Learning from life story collection and analysis with children who work in the leather sector in Bangladesh

CLARISSA (Child Labour: Action-Research-Innovation in South and South-eastern Asia) is a participatory evidence and innovation generating programme. We are generating evidence on the drivers of the worst forms of child labour (WFCL) and exploring how to address them through participatory Action Research (PAR) with children and other stakeholders in the leather supply chain in Dhaka, Bangladesh. Our main intervention modality is Systemic Action Research (SAR) (Burns, 2007), of which life story collection and analysis (LSC&A) is the first step in our participatory design that will inform child-led PAR groups which will become the engines of innovative responses to WFCL. The LSC&A methodology is a storytelling and story listening methodology and was chosen because of the universal power of stories to make sense of complex realities and seek new futures. People across the world like to tell their stories; they like to feel listened to and they are interested in how their story connects and compares to others. By collecting and analysing stories from hundreds of children in WFCL we can visualise the bigger system that each individual story is connected to. We hypothesize that through engagement in the process of telling, listening, collecting, and analysing life stories, children engaged in harmful and hazardous work will use their understanding of systems dynamics to move into creating their own solutions to the drivers of WFCL. In 2021 in Bangladesh, we collected 405 life stories from children living in Hazaribagh, Hemayetpur, Lalbagh, and Bhairab in Dhaka, with more than one hundred of these stories collected by children themselves. Following the story collection and transcription, children were supported by the CLARISSA implementation team to collectively analyse the stories through identifying critical ‘factors’ (events which have causes and consequences) and understanding how they causally relate to each other. The analysis of the 405 life stories resulted in the development of large system maps that illustrate all the causal dynamics that underpin lived experiences of WFCL. Based on the systemic analysis process the children identified themes of the PAR to be set up in their localities. Our experience with the LSC&A methodology is the first in the context of children in WFCL. The methodology has been used in one other project in Bangladesh to date. We therefore intentionally aimed to learn from the implementation process and to evaluate if and how the telling of, listening to, collecting and analysing of life stories is empowering and in turn whether it leads to increased ownership of the problems which motivates collective action (Burns, 2021). In this learning note we share our methodological learning and reflect on operational implications for designing and facilitating an LSC&A process with children which we hope will support adaptation and use of the methodology by others working in participatory programming with children.


Learning Summary
The CLARISSA Bangladesh team successfully collected and analysed 405 life stories from children in the worst forms of child labour (WFCL).Our key learning from undertaking this first part of a systemic Action Research process in the context of WFCL is as below: • Trust and rapport building, in the local neighbourhoods of intervention, was essential in order for the process to work.This was challenging, in the context of initial mistrustgiven the sensitivity of WFCL.It took six months of casual and informal visits and meetings to slowly build trust with a wide range of stakeholders including children in WFCL, their parents and gatekeepers (small business owners, house managers).Taking consent processes seriously helped us to build rapport and also showed the parents that the team was concerned about their children's safety.Compensating children for their time contributed to trust building and establishing an office in the community showed our commitment to working there in the long term.Once initial trust was established with some adults and children in the community, a positive trust feedback loop occurred, whereby children explained the process and shared their perspectives with others who were therefore more inclined to engage.Building trust with business owners was particularly challenging, given their concerns about the consequences of participation (for their businesses).
• Closely related to building trust was being intentional with our tools and processes for a child-centred approach.This included being responsive to feedback and preferences expressed by the children, such as adapting specific tools or undertaking processes at a time and place preferred by the children.This meant that team members had to be flexible and sometimes work outside of standard office hours.An enabling environment for child-centredness was one where children felt respected, loved and heard.Peer-to-peer working helped to build trust, but importantly, when it comes to sensitive topics, this worked best with children of the same gender.
• We learned about shifting our own mindset that underpins our beliefs around children's abilities.This evolved over time as we witnessed more and more how children were able to undertake the collection of stories and engage in the subsequent analysis of the stories with quality and depth.
• Life story telling and analysing has the potential to retraumatise the participant.When working with children in WFCL, safeguarding and mental health and psycho-social support processes need to be in place, especially as we were collecting stories during the height of the COVID-19 pandemic.Mental health and psycho-social support services were made available for children when they experienced emotional breakdown.All facilitators were trained to support children when this happened, and a CLARISSA social worker was central to the process.Implementing good COVID-19 safety procedures helped to build trust with parents and children by providing tests and safe spaces to undertake the work.

