CHAPTER 3 RESEARCH DESIGN
3.5 Pilot One Phase
any siblings or twins etc. This is relevant as it indicates the amount and quality of language input that the child gets.
While the UK CDI asks, “How many siblings does your child have (include full and half siblings)”, Setswana language differentiates between older and younger siblings (bomogolowe and bomonnawe respectively) and the Setswana culture has no notion of a
‘half’ sibling. The question was adapted appropriately.
The child’s mother and the child’s father
This section asks about the parents’ age, marital status, education levels and employment status. This information is important both as factors of SES, and possible indicators of language input.
The education system of Botswana is structured differently from that of the UK and even South Africa, so while the UK CDI has options like “GCSE/O level/NVQ level 1 or 2/similar; A level/NVQ level 3/similar; university degree/HND/HNC/NVQ level 4 or 5/similar” for the parents’ highest level of education, in the Setswana CDI it was adapted to
“primary school/standard 7/PSLE; junior secondary/form 3/JCSE; senior secondary/form 5/BGCSE etc”.
The household
The final section is important for determining the SES of the child’s family. It asks about the income and expenditure (amount spent on groceries) of the home. The options for amount of money for income were divided into categories that are aligned with the tax brackets recognised in Botswana. These were combined with other variables such as the parents’ education levels and the number of rooms in the house, to measure the SES of the family.
differences in infants from an urban area (a city) and a semi-urban area (a peri-urban village).
Considering the urban-rural variation found in Botswana and the implications this divide has on language acquisition, the sample in this study also allowed us to explore if a similar pattern of urban/rural differences could be found between urban/peri-urban contexts.
3.5.1 Setting
The selected urban area was the capital city of Gaborone, and I selected a clinic in close proximity to my home to conduct the study. The peri-urban area is a village about 30 km north of Gaborone. It was selected out of convenience as I was connected to one of the managing nurses at the local clinic.
I chose to use a clinic setting because the age range for this phase is not school going children and so a pre-school setting would not be suitable. Children in Botswana are entitled to services provided at Child Welfare Clinics (CWCs) across the country, from birth to the age of five years. According to official records, about 80% of Botswana’s children go to Child Welfare Clinics, and these services are provided free of charge which means all children are eligible regardless of SES (Govt. of Botswana & UN in Botswana, 2010). For these reasons, a clinic setting would allow me to find a large enough sample within the target age range, from a variety of socio-economic backgrounds, considering the geographic variable.
3.5.2 Ethical approval and recruiting
The SA-CDI team received ethical clearance from the Linguistics Ethics Subcommittee at the University of Cape Town, which covered this Setswana study (Appendix C). As I was going to be based in state clinics, I needed approval from the Ministry of Local Government and Rural Development in Botswana, which I obtained through our collaboration with the University of Botswana (Appendix D). Then, for each clinic, I had to obtain permission from the relevant District Management Health Team (DHMT) (Appendix E and F). I also went to introduce myself to the matron in charge at each clinic and explained the research I was hoping to conduct in their clinic. Although I did not need any formal permission from them,
in the culture of Botswana, the notion of showing botho6, i.e., showing respect in such a manner to elders and authoritative figures, is highly valued, and asking the matron for their blessing to conduct my research in their clinic is a sign of respect which facilitated a welcoming and cooperative environment in which I was able to work.
Once I had gained all these levels of clearance, I could then start collecting data at the clinics. I spent three working days at each clinic and had 15 physical copies of the
questionnaire with which to gather data. At the start of each day in the clinic, the nurse in charge of the CWC would introduce me to the parents sitting in the waiting room and give them a summary of my research study. Parents that had a child aged between eight and 18 months and who had an interest in the study would then indicate so, and I’d approach them individually to explain the study in more detail and show them the questionnaire. Thereafter they could decide if they wanted to participate in the study and sign a consent form
(Appendix A). Participants would either fill out a questionnaire for themselves, or I would sit with them and fill it out for them in interview format, depending on their preference. Of the 30 participants recruited, 22 opted to complete the form independently (three of whom would later be excluded; see section 4.1) It took between 45 and 90 minutes to complete a
questionnaire, and it was done while they were waiting in line to see the nurse or after they were done with the nurses.
Due to the academic calendar, the timeline I had to complete data collection in
Botswana was approximately one month. To receive ethical clearance from each DHMT and the two relevant matrons took approximately two and a half weeks. Therefore, I was unable to return to the clinics and recruit more participants after excluding some during data processing.
3.5.3 Data analysis
The CDI tool as it was used in the pilot phase (Appendix H) had four major sections.
These were the Phrases, Vocabulary, Gestures, and Family Background Questionnaire. They were compiled into a paper document for data collection, which meant that once it was collected, the data needed to be captured. For this a database was built using Microsoft Access7, and then a research assistant8 manually captured each questionnaire into the
6Also known as ubuntu: the African philosophy of humanity, literally defined as “a person is a person by other people” or ‘motho ke motho ka batho’.
7 Thanks to Yolandi Ribbens-Klein for the assistance
8 Thanks to Boitumelo Matlakala for the assistance
database. The responses were coded numerically into the database. These codes were either in binary to represent ‘yes/no’ responses such as in the phrases and gesture sections, or sequentially such as for ‘doesn’t know/understands/understands and says’ responses in the vocabulary section and the options in some questions of the FBQ.
The data was then analysed statistically using Microsoft Excel and IBM SPSS Statistics 27 software. Various types of descriptive statistics were used to measure frequency, central tendency and variation in the data. Inferential statistics were also used, namely Pearson or Spearman correlations, to measure the relationship between the different variables.
Qualitative content analysis was also used in some instances where patterns seemingly started to emerge between certain variables, but due to the small sample size a statistical analysis may not have yielded significant findings.