the quality of teaching and learning, may provide guidelines for other HODs to improve their own instructional leadership practices. Since it is a process-based model, application of the model as an approach to improve instructional leadership could prove beneficial in both well-resourced and under-resourced contexts.


Introduction
In many under-resourced schools in South Africa, learners are failing to meet the required academic standards, according to the Trends in International Mathematics and Science Study (TIMSS) and the Progress in International Reading Literacy Study (PIRLS), as well as in terms of matriculation results (Mullis, Martin, Kennedy & Foy, 2007).The South African Department of Basic Education (DBE) tracks learner performance through the Annual National Assessment (ANA), the results of which indicate that learner performance is unacceptably low (DBE, Republic of South Africa, 2016), at least for the poorest and most disadvantaged children.Although the national pass rate was 70.7% in 2015 (DBE, Republic of South Africa, 2015), the results were skewed in favour of better-resourced schools, which are not financially, geographically or linguistically accessible to the majority of children in South Africa.
One of the reasons learners are performing so poorly is that teachers are struggling, for various reasons, to provide a quality teaching and learning experience ( Van der Berg, 2008).The difficulties of working in socioeconomically challenged communities, poor initial teacher preparation, lack of ongoing professional development opportunities, and poor school infrastructure all contribute to poor teacher performance, and increase the need for ongoing support and development of teachers through effective instructional leadership (Spaull, 2013;Wood & Olivier, 2008).Although the literature overwhelmingly identifies the principal as the main instructional leader, and local research has suggested that schools where principals teach do show better academic results (Roberts & Roach, 2006), the majority of principals in so-called "township" and rural schools are too occupied with the daily challenges of just keeping their schools functioning, to fulfil the role of main instructional leader.In such cases, it makes more sense for them to delegate their task to heads of department.
According to policy guidelines (Department of Education, Republic of South Africa, 2000Africa, , 2002)), heads of department (HODs) are positioned as instructional leaders.They should (i) assist teachers in setting and achieving personal and professional goals related to improvement of school instruction, and should monitor that these goals are successfully achieved; (ii) do regular formal and informal classroom observations; (iii) do postclassroom observation conferences with teachers, with the focus on improving instruction; and (iv) provide constructive critical evaluations, making recommendations for personal and professional growth goals according to individual needs (McEwan, 2003).However, in the first cycle of this larger action research project, where data was generated from two under-resourced schools, we found that involvement of the participating HODs in instruction tended to be limited to acting as "final checkers" of teachers' reports of work covered, where they adopted a task-oriented management role, rather than working with teachers on an ongoing basis to improve instruction (Seobi, 2016).We found that the HODs struggled to interpret the prescriptions of what they should do, and to translate these prescriptions into a coherent and sustainable framework for instructional support.We also found that they adopted a hierarchical, transactional leadership style, which did not foster the trusting relationships necessary for effective mentoring, coaching, and teamwork, which are essential for the provision of quality instructional support (Wood, Seobi, Setlhare-Meltor & Waddington, 2015;Zuber-Skerritt, Wood & Louw, 2015).The findings of the first cycle of this research project clearly pointed to the need to explore ways to help the HODs to reflect on and improve their instructional leadership practices.Since the HODs would need to sustain any improvement, it was imperative that they own the process and take responsibility for ongoing development of their practice.We therefore engaged them in a participatory process, to enable them to work as a team to develop an instructional support framework that suited their particular school ethos and context.As Bush, Glover, Bischoff, Moloi, Heystek and Joubert (2006) point out, South African education needs theories of leadership that are relevant to the context of the country.Any such theory needs to take into consideration the challenging contexts in which teachers work, and needs to be able to build human capacity, despite the sociohistorical disadvantages that still impact so negatively on teaching and learning.
The question that guided the process in the cycle that we now report on was 'how can heads of department in under-resourced schools improve their instructional leadership practices?'We first explain the theory that helped us to facilitate and make sense of the emerging process, before outlining the methodology used to explore the research question.Since the process of instructional leadership development is the focus of the study, we then offer a step-by-step explanation of how participating HODs came to learn how they could collaborate so as to improve the quality of teaching and learning at their school.We conclude with some suggestions on how the learning gained from this study help develop a more context-specific theory of how to develop instructional leadership.Although based on a South African case study, the findings of this research could be applicable for education globallywhether in emergent or developed economiessince they provide general knowledge about how to improve instructional leadership practice.

