Suppose you have to run a survey about the coffee drinking habits of high school students of USA. The population of the students is about 4 million. You can not even imagine running the survey by asking each and every student to get the relevant data because of requirement of huge amount of time, money and other resources. The cost of the survey in this case would be too monumental to justify the effort. To solve these types of problem, sampling can be used.
Definition Of Sampling
Application of certain queries to less than 100% of the population(group of all items that we are trying to observe and analyze) is known as Sampling. In simple terms, sampling is the process of selection of limited number of elements from large group of elements (population) so that, the characteristics of the samples taken is identical to that of the population. In above examples, suppose you choose 1000 students among 4 millions students. then:
- 4 millions students is population
- 1000 is the size of sample
Sampling is a great tool if you have to deal with a huge volume of data and you have limited resources. When you have large population of the data, then it can also be the only option you have.
Although you do not subject all the data to your queries, the chance that you get the desired results is almost similar to that when you do thorough checking. Provided that your choice for the sampling techniques must be appropriate.
How Sampling Works?
First of all, we have to choose the basis of sampling, i.e. the rule that will determine whether a sample is chosen or not. After we are sure of the method which will be used for the process, you select the samples as specified in the previously set plan. The method used for choosing the samples as the very name suggests, is the most crucial part of the whole process, it defines whether the analysis accurately describes the entire population or not.
As shown in the figure above, sampling is done by choosing a small segment of the population. We can say that the sample correctly represents the population because the ration of white:grey:black is the same in the sample as in the population.
Advantages of Sampling
Sampling have various benefits to us. Some of the advantages are listed below:
- Sampling saves time to a great extent by reducing the volume of data. You do not go through each of the individual items.
- Sampling Avoids monotony in works. You do not have to repeat the query again and again to all the individual data.
- When you have limited time, survey without using sampling becomes impossible. It allows us to get near-accurate results in much lesser time
- When you use proper methods, you are likely to achieve higher level of accuracy by using sampling than without using sampling in some cases due to reduction in monotony, data handling issues etc.
- By using sampling, you can get detailed information on the data even by employing small amount of resources.
Disadvantages of Sampling
Every coin has two sides. Sampling also have some demerits. Some of the disadvantages are:
- Since choice of sampling method is a judgmental task, there exist chances of biasness as per the mindset of the person who chooses it.
- Improper selection of sampling techniques may cause the whole process to defunct.
- Selection of proper size of samples is a difficult job.
- Sampling may exclude some data that might not be homogenous to the data that are taken. This affects the level of accuracy in the results.