The process and design of the study will take shape as guided by the following sub-topics.
3.3.1 Sampling
Two orphanage schools in rural KwaZulu-Natal were selected for the study. Purposive sampling was used to select three learners, three teachers, a resident social worker, a member of school management and two caregivers. A total of eleven participants was purposively selected. In School ‘A’, individual interviews were held with three teachers, a social worker, and three learners. In school ‘B’, interviews were held with the deputy principal, one teacher and one caregiver. They all have first-hand experience and information in dealing with vulnerable children.
3.3.2 Data generation methods and tools.
Coghlan (2019) differentiates between data gathering and data generation. Data gathering is discreet, and the researcher is detached or inconspicuous and engages in a remote and unobtrusive observation or document analysis which all may be done from outside the research setting. However, data generation is participatory in nature and becomes a form of intervention (Terre Blanche, et al., 2006), , as both the researcher and the participants are made aware of the issues through engagements, and are invigorated to think about actions to change the situation. Data was generated through three successive levels. This was done to carry out a meaningful ethnographic investigation into the lives of all members of the school community as a learning ecology (Myende & Hlalele, 2018) Ethnographic study has its origin in sociology and anthropology (Barak, (2017). The three instruments of data generation are: Observation, Document analysis and one to one interviews.
3.3.2.1 Observation
Observation is a data generation technique whereby participants interact within their natural setting. Observational data collection strategy is an integral part of conducting an ethnographic study where lived realities are explored (Creswell & Creswell, 2017). Observation becomes a handy tool to investigate roles, relationships, ethics and values. Participants are observed
naturally working in teams, forming teams as they play games, and solving problems and making decisions (Jorgensen 2015). Observation is useful even for the purpose of confirmation and complementation or rejection of research results obtained using other methods (Powney &
Watt, 2018; Jamshed, 2014). One of the orientations of ethnographic observation is that it has its origins in sociology and anthropology, and is therefore, useful in studying phenomena of cultures such as culture of poverty and school culture, which were also investigated in this study (Merriam & Grenier, 2019).
3.3.2.2 Document Analysis
The second level of data collection was done using the technique of document analysis (family history, progress reports, health records, psychological reports). Data analysis is a systematic process of literature review and relevant documents to glean useful information (Botha &
Kourkoutas, 2016; Aspers & Corte, 2019). The data corpus was augmented and much enriched by document analysis. The first step was to go through the archived documents for each participant learner, which included their life history documents, previous psychological reports and academic progress reports. Then, each learner was asked to write a short essay about their life experiences. The authenticity of information recorded in the documents (files) was guaranteed as various health and social practitioners recorded results of scientific findings and objective health reports aimed at improving the participants’ well-being without bias in any way (Bowen, 2009; Bernard, Wutich & Ryan, 2016).
3.3.2.3 Interviews
All the participants were individually interviewed. This was an eye-opening process for the researcher as preconceived ideas, half-truths and myths were replaced by reality. Three sampled learners also gave accounts of their lived experiences. Case study research methods were used to conduct open-ended interviews as well as document analysis. A case study is flexible and allows a researcher to collect data using a range of methods (Thomas,(2015).
2011). It is particularly suitable for gathering data on life experiences occurring over a long period of time. The interviews were conducted using The Interview Schedule (Appendix 6).
Though the questions were pre-set, the interview schedule served only as a guide as the interview was allowed to flow freely with the researcher probing to get more information. The interviews were recorded and later transcribed. This methodology was deemed appropriate as it enabled the participants to freely express their views and feelings. The researcher
familiarized himself with the data and identified themes pertaining to past experiences of adversity and possible “bounce back” themes as well as developmental risk factors, protective factors and adaptation factors (Donald et al., 2010).
3.3.3 Data analysis methods
Data analysis was conducted in three coding levels. The first was a description of experiences and events as narrated by the participants. The second level was to cluster descriptions from the first level into themes. The third level involved thematic analysis; a method for identifying, analysing, and reporting themes within data (Botha & Kourkoutas, 2016, p.788).
Thematic analysis works well with theory to explain and predict various outcomes of research within a context. Clarke and Braun (2014) say that it works hand in glove with social constructivism to deepen the investigation of meanings and experiences within a context. The data was also critically distilled for findings which would directly or indirectly lead to participants self-reflection. Six phases of thematic analysis, adapted from Clarke and Braun (2014), were used. Details of data analysis are given in chapter 4.
3.3.4 Trustworthiness
The selection of participants, the natural context and the data gathering methods adopted in this study ensured credibility. The researcher enhanced credibility by adopting data triangulation of data generation methods, thereby providing confluence of evidence (Bowen, 2009).
Transferability can be enhanced by fully describing the research context. The contexts in this study were fully described. This is done so that the results can be applied to similar contexts The research is dependable because the data was authentic as the participants were drawn from various categories of professionals working with vulnerable children. Efforts were made to minimize any individual participant or the researcher’s bias in collecting and interpreting the data to ensure evaluative validity (Gay et al., 2012; Wahyuni, 2012).
3.3.5 Ethical Considerations
Participants took part in the study voluntarily and no incentives were promised. Ethical clearance was obtained from the UKZN Research Office (Appendix 5) and permission to conduct the research was obtained from all gatekeepers including the Department of Education (Appendix 4) and school management (Appendix 3). Informed consent forms were signed by
all participants (Appendix 1). Participants’ rights, beliefs and cultures were respected in that all questions were sensitive to their culture, and all participants were informed that they had the right to withdraw from the study at any stage if they felt uncomfortable. Participants’
privacy was protected, and their real names were not used. Interview questions were sensitive to private issues and the researcher refrained from soliciting graphic details of violence and/or abuse. A social worker was available to address any issues that could arise when the researcher engaged with minors. No traumatic issues arose. Data remained confidential as it was not shared with anybody except the supervisor and pseudonyms are used for the schools and participants (Creswell & Poth, 2017). Data was not made available to, or discussed with, anyone except the researcher’s supervisor (Tracy, 2019).
3.3.6 Limitations of the study
Given that the study was context bound, it was not possible to be widely generalizable (Silverman, 2011). However, Thomas (2011) states that an exploratory case study can be taken further to explore other case studies and produce context-dependant knowledge. Some of the participants did not speak or understand English very well, but the issue of language was not much of a barrier since both the participants and the researcher could switch to isiZulu.
Notwithstanding different experiences and perspectives between the researcher and the participants, very minimal information could have been lost during conversations (Hammersley 2017).I corroborated findings across data sets by analysing data collected using a triangulation of methods to reduce the effect of possible biases that can occur if the study had a single data generation method (Bowen, 2009). Qualitative research strives to collect, integrate, and present data from a variety of sources of evidence as part of any given study.
The generalization of findings was not possible because only a small sample of participants all from a similar cultural background was used.