Random and non-random sampling. Choose the ideal sample for your survey.

Random sampling or non-random sampling are expressions that you will find in any market research proposal. And the quality of your market research and the reliability of the decisions you make with it may depend precisely on the sampling chosen. Today we will tell you all the details about the different types of sampling so that you can choose the one that best suits your study.

What is market research sampling?

Sampling is the process of selecting a market research sample. Having a quality sample that has been chosen according to an appropriate method will help us to be able to draw conclusions from the research with confidence.

Types of sampling. Random and non-random sampling

There are two main types of sampling: random and non-random sampling.

• A random sample or probability sample is one in which all elements that are part of the universe can be selected and have an equal chance of being chosen for the study. This type of sampling is ideal, as it reduces the risk of any sample bias, and in principle, offers maximum representativeness. However, it may sometimes be difficult to access a random sample, or it may be economically unfeasible. Therefore, non-random sampling is often chosen.
• Non-random sampling or non-probability sampling is a method of sample selection in which not all individuals in the universe have the same probability of being selected. Unlike random sampling, this type of sampling does not depend exclusively on chance, but is accepted by the researcher because either not all individuals in the study universe are accessible, or because it has greater ease of access or other advantages that may influence the cost or speed of the study.

So while it is desirable to work with random samples, one should not be afraid to work with non-random samples when there are good reasons to do so.

Random sampling – what types are there?

There are several types of random sampling:

• Simple random sampling. The size of the universe is known, everyone is accessible, and everyone has the same probability of being chosen. It is as if all the people who could participate in the study were in a lottery draw and we chose some at random. This type of sampling is theoretically ideal, but in practice it is very difficult to achieve when you want to work with representative samples of a country’s population. Any data collection technique will incorporate a bias. For example, if personal interviews are to be conducted, it will not be possible to send interviewers all over the country to choose people at random, but the fieldwork will be concentrated in a few selected locations. Or, for example, telephone interviews will have more difficulty locating people who are unable to answer calls for much of the day, so they will be less well represented in the study. And even online research, which is widely used today, will have more difficulty reaching older people, who are statistically less digitally literate.
• Systematic random sampling. People are sorted in a list and a first record is chosen at random and then people are chosen every few records. This was very common when interviewing in households. A random decision was made as to where to start selecting people and an attempt was made to access them every few households or street numbers.
• Stratified random sampling. The study population is divided into strata, i.e. mutually exclusive population groups – for example, you want to do a study in the metropolitan areas of Madrid and Barcelona, because you think there may be differences in their way of thinking. In order to be able to analyse the results with greater confidence, a sufficient sample size is defined for each stratum, proportionally or not, and simple or systematic random sampling is followed within each stratum.

Types of non-random sampling

Let us now look at what types of non-random sampling are also commonly used in market research.

• Cluster sampling. In this type of sampling, the sampling units are not the individuals themselves, but groups of individuals. Let us imagine that we want to conduct a personal study and we think that the results will be very different between large, medium and small municipalities. The option of sending interviewers all over the country is economically unfeasible, but we can send interviewers to a sample of municipalities to interview samples of people. This will allow us to know if there are differences in opinion or behaviour among people in each type of municipality and we will be able to draw conclusions.
• Snowball sampling. In this sampling method, the individuals selected first suggest new participants for the study. This type of recruitment is used in cases where the study population is difficult to access. For example, it is often used in B2B, when we want to interview certain types of specialists or when we want to interview patients with rare diseases or people with very specific hobbies. In all these cases, it is very likely that the people initially selected will know others with similar characteristics that they can suggest, and sometimes receive a small incentive in return.
• Convenience sampling. This is based on selecting people to participate in a study because of the fact that they are available and agree to participate in the study. This type of sampling is common in clinical trials, where the researcher does not know the size or characteristics of the study population and volunteers are needed. With the spread of online research, this method has also expanded to panel studies where people sign up to participate in studies.
• Quota sampling. This sampling is similar to stratified sampling, but without complete randomisation within each stratum. In this sampling method we also divide the population into different exclusionary groups – for example age or geographical area. each of these groups will have a population size that we know. Well, in the sample we will define quotas to be obtained that are proportional to the sizes of the segments of the study population. For example, if we define a quota by gender, and we know that in the population there are 52% women and we wish to reach a sample of 500 individuals, once we have interviewed 260 women, we would not accept any more, since we would have completed the quota. In this way, although we are dealing with a non-random sampling, we guarantee a very good similarity between the target population and the sample.

Find the sample you need

Online consumer panel research is widely used for surveys around the world. Companies of all industries and sizes rely on the quality of these samples to get the data they need. The great advantage of consumer panels is the immediate availability of people who have expressed an interest in participating in market research, which makes research fast and efficient.

We are testers’ own consumer panel currently gives you access to 130,000 consumers and users in Spain and Portugal, and gains new participants every year. For research outside Iberia, we have integration with leading international consumer panels that ensure you can research on all five continents.

All We are testers have passed strict quality controls to ensure their suitability. They are motivated and incentivised to provide reliable answers in all types of studies. And each study is monitored by checks that identify any minimally suspicious behaviour and eliminate potentially affected responses before the data becomes part of the results. This way, you receive data of the highest quality.

The panel allows the creation of representative samples of the population and facilitates their segmentation by socio-demographic variables, consumption data and previously collected attitudes. In other words, even if you are looking for a needle in a haystack in terms of sample, you may find it ready for you to use. And if the specific criteria you need are not available, you can always enter filter questions to easily identify the individuals you are looking for.

The We are testers research platform also allows you to define the quotas you want for a more representative sample. For example, you can choose quotas for gender, age and geographic area, which will give you the extra confidence you need.

Contact our experts to find out all the details of the We are testers panel sample and start making more decisions supported by quality market research.