+34 616 71 29 85 carsten@dataz4s.com

Non-probability sampling

leNon-probability sampling is often turned to when resources are short. It is easier, faster and less costly than probability sampling, but it also opens for bias.




Convenience sample example

Say we wish to know about certain behavioral patterns among potential clients for our products. One of the fundamentals in being a potential client is that they are not clients. Potential clients are persons who are not buying our product. But to study potential clients, might require considerable resources, as they are likely not be all in our internal register of contacts.

In order to analyze the behavior of potential clients, and as we don’t have, they contact details, nor other “convenient” way of studying them, we can study our clients as we have these all registered. This make things much easier plus saves us time and money.

From this sample list of existing clients, we run the correct statistical methods and do the correct inferences, but the results are non-representative, as we are researching existing clients instead of potential clients. And in this case, the most interesting conditions might be the ones that are different from these two groups in order to understand how to convert non-clients into clients. Why are the potential buyers not buying?

However, it was much easier just to run a list from the existing clients from the database. The under coverage of the potential buyers is absolute. This is what’s meant by convenience sampling.

We can conduct correct statistical methods on an incorrect sample frame.

Convenience sampling



Judgement sampling

Judgements sampling is when one individual or a group of individuals choose who and who not to include in the sample. Although the individuals chosen for this assignment are considered experts in the field the individuals that they chose to be included in the sample are chosen subject to this expert’s subjective opinion, views and insight which might not reflect the real-world population.

Judgement sampling



Snowball sampling


Snowball sampling can be applied when sensitive information and/or interviewed persons are to be interviewed. For example, it can be a research on persons who are ill with a deadly virus or with Aids. These victims might not be very interested in participating. In such cases, the analysts can go through doctors or other persons close to the victims to ask them their help with the questions.

Snowball sampling




Quota sampling


The analyst pre-defines attributes of the elements in the population that should be sampled. For example, it could be all women in a company, or it could be voters in the age range of 18-22 years in a region, etc. This is a way of collecting samples in a fast way but leaves space for bias.


Quota sampling



Non-probability sampling summared

Visualizing the four sampling methods described above:

Non-probability sampling


Non-probability sampling learning resources



Carsten Grube

Carsten Grube

Freelance Data Analyst


Submit a Comment

+34 616 71 29 85

Call me

Spain: Ctra. 404, km 2, 29100 Coín, Malaga


Denmark: c/o Musvitvej 4, 3660 Stenløse

Drop me a line

What are you working on just now? Can I help you, and can you help me? 

About me

Learning statistics. Doing statistics. Freelance since 2005. Dane. Living in Spain. With my Spanish wife and two children. 

What they say

20 years in sales, analysis, journalism and startups. See what my customers and partners say about me.