Theoretical sampling

Theoretical sampling, sometimes called theory-based sampling, is a research method in which the data is analyzed as it is collected so that the researcher can develop a theory about the subject. This allows the researcher to determine what data is needed next and where and how to obtain it. It allows theories to grow and evolve and helps the researcher to make connections between different but related concepts. Theoretical sampling is important in the development of grounded theories and is often used in research involving how people and groups interact, as well as in medical research.

Background

The concept of theoretical sampling developed from the work of Barney Glaser and Anselm Strauss. In 1967, the pair of American sociologists developed the idea of grounded theories, so-called because they were “grounded” in the data that was uncovered during research. Up until that time, most research was done by developing a theory and then seeking research that proved or disproved the theory. Instead of beginning with a theory and then finding the data, Glaser and Strauss promoted the idea of developing theories from the data as it was found. This allowed the data to shape and guide the development of the theory and made it possible to adjust the theory to correspond to new data. Later, Strauss would collaborate with another American sociologist, Juliet Corbin, in continued work on grounded theories.

Prior to the work of Glaser and Strauss, researchers would develop a theory, determine how to prove or disprove the theory, and then gather information that would support it. Once the theory was proven or disproven, data gathering was finished. Grounded theories required a new way of finding data.

Overview

Through grounded theories, the researchers start with an area of interest to study and begin gathering information about it, while trying to avoid any preconceived notions of what will be learned. As data is gathered, researchers uncover patterns and relationships between different data points. These patterns and connections lead to the formation of a theory.

The researchers then use that theory as the basis for looking for more information. The theory will suggest new areas to explore and new lines of study to research. As the resulting new data is gathered, it is analyzed, and the theory is adjusted to incorporate it. The process continues until the researcher determines that additional questioning or research does not provide any new information.

Since the researcher is not starting with a preconceived idea of what the research will uncover, theoretical sampling is more open-ended than other forms of data gathering. Instead of surveys with yes or no questions, for instance, the researcher will be more likely to interview people as a means of gathering information. Open-ended questions are often used to allow the interviewees to share all the information they feel is important without any limitations that might be inadvertently imposed by the researchers.

Sometimes, other forms of research strive to incorporate randomness into its data collection methods in an effort to avoid bias. For instance, to gather information on how many people watch football, random information gathering would require researchers to intentionally go to many different areas and interview people from different walks of life. For theoretical sampling, it is more important to focus on gathering information that will address the concept under study than it is to focus on whether the sources of the information are random or not. For the researcher employing theoretical sampling, the most important first step in the process is determining who to ask to learn about the concept and where to find those people.

For example, researchers may start with an interest in homelessness in a rural community. They are not sure if the problem is larger or smaller than in a nearby big city, so they might start by interviewing people they find in areas more likely to be frequented by people who are homeless and have limited money, such as thrift shops, soup kitchens, and food pantries. They might also interview the people who work in these places to gain additional information. From this, they might develop a theory that while the problem is more noticeable in the larger city, there is a significant homeless problem in rural areas as well. The researchers might then further their study by asking more questions to determine why people are homeless, where they stay, what resources they need, etc. This will eventually help the researchers develop a theory about homelessness in the rural area. The theory might then lead to finding ways to distribute resources to address the needs of this population.

This form of research allows scientists to discover the areas that are most significant. This can lead to a greater understanding of the concept under study. Conversely, a researcher who establishes a theory first and then seeks information limits what can be learned. For instance, if researchers investigating rural homelessness decided that their theory was that homelessness was a bigger problem in cities than in rural areas, they might randomly question people in the city about where and when they have encountered homeless people and might not speak to many people with more than a passing contact with homelessness. This could limit their ability to develop a fair assessment of rural homelessness.

Theoretical sampling is often more comprehensive than other forms of data collection because it employs both inductive and deductive methods. Inductive methods start with specific bits of information and look for patterns and relationships between them to form broader generalizations. Deductive methods start with the theory and look for information to support or disprove it. As a result, theoretical sampling can be a more effective way of doing research. However, the sheer amount of data gathered in the process of developing a theory using theoretical sampling makes it a more complicated, expensive, and time-consuming method of gathering information and forming a theory.

Bibliography

Bohm, Andreas. "Theoretical Coding: Text Analysis in Grounded Theory." A Companion to Qualitative Research, edited by Uwe Flick et al., Sage, 2004, pp. 270-751.

Bradford, Alina, Mindy Weisberger, and Nicoletta Lanese. "What's the Difference between Deductive Reasoning and Inductive Reasoning." Live Science, 6 Mar. 2024, www.livescience.com/21569-deduction-vs-induction.html. Accessed 28 Jan. 2025.

Dudovskiy, John. "Theoretical Sampling." Business Research Methodology, research-methodology.net/sampling-in-primary-data-collection/theoretical-sampling. Accessed 28 Jan. 2025.

"Grounded Theory." Robert Wood Johnson Foundation, www.qualres.org/HomeGrou-3589.html. Accessed 28 Jan. 2025.

McLeod, Saul. “Theoretical Sampling in Grounded Theory.” Simple Psychology, 30 Oct. 2024, www.simplypsychology.org/theoretical-sampling.html. Accessed 28 Jan. 2025.

"Theoretical Sampling in Grounded Theory." Statistics Solutions, www.statisticssolutions.com/theoretical-sampling-in-grounded-theory. Accessed 28 Jan. 2025.

"Theory-Based or Theoretical Sampling." Robert Wood Johnson Foundation, www.qualres.org/HomeTheo-3806.html. Accessed 28 Jan. 2025.