Understanding Research Sampling
Understanding Research Sampling
When it comes to a research study, making sure you have the right sample is critical to success. If your participants are not from the population affected by your research question, the study will fail to achieve its goals no matter how good the analysis. With this in mind, it is important that you think purposefully about how you will sample your population, regardless of research methodology and design. Three major considerations when determining a sample are the population, its instances, and access.
Sampling and Population
Population can have numerous meanings depending on the context. In research specifically, population is a technical term which represents the entirety of units under study. Your population is all persons your research question affects. In other words, your sample population can be defined by the question, “What is the full population that is related to your research?” This could be all the school bus drivers in a district, all the lawyers in a city, or all the senators in congress; it just depends on the focus of your study.
For evaluation or experimental studies, your population includes both the control and the group receiving intervention. You need to be precise in describing the shared and non-shared characteristics of the population of your study in order to make sure those characteristics align with the study objectives.
Example: I am interested in the development of established group identity in emerging adulthood. I decided to study this through investigating how freshmen students pledging a fraternity or sorority develop their shared identity. I need to be specific on what would include or exclude anyone from the study. I’d want to make sure that I was actually using freshmen or recent freshmen who are actually pledged to a fraternity or sorority. In this instance talking to seniors or talking to freshmen who are not part of the fraternity or sorority system, although emerging adults, would not help.
The key is having a specific target population and then within that population a more specific sample population that has a high level of shared characteristics.
Sampling and Instances
The context or instance in which the population is connected to your project focuses the research. You will need to ensure that wherever you are planning on conducting the study has the phenomenon of interest to study.
Going back to the fraternity or sorority freshmen pledges example above, you would need to make sure the population exists at whichever institution you hold the study. If you were planning to study within a community college and they don’t have Greek-life, then the study won’t work. While that community college had a lot of male and female freshmen, who are emerging adults, without those sororities or fraternities the site won’t work for the purposes of exploring freshmen fraternity or sorority identity development.
Sampling and Access
If you have found an instance where your phenomenon of study is occurring, have found the population, and identified a sample population to pursue from within that larger group, you still have to think about access. For example, if I was doing a study of Presidents, there are lots of directions I could consider. However, if my goal is interviewing Presidents of the United States, although they are easy enough to identify, actually getting access to those leaders for the purposes of the study would be extremely difficult if not impossible, particularly without the right connections.
When it comes to access you will need to think about the people as well as the organization(s) they represent. Do you need permissions only from the participant or will you need permission from the participant’s place of employment? Do you even have the ability to contact their place of employment or the person about being a participant?
Thinking back to the example project on emerging identity study of freshman in sororities or fraternities, I’ll likely need permissions from the University I identify, the individual fraternities and sororities I want to include, as well as individual participants. The point here is that you need to be realistic about who you want to include in your study and have a plan for reaching out to both the gatekeepers and the participants. If you cannot make contact, you cannot solicit someone as a participant. However, if you know your population, know you have access to them, and know those people have the information you need, you have what it takes to go recruit your sample.
So that’s it! Taking steps early in your research to be sure you can answer your research question by sampling the appropriate participants within your population, and by being aware of how instances of the population, access to the population, and human subject protections might limit your sampling from the start, can help reduce later headache. Happy sampling!
Do you have a specific research study you are trying to develop a sample for? Share your thoughts in the comments below.
For more on recruiting human subjects, please visit our IRB Corner post on the same topic.