Conceptualising Instructional Leadership
In our conceptualisation of instructional leadership, we were influenced by Zuber-Skerritt's notion of action leadership (2011).Zuber-Skerritt (2011) describes action leadership as an inventive, pioneering, collaborative, and self-developed way to lead people.This leadership style is based on the democratic values of autonomy, equal opportunity, belonging, and self-realisation (Zuber-Skerritt, 2011).It is developed through critical self-reflection, and from being open to learning from others in action learning sets that meet regularly.Members of such groups set goals, plan to put them into action, and critically evaluate the outcomes to determine what further change is needed.Participation in such groups not only leads to enhanced goal attainment and skills development, but also builds trusting relationships among colleagues, which is necessary for ongoing individual and collective transformation (Dilworth & Boshyk, 2010).Action leadership focuses on developing an organisation's capacity to change, by encouraging the crafting of a collective vision towards the improvement of practice (Hallinger, 2003;Lee, Walker & Chui, 2012).From this perspective, HODs as instructional leaders should engage with teachers to foster shared ideas and practices for quality teaching and learning (Bush, 2008;Hallinger & Murphy, 2013;Nordengren, 2013).
Even though instructional leadership models in recent years have shifted towards distributed and transformational leadership styles (Blase & Blase, 2000;Spillane & Diamond, 2007), this change has not been translated into practice in many schools in South Africa (Hoadley, Christie & Ward, 2009).As Neumerski (2013) explains, despite substantial developments in instructional leadership theory (e.g., Biancarosa, Bryk & Dexter, 2010;Hallinger, 2005;Robinson, Lloyd & Rowe, 2008), knowledge of how such theory is enacted to improve teaching remains limited (Mangin & Stoelinga, 2008;Spillane & Diamond, 2007).We wanted to address this concern.We assumed that facilitating an action learning and action research approach with HODs would enable them to develop their capacity to become lifelong learners who can practise facilitative leadership based on collaborative enquiry within a specific school context (Zuber-Skerritt, 2011).Our aim was not only to help HODs understand their roles better, but also to translate such an understanding into action, so as to allow them to support teachers to improve the quality of teaching and learning.

Methodology
We chose a participatory action research (PAR) design, since the critical reflection central to the process enables personal and professional development (Dilworth & Boshyk, 2010).We worked with HODs from an under-resourced primary school in a peri-urban area in the Eastern Cape Province, in response to a specific request by the principal, who had been using action research and action learning to improve his own leadership practice.This school had enjoyed considerable success in setting up a community volunteer programme (see Kearney, Wood & Zuber-Skerrit, 2013), but it was still struggling to improve the academic performance of its learners.Volunteer parents were deployed in classrooms as teaching assistants, but not all teachers used them.Furthermore, the volunteer parents were generally allocated menial tasks, rather than their helping with teaching and learning.The four HODs in the school (three females, and one male) volunteered to participate in this project.We facilitated six action learning set sessions with them.Table 1 below shows how data was generated, documented, and thematically analysed (Bryant & Charmaz, 2007;Neuman, 2011) during the research process.It also demonstrates the measures taken to ensure trustworthiness of the study.The HODs signed informed consent forms, and the study obtained clearance from the ethics board of the institution concerned, which confirms the ethical integrity of the research process.

Table 1 Data generation and documentation techniques
Data-generation techniques Data-documentation techniques Qualitative questionnaires: given to teachers and volunteers by HODs to get their opinions.Transcripts of recorded action learning set meetings: done by researchers and given to HODs for verification.Analysis of group exercises: done by HODs with the facilitation of researchers Observations: HODs observed teachers, and they documented teachers' reactions and their actions during the process of research for them (HODs) to discuss those changes during their action learning set meetings.If there were no changes observed HODs would think of planning and implementing other strategies.Researchers observed how HODs progressed and changed during the process.Photovoice: We advised HODs to request teachers and volunteers to do photovoice to share their views with HODs because unlike a qualitative questionnaires, a photovoice exercise can encourage the participants to think more deeply about the context when deciding on the kind of quality they want (Baker & Wang, 2006).Prompts given: stakeholders were asked to take photographs of things that depict quality teaching and learning at their school.
Reflective journals kept and used by: Participants: for recoding discussions during meetings, that included planning, reflections on strategies applied, developed strategies for further improvement, and documentation of actions taken and/or agreed upon.Researchers: keep track of the participants' progress.

