Grounded Theory

In school, you may have been taught about the scientific method, which traditionally goes something like this: ask a question, do background research, form a hypothesis, conduct tests, and then analyze your data to decide whether your hypothesis was proven or disproven. You may remember the feeling of formulating a hypothesis and then eagerly awaiting the results to see if you were right. 

Maybe your hypothesis was something as simple as “if I drop Mentos into this soda bottle, there will be an explosion,” or more complex, such as analyzing whether pet owners are more likely to follow a vegan diet. 

Are there other ways to structure research? Absolutely. When conducting qualitative research in the social sciences, it’s important to approach complex questions about the human experience with an open mind. Grounded theory is a systematic inductive approach where, instead of working off of an existing theory, new theories are derived from the data itself, and data collection and analysis are iterative.¹

Grounded theory almost always involves qualitative research, which requires an open-ended question. Let’s say, for example, if you wanted to study the impact of pets on people’s lives, a quantitative approach might ask: are pet owners more likely to be vegan? This can be tested through quantitative methods (e.g., surveys, experiments, etc.). Notice that this is a yes or no question. In contrast, a qualitative question might ask: how do pet owners feel about veganism? This question is more difficult to answer with a single experiment. Qualitative research allows for deeper and more nuanced answers and takes many forms, such as observational studies, interviews, focus groups, ethnographies, and media reviews.

Now, let’s break down what conducting grounded theory research actually looks like: 

First, determine the initial research questions. This is not the same as having a hypothesis. Instead, hone in on what specific phenomena, experiences, or narratives you’re trying to understand. This can be grounded in the existing literature on the topic or an attempt to address a gap in the literature. Your research question (or questions, as sometimes you may have 2-3 related questions) will be the guiding force for the rest of your research, helping you determine your methods of recruitment, data collection, and analysis.

Don’t be afraid to return to your research question early and often. Part of grounded theory involves adapting your research methods and maybe even the initial question as you progress. For example, if certain themes come up over and over as you’re conducting interviews, it may be worth asking if your research question should be adapted to address these prominent themes.

The second step involves recruiting and collecting data using theoretical sampling. Gather rich and detailed data through interviews, observations, documents, or other relevant sources, and choose participants based off of their relevance to the research question. This process is iterative and continues until theoretical saturation is achieved, where no new information alters the emerging theory​.

Once you have some data, such as the records from relevant social media posts or recordings from in-depth interviews, prepare that data for analysis by turning it into a codeable format, such as transcripts.

In grounded theory, coding begins with open coding, which involves going through data line-by-line to identify discrete pieces of information and assigning labels (codes) to these pieces based on their content and meaning​.² This phase is highly exploratory and open-ended, allowing new ideas to emerge​. 

Codes should be based on the trends in the data. Going back to the example of the research question about pets, you might notice that when people refer to their own pets they use words like “smart,” “bright,” or “brilliant.” If these words keep coming up within one participant’s interview, or if many participants mention their pet’s intelligence, it may be worth coding, and you could highlight these quotes with a code like “pet intelligence”. 

Next, organize the initial codes into categories and subcategories, looking for connections between codes to form broader themes​. Maybe you’ve also created a code for when people talk about the intelligence of farm animals. It’s important to note that everything within a category doesn’t have to have the same valence. Perhaps you record one person saying ‘dumb pig,’ while someone else noted ‘pigs are as smart as 3-year old humans.’ Although the sentiments are conflicting, they could both be categorized as ‘farm animal intelligence.’ Then, the ‘pet intelligence’ code and the ‘farm animal intelligence’ category could be grouped into a theme of ‘animal intelligence.’ 

From these larger groupings or categories, identify a core category that encapsulates the main theme of your research. This category should unify all other categories and provide a central framework for your theory​. Finally, connect the core category with the other categories to form a comprehensive theory, refining the connections and ensuring that all data fits into your theoretical framework​.

As you continue to develop your theory, engage with peers to review and ensure its credibility. Reflect on your own biases and be mindful that they haven’t unduly influenced the research​. This is known as reflexivity. Just like quantitative research, grounded theory will require finalizing a comprehensive report that includes the theory, the process of its development, and its implications. Like any good research, findings should be grounded in the data (refer to your codes to support any claims or conclusions!), accessible, and usable by other researchers​​.