By: Stephanie Aditadwinanti, Jing Jia, Frank Lee, Trevor Lee & Farhan Mohamed
Importance of sampling in marketing research
Sampling is an important part of conducting marketing research. Sampling is, essentially, selecting a smaller group (the sample) to act as a representative of an entire larger group (the population). While it may not be apparent, sampling occurs when many decisions are being made.
The use of sampling is seen on a frequent basis, for example in public opinion surveys during elections.
Where a population of one million people is too large to even consider surveying, a sample of a much smaller group will be selected to come up with a generalization of the entire population. It is assumed that the results from the sample are a representation of the characteristics of the population.
When advertisers and marketers target customers, they need to know what the customers want, think and how they will react. Rather than researching the entire population, the use of a sample size of customers is a cheaper, faster, and more practical way to understand that target population.
Classification of sampling techniques
Coming up with a sampling technique is a very important step in research design. There are two sampling designs: probability and nonprobability sampling design. Probability sampling is when each member/unit in the population has an equal chance of being selected. Whereas, non-probability sampling sees assumptions made by researchers about the population that they think make the result of the research more accurate.
Probability sampling techniques:
Simple random sampling Systematic random sampling Stratified sampling
Cluster sampling
Non-probability sampling techniques:
Convenience sampling
Judgement sampling Quota sampling Snowball sampling
Probability Sampling Techniques
Simple random sampling
Simple random sampling is a procedure that assures each element in the population has an equal chance of being selected. Meaning that the population of the research is derived from dividing equally into females and males. Simple random sampling is committed when researchers investigate the equal population between females and males in a particular place. For example, assume Capilano University has 10000 people in total including staff and students, divided equally into females and males, and researchers would like to select 1000 people for further study for who prefer to use textbooks and who prefer to use e-books. Not only does each person have an equal chance of being selected, researchers can also easily calculate the probability of given people being chosen; this means every student or staff in the school as 10% or 1/10 chance of being selected using this method.
Conceptually, simple random sampling is the simplest method of the probability sampling techniques.
Advantages
Free of classification error
Requires minimum advance knowledge of the population other than the frame.
Relatively easy to interpret and collect data
Best suits situations where not much information is available about the population
Disadvantages
Difficult to conduct if the size of the population being studied is large Needs a lot of time and money
Systematic random sampling
Systematic random sampling is a statistical technique which involves the selection of elements from an ordered sampling frame. Marketing researchers usually use this method for its simplicity and its periodic quality. The first step of systematic random sampling is that researchers use numbers to select an integer being less than the total number of individuals in the population. After selecting an integer as an interval, the second process is that researchers then to pick another integer that serve as the constant difference between any two consecutive numbers. For example, researchers estimate total population of 8500 people in Capilano University and need 15 subjects and first start off with 200. Then, they pick interval of 9 and the number continues so on by adding 9. Marketing researchers find this method simple to use and be able to do manually.
Advantages
It is simple to apply for research
Allows researchers to add a degree of system into random selection of subjects It assures that the population will be evenly sampled
Disadvantages
Causes errors in sampling It requires time to apply
Stratified sampling
Stratified sampling is a procedure that is used when there is a considerable amount of diversity within the target population. The first process in stratified sampling is to divide the population into mutually exclusive groups based on a similar characteristic, which are called strata. The second step is to then do a simple random sample group that is chosen independent from each strata group.
Stratified random sampling can also be broken down into two types: proportionate stratified sampling and non-proportionate stratified sampling. For proportionate random sampling the sample size is determined by the strata population compared to the target population. An example would be if you were interviewing Capilano University students and you organized them into three different stratas based on three different faculties. The larger of the three faculties would be sampled more heavily compared to the smaller stratas. In comparison, non-proportionate sampling is independent of the proportion to the total target population.
Advantage
Gives a good understanding of each strata group, as well as its unique characteristics.
Disadvantage
It is difficult to identify the correct stratifying variable.
Cluster sampling
Cluster sampling is a procedure that is very similar to the process of stratified sampling. In cluster sampling, the first step is similar in that you also divide the population into mutually exclusive groups called clusters.
The second step is then to select a cluster using simple random sampling. The major difference in cluster and stratified sampling is that only the cluster that is chosen is used for further sampling instead of getting
samples from every single cluster. Also, the main purpose of cluster sampling is much different that stratified sampling as it puts more focus in increasing sample efficiency and decreasing costs.
One of the main ways that cluster groups are determined is by using geographic areas. Say for instance there are three different streets: Main Street, Commercial Street, and Broadway Street. All of the residents from each street would be a part of their own separate cluster.
Advantage
Reduced costs
Disadvantage
Each cluster may not be a true representation of the population
Nonprobability sampling techniques Convenience sampling
Also known as an accidental sampling, is the process of collecting non-probability data in a convenient ways.
