Data collection methods
Sampling involves selection of units from a given population and using this selected population to give a generalized view of the population at large. Non-probability sampling is one of the sampling methods which is based on the assumption that the distribution of characteristics is even within the population being researched upon (Struwig & Stead, 2001). Non-probability sampling is an effective method of sampling which should be recommended to researchers. Non-probability sampling represents the best strategy for selecting research participants.
Non probability sampling is usually based on a random selection process. This method is considered the best due to its time saving nature and the ease of carrying out the study. This method of sampling provides an alternative that is viable when dealing with research on a population that is very hard to reach or one where there is no list of population. A very good example is when conducting research on prostitutes and drug addicts. These are people who are not willing to come out. In such a case a type of non-probability sampling called snowball sampling is used (Struwig & Stead, 2001). This method is also used in research projects whereby the main purpose is to show whether there is a problem or not. A good example is exploratory research whose main aim is to achieve results in the fastest and cheapest way. Because of its inexpensive nature, it can be used to determine whether an issue or problem is worth being examined in detail. It is best used when showing that a particular trait can be found in a particular population. It can also be used by researchers who intend to focus on exploratory and qualitative studies. This method of sampling can also be used to generate results that will be used to come up with generalized conclusions pertaining to an entire population. Non probability sampling is not suitable in situations where the data needs to be correct and evenly collected over a certain area. This method is also not suitable since it could end up with biasness in the selection of samples by the researcher.
There are several methods of data collection that are used to collect data for analysis. One of them is using questionnaires. Questionnaires are a method of data collection which involves questions and answers. This method is used to collect data from a very large number of respondents and in areas where standardization is necessary. The questions may be open ended which gives the respondents a chance to answer in form of a flowing narrative while other questions can be close ended which give the respondents a chance to choose from answers that are pre-selected (Onwuegbuzie & Collins, 2007). This method requires higher levels of literacy in cases where the respondents are supposed to fill out the questions by themselves. Initially this method involved pen and paper but due to advancements in technology this is slowly changing and most questionnaires are being administered online or via computer as email attachments. The online questionnaires are proving to be easier to compile since the computer can be programmed to produce refined data based on the replies given by the respondents. Just like any other method of collecting data, questionnaires have their own advantages and limitations.
One of the strengths of using questionnaire is the fact that it can be used to reach a large number of people who have settled over a comparatively wide geographical area. In such a case, questionnaires are inexpensive to administer and also not time consuming as compare to other methods of data collection such as interviews. They also reduce the chances of the evaluator being biased because similar questions are asked to each respondent. Most people are familiar with questionnaires and therefore researchers do not need to educate. Some people are shy and therefore they tend to be more comfortable with questionnaires than interviews. Some questions can also create awkward moments when asked face to face as opposed to when they are in a questionnaire. When using questionnaires it is easy to tabulate results from close ended questions hence making the process straightforward. Questionnaires can be used to cover a relatively wide range of subjects and are good when collecting descriptive data. Since questionnaires can be online, they can be analyzed using a wide range of software.
The limitations include the fact that the size of data collected and the diversity of the information will be based on the literacy levels. If less people are educated then there will be a very limited number of responses. Most respondents may not complete the questionnaire for one reason or another and as a result there is a low rate of response. Since the questions could be close ended, there is no chance of asking for additional information from the respondents. Last but not least questionnaires are very hard to formulate and they are time consuming to create.
A potential ethical issue with questionnaires is the fact that it does not offer any protection of the respondents from harm. It can arouse painful memories especially if the questions are personal in nature. Such a questionnaire will cause stress on the respondent which may not be the intention of the researcher. To counter this, the questionnaire should be made in such a way that personal questions are made optional to avoid hurting the respondents. The contents of a questionnaire should also be thoroughly evaluated and sensitive questions removed to avoid discomfort to the respondents.
Measurement reliability refers to the level at which a measuring instrument or technique can be relied upon to produce consistent results even after being used repeatedly (Drost, 2011). Measurement validity is how successful a measuring instrument or technique is at quantifying whatever it is that it was designed to measure. A good example is when a person comes up with a way of testing memory. In order to ensure that the technique works one has to carry out the test and get relatively similar results. That makes the technique reliable.
Drost, E. A. (2011). Validity and reliability in social science research.Education Research and Perspectives, 38(1), 105.
Onwuegbuzie, A. J., & Collins, K. M. (2007). A typology of mixed methods sampling designs in social science research. The qualitative report, 12(2), 281-316.
Struwig, F. W., & Stead, G. B. (2001). Planning, designing and reporting research. Cape Town: Pearson Education South Africa.