Good Essay About Sampling Technique
Simple random sampling
All the subsets of the frame are given an equal probability in a simple random sampling of any given size (Lemeshow et al., 2013). This means that all the elements in a given subset have equal chance of being selected. By doing so the biasness will be reduced and the analysis of the results will become simple. This technique of sampling is suitable to usual and small target of the population.
In this technique of sampling, the target populations to be study will be arrange in a specific order and thereafter the elements are selected at some regular interval a long that ordered list (Wheeler, 2013). During the first stage the sampling is random when coming up with an ordered list and then the process proceeds by selecting an element after every fixed number of elements. This method of sampling is reliable when the variable of interest is within the ordered list. The stratification induced can make this method more reliable and easy to implement.
Stratified sampling is applicable to population which depicts a number of distinct categories (Wheeler, 2013). The population will be categorized into separate strata based on that distinct characteristic. Each of the stratums will be taken as an independent sub-population and from there the individual elements from each group will be randomly selected.
In the process of designing a sample, one may found an auxiliary variable or size measure which he think will correlate to the variable of interest from a given population (Wheeler, 2013). The auxiliary variable found out will be used as a basis of stratification. On the other option is probability proportional to the size where each element will be given a selection probability which is proportion to its size measure. This technique of sampling is very suitable to population where the element size varies greatly and where the auxiliary information is available.
In a cluster sampling, the target population will be clustered according to the time periods or by geographical locations (Natrella, 2013). This sampling technique is appropriate when the population is large and had spread over a vast geographical location because it will reduce travelling time, travelling and administration costs.
In this sampling technique, the target population is first segmented to create mutually exclusive sub-groups (Lemeshow et al., 2013). Judgment is then used in selecting elements from each segment which is based on a given proportion. The method is suitable if the selected element in stratification process contain the variable of interest.
This technique of sampling is applicable on an imbalanced datasets that is where the ratio of sampling will not follow the statistics of the population. The value of minimax ratio is proven to be 0.5 and can only be so under the assumption of LDA classifier with Gaussian distribution (Natrella, 2013). In this sampling technique the sampling ratio of classes will be selected in such a way that the worst case classifier error will be best over all other possible population statistics for a given class.
Accidental sampling is a type of non probability sampling where the population will be selected because it is readily available and convenient to the sampling designer (Natrella, 2013). The method is suitable for pilot testing and not for scientific generalizations.
This is a method of sampling elements in a region where an element will be given a chance of being selected if it is lying along the line of segment known as line-intercept (Wheeler, 2013). The method is suitable in a population where the size measure does not vary widely.
This is a sampling technique where at first the group of participants is randomly selected and then later asks that group information several times over a given period of time (Lemeshow et al., 2013). Panel sampling is appropriate where changes in population are likely to occur over a period of time for example to chronic illness.
Levy, Paul S., and Stanley Lemeshow. Sampling of populations: methods and applications. John Wiley & Sons, 2013.
Natrella, Mary Gibbons. Experimental statistics. Courier Corporation, 2013.
Wheeler, Dennis. Statistical techniques in geographical analysis. Routledge, 2013.