Sampling exists and is necessary because surveying the entire population is not only impractical, but also impossible. Just imagine interviewing all of the US population or all of the people living in the UK. Madness!
Since markets could comprise hundreds, thousands or even millions of potential customers, businesses do not have enough time, money and human resources to conduct market research on every single person in the society who might potentially buy the product.
Also, getting the views of every consumer in the market would be way too expensive and time-consuming. In addition, it might be impossible to identify everybody in that market due to large market size.
Large multinational businesses usually sell to the majority of the population in the country, or even several countries around the world.
What is absolutely crucial in sampling is that the method chosen must produce the sample that is representative of the whole population. The answers given by the respondents in the sample must reflect the whole populations’ opinions as closely as possible. If the sample chosen does not represent the population, then the results might be misleading and biased – simply unrepresentative.
What is a sample?
That is why sampling chooses a small group of people for primary market research purpose.
This group of people, or a sample, will represent the population for a particular market. Hence, in market research all the potential customers of a product make up the population. The term ‘population’ represents all the market.
So, a sample is the small group of people taking part in a primary market research who are chosen to represent the overall target market.
The sample size
In general, a sample size of approximately 1,000 respondents is considered to be sufficient. The sample of this size will very likely produce results that reflect fairly accurately the total preferences of the whole researched population. And, there will be much less risk of bias or any other factors significantly distorting the results.
Generally speaking, the larger the sample, the more representative of the total population it is going to be. The smaller the sample, the less representative of the total population it is going to be. Plus, larger samples will have greater confidence of the final results.
Scenario 1: Sample size too big
When the sample is too big (>1,000), it will increase the cost of market research. Because it will take much longer to organize all the people to participate in the research and carry out the research process – surveys and interviews.
Scenario 2: Sample size too small
When the sample is too small (<1,000), it will decrease the confidence about the market research results. Because of the limited number of respondents chosen, it will increase the likelihood of chance variations to occur.
Obviously, by asking fewer people in smaller samples will help the business to lower the research costs by saving on time and resources, but with smaller samples, the results may be statistically insignificant.
How to start sampling?
When designing a sample, the business organization needs to first consider the following two aspects of sampling:
- Sample frame. A complete list of all people or households in the target population from which the sample is drawn.
- Sample size. The number of people to include in the survey.
It is important to point out that the larger the sample size as a proportion of the sample frame, the more likely it is to represent the characteristics of the total population. So, if a firm can afford a larger sample, it is likely to produce more reliable results.
Basic sampling methods
Once the sample frame and the sample size have decided by the research manager, a sampling method must be chosen to select the appropriate sample.
A. PROBABILITY SAMPLING METHODS
Probability sampling methods include the following six methods. The probability of each respondent’s inclusion in the sample can be calculated.
- Simple random. The simple random sampling gives all people an equal chance of being selected. It assumes that everybody in the population is similar. The computer software is picking numbers representing each person out of a hat – randomly.
- Systematic. The systematic sampling chooses people from a larger population randomly but using a fixed periodic interval such as every 10th person.
- Quota. The quota sampling divides the whole population into different groups (segments) who share specific characteristics such as the same gender or age, similar income, social status, etc. Then, a certain number of people from each group (segment) will be selected.
- Stratified. The stratified sampling divides the whole population into different groups (segments) who share specific characteristics, like in the quota sampling method. Then, a certain number of people in each group (segment) will be selected based on the same percentage in total population, e.g. if the population has 65% of males, so would the sample. The precise knowledge about how population is divided up is necessary.
- Cluster. The cluster sampling divides the population geographically into different areas (clusters). Then, it chooses respondents within each cluster, or from multiple locations.
- Multistage. The multistage sampling divides the population geographically into different areas (clusters), like in the cluster sampling method. Then, a sample is drawn from a specific another area within that area such as country – province – city – district – county – street – block – house/apartment.
B. NON-PROBABILITY SAMPLING METHODS
Non-probability sampling methods include the following three methods. The probability of each respondent being selected into the sample cannot be calculated.
- Convenience. Convenience sampling uses those people who are easy for the market researcher to reach at the moment such as employees in the same office, all students in the classroom, or people in the coffee shop.
- Snowballing. Snowballing sampling starts with one specialized individual who is asked to recommend other people to be included next in the sample.
- Judgmental. In judgmental sampling, people are chosen only based on the market researcher’s intentions.
Sampling results
When it comes to sampling results, they must be statistically accurate, otherwise they are useless. The accuracy of results is determined by the confidence level of the results.
Most market research methods aim to produce confidence levels of .95 (95%) or higher. This means that the results will be accurate 95 times out of 100. For example, five surveys could not have been filled in completely out of every 100 surveys handed in to respondents. Confidence levels are dependent upon the number of people surveyed. There are precise formulas for calculating the confidence level.
Business managers will be interested in how accurate the market research results are going to be. All the market research results must be backed with statistics.
In other words, the firm will rely on the results representing the views of the entire target population to make business decisions. Therefore, the confidence level should be above 0.95 (95%).
Benefits of sampling
Sampling has a number of advantages over a full census of a population and these include:
- Aids decision making. When the right people are chosen as respondents, the risks of new product failures will be reduced.
- Saving time. It will take a lot less time to interview or survey a sample rather than a full census. Hence, research results will be produced faster, which will shorten the time to make decisions.
- Fewer resources required. Far less people will be involved in setting up and managing the sampling process.
- More cost-effective. Costs of market research will be relatively lower when just a sample is taken instead of involving the whole population.
Problems with sampling
- Data reliability. Human behavior is unpredictable because people do not always behave in the way they say they do and often change their mind. Human beings’ thinking, saying and doing is rarely aligned.
- Prone to bias. Market research biases including sampling bias, questionnaire bias and other biases.
In summary, sample is a small group of people representing the entire population that is selected for the market research study. The market research is testing on a sample to understand how the population will react to a product being marketed. Sampling should be done because it is normally too costly and almost certainly impossible to ask everyone in the target population.
