A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Qualitative data analysis involves systematic examination of non-numerical data (interviews, observations, documents) to identify themes, patterns, and meanings that illuminate the research question. Coding is the fundamental process of labeling units of text or behavior with conceptual categories that organize data into interpretable patterns. Thematic analysis identifies recurrent themes across participants; grounded theory builds theoretical understanding from data; phenomenology focuses on subjective lived experience. Reliability requires consistent coding by multiple coders and transparent documentation of procedures.
Code a small qualitative dataset independently, then compare codes with another coder to identify disagreements and refine operational definitions of codes.
Qualitative analysis is less rigorous than quantitative (actually, qualitative analysis requires systematic procedures and careful documentation). Qualitative findings are simply opinions (actually, systematic analysis of data produces evidence-based interpretations).
Your prerequisite on descriptive research methods established that some research questions cannot be answered by counting outcomes — they require understanding meaning, experience, and process. Naturalistic observation gave you the methodological move of systematically watching behavior in context. Qualitative data analysis is what happens after you have collected the data: interviews transcribed, field notes written, documents gathered. The challenge is that this data is rich, contextual, and non-numerical, which means the path from raw data to conclusions requires a different kind of rigor than statistical analysis — but rigor nonetheless.
Coding is the foundational operation. A code is a label applied to a unit of data — a phrase, a sentence, a paragraph — that captures what that unit is *about* at a conceptual level. In open coding (common in grounded theory approaches), the analyst reads through the data without predetermined categories, labeling whatever seems meaningful: "expresses frustration," "mentions family obligation," "uses avoidance strategy." This initial pass is deliberately exploratory. In subsequent rounds, codes are compared, merged, split, and reorganized. Axial coding identifies relationships between codes: which codes seem to cluster together, and what causes, contexts, or consequences surround them? Selective coding identifies the central theme or core category that integrates the others into a coherent account. The progression moves from raw particulars toward conceptual abstraction.
Thematic analysis is a more flexible approach that focuses on identifying, analyzing, and reporting themes — recurrent, meaningful patterns that appear across participants or data sources. A theme is not just a topic that comes up frequently; it captures something important about the data in relation to the research question. The six-phase process (familiarization, generating codes, searching for themes, reviewing themes, defining and naming themes, writing up) is iterative, not linear: you may return to earlier phases when a theme that looked coherent turns out to be two different phenomena, or when a minor code in early rounds reveals itself as central.
Establishing trustworthiness — the qualitative analogue of reliability and validity — requires systematic procedures. Inter-rater reliability is assessed by having two or more coders independently code the same data, then comparing their codes using Cohen's kappa or percent agreement. Low agreement signals that the code definitions are ambiguous and need refinement — the same process you use when refining operational definitions in quantitative research. Audit trails (detailed documentation of every analytic decision: why a code was created, why two codes were merged, why a particular theme was dropped) allow other researchers and readers to evaluate the reasoning behind the analysis. Member checking — sharing interpretations with participants to assess whether they recognize the findings as authentic — is another trustworthiness strategy unique to qualitative work. The goal is not objectivity in the quantitative sense, but reflexive rigor: being transparent about the analyst's perspective and systematic about the process.