Sampling Techniques - Assignment Example

SAMPLING TECHNIQUES By Sampling Techniques QuestionOne Definition Probability is the method employed by managers and research bodies to make surveys and come up with exhaustive data representing the entire subject matter. Sampling means taking or considering some units while leaving out others and the selected ones are to represent the whole unit. Sampling is the general research tool that involves both probability sampling and non-probability sampling. Probability sampling techniques and non-probability sampling techniques has some differences and the differences depend on the type of research and the type of sample units in use, sample size and gatekeepers that influence access to the sample units. The difference between the probability and non-probability sampling techniques is very simple. Probability sampling techniques involve random selection while non-probability sampling does not involve random sampling. This whole situation does not mean that non-probability sampling does not represent the whole population unit. The only upper hand that probabilistic sampling approach has over non-probabilistic approach is that with probabilistic we know the odds in representation and we are able to estimate confident intervals for the statistic. More researchers in general prefer random sampling to non-random sampling because they consider them accurate and rigorous. Non-probability sampling technique does not give all the individuals in a given population equal opportunities of being selected. In any form of research, true full sampling is often difficult to achieve due to some factors like fixed tome, lack of enough money and workforce to carry out the research. In contrast with probability sampling, the population in non-probability sampling is not at random but rather based on accessibility or purposive judgment of researcher. The downside of non-probabilistic research is that it leads to generalizations and many assumptions about the sample that was not in the process. There are different types of both random sampling and non-random sampling. Types of random sampling include simple random sampling, stratified sampling, systematic sampling, cluster or area sampling and multistage sampling. On the other hand, non-random sampling methods include accidental or convenient sampling and purpotive non-random sampling. Factors to consider while choosing between the two Choosing whether to use probabilistic research sampling or non-probabilistic research sampling methods depends upon some factors. Some of the factors include period for the research, the sample size, workforce, nature of the research and the amount of cash. These factors will make a researcher to employ one of the methods or both of them at some different points. For time to be involved in the research, a researcher does have a limited time within which the work should be done and submit it. This period may be long enough to enable him or her employ random research or can be short for him or her to randomize. If there is enough time, the person can go through all the population and get a clear view of the whole unit. With short research periods, only select a few units in non-randomization to save on time. Sample size is also a factor that determines the sampling method that one uses. There are over sized sample units and undersized sample units. When a sample unit size is very large, it is almost impossible to use random sampling because it is tedious and time consuming. Such a population should be non-randomized for efficiency. A big sample size may include large organizations, a bigger population of people, students from one region, animals within a game reserve or plantations within a region. In such cases then it is wise to use non-randomized selection. However, when the sample space is small, divide it up into manageable units and then employ randomized sampling. Workforce means the amount of support staff in the whole exercise and the amount of work in the entire exercise. This involves those people that a researcher sends to the field to collect the data and interview the population. If the taskforce has few people then they cannot use randomized research in a sample field that is large and comprises of many sample units. This will force the researcher to adopt non-randomization because the workload will be manageable with the available taskforce. Before a researcher decides which method o use in data collection, he or she should consider the amount of work and the taskforce he or she has in place to help in the research process. The most important factor that is out in consideration before selecting the method to use in your research method is the amount of money and the nature of the research work. Most research works receive their funding by the government and other agencies while others get fund from individuals. Those that get funds from individuals tend to use non-randomized methods because it is less expensive than the randomized methods. Nonrandomized researches do not involve employing many people to go out in the field to work with the large sample sizes to get data. Non-randomized method is cost effective and therefore most researchers who need qualitative results in a more efficient way to employ this method of research. In general, the method of sampling employed by a researcher is depends on the above-mentioned factors. The factors are not constant but vary depending on location and time. For instance, a method used in another country cannot work in some other because of geography or cooperativeness of the people. Whichever method a person uses should be able to give out full information target and not be bias. Question two. Introduction Probability is the method employed by managers and research bodies to make surveys and come up with exhaustive data representing the entire subject matter. Sampling means taking or considering some units while leaving out others and the selected ones are to represent the whole unit. Sampling is the general research tool that involves both probability sampling and non-probability sampling. Non-probability sampling technique is a sampling technique in which the population is not represented fully and not all members have the same probability of being selected. Probability sampling means the members of the population have the same probability of selection. The difference between the probability and non-probability sampling techniques is very simple. Probability sampling techniques involve random selection while non-probability sampling does not involve random sampling. Non-probability sampling technique does not give all the individuals in a given population equal opportunities of being selected. In any form of research, true full sampling is often difficult to achieve due to some factors like fixed tome, lack of enough money and workforce to carry out the research. In contrast with probability sampling, the population in non-probability sampling is not at random but rather based on accessibility or purposive judgment of researcher. Probability Samples Methods Simple Random Sampling (SRS) Simple stratified sampling considers each element in the population, and as such, each of these elements has an equal as well as known probability of selection. As such, the selection of every element is very independent of each other thereby making it the purest form of probability sample since its least biased and offers significant generalisability. Pro: each element has a known and equal chance of selection Con: it is cumbersome in provision of unique designations to every member of the population Systematic Sampling Under systematic sampling, the process of choosing a sample is by selecting a random starting point and then picking every ith element in the succession from the sampling frame. The determination of the sampling interval, I, is through dividing the population size N by the sample size, n, and then rounding it to the nearest integer. Pro: it is less expensive Con: it has a small loss in sampling precision Cluster Sampling Cluster sampling divides target population into mutually exclusive and collectively exhaustive sub-groups known as clusters. When selecting the elements of these clusters, one must ensure that they are as heterogeneous as possible, whilst the clusters remain homogeneous. A good example of cluster sampling is area sampling, whereby a geographic location is divided into clusters. Pro: has economic efficiency as it is faster and less expensive compared to simple random sampling Con: can lead to cluster specification error, in the sense that the more homogeneous the clusters are, the more precise the sample results Stratified Sampling Stratified sampling involves two steps in the process of partitioning the population into sub-populations, or strata. The strata in this case have to be mutually exclusive and collectively exhaustive to enable assignment of every population element to one and only stratum, not omitting any population element. Pro: it provides a more accurate overall sample of skewed population Con: it is a more complex sampling plan that requires different sample sizes for each stratum Non-probability Samples Methods Convenience Sampling Convenience sampling tries to obtain a sample composed on convenient elements. As such, the selection of respondents bases on their presence at an event, such as being in the right place at the right time. A good example is use of students, and members of social organization Pro: consumes very little time Con: some samples may not be representative Judgment Sampling Judgmental sampling is part of convenience sampling whereby selection of population elements bases on judgment or an educated guess by the researcher. Examples include test markets, and expert witnesses in courts Pro: low cost sampling Con: it is subjective Referral Sampling Referral or snowball sampling entails selection of an initial group of respondents at random. The respondents have to provide contact details of additional respondents belonging to the target population of interest. Members who are not famous or liked or give conflicting responses have a lower probability of their ideas being considered. Pro: can estimate rare characteristics Con: it is tedious and time consuming Quota Sampling Quota sampling sometimes appears as a two-stage restricted judgmental sampling. The first stage comprises of developing quotas or control categories of population elements, whiles the second stage involves selection of sample elements based on judgment or convenience. Pro: sample can be controlled for certain characteristics Con: has no assurance of representativeness References Beall, A. (2010). Strategic Market Research: A Guide to Conducting Research That Drives Businesses. Bloomington, Indiana: iUniverse Publishers. Read More