Grounded Theory

Glaserian GT (Glaser, 1992) emphasizes allowing theory to emerge naturally through open coding and memo-writing without forcing structure; researcher assumes neutral stance. Straussian GT (Strauss & Corbin, 1998) formalizes axial coding, mapping causal relationships, conditions, and consequences; more procedural and structured. Constructivist GT (Charmaz, 2006) explicitly acknowledges that researchers co-construct theory through dialogue with data and participants; emphasizes reflexivity and researcher positioning. All three use constant comparison and theoretical sampling; choice depends on research goals and epistemological preference.

GT is not driven by target sample size but by theoretical saturation. Homogeneous, straightforward phenomena may saturate at 12–15 interviews; complex, heterogeneous phenomena may require 30+. The first few interviews may generate many codes; later interviews yield fewer new codes until saturation. Document saturation by tracking code frequency and novelty; when three consecutive interviews add no new categories, saturation is likely approached.

Purposive sampling recruits a defined sample upfront based on inclusion criteria. Theoretical sampling, unique to GT, is sequential and data-driven: early data guide the selection of subsequent participants to fill gaps, test emerging concepts, or explore variations. For example, if initial interviews suggest role differences in coping, you theoretically sample nurses vs. physicians vs. family members to explore how role affects experiences. Theoretical sampling continues until saturation is achieved.

GT generates theory inductively from data; hypothesis testing validates pre-specified predictions. GT is not hypothesis-testing research. However, theories generated by GT can subsequently be tested in quantitative studies (e.g., survey or RCT). GT begins without hypotheses; findings are presented as emergent theoretical propositions, not hypothesis confirmations.

Document constant comparison systematically through memo-writing, explicitly noting decisions and contradictions. Maintain coding consistency by revisiting earlier codes as new data emerge and refining definitions. Use peer debriefing—discuss emerging themes with a colleague to test assumptions. Conduct member checking if feasible (sharing preliminary findings with participants). Ensure audit trail: preserve raw data, coding decisions, memos, and theoretical development process for potential review. Report sample characteristics, sampling method, saturation indicators, and researcher reflexivity (acknowledgment of bias, prior experience).