Sampling procedure Sample selection

According to Lohr (2009), a sample is a subset of a population. A perfect sample of the above, as the aforementioned notes, would be a scaled-down version of the population, mirroring every characteristic of the whole population. Thompson (2012) notes that sampling consists of selecting some part of a population to observe, so that one may estimate something about the population from which it was taken. The multiple sampling method was employed in the present study in order to allow for the triangulation of methods. Both the convenience and the stratified random sampling methods were employed in the current study. In addition, the observation method was also employed that will be discussed further on in the study.

Systematic stratified random sampling

Babbie et al. (2007:191) propose that “stratified random sampling is a method for obtaining a greater degree of representativeness-decreasing the probable sampling error”. The authors note that using such a method enables the researcher to obtain a gross estimate of the results, without incurring the amount of cost or time that is required to select a random sample, and to

ensure that relevant subgroups are adequately represented in the sample, by means of dividing the population into strata that are based on such important variables as historical racial classification. Use of the above-mentioned method helps to ensure the proper representation of the stratification variables, so as to enhance the representation involved to the other variable(s) related to them (Babbie et al., 2007). The application of this method in the study is explained later below.

Convenience sampling

According to Denscombe (2010), convenience sampling is built upon selections that suit the convenience of the researcher, and that are ‘first at hand’. In addition, Gravetter and Forzano (2011) highlight that, in convenience sampling, the researcher uses those participants who are relatively easy to access. The aforesaid authors further indicate that people are selected on the basis of their availability and willingness to participate in the study. Denscombe (2010) highlights that convenience sampling is mostly used in cases where the researchers have limited time and resources at their disposal. With the use of this method, the researcher was able to reach those respondents who attended the evening shows. This was done to ensure a fair representation of the respondents who attended the late shows. Moreover, the researcher did not have to stay until the end of the show, but, rather, until the required number was reached.

Attendees of the MACUFE event formed the sample for the current study. In terms of numbers, Hunt and Lathlean (2015) note that a required sample size must be calculated during the design and planning of a research study, rather than being guessed at, or calculated retrospectively. Determining the sample size required for the research work during the planning phase is important and necessary for the following reasons: (i) a sample that is too large may waste valuable resources; (ii) a sample size that is too small can be unethical or inappropriate, as it might not be big enough to allow for the detection of differences among the different participants (Hunt & Lathlean, 2015).

In determining the sample size of the current study, a sample size calculator was used.

Denscombe (2010:43) states that “a sample size calculator demonstrates very clearly how the required sample size varies according to what the researcher wants in terms of level of confidence and margin of error”. The aforementioned author illustrates the fact that the greater the level of confidence, and the greater the level of accuracy, the bigger the size of the sample must be. With a population size of approximately 140 000 in terms of the audience at

the MACUFE event (South Africa, 2013b), and with a 95% confidence level, the sample size for the current study was determined as being 385. After conducting a data collection, cleaning and quality check, all 385 questionnaires were completed, giving the study a 100%

success rate.

In administering the questionnaire, both stratified random and convenience sampling procedures were followed. The stratification was based on the various venues (see Figure 4.2 below) from which the sub-events of the MACUFE event were hosted. The attendees of the MACUFE event, both men and women aged 18 years and above, were randomly selected at the various venues to participate in the interviews conducted. Thereafter, the respondents were systematically chosen, in terms of which every third person encountered was interviewed. Tustin, Ligthelm, Martins and Van Wyk (2005) suggest that, if the selected respondent does not wish to participate in the survey, any available and willing respondent be selected to participate in the study, with the same process possibly being repeated. In addition, those respondents who attended the evening shows were conveniently selected to participate in the study. Data collection took place during the entire period of the festival (from 30 September to 9 October 2016). Not all the venues are shown on the map that is shown in Figure 4.2 below, as some were further afield than the map indicates.

Figure 4.2: Map of the study area, showing some of the venues where the sub-events of the MACUFE took place

Source: Modified by author from Google Maps (2017).