Population and Sampling
What Do I Need To Know About Getting The Data For My Research?
As a graduate student, what do you need to know about collecting your data? How will you determine where to look, who or what to use for the data sources, and what you can expect to do with the collected data. Doing all of these things correctly can determine whether you will have quality scholarly work or just another worthless paper; the latter also means you will still be ABD, i.e., “all but dissertation.”
Basic Definitions
What is a population? It is a collection of those things about which to make inferences. It could be a group of people such as a teacher’s organization, nurses, or coal miners. A population could be a group of many other living things or inanimate objects. Examples would be all dogs of a certain breed or all automobiles of a certain make and model. Populations are quite large and extensive. Researchers often seek to explore populations to learn more about certain aspects of that population.
What is a sample?
A sample is a smaller portion of a larger population with which to make inferences about the population. In research, samples are selected in many ways and how the sample is selected can determine how well the collected data represents the population.
Why Sample?
Sampling methods in both quantitative and qualitative research should get a part of the larger population that will represent that population.
If I am conducting a quantitative study, I must calculate my sample size so it will have a greater chance of representing the total population. This is required if I am to achieve statistically significant results. If I do it correctly, my results can be generalizable to the larger group. Statistical analyses are applied to quantitative data, intending to test predetermined assumptions about the population and the likelihood that the assumptions are accurate. Because the assumptions can be generalized to the larger population, the sampling method may vary with the population. Certainty lies in the statistical analyses performed on the data.
It is different when conducting qualitative research. Qualitative sampling focuses on selecting the data sources most likely to provide a wealth of accurate information about the phenomenon being studied. I am gathering large amounts of data from each participant and would seek what qualitative researchers call “saturation,” the point at which the data I am collecting approximates what I already have. From the collected data, I would look for similarities and develop themes from those. A goal of qualitative research is to provide a theoretical generalization to expand the lens through which the larger population can be viewed. I would get a taste of how other participants would respond, and while I cannot generalize to other samples, I can expand on theories related to the problem. While it is difficult to generalize to other similar groups, the results can provide insight into the larger population’s characteristics. This is the essence of sampling in research, to use the smaller group to understand the larger group.
Relationship of Population, Sampling Frame (Target Population), and Sample
Research requires something or someone from which to gather data. Further, I should be able to get the data I need from the source(s) and need to do it properly.
Using the population
For example, if I wanted to investigate the work habits of nurses with ten years of experience in pediatric ICU, the population would be all nurses with those characteristics. Those would be the participants able to provide the information I need to answer my research question. If I wanted to be sure of my findings, my best alternative would be to interview or survey every last nurse who met the above qualifications, but there are many obstacles in attempting such a feat; the main one is access.
Using a part of the population
I could narrow the larger population to just the hospitals in my state and then attempt to interview or survey all those nurses who met the qualifications (10 years of experience in pediatric ICU). That would be a sampling frame or target population, an accessible part of the larger population. I would have fewer people to interview, but the time, effort, and coordination needed would most likely be excessive. What would probably happen is I could only reach a part of those nurses.
Narrowing Even Further- The Sample
Using my sampling frame, I could ask some of the eligible nurses from the selected hospitals to participate and draw conclusions about the population using the responses. My goal would be to use a manageable number to approximate the entire population while still ensuring an adequate number of participants to maintain a credible study.
How Do I know the Pie is Tasty?
I like pie and am always looking for the tastiest one. My ideal situation would be to be a judge at a pie-eating contest. As much as I like pie, if there are 10 of them in front of me and I must find the tastiest one, eating all ten is not a realistic (or a desirable) alternative. Instead, I would taste each one and decide. The key is getting enough pie to make an accurate decision. If I stick a toothpick in the pie to taste it, my odds of making a good decision are poor. But, if I cut a small slice of the pie and eat it, my odds are much better because I have enough to make an accurate decision.
Summary
Think of Population, Sampling Frame, and Sample like the Russian Matryoshka dolls, one inside another yet each representative of the other. This is the essence of sampling, to investigate the smaller group and draw conclusions about the larger group. How I do that will vary with the method I select. Your goal is to get a sample that will represent the larger population.
