No doubt, too many articles, blogs, videos, and any other resources have published materials related to sampling techniques. Yet, confusion remains, especially about the difference between convenience and purposive sampling.
I write this blog because I found this still needs further explanation, seen from the manuscripts submitted in Belitung Nursing Journal. I will not describe both sampling techniques in detail, just rationales and principles to use each of them.
Yes, “convenience”! Easy to imagine, we conveniently select the samples for our study. But wait, we still need rationales or reasons why to use it. When you conduct a study and select samples just because you stay in there, that’s one of the best reasons, but absolutely you cannot write it up like that. You need to find the reasons that make sense.
In addition, let’s say you cannot find the list of patients with hypertension in public health centers. So, you just see them based on their visits in the setting at the time of your data collection. This makes sense. Either you want to conduct an interview or a questionnaire survey, it’s all up to you. It can also be called accidental sampling.
Another example, you conduct an online survey to explore the health behavior of adults regarding COVID-19. You have some inclusion criteria to define adults. However, you do not group them based on location or socio-demographic criteria during data collection; you just accept all results and group them later.
And, if you choose the samples because of limited time and budget, that’s convenience too. Be careful using this reason, others will just ask you, “how if we give you more time and budget, can you do probability sampling?” (if students want to complete study faster, forget about to write this kind of reason).
The other example – you do an interview study of nurses to understand their lived experience during the COVID-19 pandemic. Don’t get confused; as long as the samples were homogeneous, then pick convenience sampling. For example, let’s say you just focus on staff nurses (all bachelors) in a public hospital. That’s convenience. Check the purposive sampling explanation below to extend this point.
Following up on the explanation of the interview study example above. The difference between purposive sampling and convenience sampling is that we use the purposive technique in heterogenic samples. We do not focus on just bachelor nurses but also diploma nurses, one nurse of each unit, and private hospital. We also select the nurses based on their experience in the units, how long they struggle with COVID-19, not just pick up any nurses. You definitely have strong reasons why choosing the sample, particularly to make your data richer. Not all inclusion criteria can be used to define purposive. It is somehow a bit tricky. You need to think of it critically.
The majority of the use of purposive sampling is for many kinds of participants, such as nurses, directors, nurse managers, nurse leaders, representations of nurse associations, and some others who agree to interview or FGDs or any other qualitative data collection methods.
In addition, purposive sampling is mostly used for qualitative study only, not for a quantitative study. So, it is confusing if you use Quasi-experiment with purposive sampling. I found several articles that use this kind of method, with the reasons according to inclusion criteria. But all samples are homogenous. It’s weird anyway. That must be Quasi-experiment with convenience sampling, at least according to my perspective.
Let’s check the meaning beyond “purposive”; it certainly has a particular purpose or is adaptive to a given purpose. It is a judgmental sampling that relied on your subjectivity.
In conclusion, reasons to use purposive sampling:
- Expert – you choose the samples because of their knowledge or their expertise in a certain area
- Critical case – you select samples based on the critical case you find in the setting and certainly will give you more data
- Extreme case – you choose samples based on the extreme of the data with unique characteristics
- Maximum variation – you choose various samples, similar to what I said above (nurses, nurse managers, nurse leaders, associations, etc.)
Despite the difference between both techniques, it still has many weaknesses compared to probability sampling. So again, do not forget to provide clear reasons why in your study you can’t conduct probability sampling. And most important point is that do not forget to provide evidence in any rationales you make, and of course, cite them correctly.
In the end, Chok Dee Khrub!
Cite this article as: Gunawan, J. (2021). Convenience Sampling Vs. Purposive Sampling. Retrieved from https://jokogunawan.com/blog/2021/07/28/research/853/.
This is an open access article distributed under Creative Commons Attribution 4.0 International (CC BY 4.0)