A doctoral student spends three weeks at the back of a primary classroom, studying how a teacher handles pupils who fall behind. When she starts analysing, her notebook reads “teacher very patient”, “positive atmosphere”, “one boy looks bored”. No times, no words spoken, no idea which boy. Three weeks of work are close to unusable.
Observation is the one method that shows you what people do rather than what they say they do. But its value lives entirely in the notes. Anything not written down properly, as far as analysis is concerned, never happened.
When to observe rather than interview
Observation pays off when behaviour differs from accounts (a doctor says she always explains the diagnosis, yet the consultation you watch lasts four minutes), when people are unaware of their own routines, or when interaction between several people is the point. It suits you less when you need thoughts, motives or past experiences.
Most strong projects combine the two: observe to see, interview to ask why. Observing first makes your interview questions far more concrete.
Choose your observer role
| Role | What it looks like | Gain | Cost |
|---|---|---|---|
| Complete observer | No interaction; people may or may not know | Minimal interference | Cannot ask what actions mean |
| Observer as participant | Openly present, some casual conversation | A workable balance | Stranger effect in the early sessions |
| Participant as observer | You work alongside them, for example as a teaching assistant | Deep insider understanding | Hard to write on the spot, easy to lose distance |
Choose a role and state it in your methods section. Readers need to know where the data were collected from in order to judge them.
What to watch: do not try to watch everything
Newcomers sit down and try to record it all. After an hour they are exhausted and have recorded what was most striking, which usually means most unusual rather than most important. Prepare an attention frame in advance, for instance:
- Space: layout, who sits near whom, movement routes.
- People: who is present, their roles, who is absent.
- Activities and sequence: what happens in what order and for how long.
- Verbatim talk: key sentences, word for word.
- Objects: how documents, phones and tools are used.
- What does not happen: the question nobody asks, the pupil who is never called on.
This is not a tick-box form. It keeps your attention balanced in the first sessions, and then you narrow down to what matters for your question.
Jottings on site: fast, short, time-stamped
While observing you only have time for jottings: keywords, verbatim phrases, times and a sketch of positions. A few conventions help:
- Write the time in the margin whenever something new starts: 09:14, 09:17.
- Use codes rather than names: T, P3 (third desk, blue jumper). Faster, and it protects identities.
- Use quotation marks for verbatim speech and square brackets for your own thoughts: [teacher skipping P3 because time is short?].
- Sketch the room on the first page of each session.
Where a notebook draws attention, type a few words on your phone or step out for a minute to write, then return.
Write up full notes within 24 hours
This is the decisive step and the one most often skipped. Memory for detail fades fast after a session. A practical rule: do not sleep before writing up. Experienced ethnographers often spend as long writing as they did observing, sometimes longer.
A full write-up has four clearly separated parts:
- Session details: date, start and end time, place, who was present, your role.
- Description: what you saw and heard, in time order, concrete enough for someone else to picture it.
- Analytic notes: ideas, emerging patterns, links to theory.
- Personal and methodological notes: your feelings, anything you did that may have affected the scene, things to do next time.
Keep description and interpretation apart
The most common error in field notes is writing conclusions in place of observations. Compare:
- Interpretation: “The teacher is very patient with weaker pupils.”
- Description: “09:22, P3 gets the division wrong for the second time. T walks to the desk, crouches beside him and says, “Let’s try it with 12 sweets.” The exchange lasts about three minutes while the rest of the class works on page 34.”
Description lets readers, and you six months later, judge for themselves what patience looks like. Interpretation closes off every other reading. Write interpretations by all means, but put them in the analytic notes.
Ethics and the observer effect
Observing in public spaces and observing inside classrooms, hospitals or workplaces carry very different obligations. In organisational settings you need permission from whoever is responsible, information for participants and ethics approval where your institution requires it. With children you need consent from parents and the school. Never photograph or record without explicit permission.
People change their behaviour in the first sessions. To reduce this, visit regularly, treat the first few sessions as familiarisation and flag their data, and compare behaviour early and late in the fieldwork.
A sample page of field notes
Session 7 – Class 3B – Tuesday – 08:45 to 10:10 – Role: observer as participant, back desk by the window.
Description: 08:47 T hands out worksheets and says, “Pairs today.” Pupils pair by seating, except P3 and P11 who have no partner; at 08:50 T puts them together...
Analytic note: third time I have seen pupils who get individual attention being paired with each other; possibly an informal ability grouping.
Method note: T asked whether I write down names; I need to restate the anonymity agreement.
What to do next: before your first session, draft a one-page attention frame and the four-part template above. After that first session, reread your write-up and strike through every interpretive sentence hiding inside the description. Building that habit early decides the quality of your whole dataset.
Câu hỏi thường gặp
What are field notes in qualitative research?
Field notes are a systematic record of what a researcher sees and hears in the field, made up of quick jottings on site and a full write-up afterwards. Good write-ups keep description separate from analytic and personal notes.
Should I record audio or video during observation?
Only with explicit permission and when recording will not change behaviour too much. Even with recordings you still need field notes to capture context, analytic ideas and things the device cannot pick up.
How many observation sessions are enough?
There is no fixed number. Observe until new sessions stop revealing patterns relevant to your question, and long enough to get past the period when participants behave differently because a stranger is present.
What is the difference between participant and non-participant observation?
In participant observation the researcher takes part in activities to understand them from the inside. Non-participant observation keeps a distance and does not intervene, which reduces interference but makes it harder to ask what behaviour means.
How do you reduce the observer effect?
Visit regularly over a long enough period, treat early sessions as familiarisation, stay unobtrusive and compare behaviour at the start and end of fieldwork. Explaining your purpose clearly also helps participants relax.