Introduction
CLARISSA (Child Labour: Action-Research-Innovation in South and South-eastern Asia) is a participatory evidence and innovation generating programme.We are generating evidence on the drivers of the worst forms of child labour (WFCL) and exploring how to address them through participatory Action Research (PAR) with children and other stakeholders in the leather supply chain in Dhaka, Bangladesh.Our main intervention modality is Systemic Action Research (SAR) (Burns, 2007), of which life story collection and analysis (LSC&A) is the first step in our participatory design that will inform child-led PAR groups which will become the engines of innovative responses to WFCL.
The LSC&A methodology is a storytelling and story listening methodology and was chosen because of the universal power of stories to make sense of complex realities and seek new futures.People across the world like to tell their stories; they like to feel listened to and they are interested in how their story connects and compares to others.By collecting and analysing stories from hundreds of children in WFCL we can visualise the bigger system that each individual story is connected to.We hypothesize that through engagement in the process of telling, listening, collecting, and analysing life stories, children engaged in harmful and hazardous work will use their understanding of systems dynamics to move into creating their own solutions to the drivers of WFCL.
In 2021 in Bangladesh, we collected 405 life stories from children living in Hazaribagh, Hemayetpur, Lalbagh, and Bhairab in Dhaka, with more than one hundred of these stories collected by children themselves.Following the story collection and transcription, children were supported by the CLARISSA implementation team to collectively analyse the stories through identifying critical 'factors' (events which have causes and consequences) and understanding how they causally relate to each other.The analysis of the 405 life stories resulted in the development of large system maps that illustrate all the causal dynamics that underpin lived experiences of WFCL.Based on the systemic analysis process the children identified themes of the PAR to be set up in their localities.
Our experience with the LSC&A methodology is the first in the context of children in WFCL.
The methodology has been used in one other project in Bangladesh to date.We therefore intentionally aimed to learn from the implementation process and to evaluate if and how the telling of, listening to, collecting and analysing of life stories is empowering and in turn whether it leads to increased ownership of the problems which motivates collective action (Burns, 2021).In this learning note we share our methodological learning and reflect on operational implications for designing and facilitating an LSC&A process with children which we hope will support adaptation and use of the methodology by others working in participatory programming with children.

LSC&A design and implementation process
The design of the LSC&A includes three phases with multiple steps 1) Life story collection by adults (10 steps), 2) Child-led life story collection (12 steps) and 3) Life story Analysis (20 steps) (Shown in Figure 1).In what follows we provide a detailed description of the implementation process.Phase 1: Life story collection From October 2020 to August 2021, a total of 405 life stories (see Table 1 for demographic information relating to storytellers) were collected from children engaged in WFCL.The type of work that child storytellers were engaged in included different kinds of leather processing, leatherwear making, garment work, waste picking and construction work, among others.Life stories were collected from two locations in order to capture different types of stories; stories of children from a specific neighbourhood (the wider Hazaribagh area), and stories of children working in leather supply chain locations (Lalbagh, Hemayetpur, Bhairab and Hazaribagh).Among 405 stories, 104 were collected by children aged 13-17, currently engaged in WFCL and trained in story collection, and the rest were collected by adult staff of the Bangladesh team.

Development of life story collection toolbox and co-design
We first developed a life story collection toolbox to support systematic documentation of all activities.The toolbox was developed and refined based on feedback from all team members and includes personal information forms, consent forms, feedback forms, a learning documentation template, a safeguarding checklist, a personal information database, a feedback database and a story documentation template.We also engaged in a series of planning meetings to co-design the details of the process and become familiar with the final toolbox.

