Sampling Techniques and Methods - Essay Example

Research Methods Paper March By population we mean the complete count of all elements in a certain frame. An example is the total number of citizens in a country or the number of doctors in a certain country- a population of doctors. In general according to Argyrous (2005), a population represents the total elements in a certain sample space (p. 36). On the other hand, a sample is a subset of the population. Normally a sample is taken when one wants to study a certain population and then he/she has limited finances (Bryman 2008, p. 23). In such a case, he/she takes a representative sample to make inferences about the population. According to Bryman (2008, p. 25), a sample has advantages than a complete count in that; It is cheap in terms of time and resources to use a sample and make inferences a bout a certain population; A sample which is not biased has precision as a complete count; Sometimes it may not be possible to get all elements in a certain population and thus a sample may be used in such a case without losing the degrees of freedom. To Bryman (2008, p. 27), the size of a sample if determined by the following factors; Resources available (time and money); Error on the bound of estimation; The population size; Different sampling techniques exist and they are divided into two categories namely formal and informal (non probability sampling). The formal sampling techniques operate on a known probability of choosing and they include; simple random sampling, stratified sampling, cluster sampling, systematic sampling and multistage sampling. Simple random sampling is a technique where all the elements are said to have the same chance of being selected. The probability of selecting any elements in the population is equal (Green and Salkind 2008, p. 57). This technique is applied in cases where all the elements in that certain population have same traits (characteristics). Such populations are said to be homogenous. Another formal sampling technique is stratified sampling and is applied in cases where the population is made up of elements of different traits. In this technique, the population is sub-divided into non-overlapping sub-groups called strata (each is a stratum) each made up of elements with the same traits. Once this has been done, a simple random sample is selected from each stratum and then combined for final analysis. This technique is advantageous in that it eliminates biasness which is present if simple random sampling is used and also leads to higher precision. Cluster sampling on the other hand is a probability sampling technique in which the population is divided into clusters and then the researcher selects randomly the clusters to be included in the final analysis. It is mostly used in cases where getting the entire population for the research study is impossible or where the study population is concentrated in regions e.g. schools, churches, counties etc. The only disadvantage of this technique is less precision than even simple random sampling and stratified sampling. Systematic sampling is the selecting of sample elements in a systematic manner. For example, a researcher may decide to pick the every third element in the population. In such a case then, all the third elements form the sample. The advantage is simplicity in selection while its disadvantage is biasness. The final formal technique is the use of multistage sampling where the sampling units are exposed to a certain criteria. In this case some of the units are eliminated while others are selected. The selected elements are then exposed to another criteria and this is repeated until the requited sample size is attained. For example if a research would like to get a sample of household heads from a number of villages. Then the first step will be to select the villages either using random numbers or any other method. After which the probability of choosing each household from the said population is determined and used to determine the households to form the sample. The informal techniques assume no known probability of selection and include quota sampling, purposive sampling, accidental sampling and snowballing technique. In quota sampling, each region or population is allocated a certain slot of the required sample size. For example, if a sample of 520 is required for a certain study in the US then each state in the may be requested to selected 10 people of its entire population to take part. In purposive sampling, the research selects his/her sample according to the purpose of the study. The researcher makes the choice as he/she likes. This technique is related to accidental sampling where the selection is by the liking of the researcher. The last one is snowball sampling where the researcher uses a network of participants. The first participant determines all the others who will participate in the study. In this technique the researcher requests for information on how to get another participant from the previous participant. All informal sampling techniques suffer biasness and low precision of the analyzed results. When design questionnaires, the following issues arise; Clarity; Leading questions; Phrasing; Hypothetical questions; and Prestige bias. The clarity of questions is very important and determined the kind of data which a researcher will get. The problem here is getting an ambiguous question which will have different meanings to different participants (Punch 2005, p. 62). This problem can be avoided by minimizing the use of unnecessary adjectives. Further, there is need to avoid the use of ethnic vocabularies which may be not understood by other participants. Another problem that arises is the use of leading questions both during the question development of choices of answers. To correct this, all answers must be in such a way that they are equally likely and not giving choices as if to suggest the answers. Another problem may arise due to phrasing since different adjectives give different meanings while others are compliments of others. Using both in the same sentence means distorts the meaning of the entire question. Such situation can be corrected by using only terms whose meaning is as the intention of the researcher. Further, questions dealing with personal/private matters need to be avoided as much as possible. Remember that confidentiality of data and research information is one of the tenets of research ethics. Participants have trust if the information they give is confidential and used for the sole purpose of the study. Embarrassing questions should be avoided as much as possible. Asking respondents for answers for hypothetical questions means that they give answers to questions they have never given thought to. For example, asking a participant of what he/she will do if he/she was the president. This question is more tasking and means more thought than anticipated. Avoiding such questions solve the situation. Prestige bias is all about respondents giving information as if to portray themselves as special .The only solution to such is by ensuring that the questionnaire is a s personal as possible. References Argyrous, G. (2005). Statistics for Research: With a Guide to SPSS. 2nd ed. London: Sage. pp. 11-67 Bryman, A. (2008). Social Research Methods. 3rd ed. Oxford: Oxford University Press. pp. 7-81 Green, S.B. & Salkind, N.J. (2008). Using SPSS for Windows and Macintosh: Analyzing and Understanding Data. Pearson international ed., 5th ed. London: Pearson Prentice Hall. pp. 12-100 Punch, K. (2005). Introduction to Social Research: Quantitative and Qualitative Approaches. 2nd ed. London: Sage. pp. 9-102 Read More