Grounded Theory Methodology
Grounded theory is a qualitative research methodology that develops new theories or conclusions directly from data, rather than testing one that already exists. The process is inductive, meaning the theory emerges from the numbers as they accrue. Here’s a general overview of how this shakes out:
Data collection - in the form of interview transcriptions or transcribed observations
Open coding - researchers meticulously read through raw data and break it down into discrete, conceptual parts. Each part is then assigned a descriptive label, or "code," to identify key ideas and patterns as they emerge directly from the data itself
Axial coding*** - researchers systematically connect and relate categories to build a more comprehensive theoretical framework (***See Pitfall #6 below)
Selective coding - This is the final stage of analysis where you identify the core category—the central phenomenon or main story that all other categories revolve around. You'll then refine and integrate all your categories to build a cohesive theoretical framework that explains the process or phenomenon you've been studying.
Success with the grounded theory methodology largely depends on investigators maintaining self-awareness throughout the process, continuously checking in with their own emotions, biases, and assumptions. This can be done via memos written to themselves or co-investigators, or by debriefing with third-party "devil's advocates." These strategies are referred to as reflexivity.
Pitfalls to grounded theory methodology include:
Sampling Bias: Sometimes, researchers may use theoretical sampling—letting the emerging analysis guide who to talk to next. This becomes bias when a researcher, often unconsciously, samples new participants they expect will confirm their budding theory, rather than those who might challenge it. They might stop sampling from a "difficult" or "confusing" group, leading to a theory that is neat but incomplete.
Leading Questions: This involves the researcher unintentionally leading participants or themselves. Because GT interviews are open-ended, the researcher's probes—either spontaneous or standardized—are critical. For example, "tell me about how stressful it was..." plants the concept of "stress".
Elitism (or "Eloquent Participant" Bias): A researcher may give disproportionate weight to data from one or two participants who are particularly articulate, insightful, or dramatic. The resulting theory then reflects that one person's experience rather than the collective pattern or process.
Cultural or Professional Bias: The researcher's own background acts as an unexamined filter. For example, a medical doctor conducting GT on "the patient experience" might interpret all data through a clinical, diagnostic lens, completely missing the social, financial, or emotional dimensions that are central to the patient's actual concern.
Premature Saturation: This is a bias driven by convenience. The researcher claims to have reached "theoretical saturation" (the point where no new insights are emerging) simply because they are tired, out of time, or out of money. They stop collecting data before the theory is fully dense and nuanced, resulting in a superficial or "thin" theory.
Procedural Bias (The "Glaser vs. Strauss" problem): Glaser and Strauss were the initial developers of this technique. The Glaserian approach (or "Classic GT") is a purely inductive method where the researcher remains passive, allowing theories to emerge from the data with no preconceived framework. The Straussian approach is more prescriptive, providing the axial coding step above. We’re getting into the weeds here, but basically, these guys couldn’t agree about whether axial coding was a strength of GT or antithetical to its spirit. To put it simply, the Glaserian approach might deviate significantly depending on the investigators’ freewheeling, whereas the Straussian approach risks forcing conclusions where none exist. Papers don’t commonly state which approach was taken, so pay attention to the methods section for clues.
Alternative approaches to studies requiring inductive reasoning include thematic analysis, which is more flexible and focuses on identifying patterns (themes) rather than building a dense theory, and Interpretative Phenomenological Analysis (IPA), which specifically explores how individuals make sense of their personal, lived experiences.