Rapport building and getting to know the neighbourhoods
We then moved into the field locations (informally) in order to build rapport with people living and working in the selected neighbourhoods.As well as getting to know members of these communities, we also wanted to become known -in order to facilitate local access.A transect walk around the community enabled this two-way visibility.We also had informal talks with influential people from the communities (such as slum managers, parents, religious leaders, and other stakeholders) and with children engaged in WFCL.The rapport building process was not uniform across all locations, for instance in Bhairab and Hemayetpur we were unable to spend sufficient time in the neighbourhoods, and so relied on gatekeepers (such as slum managers and small factory owners) who later helped identify storytellers.Community Mobilisers 1 and Field Organisers 2 also contributed to identifying children to participate in these locations.In some locations we had existing contacts through conducting early scoping studies.In some locations, however, we had to build new relationships.After spending sufficient time engaging directly in the neighbourhoods, we felt ready to begin the life story collection process.
1 Community Mobilisers are responsible for mobilising people living in Hazaribagh neighbourhood as part of a CLARISSA social protection intervention.They collect real-time household data, are involved in casework and in the future will be involved in collective community organising.They were in post prior to the LSC&A process, and so were able to support in the identification and selection of children storytellers.
2 Field organisers were specially recruited by Gram Bangla Unnayan Committee to support the facilitators and documenters during the LSC&A process.They helped them by finding children from the community and provided logistical support.

Identification of potential storytellers
We first identified potential storytellers from the broader Hazaribagh area where we had been engaging for longer, and then moved to other areas.We used two criteria to identify potential storytellers: 1) must be aged between 8-17 years (which was later modified to 11-17 years as the team felt that children below the age of 11 do not have much experience to share in their life stories); 2) must be currently engaged or was engaged previously in WFCL for at least a year.Local gatekeepers were central to identification of storytellers, as was the use of the snowballing method3 .
Adult-led life story collection

Piloting of life story collection
In October 2020, we collected 11 life stories from children in Hazaribagh in a pilot of the life story collection process.The objective of the pilot was to gain practical experience of the life story collection process to learn and adjust the planning to strengthen the design.We also wanted to apply our learning from the online training received (due to COVID-19 the training was virtual) in order to make necessary changes in our story collection plan.An After Action Review (AAR) was held to harness learning experiences from the pilot, to understand what went well and what needed to be improved.The life story collection tools were adapted accordingly.
The process of adult-led life story collection From October 2020 to March 2021, a total of 301 life stories were collected by the team.Among these 301 life stories, 200 were collected from four supply chain locations (Hazaribagh, Lalbagh, Hemayetpur and Bhairab) and 101 life stories were collected from neighbourhood locations (wider Hazaribagh area).Halfway through this process, after realising how resource intensive it would be, six Documentation Assistants4 (DA) and two Field Organisers (FO) were recruited.DAs were recruited to speed up the story collection activities by helping with facilitation, documentation, and transcription of the life stories.The FOs were recruited to identify potential story collectors and to support mobilisation which included finding space for the story collection and talking to parents.
The team was divided into pairs, with one facilitator and one documenter in each, with one female and one male.During the collection process the facilitator asked the questions, and the documenter documented the conversation through recordings and taking notes.Before the start of the story collection process, the storytellers were always asked if they were comfortable talking to someone of the opposite sex (if the storyteller and the facilitator were of opposite sexes).If the storytellers said they were not comfortable with this, the facilitator and documenter changed roles.
The pairs first approached the children they had identified and talked with previously and verified whether they wanted to share their story.If the child confirmed their participation, the team then explained thoroughly the purpose of story collection and obtained formal written consent from the child and their parents/guardian.A child friendly sitting space was always ensured during the time of the life story collection and the wellbeing of the storyteller was kept in mind at every step of the process.In the process different types of questions (prompt questions, encouraging questions, clarifying questions, deepening questions) were asked, such as: • Tell me about your life and the most important things that have happened to you?
• Tell me about your childhood and all the things that led to you working in this place?
• Tell me about yourself?
• How do you see your future?
• Can you tell me what you worry about when you go to sleep at night?
• How have you been managing your life?
• How do you think you can improve your life in the future?
These questions aimed at starting a discussionand collectors then followed up with questions according to the context and discussion.Immediately after the storytelling process the children were asked to complete a child-friendly feedback form.
Phase 2: Child-led life story collection A total of 104 life stories (see Table 2 for demographics) were collected by children.The intention of the child-led life story collection was to explore how the children perceive their life when they talk to their peers and so to understand the difference between adult-led life story collection and child-led life story collection.

Preparatory Activities
Informal meetings with children's groups from the community To prepare a longlist of children who could become life story collectors, the team, with the help of community gatekeepers, organised informal yard meetings with several children from local neighbourhoods who had been storytellers, as well as their friends who were in WFCL.
Children were asked how they felt about the story collection process and whether they thought they would be able to collect stories from their peers.

