6.1: Identifying a sample

Although a market researcher might want to include every possible person who matches the target market as part of the research project, it’s often not a feasible option, nor is it of value. If one does decide to include everyone, it would be a census of the population. Getting everyone in a population to participate is time-consuming and highly expensive, so instead marketers use a sample, whereby a portion of the whole is included in the research. It’s similar to the samples you might receive at the grocery store or ice cream shop; it isn’t a full serving, but it does give  a good taste of what the whole thing would be like.

So how does one know who should be included in the sample? Researchers identify parameters for their studies, called sample frames. A sample frame for one study may be college students who live on campus; for another study, it may be retired people in Ottawa, Ontario, or small-business owners who have fewer than 10 employees. The individual entities within the sampling frame would be considered a sampling unit. A sampling unit is each individual respondent that would be considered as matching the sample frame established by the research. If a researcher wants businesses to participate in a study, then businesses would be the sampling unit in that case.

The number of sampling units included in the research is the sample size. Many calculations can be conducted to indicate what the correct size of the sample should be. Issues to consider are the size of the population, the confidence level that the data represents the entire population, the ease of accessing the units in the frame, and the budget allocated for the research.

The identification of the target population confirms the number of respondents and the way they are selected. If done correctly, general conclusions can be drawn about the views of the target population based on a small number of respondents. For example, when properly selected, a survey of 1,000 citizens can allow a researcher to draw conclusions about the views of all citizens in a country. If, on the contrary, there are mistakes in the selection of respondents, the results of the survey can be biased to the point of being useless.

The process of choosing participants for a market research study is known as sampling methodology in statistics. If the project is being managed by a market research firm, then a team of highly skilled statisticians will assist in choosing a sample from the target population by performing a series of mathematical calculations. The results of the validity and accuracy of professional surveys, for instance, are typically presented as part of the findings presentation.

For organizations and businesses who may be conducting the market research themselves, there are some key considerations to ensure that the sample selected is representative of the target population.

Market research project designs that indicate the right sample size often assume a 100% response rate. If a response rate is suspected to be lower, the sample size needs to be adjusted upwards.  “A high response rate is important for drawing valid result conclusions. This is particularly the case if those who ignore a survey and would have answered differently than respondents. For example, in customer satisfaction surveys, those who are unhappy with the service may answer the survey to channel their anger and to ask for change, while those who liked the service may not bother responding. In this case, survey results are biased and the bias will be more important if the response rate is low,” (OECD, 2012).

References

Albrecht, M. G., Green, M., & Hoffman, L. (2023). Principles of Marketing. OpenStax, Rice University. CC BY 4.0