Doing the Literature Review in Grounded Theory studies (2 dilemmas)
Based on Qualitative Researcher Dr Kriukow's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.
Treat prior knowledge as framing and critical orientation, not as a set of concepts to impose on data during coding.
Briefing
Grounded Theory research doesn’t require a “blank mind,” but it does demand tight control over how prior knowledge shapes analysis. Earlier versions of Grounded Theory treated previous knowledge as something to avoid; newer approaches increasingly accept that researchers inevitably bring concepts from reading and experience. The key distinction is between using prior knowledge as a starting point—enough to critically frame what needs studying—and using it as a template to force data into pre-existing theories. In the example of a study on Polish migrants’ English language identity in Scotland, the researcher built the initial construct of “English language identity” through reading across psychology, social psychology, and social linguistics, drawing connections among ideas like self-concept and self-esteem. During data collection and analysis, however, additional reading stopped to reduce the risk of later theories steering interpretation. The researcher also used reflexive memoing to track how earlier concepts might influence coding and conceptual development, explicitly entering analysis with a “blank mind” and avoiding the use of literature-derived frameworks as models for what to look for.
That approach doesn’t eliminate the literature—it changes when and how it’s used. After developing the emerging explanation from the data, the researcher returned to the literature using a constant comparison logic: newly formed concepts were compared not only with more data, but also with what others had already described. The result was not a sense of failure or discouragement when similar concepts appeared under different names. Instead, finding overlaps strengthened the credibility of the findings, because the concepts that were “grounded in the data” had later been supported by prior work. The takeaway is practical: similarity to existing scholarship can validate the analysis, especially when the researcher can show that the concepts were not imported during initial coding.
A second dilemma follows from a common misconception about timing: many people assume the literature review must be written only after the Grounded Theory analysis, once new concepts are fully developed. Earlier Grounded Theory traditions sometimes encouraged postponing the literature review until after analysis, but that expectation collides with academic realities. Students are typically required to submit proposals or early reports, demonstrate familiarity with the literature, and justify a research gap before fieldwork begins. In practice, the literature review often has to be drafted early, then revised later.
Using the same English language identity example, the researcher had already written a literature review before the final conceptual framework emerged. Once the analysis produced new concepts, the literature review was revisited and restructured so that the newly identified constructs appeared as if they had been part of the original framing. That can raise a concern that the study looks less inductive than it was. The response is methodological transparency: if the methodology chapter clearly explains the sequence—how concepts were developed from data and only later integrated into the literature review—readers can understand that the final write-up reflects an iterative process rather than a pre-planned fit to existing theories.
Cornell Notes
Grounded Theory research can start with prior knowledge, but it should not use that knowledge as a set of concepts to impose on data. The practical model is: read enough to frame the problem and critically assess what matters, then limit further literature influence during coding and analysis, using reflexive memoing to monitor bias. After concepts emerge from data, returning to the literature through constant comparison can reveal that similar ideas already exist—often under different names—and that overlap can strengthen validity. A related timing dilemma is the belief that literature reviews must wait until after analysis; academic requirements usually force early drafting, followed by later restructuring to align the review with concepts that emerged from the study.
How can a Grounded Theory study use prior knowledge without letting it control the analysis?
What does “constant comparison” look like when revisiting the literature after analysis?
Why isn’t finding that “nothing is new” necessarily discouraging in Grounded Theory?
Why is postponing the literature review until after Grounded Theory analysis often impractical?
If the literature review is written early, how can it be aligned with concepts that emerge later?
Review Questions
- What is the difference between using prior knowledge as a starting point versus using it as a template during Grounded Theory analysis?
- How does returning to the literature after developing concepts from data strengthen (or threaten) validity?
- What practical steps can be taken when an early literature review must be written but later concepts emerge from Grounded Theory coding?
Key Points
- 1
Treat prior knowledge as framing and critical orientation, not as a set of concepts to impose on data during coding.
- 2
Stop or limit additional reading during data analysis to reduce the risk that new literature claims steer interpretation.
- 3
Use reflexive memoing to track how earlier concepts might influence conceptual development and coding decisions.
- 4
Apply constant comparison not only to data, but also to the literature after concepts emerge, to check for overlaps under different names.
- 5
Expect to revise an early literature review to incorporate concepts that emerge during Grounded Theory analysis.
- 6
Maintain methodological transparency so readers understand that literature integration can occur after concepts are grounded in data.