Screening workshop for identifying potential children story collectors
A screening workshop was held to identify potential child story collectors from the longlist.A total of 50 children (23 male, 27 female) were invited in three batches to the screening workshop.The participants were divided into pairs, with 7-10 pairs in each batch.The pairs were given 10 minutes to chat with each other.The topic of the discussion was "to know each other's life".After that, they were given the floor to introduce themselves to each other based on the discussion.A score sheet was used by the team members to assess the children based on their skills in questioning, storytelling, and presentation.Using this assessment, 20 children were selected initially to work as story collectors.
Meeting with parents of the potential children story collectors A meeting was held with the parents of the potential child story collectors to inform them of the work modalities involved.Parents of all 20 children were invited to join one meeting.A total of 14 parents joined the meeting.As well as the details of the work the children would do, they also were informed of the benefits they would receive for taking part in the story collection activities.A terms of reference detailing the role of child story collectors was shared.Parents asked questions for clarification and their consent was sought.Consent from parents who were absent from the meeting was taken later at their homes.

Risk assessment and risk mitigation plan
After the parent's meeting, risk assessment and risk mitigation plans were formulated by the CLARISSA Social Worker.A list of possible risks (safeguarding risks during commuting, risk of facing problems with the employer in relation to joining CLARISSA activities, risk of peerto-peer abuse, breach of confidentiality, conflict within the family etc.) and the mitigation of those risks was developed based on group discussions.

Training for children life story collectors
During the last week of March 2021, a six-day training was provided to the selected story collectors, which focused on the process itself as well as safeguarding and ethical considerations.The training was delivered through lectures, group work, group discussions and several games.Out of the initial 20 children selected, 19 (five male, 14 female) participated in the training (one child had left Dhaka so could no longer participate).Following the training, semi-structured phone interviews were completed with the participants to assess if and how the training had contributed to developing knowledge of the process and skills required.Given the sudden COVID-19 lockdown around this time, it was also an opportunity for the team to check in on how the children were coping.Interviews were recorded, transcribed and thematically analysed and a report was written.The short assessment helped us to understand the children's level of knowledge of the life story collection process as well as how they were coping with the sudden lockdown.

Refresher session with children story collectors
The COVID-19 lock down led to a two-month gap between training and implementation, so a refresher session was held.It took place in a community meeting place and the facilitator tested the knowledge of the children by engaging them in group discussions regarding different steps of the life story collection process.We found that children did not forget their learning from the earlier training, and they seemed ready to move to the field to work as story collectors.Out of 19, one child could not join the programme as she had started working in a factory during the lockdown, leaving 18 in total.

Child-led story collection process
From July to August 2021, the 18 child story collectors collected a total of 104 life stories from children engaged in WFCL in the wider Hazaribagh area.Children collected all the stories by themselves.They also recruited the storytellers using the criteria described above.
Each story collector was accompanied by an adult documenter from the team.Though there was always an adult present, they played little to no role in the process except in documenting the story.The entire process was led by the child story collectors.

Phase 3: Narrative Analysis
The basis of the claim that the LSC&A process enables deep participation is that participants analyse their own stories.Further, in keeping with the child-centred approach, the stories were analysed by the children themselves in a participatory manner, with adults playing the role of facilitators.From 12 th September to 24 th November 2021, all 400 stories were analysed by a total of 53 children (see Table 3).Among these 53 children, 37 had already engaged as either a storyteller or a story collector (in some cases both).16 were new participants who had not been involved previously.Age and having been engaged in WFCL were the selection criteria.

Preparatory activities
Before the start of the narrative analysis process, a refresher session was conducted virtually by the international programme team covering the facilitation process and how to support children to conduct analysis.The main implementation team then provided training to the field team who would support narrative analysis workshops (NAWs).As with the previous steps, a toolbox was developed, including an attendance sheet, a story analysers database, a template for documenting plenary discussions, a template for observing group work, a template for facilitator's reflections, an in-depth interview guide, a template for reflections and learning, and a Mental Health and Psycho-social Support (MHPSS) debrief.

Pilot Analysis workshop
A pilot workshop was held to test our skills through practicing the facilitation process with children.13 child story collectors were invited to participate in the pilot.The children were briefed thoroughly regarding all steps in the process and a total of 13 stories were analysed by the children.The team adapted the methodology accordingly.

