RESEARCH DESIGN AND METHODOLOGY 4.1 Introduction
4.3 Theoretical framework
4.4.2 The researcher as a data-gathering tool
It was typical of this research study that I assumed a data-gathering role because in qualitative research, the researcher becomes the data-gathering instrument. In field studies, data are generated through the researcher's participation in a social setting (Allen 2004:7). In quantitative research, questionnaire surveys are the preferred methodological tool, whereas in observational studies, the field worker becomes the research instrument (Ibid). Information on FA provisioning in schools was gathered by me as the data-gathering instrument through interviews, observation and FDGs.
4.4.3 Data collection methods and tools
According to Jenne and Green (1976:33), data for evaluation are collected from existing records or by use of a wide variety of instruments. For this study, data were basically gathered through observation, face-to-face interviews with school managers and FDGs with learners and teachers.
Face-to-face structured interviews were held with selected school principals, FA appointed persons, pre-school managers and education officials responsible for health and safety at the provincial offices of the GDE. Babbie and Mouton (2001) argue that FDGs and interviews have high face value, as one can observe body language. I preferred to use interviews and group discussions instead of questionnaires for the reason that the former allowed me to observe gestures, body language and facial expressions that I could not get from questionnaires. I could also follow up on answers immediately, something that would not have been possible with questionnaires. If I needed further clarification on a point, I could ask the participants to elaborate. In interviews I could also observe non-verbal cues and gestures. The research questions demanded that I had to take the qualitative route because the evaluation of the administrative efficacy of the FA provisioning according to NEPA provisions could not be quantified, but only described using words.
According to Polkinghorne (2005), data sources for qualitative research are interviews, observations and documents. Interviews sought to elicit information that was subsequently analysed in the light of the research topic. Focus group interviews with learners were carried out in a classroom and a gymnasium hall after school hours so as
not to interfere with teaching and learning. Focus group discussions with teachers were held in the staff room after school hours. Permission from parents was sought by means of assent letters in the case of minors. Deviations from the interview schedule were allowed to accommodate individual situations and circumstances. Probing questions were used to dig for detailed information.
The primary sources of data were audio-recordings of interview transcripts, FDGs and observations. It was crucial that the data should be in a form that enabled me to understand and interpret it for subsequent coding.
4.4.4 Population and sampling procedures
It was not feasible to include all schools, learners, teachers, school managers and principals in Gauteng province. The researcher purposefully selected six different schools for study. The sampled schools ranged from Grade R to high school. The study population comprised all pre-schools, primary and high schools in Gauteng province.
Two pre-schools, two primary schools and two high schools were selected for study, using the purposive, or judgemental sampling method. Schools that were information-rich were selected for study.
Two pre-schools were selected for this study. There are different types of pre-school ownership models in Gauteng province. I selected one privately owned pre-school for my study because many of the deaths and injuries reported in the printed and electronic media in Gauteng province occurred in privately owned schools. The other pre-school I selected was well-resourced. This pre-pre-school belonged to a well-established public institution of higher learning and was established for the sole purpose of providing day-care for the children of the institution’s workforce. The pre-school was run by a principal who was accountable to a board of governors. Pre-school ownership in Gauteng province falls into one of these two categories. The GDE does not own pre-schools per se but regulates and provides space for pre-pre-schools in public pre-schools.
Two primary schools were selected for this study. The first primary school was a former model “C” public school catering for middle-class and affluent learners. It is important to note that the post-1994 migration of learners from township schools to the former model
“C” schools for better education has resulted in these schools enrolling learners from
poor backgrounds as well. Historically, public schools in the South African landscape were grouped into quintiles, depending on location and how well-resourced they were.
During the apartheid era there were five quintiles, with quintile one being the poorest schools and quintile five being the affluent schools. Former model “C” schools catered for white children during apartheid and were generally better resourced than the township schools. According to Govender (2015:4), “prior to the new dispensation in 1994, the term ‘Model C’ was coined to describe former semi–private and mostly white schools that fell under the control of the House of Assembly.” These schools received less funding from the government but had greater autonomy. Although the term was scrapped, it continues to be used to describe wealthy schools (Ibid).
I selected the former model “C” school to serve as a model case of FA provisioning level because of its relative affluence when compared to township schools. The second primary school selected was from a township. The township primary school was purposively selected as an extreme example at the other end of a continuum to provide a comparison with the relatively well-resourced former model “C” primary school. This township school was given the label “D”.
Two high schools were selected for this study. One school was an exclusive “top-notch”
private school and the other one was a former model “C” high school. The former model
“C” school was identified as school “E”, while the affluent school was labelled “F”. The different schools were selected for the sake of comparison and to ensure adequate coverage of the different types of school in the GDE’s landscape. Gauteng province is a relatively “urbanised” province with fewer rural or farming community schools. The province is the economic and commercial hub of South Africa and I did not select a far-flung rural or farming community school for this reason.
4.4.5 Observations
According to Smith, Sinclair, Raine and Reeves (2005:165), there are two main types of qualitative methods that may be useful in health care evaluation: observation and in-depth interviews. Observation techniques seek to examine what people do as well as what they say. Qualitative observation may be non-participant or participant; it may also be covert or overt. I did not opt to do participant observation because it raises “important
ethical considerations as the participants are usually unaware that they are being observed” (Smith et al 2005:165). For this study, I used non-participant observation.
