73 narrative, qualitative information (Marshall, 2005: 132, cited in Mugabe, 2011: 61). Reja et al. (2003: 161) explain that close ended questions limit the respondent to the set of alternatives being offered, while open ended questions allow the respondent to express an opinion without being influenced by the researcher (Foddy, 1993: 127, cited in Reja et al., 2003: 161). The consequence of this is that the quality of the survey is compromised. The advantages of the open ended questions include the possibility of discovering the responses that individuals give spontaneously, and thus avoiding the bias that may result from suggesting responses to individuals, a bias which may occur in the case of close ended questions. Reja et al. (2003: 161) continue to explain that open ended questions do however, also have disadvantages in comparison to close ended, such as the need for extensive coding and a greater degree of non-responsiveness. The researcher’s findings are that open ended questions were suggested for use decades ago, at the initial stage of questionnaire design in order to identify adequate answer categories for the close ended questions. In the later stages of the questionnaire design, open ended questions can be used to explore deviant responses to the close ended questions.
The surveys of this study utilised open ended questions to encourage free expression with regard to what the communities believe to be the prevailing issues in environmental management, service delivery, public participation processes currently employed and recommendations that can be made to improve public participation. Open ended questions were used to a significantly greater extent in the EMF team survey, with the intention to gain an understanding of the lessons learnt from the iLembe EMF. Open ended questions in the officials and stakeholders’ surveys were used mainly for recommendations to improve the public participation process. For the most part, in all surveys, close ended questions were used.
74 the extent rural communities of the geographical area applicable to the EMF are given the opportunity to voice their concerns or opinions or local knowledge and to examine to which extent community needs and concerns are reflected in the EMF. The data were collected by means of surveys, both questionnaires and key informant interviews. The questionnaires were completed in 400 households while the key informant interviews were conducted for five categories, namely the EMF team, participating stakeholders, NGO representative, IDM public participation officials and NLM public participation officials. According to Sibande (2009: 14) sample size should be determined at the design stage.
4.5.1 Sampling Framework
Quantitative research ideally involves probability sampling to permit statistical inferences to be made (Sandelowski, 2000: 248). Mujis (2004: 8) states that when data are collected in quantitative research, we have to collect them from someone or something. The people or things we collect data on or from are known as units or cases and the data collected are referred to as variables. Units or cases are also known as samples, which are defined by Ladner (2008:3) as:
A sample is a subset of the population selected by either probability or non- probability methods. If you have a probability sample, you simply know the likelihood of any member of the population being included (not necessary that this is “random)
(Ladner, 2008: 3) A variable is any measured characteristic or attribute that differs for different subjects.
Quantitative variables are measured on an ordinal, interval, or ratio scale (MacDonald and Headlam, undated: 24). Variables are any characteristic of the unit we are interested in and want to collect (e.g. gender).
According to Rubin and Babbie (2005: 247), purposive sampling is referred to as choosing a sample by regarding a population who may represent best the general opinion. In addition, purposive sampling is a deliberately non-random method of selecting participants for research, which allows individuals to be selected because they have knowledge relevant to the research. Despite its inherent bias, purposive sampling can provide reliable and robust data (Tongco, 2007: 154). Furthermore, the description given by Ladner (2008: 5) above leads to the choice of probability sampling, which allows for the results to be generalized to
75 the population and is not limited to make inferences only about the group of individuals participating in the study (Mugabe, 2011: 64; Tongco, 2007: 147). Tongco (2007: 153) states that whenever possible and deemed efficient, random or probability sampling is recommended as a means of informant selection because randomization reduces biases and allows for the extension of results to the entire sampling population. The sample for this study was therefore selected by purposive sampling in the selection of Ndwedwe and then stratified random sampling to select households within the wards identified through selection criteria. This was done in order to select communities with the indigenous knowledge required and whose views are required to satisfy the aim of this study. In this regard a systematic sampling design or mixed design was used. However, Tongco (2007: 147) warns that random sampling is not always feasible and efficient. A high dispersion of samples may induce higher costs for a researcher. Furthermore, not everybody is willing to participate, and possibly not be available.
Purposive sampling is an informant selection tool widely used (Tongco, 2007: 147). Tongco (2007: 147) further states that the purposive sampling technique, also referred to as judgment sampling, is the deliberate choice of an informant or informants 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. Simply put, the researcher decides what needs to be known and sets out to find people who can and are willing to provide the information by virtue of knowledge or experience. Purposive sampling is especially exemplified through the key informant technique, as undertaken in this study. In stratified random sampling, the population is subdivided into subpopulations called strata (Kitambara, undated: 10), as this case, Ndwedwe was subdivided into the wards. Within these strata, each has its own sample size, which refers to the 400 households randomly selected (Kitambara, undated: 10). The advantages of such a sampling method are increased precision, being flexible in the choice of the sample design for different strata and the able to get estimates of each stratum in addition to the population estimates (Kitambara, undated: 10).
While Ndwedwe was selected based on being an in-land municipality and having the highest rate of rural settlements, in order to narrow the scale of the study area, Census 2011 Interactive data in SuperCROSS was used to select wards which recurring emerged as indigent in the following criteria:
76 a) Unemployment;
b) Predominant age group between 20 – 30 years of age;
c) Low levels of education or school attendance;
d) Annual income levels e) Predominant language;
f) Service delivery, where focus was placed on electricity, water and sanitation.
These criteria resulted in the selection of wards three, twelve and seventeen for the study area.
4.5.2 Ethical Considerations
With regards to ethical considerations, as this study was interested in the input from both youth and the elderly, ethics were abided by, by ensuring that respondents were older than the age of eighteen. Furthermore, the facilitators were briefed and requested to treat all respondents with respect, accepting their responses without insistence if they wished to decline to partake at any point in the survey. Respondents were not questioned on their health status. The study did not involve the collection of confidential information, participants being required to commit an act which might diminish self-respect or cause them to experience shame, embarrassment, or regret; participants being exposed to questions which may be experienced as stressful or upsetting, or to procedures which may have unpleasant or harmful side effects; the use of stimuli, tasks or procedures which may be experienced as stressful, noxious, or unpleasant; or any form of deception. Furthermore, the study proposal underwent an ethical clearance process prior to conducting the study. During data collection, each respondent was provided with a consent form (refer to Appendix A) which detailed the nature of the research, identified the researcher and the institution at which the research was being undertaken, and the voluntary, anonymity, options for withdrawal and confidentiality assurances of the research.