Narrative Analysis Workshops
The main objective of the analysis process was to identify essential 'factors' (events with causes and consequences) and understand how they causally relate to one another as a system map.This was accomplished by reading the stories one by one, identifying the factors, and connecting them with arrows in small system maps, which were then incorporated into larger system maps.
Creating small system maps from the collected life stories A total of 53 children took part in this first step of the process.There were 22 male children and 31 female children, aged between 12-18 years.Among these 53 children, 37 were either storytellers or story collectors (and sometimes both).At this point we had already developed a strong relationship with the children and knew their skills well, so we selected the children who participated in the analysis based on their communication and critical thinking skills.
These children created 405 small maps from 405 life stories over the course of six workshops which took place in Dhaka, Bhairab and Hemayetpur.There were three workshops each for neighbourhood and supply chain locations.At each small map workshop, participating children (normally 15) were divided into five trios with an adult facilitator assigned to each trio.There were two documenters present at each of the workshops, who observed and documented the entire process.
The small map creating process in trios started with the facilitator or one of the members of the trio reading out the story.The trio noted down the key factors from the stories, with the help of the facilitators.The trio then drafted small maps on paper portraying the causal relationships between the key factors in the story.Once the trio finalised their draft map, they transferred it to a larger sheet of paper.Next, they identified the two most important factors from the stories, wrote them on sticky notes and stuck them at the top of their final small system map.Figures 2 and Figure 3 show examples.Creating a large system map from the small system maps After the creation of 405 small system maps from 405 life stories, two large system map workshops were held to create two large system maps -one from the small system maps of the analysed neighbourhood stories and one from the small maps of the analysed supply chain stories.30 children, who were part of the small system map workshops, participated in the large system map workshop.Like the small system map workshops, at each large system map workshop, 15 children were divided into five trios and each trio was accompanied by one adult trio facilitator.Apart from the trio facilitator, there were also two main facilitators and two documenters for each of the workshops.
The key steps of the large system map workshops included: • Clustering of sticky notes -children called out an issue on one sticky note (sticky notes containing most important factors in the small system maps) and placed it on the wall.After that they asked for any other sticky notes that were similar and clustered them together.They repeated the process until all the sticky notes from the small system maps had been transferred to the wall.Next, they drew lines around the clusters and titled them.Lastly, they had a plenary discussion on (a) what are the big trends they see and (b) what they find surprising in what they see.
• Producing a big system map -children drew all the linkages between the factors on the small maps.
• Identifying the top 20 or so relationships, verifying the actual number of stories that these relate to, and thickening the lines on the map to enable the important patterns and dynamics to emerge.
• Discussing the key drivers of the outcomes that the groups thought were most strongly represented in the stories.
• Identification of areas where action might influence decisions regarding the key issues for Action Research groups.
Figure 4 shows the large system map from the neighbourhood locations.

Learning from and about the Life Story Collection and Analysis process
Central to the program's evaluation and learning agenda is our intention to learn from and about our participatory process innovations and methodologies5 .Throughout the LSC&A process we used a range of monitoring, evaluation and learning (MEL) activities to collect data for learning and evaluation.This included seven after action review workshops (three during the life story collection process and four during the narrative analysis workshops), reflection sessions with children after the training of child story collectors, focus group discussions with children story collectors and in-depth interviews with 27 children engaged in the narrative analysis workshops.As a programme team we collectively analysed this data to respond to our evaluation agenda and synthesized key process learning.