Data analysis and interpretation
Thematic analysis and interpretation (Bryant & Charmaz, 2007;Neuman, 2011) done by participants and facilitators together.

Credibility
Triangulation of datageneration methods; independent recoding; member checking.

Transferability / process validity
Rich description of research process.

Dependability
Inquiry audit trail.

Catalytic validity
Evidence of how the study stimulated enthusiasm for change in the participants and others.

Rhetorical validity
Critical feedback from the validation group on the research report.

Ethical considerations
Informed consent forms were given to participants then discussion of contents of the forms (participants' roles and the meeting timeframes) took place among the participants and the researchers before they volunteered their participation by signing.Values like transparency, safety, confidentiality, trust and privacy were guaranteed (Stringer, 2007).

Process of the Research and Discussion of the Learning Gained
We facilitated initiation of an action research enquiry by the HODs.The process is detailed in Figure 1.Of course, in reality, the process was not as linear as it appears to be, with the steps from Action 4 onwards overlapping and occurring concurrently at times.Since it would not be possible to discuss the whole process in depth in one article, for now we will concentrate on discussing what the HODs learnt, which enabled them to answer the questions that they had developed, and how they came to find new ways to implement the learning gained, so as to improve their instructional leadership.

Figure 1
Iterative cycles of action taken by HODs to improve their instructional leadership in order to improve the quality of teaching and learning (Seobi, 2016:102) The regular action learning set meetings (the HODs met weekly, in addition to the times that they met at the six action learning set meetings that we facilitated) gave the participants the opportunity to meet to develop plans, agree on strategies to be implemented, and reflect on the situation after implementation of the strategies.Such collaborative learning is similar to the process recommended in McCombs and Miller's (2009) six steps to continuous improvement of quality teaching: 1. Identifying and clarifying core beliefs about the school culture.2. Creating a shared vision of what these beliefs look like in practice.

Collecting accurate and detailed information about
the gaps between the vision and the current reality.4. Identifying what innovations will help to close existing gaps. 5. Developing and implementing an action plan that supports teachers through the change process.6. Embracing collective autonomy and collective accountability for closing existing gaps.
To facilitate a clear understanding of how the research process contributed to the HODs' understanding of how they could improve their instructional leadership of the process, we will refer to the following steps in our discussion.In the discussion of the steps, these codes are used: P = participating HODs; T = teacher; TA = teaching assistant.

Action 1: Identification of issues at the school in terms of the quality of teaching and learning
The first step was to enable discussion of the culture of the school in terms of teaching and learning.However, first we had to build relationship among the participating HODs.Collaboration and commitment towards attaining any goal has to be based on a relationship of trust and respect.
Although the four participating HODs in this study had a decent collegial relationship, it was important to enable them to deepen their understanding of each other as people, rather than just as colleagues.
Relational wellbeing, which is characterised by care, compassion, and mutual support, is at the heart of any successful collaborative effort (Evans & Prilleltensky, 2007).Relationship is also one of the core principles of action learning and action research (Zuber-Skerritt, 2011), and so we facilitated the creation of a humanising space, through engaging the participants in experiential exercises designed to allow them to express care for one another and to learn about "the person behind" the "colleague".Participants were able to appreciate how such exercises could also be used to facilitate relationship building among the teachers and teaching assistants they were supposed to support.The first shift in thinking was evidenced by the realisation of the HODs that they could achieve so much more if they worked as a cohesive and self-reflective team, rather than working independently as either Foundation Phase or Intermediate Phase staff, as they had been doing up until then: Having one voice was something that was lacking before at our school (P4).
Within the safe space of the action learning set, we facilitated dialogue to help the participants explore how they conceptualised their roles as instructional leaders, and how they experienced their practice within their specific school context.The four questions listed in Figure 1 (Action 1) were crafted by the participants to explore their concerns, and we now explain how they came to develop these questions.
Q1: "How do people perceive quality in our school?" The participants thought that it was important for them to have a common understanding of the type of quality education teachers and teaching assistants want, before they can have any influence on that quality.They realised that quality is a multifaceted concept (Runciman, Merry & Walton, 2007), and that it may be understood differently by each stakeholder:

