Qualitative Research

What makes a study qualitative research

Qualitative research is an approach to answering a research question, not a single method or a synonym for “uses interviews.” A study is qualitative when four things hold together across its whole design, not just in one part of it:

• Non-numeric primary data — words, images, observed behavior, or documents, rather than counts or scale scores.
• A meaning- or process-focused question — how or why something happens, or how people experience or make sense of it, rather than how much, how many, or whether a measurable effect exists.
• Iterative, inductive analysis — codes and themes are developed from the data itself, often while more data is still being collected, rather than tested against a hypothesis fixed in advance.
• Contextualized, “thick” reporting — findings are described in enough detail that a reader can judge how they might apply elsewhere, rather than presented as a statistical estimate with a margin of error.

Qualitative research most often draws on one of a handful of established traditions — narrative research, phenomenology, grounded theory, ethnography, and case study are the five most widely cited (Creswell & Poth’s framework) — but the traditions are implementations of the approach, not the definition of it. For a full side-by-side of how qualitative research differs from quantitative research on data, sample-size logic, rigor criteria, and IRB review, see Qualitative Research vs. Quantitative Research.

Two worked examples

Example 1: An interview study of a lived experience

A researcher wants to understand how postdoctoral researchers experience the transition into an independent investigator role. They conduct 15–20 semi-structured interviews, ask open-ended questions about that transition, transcribe the recordings verbatim, and analyze the transcripts through iterative coding — reading early interviews while still scheduling later ones, refining the interview guide as themes emerge — to describe the shared structure of that experience. This is qualitative research on all four counts: the data is text, the question is about meaning and process, the analysis is inductive and concurrent with data collection, and the output is a thematic, contextualized account rather than a numeric estimate.

Example 2: A document analysis with no interviews at all

A researcher wants to understand how “data sharing” is defined differently across funder policy documents. They collect a set of published data management plan templates and policy guidance documents, code them thematically for how each source frames sharing obligations and exceptions, and describe the patterns and contrasts across sources. No interviews, no human subjects, and no numeric data are involved — it is still qualitative research, because the object of analysis is textual, the aim is interpretive (how meaning is constructed across documents) rather than measurement, and the coding is inductive rather than a fixed scheme applied to confirm a prior hypothesis.

Counter-example: when text data doesn’t make a study qualitative

A survey asks respondents a few open-ended questions, then every response is sorted into a fixed set of categories decided before data collection began, and the results are reported as frequencies — “62% of respondents cited time constraints as a barrier.” This is quantitative content analysis, not qualitative research, even though the raw data started out as words. The coding scheme is confirmatory and fixed in advance rather than emergent from the data, and the reported finding is a numeric proportion rather than an interpretive account of meaning or process. Researchers sometimes label this kind of study “qualitative” in a proposal or manuscript simply because the raw material was text — the methods literature calls the underlying operation “quantitizing” qualitative data, and the resulting study should be described and reviewed as quantitative (or, if a genuinely separate qualitative strand exists alongside it with its own inductive analysis, as mixed methods), not qualitative.

Why the distinction matters for research administration

Getting this right affects three concrete decisions a research office makes: how a study should be described in a grant proposal’s Approach section, which IRB review pathway (exempt, expedited, or full board) actually fits the design, and what a manuscript’s methods section needs to report to satisfy reviewers and journal guidelines. Those procedural details — sample-size logic (saturation vs. statistical power), the applicable rigor framework (trustworthiness vs. validity/reliability), and how each pathway maps to 45 CFR 46 — are covered in full in Qualitative Research vs. Quantitative Research: Key Differences, rather than repeated here.

Related CASRAI terms

• Quantitative-qualitative balance — the responsible-assessment principle that indicators should support, not replace, qualitative judgement.
• AI in qualitative coding — reporting expectations when AI tools assist with qualitative analysis.
• Informed consent — a human-subjects requirement that applies to interview- and observation-based qualitative studies the same way it applies to quantitative ones.

References

• Creswell, J. W., & Poth, C. N. Qualitative Inquiry and Research Design: Choosing Among Five Approaches (4th ed., 2018).
• Lincoln, Y. S., & Guba, E. G. Naturalistic Inquiry (1985) — the trustworthiness framework (credibility, transferability, dependability, confirmability).
• The SAGE Encyclopedia of Qualitative Research Methods and The SAGE Dictionary of Qualitative Inquiry, SAGE Research Methods.
• 45 CFR 46 (the Common Rule), Subpart A — the human-subjects review framework applicable to interview-, focus-group-, and observation-based qualitative research in the U.S.