Probability sampling involves the use of chance in selecting a sample. The human mind cannot control the selection or non-selection of any of the units; therefore, the units have identifiable nonzero probability of being used for the sample. The sample in non-probability sampling is not dependent on chance. It is determined by a person. As explained by Smith (1983), some person may use personal judgment in selecting the sample. The study used probability sampling because it is free from personal influence. Non-probability cannot be viable to determine an error. It is also not applicable to statistical methods.
Using qualitative research calls for the exploration and understanding of the meaning groups or individuals that assigned to a specific human or social problem. The research process needs lots of emphasis on the meanings and processes that are not thoroughly measured or examined in terms of intensity, amount, quantity or frequencies. According to Neuman (2000), qualitative research seeks to ascertain the socially built nature of reality, to emphasize the connection between the researcher and the main purpose of the study, and to stress the value of the whole inquiry.
Use of quantitative research, on the other hand, would mean testing the objective theories by looking into the existing correlation among all variables. These variables are then measured on certain instruments and the numerated data analyzed by use of laid down statistical processes. Quantitative research rarely employs process investigation but stress the dimension and analysis of casual relationships existing between variable within a certain value.
Another approach, mixed methods research, combines both quantitative and qualitative forms. It incorporates the use of the two approaches, philosophical assumptions, and mixing of both quantitative and qualitative approaches. This study utilized this approach because it takes advantage of the positive aspects of the two approaches. As Neuman (2000) confirms, it does more than just evaluate objective data that is made up of number. The general strength of the study will be grater that just quantitative or qualitative research.
Data Collection Tools
Data collection is a critical part of any research. It has to be accurate since inaccuracy may have certain impacts of the results of the study, hence end in invalid results. Quantitative methods of collecting data depend on random sampling, as well as, planned data collection instruments, which have to fit different experiences. The end results are easy to generalize, compare and summarize. This study used face to face interview, telephone interview, in depth interview and questionnaires.
It used face-to-face interview because it is easy to create rapport with expected participants hence gain their trust and cooperation. According to Jean-Claude (1991), they bring the highest rates of response. Additionally, they give room for clarification on any ambiguity. However, it is impractical when large sample are used because it can be expensive and consume much time. Telephone interviews were done on individuals who could not find time to sit down for a face to face interview. It takes less time and does not need travelling. The problem with this is that it has a low rate of response.
The study also used two questionnaires: the paper-pencil and web-based questionnaires. Paper-pencil questionnaire was employed because it can save cost and time. It can also be sent to many people who agree to take part in the study. Those who use this type of questionnaires are more truthful because the researcher is not present when they give answers to the questions asked. However, not everyone returns the questionnaires sent to them. Web-based questionnaires were sent to those connected to the internet. It is quick and involves less detail when compared to the other type of questionnaire. However, the problem with it is that it excludes those who cannot access the Internet. In-depth interview was used in instances that involved a lot of data, which required much time consuming response. It is a good option because it allows the collection of detailed data. The problem is that it consumes a lot of time.