I was wondering whether, if we want to know what quality education is, shouldn't we know what teachers think about it? (P1)
They decided to involve the teachers and the teaching assistants, to ensure that the vision they crafted would be based on the voices of all the stakeholders.
Q2: "How can we give effective support to the teachers?" The participants did not have time to meet with individual teachers face to face; therefore, they merely monitored the portfolios of the teachers and made brief written comments.They were thus worried that actual classroom practice could differ from what teachers claimed to be doing in the classroom.Their concern was that if teachers are not maintaining quality teaching and learning, moderation of their portfolios will not help to improve teaching and learning.Moderation has little impact on the quality of teaching and learning, as it cannot address the complexity of a whole teaching programme (Horsburgh, 1999), unless it is conducted with the teacher, where improvements can be discussed, and assistance to implement improvements can be negotiated.The participants were concerned that they did not have a system in place that they could use as a guide to support teachers in their everyday practicethey merely moderated the quality of the work, without any developmental input.
Q3: "How can we get all the teachers on board?"Some of the teachers were resistant to allowing the teaching assistants in their classroom; others were more open to using them, but utilised them only to clean, do other menial tasks, or watch the class when they were called out.The HODs expressed the opinion that the teaching assistants could be developed to offer more instructional support, but they realised that they would need to convince the teachers of this.
Q4: "How can we mobilise the teaching assistants to help the teachers more in the classroom?" The participating HODs realised that the teaching assistants would need help in developing the skills necessary to make them effective in supporting classroom instruction, since none of them had received any training in this regard.
Action 2: The HODs gathered data from the teachers and the teaching assistants In order to create a shared vision of what quality education at their school might look like, the HODs constructed a questionnaire with open-ended questions, to obtain the views of the other teachers and the teaching assistants on "what is quality education, its appearance, and also their theories on management" (P2).The purpose was to encourage everyone involved in teaching (both the teachers and the teaching assistants) to be involved in deciding what they wanted, and to take collective responsibility for any actions towards change in teaching practices in the school (Leach, Pelkey & Sabatier, 2002).We suggested that the HODs also use photovoice (see Table 1), as we thought that this might provide an alternative means of expression for those who were not comfortable with completing a questionnaire (Baker & Wang, 2006).Through discussion in the action learning set, the participants identified themes from the visual and the written data, to help them to craft a collective vision (see Figure 2).
Theme 1: Quality education must be holistic The data analysis conducted by the participants indicated that the teachers and the teaching assistants valued education highly, and saw it as a passport to improved life opportunities: Quality education to me means a better future for all (T6).
The kind of quality education that the stakeholders wanted for their school was to enable learners to develop their talents to the full, to enable them to realise their creative potential, and to guide them to take responsibility for their own lives and the realisation of their personal goals.Participants concluded that "results are necessary to compile an academic report" (P3) for learners, but that these are not the only variables that should be focused on when looking at improving education.
The responses of the teachers and teaching assistants echo much of what is known about quality education, namely that it offers the possibility of inclusion in the larger fabric of society, and that it is often associated with training for jobs and capacity building, in order to break the cycle of poverty (Ngomedje, 2006); the curriculum and pedagogical practices should prepare learners to thrive in their social contexts, rather than just empowering them to achieve academically (Penney & Chandler, 2000; Van der Berg, Taylor, Gustafsson, Spaull & Armstrong, 2011); and pressure to achieve high marks at all costs ignores the fact that education must have some direct bearing on the wellbeing of the social and ecological environments people actually inhabit (Gruenewald, 2003).
Failure to consider how education will impact on the future of learners seems to be one of the issues that has been overlooked and/or marginalised in many contemporary discussions (Penney & Chandler, 2000).This raises concerns about the adequacy and appropriateness of the ways in which education is currently developed, structured, and taught in schools.The stakeholders in this school wanted "education that will make the learners to be [sic] free from the bondage of mental inferiority" (see Figure 2).They were adamant that learners should not see themselves as inferior because of social circumstances.They felt that learners should be educated to believe in themselves, to feel proud of their background.They were of the opinion that education should provide learners with "equal opportunities, devoid of major disparities" (TA 2).
Theme 2: Quality education must be supported by a favourable school and classroom environment The teachers and the teaching assistants felt that quality education could be realised only in a favourable school and classroom environment, meaning that "learners should enjoy themselves at school and teachers must give their best to educate learners" (T10).They wanted teachers to "be more involved in the children they teach at school" (T4), and to be "able to motivate learners and ensure excellence in the classroom by requiring learners to work hard" (T7).These responses highlight the important role of the teacher in producing quality education.
Definite ideas of required teacher conduct were voiced: "teachers should be on time for their classes" (T8); "they must ensure that the syllabus is completed within specified time frame" (T2).They expressed a desire to have teachers that are passionate and have a love for learners and the community that they are serving.They wanted "teachers who are committed, productive, present all the time, clued up on all aspects of education (curricular and co-curricular)" (TA12).To reach that level, teachers need continuous development, "to develop learners' knowledge, skills" (T10).Furthermore, teachers were required to "use teaching and learning strategies that cater for a variety of learners" (T6), and with those strategies, "learners must be interactive at all times" (T5) in their learning processes.
The stakeholders also recognised the need to have adequate resources in the school, but they felt that they could still do much to encourage learners, through recognition of their efforts.
The second photograph in Figure 3 shows print-rich walls, filled with learners' work, so as to enhance the quality of teaching and learning at the school.The participants thought it was important for teachers to display learners' work on the walls of the classroom, to show "recognition of learners' efforts", and that this "constantly encourages learners to perform better" (T7).The affective outcomes of education are at least as important as the cognitive results, and the acknowledgement of learner efforts is reflected in learners' increasing motivation to learn (Kim, Fisher & Fraser, 2000).
Theme 3: Quality education must demonstrate a link between the school and the community Another narrative that emerged from the data was that education should be contextually relevant, enabling children to be productive citizens who can contribute meaningfully to their community (see Figure 4): "quality education has to be responsive to the ethnic group it serves" (TA3).
It was felt that one way of creating a link between the school and the community was to involve parents in the education of their children.The school in this study prides itself on being a community school (Kearney et al., 2013).Community schools collaborate with many partners to offer a range of support and opportunities to children, youth, families, and communities (Epstein, Sanders, Simon, Salinas, Jansorn & Van Voorhis, 2002).In this school, the use of parents as teaching assistants is one way that the school involves parents in the education of their children.This school strives to be an enabling space that promotes health on all levelsphysical, social, and emotional.The data generated by the teachers and the teaching assistants seems to echo the findings in the literature, as the stakeholders see education as a "social rather than an isolated process" (Osterman, 2000:324).They also emphasised that quality education at their school can only be achieved with community support.