Designing and facilitating a child centred process
Child centeredness is a core principle of the programme which is built upon the belief that children are experts of their own lives.Therefore, our participatory design was based on the perspectives, opinions and lived experiences of children.We intentionally gathered and used feedback from the children and aimed to build and use child-friendly tools, spaces and facilitation processes.
We employed a range of methods to gather feedback from the children on their experience of the process, such as feedback forms, in-depth interviews and focus groups.Feedback received included suggestions to make the discussions and sessions more participatory, enjoyable and to use frequent energiser.We responded to this feedback directly by paying more attention to creating a fun and interactive space.For instance, at the beginning of story collection the consent form was a bit long, and the children and their parents had to sign in three different places, and after receiving this feedback we modified the form to make it more user friendly.We used an array of child friendly tools to create an informal setting and fun process, including games and energisers.Understanding the lived experiences and specific dynamics faced by the children involved helped us to tailor our approaches to them.The use of formal and informal tools and child friendly approaches made the activities amusing and intriguing for the children.This helped the children to open up to us, to share their stories and engage in the analysis processes.
We recognised, however, that quality in a participatory process is not based on using tools alone, rather it is supported through creating an enabling environment for effectiveness in working with children.An enabling environment refers to an atmosphere where children feel comfortable to express their thoughts and opinions freely.We paid particular attention to our ways of engaging, such as being respectful, transparent, patient, empathetic in listening, and we strove to be aware of inter-generational power dynamics.For example, being respectful of the children and shifting the roles of the facilitator and the documenters in response to their comfort levels and respecting the opinion of the children (storytellers and story collectors) when choosing a suitable venue and time for story collection.We found that children opened up more when the life story collection process took place in a venue chosen by the child storyteller.Moreover, all the participating children are working, and so it was difficult for them to allocate time for the project activities given their long working hours.Some of the children were free only in the evening, outside of our standard office hours.Our response was to collect stories in the evening and sometimes late at night, and whilst working at night took us out of our comfort zone, it created a relationship of trust between us and the children when they saw that we were working late hours to listen to their stories.Flexibility in our ways of working was critical for building this enabling environment.Further flexibility was required within activities and workshops, so that facilitation was not overly structured to allow children to feel comfortable and willing to open up and express themselves.Beyond our ways of working, we also involved the children directly in creating their own enabling environment.For example, during the training, the children set the ground rules for how they wanted to engage with each other.
We found that children felt more at ease sharing their stories with their peers, particularly to those with the same background.When the children talked to their peers, they felt more comfortable than adult team members and as a result the rapport building was faster.This was aided by children being involved in identifying storytellers and so collecting stories from peers they already knew.However, we also found that children storytellers were uncomfortable talking to a child story collector of the opposite sex.As a result, sometimes the storytellers and the story collectors had to be adjusted accordingly.Evidence of the effectiveness of the peer-to peer approach came through the narrative analysis workshop where we found that children who had been engaged over a longer period could explain the process and support children who were just beginning their engagement.This further helped the adult facilitators to step back and allow the children to take the lead in the analysis process.
Reflecting further on this deep experience of engaging children in analysis we as adult facilitators had to shift our own unconscious biases about the children's ability to collect and analyse stories.Despite our experiences of working with children, we still, initially, were not sure about the children's capacity to collect and then analyse stories.We experienced 'thinking paralysis' around how the children would be able to do this work.We were able to overcome this gradually through taking time to reflect on the process and our learning along the way and having critical discussions in various platforms such as team debriefs and afteraction review workshops.But seeing and experiencing directly the children working in practice was the most crucial aspect of shifting our own beliefs.Our journey has been an experiential learning journey, and we acknowledge that there is more to learn about childcentredness, such as how to do better with including children with disabilities.

Building rapport and trust
When we started working in Dhaka neighbourhoods in 2020, Bangladesh was going through a turbulent time due to COVID-19.Many children had lost their jobs as a result of the pandemic and the economic conditions of families in general were unstable.For us it was challenging to undertake a research-based project with people who were suffering from the impact of COVID-19 so directly because in this early stage of work the tangible impact for participants was not clear.A focus on building trust we feel was central to overcoming this challenge and enabling us to complete the process successfully.Our focus was not only on working with children, but also in building trust with other people in the neighbourhoods including their parents and guardians.We made sure to remain in contact throughout the implementation period.Children were also compensated for their time which is likely to have contributed to their motivation to engage.Finally, setting up an office in the neighbourhood provided a safe physical space for the children to talk, and to enable the team to have a presence which was evidence of our longer-term commitment to the community.
Trust building was an explicit step that initiated the LSC&A process.In the Hazaribagh area we had already engaged in scoping and so had previous connections which meant rapport already existed.But this was not the case in other neighbourhoods where we had not previously worked.However, the team took time to make just enough rapport before starting the process of story collection.Our main reflection on building trust is that a participatory research team first needs to familiarise itself with the local reality and people.We engaged in casual talks with many stakeholders of the community (including house managers and slum leaders) as well as the children involved in WFCL.Building this rapport informally with gatekeepers later facilitated access to the community to identify storytellers.Evidence of this rapport was that house managers themselves provided appropriate venues for story collections and sometimes helped us to find storytellers.Sometimes, when the parents of the storytellers were skeptical of our motives, it was the community stakeholders who convinced the parents to let their children talk to us.
Consent-taking is an integral part of trust building.We made certain that children gave their active consent to take part in the story-telling process.To do so, the researcher read out the consent form, taking the time to explain the research's objective and giving children the option of participating or not.To guarantee that children who cannot read or write were able to participate, the researchers requested written or verbal agreement in the presence of their parents or guardians.The consent taking process helped build trust both with the parents and the children.As expressed by one of the parents, This is an unusual experience for us to see how seriously you follow and respect the consent process.Normally people like you visiting us do not bother much with our formal consent and respect our consent.They get straight to business as it is taken for granted.
As noted above in relation to peer-to-peer work, the children who were engaged early in the programme, helped us to build relationships with the parents of new children.At the narrative analysis workshops for example, children even helped us in the consent-taking process.
While we expected trust building to take time, we experienced how trust deepens through time in different ways for different people.We noticed that initially there was mistrust towards the team because they were unknown in some neighourhoods and there was a sense of "why would we talk to you?" Given the sensitivity of the issue of WFCL there was fear of losing business and jobs as outsiders came in to talk about child labour.Whilst we saw a shift over time, distrust did not shift for everyone.Some children did not engage because they were afraid of their employers getting angry, they were experiencing pressure within their household, or they did not want to talk openly in front of their guardians.We experienced a greater level of mistrust amongst male children.

