Where quantitative research counts and measures, qualitative research seeks to understand how people make sense of their lives, in their own words and contexts. It works with text, talk, images and observed behaviour — not numbers.

A wink is not just an eye movement: it could be a twitch, a flirtation, a signal or a parody. Thick description records enough context to tell which — that interpretive layer is the point.

Behind every study sits a paradigm — a set of beliefs about what reality is (ontology) and how we can know it (epistemology). These quietly decide what counts as good evidence.

Developed by Glaser & Strauss (1967), grounded theory builds theory inductively from data rather than testing a pre-set hypothesis. Coding and data collection proceed together until a theory emerges.

Phenomenology asks what an experience is like from the inside — the essence of, say, living with a disability, or of a mother's first contact with an anganwadi. It brackets the researcher's assumptions to centre the participant's world.

Ethnography immerses the researcher in a community over time — living among, observing and participating — to understand culture, norms and everyday practice from within. Its roots are in anthropology.

Strong for how and why questions about contemporary events you cannot control. A single SHG federation, studied through interviews, records and observation, can illuminate a whole model.

Qualitative research is flexible, but flexibility is not the same as improvisation. A clear design — question, approach, participants, methods, analysis plan — is what makes the study rigorous and defensible.

It is a working map, not a hypothesis to confirm. Expect to revise it as fieldwork teaches you what you missed.

Unlike a survey fixed before launch, qualitative design is often emergent: early interviews reshape later ones, and the sample grows in the direction the data points. This is a feature, not sloppiness.

Qualitative sampling does not aim for a statistically representative slice. It aims to select cases that are information-rich — the people, sites or events most likely to illuminate the question.

In grounded theory, theoretical sampling means deciding whom to approach next based on what your emerging analysis needs — chasing the gaps and testing the categories, not filling a quota.

Saturation is a judgement, not a magic number. You reach it when additional cases confirm rather than extend what you already understand.

Qualitative studies are often built on a dozen interviews or a single case — and that is legitimate. The goal is analytic generalisation (to theory and concepts), not statistical generalisation (to a population).

Most development qualitative work rests on the semi-structured interview — a guided conversation that follows a flexible list of topics while letting the participant lead where it matters.

A probe gently deepens an answer without leading it. The richest data usually comes after the first reply, when you ask the participant to go further.

The researcher usually holds more power — of class, language, education, institution. Participants may tell you what they think you want to hear, or what is safe to say. Naming this is the first step to managing it.

The data is not just what individuals say — it is what the group produces together: the consensus, the debate, the silences.

Group members should be similar enough on what matters (gender, age, status) that everyone feels safe speaking — but varied enough to spark discussion.

When you analyse an FGD, attend to the interaction: where did consensus form, where did it crack, what could not be said? The disagreements are often the richest data.

People cannot always articulate what they do — and sometimes what they say differs from what they do. Observation captures practice, routine, ritual and the taken-for-granted that no interview surfaces.

Sitting through gram sabha meetings month after month, or working alongside ASHA workers, reveals how things actually run — the workarounds, hierarchies and informal rules no document records.

Your gender, caste, class, language and institution change what people show and tell you. Reflexivity does not remove this — it makes it part of the analysis.

Participatory methods hand analytical power to communities themselves — people map, rank, diagram and interpret their own reality. The researcher facilitates rather than extracts.

Associated with Robert Chambers and the IDS at Sussex, PRA grew through extensive practice across India and South Asia in the late 1980s and 1990s, deeply shaping NGO and government fieldwork.

A transect walk is a structured stroll across the community with residents, observing and discussing land, water, housing and problems as you go. The walk surfaces what a seated meeting misses — and reaches the hamlets on the edge.

When community members draw their own map — on the ground with sticks, seeds and chalk — they include what matters to them and place themselves in it. The map becomes a conversation, not just a record.

Powerful for engaging youth, women and others whose perspectives are usually filtered through outsiders — and for advocacy, where the images speak directly to decision-makers.

Qualitative projects drown in material — recordings, transcripts, field notes, photos, consent forms. Disciplined data management from day one is what makes analysis possible and ethical.

Transcription converts recordings into text you can analyse. It is slow — roughly four to six hours per hour of audio — and full of interpretive choices.

A code might be a phrase like 'fear of travel' or 'pension delay'. You apply it everywhere that idea appears, so you can later gather and compare all instances.

A codebook is your living dictionary: each code, its definition, when to apply it (and when not), and an example. It keeps you consistent — and lets a second coder work the same way.

A theme is not just a frequent topic; it is a pattern of meaning that says something important about the research question.

Coding organises; interpretation explains. The analytic leap is asking: what is going on here? Why this pattern? What does it tell us about the question — and about what stays unsaid?

A rigorous analyst actively seeks negative cases — data that does not fit the emerging story — and either revises the explanation or accounts for the exception. Confirmation alone is not analysis.

Qualitative research cannot be judged by validity, reliability and generalisability in the statistical sense. Lincoln & Guba (1985) offered four parallel criteria for trustworthiness.

Stating your positionality and tracking your influence is central to confirmability. It is not navel-gazing — it is how qualitative research demonstrates that findings come from the data, not merely from the researcher.