Design of Experiments > Grounded Theory
What is Grounded Theory?
Grounded theory involves the collection and analysis of data. The theory is “grounded” in actual data, which means the analysis and development of theories happens after you have collected the data. It was introduced by Glaser & Strauss in 1967 to legitimize qualitative research. However, it’s use isn’t limited to qualitative studies; it is a general method that can be applied to many areas of research.
Watch the video for an overview:
Can’t see the video? Click here.
To start the grounded theory process, you should:
- Identify the area of interest.
- Avoid preconceived theories and focus on the data only.
- Use theoretical sensitivity— an awareness of subtle messages and meanings in data.
Research stops when you have reached theoretical saturation: the point where you have sampled and analyzed your data until you have exhausted all theories and uncovered all data.
Grounded theory commonly uses the following data collection methods:
- Interviewing participants with open-ended questions.
- Participant Observation (fieldwork) and/or focus groups.
- Study of Artifacts and Texts
The general theory can be broken down into two parts: methods and products.
Grounded theory provides qualitative researchers with guidelines for collecting and analyzing data. Although there are “probably as many versions of grounded theory as there were grounded theorists” (Dey, 1999), all of the versions have the following aspects in common (Charmaz, 2006):
- Coding (labeling and categorizing) from collected data instead of relying on theories not grounded in data.
- Social processes are discovered in the data.
- Abstract categories are constructed inductively.
- Categories are refined using theoretical sampling.
- The gap between coding and writing is bridged with analytical memos.
- Categories are integrated into a theoretical framework.
In order to say that your research is based in grounded theory you must follow the explicit, sequential guidelines. Employing just one or two methods does not make the study “grounded.”
Data analysis should happen at the same time as data collection. In other words, you shouldn’t wait until all your data is collected before analyzing it; these methods should be fluid and change if your data uncovers a new theory or potential direction. This type of concurrent data analysis and data collection is often referred to as constant comparative analysis and theoretical sampling.
Coding should be line by line, open coding: read through data several times, creating summaries for the data using preliminary labels. Axial coding is used to create conceptual families from the summaries, followed by selective coding which turns the families into a formal framework with a variable that includes all of the collected data. See this blog post for some great examples of these coding types.
Beyer, W. H. CRC Standard Mathematical Tables, 31st ed. Boca Raton, FL: CRC Press, pp. 536 and 571, 2002.
Agresti A. (1990) Categorical Data Analysis. John Wiley and Sons, New York.
Kotz, S.; et al., eds. (2006), Encyclopedia of Statistical Sciences, Wiley.
Wheelan, C. (2014). Naked Statistics. W. W. Norton & Company