Safeguarding and providing mental health and psycho-social support services
Ensuring effective safeguarding and mental health and psycho-social support services (MHPSS) is a fundamental pillar of child-centered working but has been even more important during the COVID-19 ('new normal') pandemic.Throughout the process we followed national public health guidelines for in-person meetings, including social distancing and travel.In addition, we also contacted all children beforehand and discussed the risks they would undertake when choosing to participate.
We received training on safeguarding and MHPSS, to prepare us to respond to children should they break down emotionally, as well as being better able to understand and manage our own emotions.Further, the social worker provided hands-on support to children who experienced an emotional reaction.A follow-up mechanism was also put in place to visit the children a few days after story collection, to check on their wellbeing.Reflection and learning workshops helped us to constantly reflect on how to apply the safeguarding and MHPSS principles in practice and improve on these processes in real time.
During the small map workshop, we noticed that some children became sad while reading and re-reading the stories, as some of these stories included very personal and difficult lived experiences they could relate to.We introduced an MHPSS debrief tool to run a quick check with children to understand their feelings and be able to act accordingly.The process included pausing, walking outside and offering a drink of water, and we found that this relieved the sense of sadness in some children.Our learning is that safeguarding mechanisms are crucial to the LSC&A methodology when working with stories that include trauma.

Documentation to support reflection and learning
High quality and systematic documentation of Action Research processes ensures the process is recoverable, which is central to validity claims in Action Research.Not necessarily all Action Research processes, however, are documented.The systematic and detailed documentation process is an innovation in our Action Research programming making it a key priority throughout the LSC&A processes.We have used a set of documentation tools for each step of the process.The data collected through use of the tools supported the afteraction review workshops, consolidating learning in real time.We learned that being intentional with documentation and reflecting in real time supports quality.The team reflected that the documentation processes set up were easy to use.They also expressed that documentation from the reflection meetings helped them change their existing strategies or adopted new strategies in the field such as improving the consent forms, setting up a new field office in Bhairab and Hemayetpur and being flexible with roles during story collection.
While the documentation was systematic and intentional it was not without challenges.While there are dedicated 'documenters' as part of the team, given the need to work flexibly and the intensity of the implementation process documenters sometimes completed different roles on top of their formal documentation role, making it hard to complete thorough documentation.Furthermore, the number of templates in the documentation toolbox meant that documenters felt restricted.This limited their ability to 'think outside of the box' and react to what was happening on the ground and produce documentation that may have been more appropriate to the situation.Balancing the needs of systematic documentation with flexibility and creativity is something the programme will explore in the next phase of Action Research.

