Sample size. Choose the right sample size for your survey

Base muestral. Calcula el tamaño.
Redacción 6m of reading

Selecting a sample for a market study is the first step to obtain reliable results. The larger the surveyed population, the more reliable the data in our analysis will be; therefore, the ideal would be to conduct the study on the entire population. Since this is not possible, a subset of society is selected, which is known as a sample of the survey.

Sample of the survey

The sample of a survey is a portion of the selected population and it has to be valid, appropriate, and representative of the entire population. Through a representative sample of the population, the person responsible for conducting the study can get an idea of their opinions, which is known as extrapolating the results, without the need to invest the significant amount of money it would take to conduct a study of such dimensions.

Ensuring that the sample for a market study is valid is not an easy task, but at We are testers, we have solutions to your doubts. For the sample to be extrapolatable, meaning that the conclusions drawn from the sample can be applied to the entire population, its proportion should be greater or lesser depending on various factors such as the techniques used to select the sample, the variety of population profiles, and other variables that we explain below. The size of the sample is referred to as the sample base.

Stay tuned!


Types of Survey Samples

Depending on the techniques used to select the sample, we have probabilistic or random samples and non-probabilistic, directed, or judgment samples.

Probabilistic or Random Sample

Probabilistic samples are those in which all elements of the population have the same chances of being chosen. Probabilistic sampling is carried out using statistical tools, which we do at We are testers. In principle, if the sample is random, there will be more confidence that the most representative characteristics of the population are represented.

Non-Probabilistic, Directed, or Judgment Sample

Non-probabilistic samples are those in which the choice of sample elements does not depend on probability, but on who conducts the sample or the research criteria. Non-probabilistic sampling is used when the person responsible for the study knows the characteristics of the population and knows which profiles are of interest for their sample, based on the research objectives.

Access Panel for Surveys

At We are testers, we have our own consumer panel with over 120,000 testers, with a wide range of sociodemographic and socioeconomic characteristics, segmented by more than 400 criteria. This allows us to obtain samples for market studies with general consumer profiles or specific niches using available segmentation criteria. The over 400 segmentation criteria are divided into categories (‘About Me,’ ‘Shopping,’ ‘Home,’ ‘Technology’…) which provide comprehensive information about consumer profiles. Within each category, we have a wide variety of specific criteria, from employment status to the person responsible for purchasing decisions.

In addition, we conduct international-level studies and have the capacity to conduct research in more than 80 countries.


Stratified Sampling and Quota Sampling

Whether it is random or non-random sampling, it is crucial to divide the population into different strata or segments to make it as representative as possible, especially when dealing with a large population. Depending on the characteristics of the study, it will be interesting to divide it into different segments proportionally.

Depending on whether the selection is random or not, the sampling will be stratified or quota-based. In stratified sampling, subjects are randomly chosen within each stratum, while in quota sampling, the selection is not random and is influenced by the person conducting the study. In terms of the possibility of extrapolation, stratified sampling is more recommended, where the selection of subjects is random.

The sample for a market study can be segmented into a wide variety of strata – by gender, age, education level, income, and many others – and in different proportions depending on the study’s objectives. Depending on the research goals, We are testers offer the best recommendations for selecting an effective sample for a market study.


Calculating the Sample Size for a Survey

The sample size for a survey depends on the needs of the person conducting the study, but there are certain factors to consider in any market analysis.

1st The confidence level (confidence level) or the risk we are willing to accept of making a mistake.

2nd The estimated variance or diversity of opinions in the population (response distribution):

Indicates how varied the responses are estimated to be. As a rule of thumb, it is usually assumed that the variance is equal to 50%, which means that the population has the most varied opinions possible, i.e., half would respond yes, and the other half would respond no.

3rd The margins of error that we are willing to accept (margin of error).

4th The population size (population size).

Once the variables of confidence level, variance, margin of error, and population size are selected, a mathematical calculation is performed to determine the optimal sample.


Error Margin in Market Research

A factor to consider when calculating the acceptable margin of error is that it does not decrease proportionally; rather, the margin of error decreases exponentially.

In other words, the larger the sample size for a market study, the smaller the margin of error. To have a margin of error of 0, the sample size would need to be equivalent to the entire real population. However, at a certain point, the reduction in the margin of error will not be significant no matter how much the sample size increases.

Additionally, depending on whether the data will be analyzed in its entirety or by quotas, the margin of error will be different. When dividing the sample size by two, for example, the sample is halved, so if the sample size was 2000 people, it becomes a sample of 1000 people. Therefore, the sample size must be adjusted to the specific case.

Do you want to conduct a market study and don’t know where to start? Do you need to choose the right sample for a market study? Do you want to calculate the necessary sample base size? Contact us through our contact form.

Shall we talk?

Update date 22 December, 2023

Get in touch with our experts and discover how to take your research further.