School as a favourable teaching and learning environment, with parents involved in school activities
They are helping other community members to think positive about their teachers because people think that toilets should be a dirty place.But these learners are cleaning and they are enjoying their work.There after we can sit and relax in our toilets.And the learners also are taught to respect their toilet, not to mess but to keep them clean.[sic] It is helping community to be an exemplary they will get crops to plant in their gardens.They can get vegetables so that they can be healthy.Even those who are sick they can get vegetables so that they can get balanced diet per meal per day.The fourth theme indicated that the teachers were aware that quality education depends on continuous development and learning.However, they did not mention their own developmental needs, tending to focus on the needs of the teaching assistants.There was recognition that the teaching assistants could do so much more if they were trained and mentored.The teachers indicated that "TAs [teaching assistants] can help with auxiliary duties of photocopying, and data capturing, which involves capturing learners' marks, recording them and compiling class lists" (T2).The suggestion was also made that the TAs should help support teachers instructionally, by "identifying learners' needs" (T11), and through helping slow learners.Reflecting on this data, the HODs realised that it would be important to create "opportunities for TAs to be empowered in understanding the curriculum" (P1), so that they could help teachers more in the classroom, and assist learners with their homework.This is in line with current international trends, where TAs are currently perceived to have a more professional role, recognised through the provision of several new accredited training initiatives and qualifications, such as the Professional Standards for Higher Level Teaching Assistants (Groom & Rose, 2005).Training puts TAs in a position to work specifically with learners with social, emotional or behavioural problems, and they can also play a major role in supporting the teacher in the management of behaviour in the classroom (Austin, 2002).The HODs also pointed out that teachers would also need to be helped to identify their own needs for improving teaching and learning, as there was little evidence of such needs from the data generated.

Action 3: The HODs crafted a vision for improving instructional support for quality teaching and learning
Based on data generated in Action 2, the HODs developed a vision to guide their actions when working with other stakeholders, to effect change so as to realise the desired quality of teaching and learning at their school (see Figure 5).The crafted vision linked to the themes that emerged from the stakeholders' generated data, and allowed the HODs to begin to identify gaps between the collective vision for quality education and the reality of what is actually happening.As shown in Figure 5, love was the central theme of the vision.