Working towards disability inclusion
This was perhaps the biggest area of weakness in our process.We lacked intentionality in the early stages when it came to ensuring the participation of persons with disabilities, resulting in few participants with disabilities.We also experienced challenges with ensuring meaningful participation when individuals with disabilities have been included.For example, we invited one child with a disability to the narrative analysis workshop but then struggled to ensure the child's meaningful participation.We reflected that there was a lack of sufficient technical expertise required to ensure effective participation of children with disabilities.As most of the children were selected through a snowballing process, we realised that it was not always possible to access children with disabilities due to their restricted participation in society.Many of them also do not have opportunities to work, which limited our options to access them through our visits to the workplaces and our criteria required children to be engaged in WFCL.We became aware of this gap in the middle of LSC&A processes and in response we included critical learning questions in the reflection workshops to unpack our learning around disability inclusion.This led to a programme wide ongoing discussion on how to meaningfully include children with disability resulting in a more intentional strategy for disability inclusion in the upcoming processes and training to build specialist capacity in the team.

Operational implications
The purpose of sharing our learning through this note is to support and inspire others designing similar processes.The phase-by-phase description of the processes provides the detail of how we have designed, prepared for and implemented the LSC&A methodology to inform new designs.
Operational implications from our learning about the time it takes and the ways in which trust is built early on and throughout the process are significant.The care required to build trust is compounded further when working on sensitive issues such as WFCL where mistrust towards outsiders exists amongst the participating communities.There is a need to take extra time and build contextually specific strategies to work with a wide range of different groups in the localities in order to explain the processes and the purpose of the work at the outset, and to then continue to engage with local gatekeepers, parents, business owners and the children themselves.
Sharing, listening to and analysing life stories has the potential to re-traumatise participants, making it therefore essential to have strong safeguarding and mental health and psychosocial support processes in place and detailed risk assessments and mitigation plans.The inclusion of a social worker and specialised training for the facilitation team are important operational preconditions to managing risk.
Engagement of participants early in the design of specific moments, such as the analysis workshops not only makes the process more participatory, but also contributes to building trust, participants opening up to share their own experiences, and ultimately generating good quality stories and rich, nuanced analysis of them.Finally, building intentional and ongoing feedback processes, such as embedding feedback forms and taking other opportunities to inquire into the experiences of participants (e.g., focus groups, interviews, reflection sessions) enables real time learning for flexible working.If teams can embrace that they are on a learning journey themselves, then this real time reflection can support a shifting of unconscious biases that can hold us back from more transformative work.

Figure 1 -
Figure 1 -A journey of CLARISSA LSC&A activities generated using the river of life tool

Figure 2 -Figure 3 -
Figure 2 -Small map of story collected by children

Figure 4 -
Figure 4 -Large system map from the neighbourhood locations.
conflict and violence, family separation and negligence to children Workplace environment and child abuse Financial management, savings, and debt Creating more opportunities for education Long working hours Health hazards in workplace Physical and mental abuse in workplace Lack of workplace safety Parents' pressure on children to join workforce Pressure to repay debt Drug addiction Child marriage Corporal punishment at school Inability to afford the cost of education Inability to afford the cost of treatment Author(s) information and contact details.Mashrique Sayem, Research & Documentation Officer, Terre des hommes in Bangladesh, Email: mashrique.sayem@tdh.chSukanta Paul, MEAL Coordinator, Terre des hommes in Bangladesh, Email: sukanta.paul@tdh.chDr Marina Apgar, Research Fellow, Participation, Inclusion and Social Change Cluster, Institute of Development Studies, Email: M.Apgar@ids.ac.ukDr Mieke Snijder, Research fellow, Participation, inclusion and social change cluster, Institute of Development Studies, Email: M.Snijder@ids.ac.ukThis report has been funded with UK aid from the UK government (Foreign, Commonwealth & Development Office, formerly the Department for International Development).The opinions expressed are those of the authors and do not necessarily reflect the views or policies of IDS or the UK government.This is an Open Access paper distributed under the terms of the Creative Commons Attribution 4.0 International licence (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are credited.T +44 (0) 1273 606261 T +44 (0) 1273 606261 E clarissa@ids.ac.ukE ids@ids.ac.ukW clarissa.global W www.ids.ac.ukT @ChildLabourActn T @IDS_UK F #ChildLabourAction F facebook.com/idsuk

Table 1 -
Demographics of storytellers

Table 2 -
Demographics of the child story collectors (Source: Authors' own)

Table 3 -
Demographics of the story analysers (Source: Authors' own)

Table 4 -
Action Research group themes identified from the large system maps