Sampling is a more convenient way of identifying and picking the most suitable representation of a population in a survey that could be daunting or impossible if the whole population had to be studied. In that respect, the concept is widely applied in market research where different sampling techniques are applied for different types of population. Thus, this analysis seeks to demonstrate the application of different sampling techniques using a case of Nevada National Bank that seeks to use its list of 400,000 credit card users who are spread all over the United states to test for the relationship between socio-economic characteristics and use of credit cards. To achieve the objective, the discussion begins by explaining how the bank could identify the population and a sampling frame. Further, the discussion explains how simple random, strata and cluster sampling techniques could be applied to select a sample for the test. Finally, the discussion concludes by explaining the most suitable sampling technique that should be applied for the test.
- Identification of population and sampling frame
Nevada National Bank could use the list of all its credit card users and their information provided in the cards application forms. In that respect, the bank could be using a list type of sampling frame by getting a complete list of all the 400,000 card users from its customers’ database. The list could then be used to provide the socio economic characteristics of each credit card user hence the bank’s ability to classify card users by their characteristics. (Cochran, 2007)
- Simple random sampling
Simple random sampling method is applied in selecting a sample size n from a population N in a way that every distinct sample unit N C n has an equal chance of being drawn. In that respect, a simple random sample of the credit card users could be drawn a unit by unit from the 400,000 users. This could be done by naming the number of the card users from 1 to 400,000 and choosing the sample by a means of a computer program or use of a random table in a way that gives every undrawn card user an equal chance of being drawn. (Lacobucci & Churchill, 2010)
- Stratified sampling
Selecting a sample for the card users with stratified sampling could involve classifying customers into different populations Ni depending on their social-economic characteristics. In that respect, the total population would be a combination of the strata’s populations shown as N1 + N2 + N3 + + N k = N where 1, 2, 3, k represent the socio-economic characteristics by which customers are classified. Then samples ni could be selected from each population by random sampling to get the total needed sample of n1 + n2 + n3 + + n k = n. (Cochran, 2007)
- Cluster sampling
This method is applied where it is not possible to get a precise list of the population as a result of lack of such listing of such population units as well as from a wide geographical coverage that could prove to be expensive. (Lacobucci & Churchill, 2010) In that respect, the method could be applied by dividing United State into geographical clusters within which the 400,000 credit card users are located. Then each cluster would be a population on its own from which a sample could be picked. (Cochran, 2007)
- Preferred sampling method
The most suitable method to select a sample for the 400,000 credit card users is the strata sampling which is appropriate for application in a situation where a precise population is known. In addition, the method is suitable for its ability to classify the 400,000 credit card users into different socio-economic status groups, whose status’ relationship with card usage can then be estimated. (Lacobucci & Churchill, 2010)
In view of the discussion, it is clear that different sampling techniques are suitable for different types of data and population under review. In that respect, simple random sampling is suitable where all population units have an equal chance of being drawn. On the other hand, strata sampling is appropriate where the population can be classified into different groups from which samples can be randomly drawn. Finally, the cluster sampling method has been demonstrated as being appropriate for data that is not precise and where the population is spread across a wide geographical area hence a need to divide into clusters from which the sample can be draw. In that consideration, the most suitable method for Nevada National Bank to apply in its test for a relationship between the customers’ socio-economic characteristics and their credit cards use has been found to be strata sampling. This is given that the population under review is precise and marked by unique characteristics that can be easily grouped.
Cochran, W. (2007). (3rd Ed.). Sampling Techniques. New York: John Willey & Sons Inc.
Lacobucci, D. & Churchill, G. (2010). (10th Ed.). Marketing Research: Methodical
Foundations. Ohio: South-Western College Publishers.