[Essay] Random Samples

Sampling may be defined as measuring a portion of something and then making a general statement about the whole thing (Bradfield and Moredock, 1957 p.38). For instance, one is buying mangoes. He picks up one mango from crate of mangoes and tastes it. If the mango is sweet, the he concludes and says that the mangoes in the crate are sweet although he has tasted only one mango. In this case, we may say that sampling saves the sources of data from being all consumed since it make possible the study of a large, heterogeneous population. Thus, sampling method helps us to save time and cost since it is for speed, accuracy and economy.

There are two types of sampling, the random and non-random sampling. Most researchers preferred to use random sampling than non-random sampling in order to avoid bias in the study. In random sampling, the sample is proportion (a certain percent) of the population and such sample is selected from the population by means of systematic way in which every element of the population has a chance of being included in the sample. On the other hand, non-random sampling simply selects sample which is not a proportion of the population and there is no system in selecting the sample and the selection depends upon the situation.

Apparently, techniques in selecting a sample should be considered to have a meaningful analysis of data. Now, let’s try to consider the advantages and limitations of such techniques (e.g. simple random sampling, stratified sampling, cluster sampling, and multi-stage sampling).

In simple random sampling, everyone in the population of the inquiry has an equal chance of being selected to be included in the sample. This is also called the lottery or raffle type of sampling. This may be used if the population has no differentiated levels, sections, or classes. The main advantage of this technique of sampling is that, it is easy to understand and easy to apply too. The disadvantage is that, it is hard to use with too large population because of the difficulty encountered in dealing with too large samples.

Stratified random sampling is the process of selecting randomly samples from the different strata of the population used in study. The procedure of stratified random sampling follows. Determine a common stratum or class to which all the elements of the population belong. Then divide or group the elements of population according to characteristics inherent in the whole class or stratum that makes the element of the population different from one another. After stratification has been completed, apply the simple random sampling in the actual selection of the samples in every class or stratum and the selection must be proportional, that is, the same percent is used in every class. Thus, this is used when the population of the inquiry has class stratifications or groupings. Its advantage is that it contributes much to the representation of the sample. It is also easy to apply. However, stratified random sampling also shares the disadvantage of simple random sampling.

Basically, cluster sampling at start is almost similar to stratified random sampling since we have to divide the population into discrete groups prior to sampling. The groups are termed into clusters in this form of sampling and can be based on any naturally occurring grouping. For cluster sampling your sampling frame is the complete list of clusters rather than a complete list of individual cases within the population. You then select a few clusters, normally using simple random sampling. Data are then collected from every case within the selected clusters. Selecting clusters randomly makes cluster sampling a probability sampling technique. However, the technique normally, results in a sample that represents the total population less accurately than stratified random sampling. Restricting the sample to a few relatively compact geographical sub-areas (clusters) maximizes the number of interviews you can undertake within the resources available. However, it may also reduce representativeness of your sample. For this reason you need to maximize the number of clusters to allow for variations in the population within the available resources. Your choice is between a large sample from a few discrete subgroups and smaller sample distributed over the whole group. It is a trade-off between the amount of precision lost by using a few subgroups and the amount gained from a large sample size.

Similar to cluster sampling, multi-stage sampling is used when the population is so big or the geographical area of the research is too large. The general procedure is to divide the area or population into clusters or blocks and then within the final cluster apply any of the different method of selecting a sample. The advantage of the multi-stage sampling is its efficiency. Its main disadvantage is its reduced accuracy or representativeness, on account of the fact that on every stage there is a sampling error.

Reference:

Bradfield JM & Moredock, S (1957) Measurement and Evaluation in Education. New York: The Macmillan Company Inc.