03 · What You Need to Know
What Analysis Planning Means in Qualitative Research
Planning Analysis Is Not the Same as Predefining the Findings
A qualitative analysis plan does not need to predict which themes, categories, explanations, narratives, or theoretical insights will emerge.
Its purpose is to establish how the research team intends to move from raw material to a defensible interpretation. That can include the analytical tradition, how material will be prepared and organized, when analysis will begin, how coding or interpretation will proceed, how analytical decisions will be documented, and how emerging insights may influence subsequent data collection.
The distinction is important. Planning the process supports rigor. Predetermining what the process must find would undermine the openness that many qualitative methodologies require.
Qualitative Analysis Often Begins While Data Collection Is Still Underway
In much qualitative research, analysis is not a separate phase that begins after the final interview or observation.
Methodological literature commonly describes qualitative collection and analysis as iterative. Early material is read, interpreted, coded, compared, or otherwise examined while further data are still being generated. Emerging insights can then influence what researchers ask, whom they sample, which cases they seek, and what phenomena require deeper investigation.
This relationship varies among qualitative traditions, but it means that postponing all analytical thinking until collection is complete can remove one of the methodological advantages of qualitative inquiry.
Flexible analysis
The analytical process can respond to emerging evidence according to the logic of the methodology.
Unplanned analysis
The researcher collects material without a defensible account of how it will be organized, examined, interpreted, or connected to the research question.
The Methodology Should Shape the Analysis Plan
There is no generic qualitative analysis procedure appropriate to every study.
Grounded theory, phenomenological approaches, ethnography, qualitative content analysis, thematic analysis, narrative analysis, discourse analysis, case-study approaches, and other traditions differ in their aims, assumptions, relationship to theory, and analytical procedures. Even methods sharing terms such as “coding” or “themes” may use them differently.
A study concerned with lived experience should not automatically adopt an analytical procedure simply because another qualitative study used it. Likewise, a study seeking to construct theory from data requires a different analytical logic from one intended primarily to summarize patterns of meaning across a dataset.
The analytical approach should therefore be selected because it fits the research question, methodology, data source, and intended form of knowledge.
The Research Question Still Guides the Analysis
Qualitative openness does not mean the analysis has no focus.
The research question identifies the phenomenon the study is trying to understand. Analytical procedures should help researchers interpret the material in relation to that question rather than merely reproduce the structure of an interview guide.
Recent methodological guidance notes that qualitative analysis should remain centered on the research question rather than treating participants' responses to individual interview prompts as though each prompt were an analytical category.
This is another version of the broader principle that every research question needs a defensible analytical path, even when that path is interpretive rather than statistical.
Decide What Counts as Data
Before collection, specify what material will enter the analysis.
Depending on the study, this may include interview transcripts, focus-group transcripts, field notes, documents, photographs, online interactions, recordings, artifacts, researcher journals, or other sources. The analytical significance of these materials may differ.
If interviews are recorded, decide how recordings will be transcribed and checked. If observations are central, determine how field notes will be produced and expanded. If contextual notes or researcher reflections contribute to interpretation, distinguish their role from participants' accounts.
These decisions affect what information survives the transition from fieldwork to analysis.
Plan How Data Will Be Prepared, Organized, and Protected
Qualitative projects can generate substantial amounts of sensitive and difficult-to-manage material. Analysis planning therefore overlaps with data management.
Before collection begins, researchers may need procedures for transcription, file naming, version control, de-identification, secure storage, linking field notes to recordings, tracking participant identifiers, and controlling access within the research team.
Ethical and institutional requirements may determine some of these decisions. They should not be improvised after dozens of recordings and transcripts have accumulated.
Decide Whether Coding Will Be Inductive, Deductive, or Combined
Qualitative coding does not always begin from a blank page, nor does it always begin from a fixed codebook.
An inductive approach develops codes or categories through engagement with the data. A deductive approach may begin with concepts informed by theory, prior research, a framework, or predefined analytical interests. Many studies combine the two, beginning with some sensitizing concepts while remaining open to material that does not fit them.
The appropriate choice depends on the methodology and purpose. What should be avoided is claiming that analysis was wholly inductive when the researcher actually imposed a substantial pre-existing framework, or claiming that a codebook was fixed in advance when it changed repeatedly during analysis.