Figure 5 The HODs vision of quality teaching and learning based on data generated with teachers and TAs
The HODs emphasised that love should be central to all teaching and learning at the school: "when you have love for people whom you are working with, no matter where they are from or what challenges they come with, love helps you to overcome all those challenges" (P1).The HODs believed that if the teachers teach the learners with love in their hearts, their teaching will be more effective, as they will put more effort into it.The participants wanted passionate teachers, with an ability to confront the realities of schools in disadvantaged contexts.The HODS also believed that when the teachers care, their energy will begin to flow towards satisfying the needs and desires of the learners.
Values such as respect, love, equity, peace, commitment, and trust were also highlighted as essential to be able to provide an environment that is conducive to teaching and learning (Theme 2).The text outside the heart in Figure 5 represents the need to support teachers to work together with the wider community (Theme 3).One example of this was the training of volunteers to support teachers in the classroom with instructional tasks, such as homework, and to liaise with parents to do this.The vision also represents all stakeholders, namely the teachers, the principal, the subject heads, the grade heads, and volunteers, as lifelong learners, engaged in continuous development (Theme 4).The ultimate aim of this vision is for learners to be successful and to reach their potential; hence, education must be holistic (Theme 1).
Action 4: The HODs worked together as a team to develop strategies that they used to support the teachers to ensure quality teaching and learning at their school The next step for the HODs was to devise strategies to support the teachers through the change process towards realising their idea of quality education, derived from the data generated by them.The HODs realised that teamwork would be of benefit at all levels, since it would enable a distributed form of leadership, freeing them to concentrate on supporting and mentoring, rather than just controlling and monitoring.They realised that as instructional leaders, collaboration would need to start with themselves: "it came up that we need to sit together as a school, from Foundation phase and Senior phase" (P3).Previously, the HODs had functioned independently, as two teams of HODs in two different phases: "It was more of the HODs in the Foundation phase that worked together and the HODs in the Senior phase that worked together" (P4).
After experiencing the benefits of collaborative work for themselves, the participants decided to convince the rest of the teachers and the teaching assistants to also work in teams.They brainstormed some strategies that they could develop to support the teachers instructionally, identifying gaps between the vision and the existing situation.They decided to create teams at various levels, to enable the development of focused working groups, where all voices could be heard.Each HOD would work with their subject heads, the subject heads would work with their teachers, and the teachers would work with the TAs in subject groups.The HODs found that "working in groups like this is a relief" (P3), as it helped them to feel supported, and they were able to devote more time to mentoring the teachers, rather than just monitoring the work of the teachers.
Action 5: The HODs worked with the TAs, so that the TAs could better support instruction The HODs prioritised working with the TAs, because they realised that several teachers were resistant to involving the TAs in classroom matters.One HOD volunteered to mentor the TAs until the subject heads could convince the teachers to include the TAs in the subject group meetings.The TAs were trained to do the classroom administrative tasks of photocopying documents, such as question papers and work schedules, recording learners' marks, and designing mark sheets.
The HODs also encouraged individual TAs to visit the homes of learners that were not doing their homework, to try find out how the learners could be supported to do their homework.Besides giving homework support, the HODs said that the TAs could also "replace teachers, in classes where there were no teachers because of absenteeism" (P1), to avoid having to put learners in other classes, and so prevent overcrowding.The HODs also worked with the TAs to help them learn how to maintain order in the classroom when the teacher or the HOD was involved in other school tasks: "for myself as Deputy Principal, when the principal is not here, I have to come to the office for administrative issues, or attend to parents, I leave my class a lot, and I use the TAs to attend to my class whilst I am out" (P1).
Action 6: The HODs worked with the subject heads to empower them to work with the teachers and the teaching assistants to improve teaching and learning The HODs decided to work with the subject heads to empower them to work collaboratively with the teachers to improve teaching and learning.Before their engagement in this research, the HODs had been doing the work of a subject head, rather than delegating.As one participant remarked, "[i]t was only when I attended a DBE [Department of Basic Education] training last month, that I realised I did not even know what a subject head was supposed to do" (P4).By delegating moderation of portfolios to the subject heads, the HODs now had the opportunity to support the teachers on an individual basis.Once the subject heads had moderated the work, the HODs could just check it, which enabled them to see if: (a) the subject head was in need of support; and (b) what the developmental needs were of the teacher in question.The HODs and subject heads thus started to embody values characteristic of distributed leadership, namely coperformance, interdependence, and full commitment from both parties (Gronn, 2002;Spillane, 2012).This experience also helped the subject heads to learn how to collaborate with their respective teachers, and to encourage them to take responsibility for improving teaching and learning, thus creating multiple layers of leadership for instructional improvement, something that is highlighted as important in the literature (Spillane, 2012).
Action 7: The HODs worked with the teachers on a one-to-one basis, in order to provide space for individual development By delegating work to the subject heads, the HODs created some time to work with the teachers individually.We asked the HODs to document their actions when they worked with the teachers, and to reflect on their learning in the action learning set discussions.Their reflections highlighted that they had learnt about the importance of establishing good relationships with individual teachers.A good relationship helped the teachers to be open to input from the HODs, giving both the teachers and the HODs space to collaboratively discuss issues, develop plans to be implemented, and agree on ways to implement the plans so as to achieve quality education.Mentoring was one of the ways in which the HODs engaged with the teachers, particularly those that were newly appointed.For example, Participant 1 stated that they had "worked with one teacher, drawing a lesson plan together and discussing it together." The participating HOD sat with the teacher after school, and they worked together to develop five lesson plans.The HOD said that the discussion with the teacher lasted "until after 5 o'clock", and that they "did not realise that it was so late" (P1).By talking to the teachers and finding out their needs, the HODs were better able to support the teachers, rather than just berating them for poor performance.For instance, when a couple of teachers failed to meet the deadline for submitting marks on a USB flash drive, the HOD in question chose to have a conversation with them to find out why they had not met the deadline.This conversation afforded the teachers the opportunity to admit that they did not have the skills to enter the marks, and so the HOD was able to train them.Through this conversation, the teachers were also able to provide input, which helped the HOD to improve the template for mark entry.Thus, mutual learning occurred, and both parties felt respected and valued.The HODs gave several examples of instances where such professional conversations had not only led to finding ways to improve teaching and learning, but had also strengthened collegial relationships in the process.