Non-participant observation describes the process of visiting one or more places where health care is occurring and observing what patients and staff do and say (Smith et al 2005:165). It was covert non-participant observation “where the people being observed were unaware of the observer” (Ibid). This refers to a situation where the researcher simply watches without being noticed. I did not opt for overt observation where the subjects are aware that they are being observed because “if the subjects are aware that they are being observed they may alter their behaviour, perhaps from increased self-consciousness, or to present themselves in a favourable light; or to adhere to official rules of comportment” (Ziebland & Wright 1997, in Smith et al 2005:166). I observed learners during break and during sport activities in the afternoon at the particular school where I was teaching. I could not do the same observation at the other schools for the reason that I had to be on duty during working hours. I could only observe the provisioning of FA by learners and teachers during tea-breaks and sport events. I had obtained permission to observe learners from the principal and the chairperson of the SGB. All the observation parameters were set out in the observation protocol (Appendix G).
I happened to be involved with the school’s football team as a manager and occasionally travelled to other schools on soccer tournaments. This gave me an opportunity to observe FA provisioning during such tournament matches as well. After obtaining permission from the school authorities and the SGB I observed FA activities over four months using an observation record sheet. I included a specimen of one month’s observation record findings in the next chapter on research findings. A copy of the observation sheet is also included in this thesis (Table 5.5). During the single tea break in the school, learners were restricted to the school area between classrooms and the school soccer fields, netball and basketball courts. It was part of teachers’
responsibilities to take turns to monitor learners during the tea breaks in what was called ‘ground’ or ‘break’ duty. I had to move around this area recording the provisioning of FA.
Sport activities took place after school hours. The sport disciplines that I observed at the school included football, netball, hockey, basketball and cricket. Observation as a research tool entails consideration of a range of factors (Allen 2004). I chose to assume a detached role rather than an ethnographic one. Established on the relative merits of familiarity and detachment in observational research, the ‘insider’ position is founded on the case for developing an affiliation with research subjects in order to ensure an authentic account while the ‘outsider’ position is built on the claim that a detached stance can ensure freedom from bias arising from too deep an immersion in the field (Bonner & Tolhurst 2002, Gerrish 1995, Reed 1995, in Allen 2004). This was necessary to avoid bias because I was a member of the school’s teaching staff. An insider participant observer can sometimes be desensitised to those features of the setting that would be interesting to others with no prior association. This insight is encapsulated in the notion of researcher reflexivity, which recognises that rather than devising strategies to limit ‘bias’ it is more productive to work towards a better understanding of the fieldwork role and its impact on the research findings (Allen 2004).
Evidence of careful attention to the fieldwork role and analytic sensitivity to how study data were generated was an important quality indicator in my observations. There was a need to separate what I knew from what actually played out on the playground. My own background is in teaching and relationships in the field therefore required careful management in order to avoid the pitfall that I was more closely aligned with what I knew as opposed to what had actually happened in the field. Managing such ‘closeness’
or ‘detachment’ was consciously attended to and accomplished by the way I talked and acted (Hunt & Bedford 1994, in Allen 2004:10). There were incidents of FA that had happened before I embarked on my study that I excluded from my observation simply because they happened outside the timeframe of my research study. In analysing the data, I did not allow my prejudices to influence data findings in other schools. The research findings moulded and informed the study.
During observations I did not suggest or influence FA provisioning actions. Another issue that warranted attention in planning the observation role was the extent of participation in the setting. In my observation protocol, I had stipulated the parameters
of what I was going to do during the observation. No camera or video recorder was to be used during observation for ethical reasons. My observation was limited to transcription and recordings on a sheet of paper (Table 5.5). In this study I wanted to assume a complete observer role by not taking part in the FA provisioning except where the situation demanded that I intervene to act in an emergency as demanded by my duties and responsibilities as a teacher acting in loco parentis. I was also careful not to
“cherry-pick” FA incidents that validated or confirmed reviewed literature findings.
Methodology textbooks often describe different modes of participation in terms of Gold's (1958) continuum, which extends from complete observer to complete participant (Allen 2004). In actual practice, however, within any one study, a researcher may adopt a number of positions along this spectrum, depending on the aims of the research and circumstances in the field. I sometimes participated in the work, but on other occasions I adopted more of an observer role, positioning myself in a strategic spot to observe the ebb and flow of playground life, observing discrete activities, such as FA administration or patient processing (Allen 2004). It became problematic where I was expected to let the incident play out without my interference for the sake of my research on the one hand, and the need to give FA in fulfilment of my obligatory duties as a teacher acting in loco parentis on the other. The observer's level of participation in a setting or social group has important implications for the type of field data it is possible to generate and consequently the analytic possibilities of the research (Ibid). To achieve this, it was necessary to generate detailed field notes of naturally occurring interaction. The findings on the observations were illustrated by Figure 5.1 in the next chapter on the analysis of findings.
According to Allen (2004:11), observation, which he prefers to call fieldwork, is labour-intensive, demanding of resources, and requires a degree of flexibility and responsiveness on the part of the researcher. Doing observation ‘at a distance’ presents some interesting personal and intellectual challenges that are rarely discussed in the methodological literature and that need to be built into all aspects of the research design (Ibid). Field researchers, however, are often especially reluctant to impose prior assumptions on the people or settings being studied, particularly when researching issues or social groups that are poorly understood and/or relatively unexplored,
preferring instead to allow phenomena of interest to be uncovered (Allen 2004:11). This important factor formed the hallmark of my study.