Do Not Assume That a Codebook Must Be Final Before Data Collection
Some qualitative projects benefit from an initial coding framework, particularly when analysis is deductive, highly structured, or conducted by a larger team. In other studies, developing a detailed codebook before encountering the data would conflict with the analytical approach.
A plan can therefore specify how codes will be generated, reviewed, defined, revised, merged, divided, or retired rather than listing every final code in advance.
The question is not whether the codebook changes. It is whether the process of developing and changing it is methodologically coherent and documented.
Plan How Iteration Between Collection and Analysis Will Work
If early analysis is expected to influence later data collection, state how.
For example, emerging categories may lead researchers to revise interview prompts, seek participants with particular experiences, investigate contradictory cases, or pursue a developing theoretical idea. In grounded theory, constant comparison and theoretically informed sampling can make this interaction particularly central.
Iteration should nevertheless remain connected to the methodology and ethics approval. Substantial changes to recruitment, data collection, or study procedures may require institutional review or amendment depending on the context.
Plan How Analytical Decisions Will Be Documented
Qualitative interpretation involves judgment. Making that judgment visible can strengthen transparency.
Analytic memos can document emerging interpretations, questions, relationships among concepts, contradictory evidence, and decisions about codes or categories. Team meeting notes, codebook versions, decision logs, diagrams, matrices, and other records may also preserve the development of the analysis.
Such documentation does not transform interpretation into an automatic procedure. It provides an account of how the interpretation developed and helps researchers examine their own reasoning.
Reflexivity Should Be Considered From the Beginning
The researcher is not external to qualitative knowledge production. Background, assumptions, disciplinary position, relationships with participants, theoretical commitments, and role in the field may shape what is asked, noticed, interpreted, and emphasized.
Reflexivity involves examining those influences rather than claiming that the researcher has somehow removed them entirely.
Planning may therefore include reflexive journaling, team discussions, positionality considerations, or other practices appropriate to the methodology. These should serve genuine analytical purposes rather than becoming ceremonial statements added after the analysis is complete.
Team-Based Analysis Requires Decisions About Roles
If several researchers will analyze the material, determine how responsibilities will be distributed and how interpretations will be discussed.
Will researchers independently code some material? Will one researcher lead coding while others challenge or refine interpretations? How will disagreements be handled? Who maintains the evolving coding framework? How will analytical decisions be recorded?
There is no universal requirement that qualitative coders must achieve numerical agreement. Whether measures of intercoder agreement are conceptually appropriate depends on the methodology and analytical purpose. Some approaches use multiple analysts to seek consistency; others use differences in interpretation as material for reflexive discussion rather than treating one coding decision as objectively correct.
Software Is Not the Analysis Plan
Stating that interviews will be analyzed “using NVivo,” “using ATLAS.ti,” “using MAXQDA,” or in a spreadsheet does not identify the analytical method.
Qualitative software can support storage, retrieval, coding, annotation, comparison, visualization, and organization. It does not determine the epistemological position, decide what constitutes a theme, interpret participants' experiences, or establish whether an analytical claim is warranted.
The analysis plan should describe what researchers will do intellectually and methodologically with the material, not merely which application will hold it.
Plan Enough to Know Whether You Can Collect the Right Material
Thinking about analysis early can expose problems in qualitative data collection just as it can in quantitative research.
If you intend to compare perspectives across different participant groups, the sampling plan must provide access to those perspectives. If analysis requires detailed accounts of a process over time, a short structured interview may not generate sufficiently rich material. If contextual information matters to interpretation, the fieldwork procedures need to preserve it.
This is why the question is not whether qualitative studies should plan analysis before collecting data. They generally should. The more useful question is what kind of advance planning is appropriate for the methodology.
Some Analytical Decisions Should Remain Open
Qualitative analysis often develops as researchers become more familiar with the phenomenon and the material. Emerging codes, categories, relationships, theoretical interpretations, and decisions about where to pursue depth may therefore remain open at the beginning.
This is not a methodological weakness when openness is part of the design. The analysis plan should distinguish deliberate methodological flexibility from decisions that could reasonably have been made earlier.
The next step is therefore to determine what a qualitative analysis plan should specify in advance and which details should be allowed to evolve.
