• No results found

Research Sample and Sample Size

METHOD OF DATA COLLECTION FOR THIS RESEARCH

3.8 Research Sample and Sample Size

Initially, I had planned to contact and interview between 400 and 600 respondents for this research. But I soon discovered that this was too ambitious for a qualitative research project at doctoral level. Apart from the difficulties of obtaining these interviews, it would take a long time to actually speak to so many people, and to transcribe and analyse so much data. It was therefore more reasonable and realistic

to aim for about 100 interviews and FGDs (See Table 1) in the areas mentioned considering the time frame needed to complete this research.

Respondents were basically identified in relation to their position within society, and mainly in relation to inter-faith marriage, markets, professional groups, politics as well as religious groups and traditional institutions. The location of respondents and their source of livelihood are of central importance because it forms part of the central thesis of this work, particularly in regard to networking in conflict and post conflict Jos.

Table 2: Total Number of Respondents by Location and Topics Discussed

S/N Categories Location Total 4 Political Party Leaders Jos, Makurdi, Nasarawa,

Kaduna

8 Non-Governmental Organisations Jos 6

Total 102

experts within’ (Tongco, 2007: 147). This method is very effective in sampling respondents who are thought to be of direct relevance to the research. Tongco (2007:147) points out that:

The purposive sampling technique, also called judgment sampling, is the deliberate choice of an informant due to the qualities the informant possesses.

It is a non-random technique that does not need underlying theories or a set number of informants.

The adopted sampling technique can be described as ‘a random selection of sampling units within the segment of the population with the most information on the characteristic of interest’ (Guarte & Barrios, 2006: 277). This procedure is quite effective in qualitative research because it helps the researcher to choose subjects carefully and ensure that they only talk to those with knowledge relevant to the research topic and thus not waste time. I also used the snowball sampling method alongside the purposive sampling method, by taking note of names and sources of information mentioned by other respondents. Tran & Perry (2003) point out that

‘Sometimes snowball sampling involves asking an informant to suggest another informant.’ Tongco (2007: 153) explains that “this method might not be free from bias…” (See also, Lopez et al, 1997; Seidler 1974; Smith, 1983; Zelditch, 1962). So it is important to state the bias clearly when the results are analysed and interpreted so as not to mislead people into inferring general conclusions (Bernard, 2002; Godambe, 1982; Snedecor, 1939). Despite its inherent bias, purposive sampling can provide reliable and robust data. The strength of the method actually lies in its intentional bias (Bernard, 2002; Lewis & Sheppard, 2006; Poggie, 1972; Tremblay 1957).

As a way of ruling out any form of bias and misrepresentation, incoherent and unclear responses were cross-checked and respondents contacted in a follow-up visit in August and September 2014. This follow-up visit enabled me to clearly understand the sense in which certain responses were made at a time and particularly to present the arguments exactly in the way they were intended.

Respondents for this study were carefully selected to respond to the central areas of research that emerged as important for women’s everyday lives, namely marriage, market relations, politics and the formal sector. The table below shows the categories and the number of interviews and focus group discussions conducted under each theme. As set out above, most of the interviews were recorded and transcribed. My field notes also served as an important part of the interview process, and I referred to them when I reflected on the research and wrote up my findings. As information obtained through interviews may not always be accurate, I paid attention where necessary to having the right contact points and triangulating the data, bearing in mind the confidentiality of the respondents and in line with the ethics I mentioned above.

Conflicting kinds of response are expected in humans when they relate in a group.

Krueger (1994: 31) pointed out that “human conditions can be distorted intentionally or unintentionally. People are not always truthful, and sometimes they give answers that seem best for the situation. Other times people hold back important information because of apprehensions or social pressure. Experts who work with small groups testify about the unpredictable nature of groups, and that group leaders or moderators can skilfully or unwittingly lead groups into decisions or consensus.”

This is illustrated by an experience I had at the Farin Gada market when the leader of the market tried to ensure that he controlled the discussion. He clearly said that the conflict had never affected the market. However, when I continued with the FGD and encouraged a wider range of respondents, including women, to speak, other views emerged. One of the female respondents in the FGD said that they had often narrowly escaped death and would have been killed if it had not been for the intervention of some of the people who traded within the market. Bearing in mind the complicated nature of interviews and of what people considered to be the truth, my collection of the data and my understanding of it both relied strongly on my own knowledge of local forms of behaviour and social interaction.