Open coding and Axial coding in Qualitative data analysis

Open coding and axial (focused) coding are practical, two-stage ways to analyze qualitative data: first tag the text with descriptive, data-driven codes, then group those codes into connected categories that can later become themes. Although the terms come from grounded theory, the workflow fits many other qualitative approaches—case studies, thematic analysis, and other coding-based designs—because it’s fundamentally about moving from raw text to organized meaning.

Open coding is the initial, exploratory phase. As researchers read transcripts or documents, they avoid imposing prior frameworks or preconceived ideas. Instead, they create codes that act like “sticky notes” attached to specific segments of text, describing what the segment is about and what it suggests. This stage is often detailed—sometimes even line-by-line or sentence-by-sentence—because the goal is to code a large portion of the material while the analyst is still unsure where patterns will emerge. The result is typically a high volume of codes, often numbering in the hundreds.

In an example scenario about online learning, open coding might generate codes such as “technical issues,” “distractions at home,” and “lack of engagement,” alongside positive codes like “convenience of learning” (no commute) and “more engaging learning context.” The specific labels depend on what appears in the data, but the principle stays the same: codes should be descriptive and grounded in the text.

Axial coding—also called focused coding in some constructivist grounded theory approaches—comes next. This stage reorganizes the initial codes by grouping them into categories and looking for connections among them. For instance, the “technical issues,” “distractions at home,” and “lack of engagement” codes could be grouped under a category such as “challenges of online learning,” while “convenience of learning” and “more engaging learning context” could be grouped under “advantages of online learning.” The “connections” part is not treated as a mysterious analytic leap; it’s essentially the act of sorting and categorizing codes based on how they relate.

As the categories take shape, the analyst becomes more familiar with the dataset, which supports the next step: developing themes. The workflow is therefore staged—open coding produces many descriptive codes, axial/focused coding organizes them into connected groups, and that organization helps themes emerge.

A final practical question is whether axial/focused coding involves additional coding or only organizing what already exists. Different approaches exist. One option is to complete open coding for all transcripts first, then only group and reorganize in the second stage. Another option is to code part of the data, pause to organize those codes, and then continue coding the remaining material using the more focused, grouped structure. Either way, the core idea remains: axial/focused coding makes the analysis more targeted by turning a long list of initial codes into an organized structure that can lead to themes.