Action 8: Development of a framework to ensure quality teaching and learning
In order to encourage collective accountability for improving teaching and learning at the school, the HODs realised that they would have to develop a framework, or model, to encapsulate the collaborative approach to instructional leadership that they had developed through this study.They did this by reflecting on their learning throughout the action research process, and they decided that the two factors that had led to an improvement in their instructional leadership were:


The importance of teamwork and participation of all stakeholders to improve instructional support to teachers, and  The importance of forming good relationships to enhance collaboration among all stakeholders.
They constructed a diagrammatic representation to use as a framework for involving all teachers, TAs, and parents in improving instruction at the school.However, it is not within the scope of this article to present and explain the model.

Conclusion
In this article, we described the process followed by the heads of department (HODs) in one school to improve their instructional leadership.Through critical reflection on their actions, within the safe space of a participatory group, the HODs came to learn that they could influence the teachers and the teaching assistants at the school to take responsibility for their own learning and development, by introducing them to the concept of working in democratic, focused, collaborative teams.By creating such dialogical spaces, collegial relationships were improved, and teachers at all levels began to feel valued.This can only help to improve teacher motivation and commitment.Although the process was not without its challenges, participation in the action research project has changed the way the HODs at this school provide instructional leadership, and the model that was developed will continue to be used to entrench the values and the vision that were crafted to support improvement of teaching and learning in the school.The main challenge experienced by the participants was finding time to work in a more relational way.However, by offering an explanation of the process followed, we have provided an answer to the research question we posed earlier in this article, namely "how can heads of department in under-resourced schools improve their instructional leadership practices?"The findings of this study can be applied to improve instructional leadership in both disadvantaged and more advantaged school contexts, since they provide a general framework for professional development in this regard.It is hoped that other school leaders will be able to learn from the process and adapt it, to improve their own approach to instructional leadership.holder acknowledges that opinions, findings, and conclusions or recommendations expressed in any publication generated by NRF-supported research are those of the author(s), and that the NRF accepts no liability whatsoever in this regard.
We express appreciation to all the Heads of Departments, the teachers, the Teaching Assistants (TAS) who took part in this study.We also extend our sincerest gratitude to the Principal of the school who gave us the permission to conduct this research at his school.This paper would have never been a success without your valuable contribution.

Note
i.
Published under a Creative Commons Attribution Licence.

Figure 2 Figure 3
Figure 2 Quality education should liberate learners from mental inferiority and broaden their horizons

Figure 4
Figure 4 Parents and members of the community should be involved in school activities