Means the data collection is already available and the population is close to hand. The researchers are randomly choosing people who are available to participate in the research study without having specific characters. Convenience sampling is usually used in field study of sociology, psychology, and political
interested in studying the use of cell phone during class time among students in Capilano University. He asks permission to one of his professor in Capilano University to spare 15 minutes of the class session and he uses only the students in that class as the sample. His research is purely using subjects that are convenient and readily available. The result of this study would not be represent all students in Capilano University and consequently the researcher would not be able to get his findings to all Capilano students.
Advantages
The data can be gathered faster than other sampling process
Easy to get the respondent as it is a random data collection process without any specific characteristics
Allows the researcher to use basic data from what they have been studied
Disadvantages
The result might not represent the whole population There is a possibility of a biased outcome
Judgement sampling
Judgement sampling is a non-probability sample, selected base on the opinion of the researchers. This type of sampling usually used when the researchers need to collect fast data from specific population. The
researcher’s judgment of the selected population make this method has high liability of error and to bias with the response. Judgment sampling is normally used in various studies. For instance, a researcher would like to make and experiment on Internet usage. It would be difficult to get random participants in general public. In this case the sample is taken from a group of people with an offer payment to give their time for the sampling process.
Advantages
Less time consuming and cost of sampling
People who are participate in the research know the material of the research so it will be easier to have the expected answer
Disadvantages
The result can be bias and stereotypes which may misrepresent the result Selected group of people may not represent all the population
Quota sampling
Quota sampling is a non-probability sampling technique. Quota sampling select the sample has the same proportions of individuals as the entire population with same characteristics (race, gender, age) or focused phenomena (frequency purchase, satisfaction level). Quota sampling is a two-stage restricted judgment sampling. In first stage, researcher need set up a restricted criterion. And in second stage, researcher select on the respondents basic on their judgments and case requirements. Quota sampling is an ideal technique for a study to investigate a trait or a characteristic of a certain subgroup. Quota sampling is also committed to observe the relationship between subgroup. For instance, some studies focus on research the traits of a certain subgroup interact with other traits of another subgroup.
Quota sampling is best choosing for such studies. Quota sampling is used in interview selections, product selections, marketing strategies and most elements of business running. For instance, we want to know if the instructors ‘attitude about using e-textbook relate to their generations. Assuming Capilano University has total 60 instructors. We divide the 60 instructors to three different generation groups and each subgroup has same members. After completing interview, we can know the relationship between instructors’ generation and attitude of using e-textbook.
Advantages
Preventing decisions to be polluted by unnecessary input Saving money when time is an issue
Be quick and easy to arrange.
Disadvantages
Limits your decisions Not allow much variety
Not possible to assess sample error as it is not random
Snowball sampling is a non-probability sampling technique. Researchers usually use this type sampling to identify potential subjects in studies where subjects are hard to locate. In other word, if the sample of the study is quite rare to find, snowball sampling can help the researcher observe the initial subject of the study.
The process of snowball sampling can be worked like chain referral. The researcher selects one subject and asks for this subject nominates another subject with the same trait as the next subject. The research will repeat this step until obtaining sufficient number of sample.
Advantages
Allows the researcher to reach populations that are difficult to sample when using other sampling methods
Cheaper, simple, cost-efficient
Disadvantages
Be little control over the sampling method, easy to out of control.
Representativeness of the sample is not guaranteed Easily cause Sampling bias
Selecting an appropriate sampling technique
There are many variables that come into play when considering which sampling technique should be used when conducting research. All of the techniques have their own advantages and disadvantages so it is up to the market researcher to determine which is more suitable for each situation. The selection process for choosing an appropriate sampling technique also depends on what the objectives are for the research;
whether it is qualitative or quantitative data. The timeframe and the accuracy of data have an effect as well on what kind of techniques should be used. In more accurate sampling methods, researchers are more likely to use probability techniques to compile data. By looking at all of these factors it is up to the researcher to select the sampling technique that best suits their needs.
Conclusion
Throughout this chapter, focus was put upon one of the most important issues in marketing research.
Sampling affects everyday decision making, therefore, it is important for researchers to understand all of the processes that go into this technique.
Sampling can be split into two different techniques: probability sampling techniques and non-probability
sampling techniques. Probability sampling techniques can be viewed as more accurate in comparison with non-probability sampling techniques, but are more intensive with cost and time involved. Probability can be further divided into simple random sampling, systematic sampling, stratified sampling, and cluster sampling.
Non-probability can be divided into convenience sampling, judgement sampling, quota sampling, and snowball sampling.
The selection of the most appropriate sampling technique depends on many different variables, such as the available resources, the degree of accuracy, the timeframe, and the importance of the decision making process.
Bibliography
Castillo, Joan Joseph. What is Sampling? 21 February 2009. . Easton, Valerie J. & McColl, John H. Sampling. 5 June 2013 .
Shukla, Paurav. Essentials of Marketing Research. Paurav Shukla & Ventus Publishing, 2008.