02 Section A, Key question 3 - Observation and ethnography (part 3)

Participation and disclosure

An observer's involvement in an activity and the participants' knowledge of the research are two separate decisions.

Observation is the systematic study and recording of behaviour in a social setting. A researcher might watch interactions at an employment advice centre to explore how people seeking work are treated. This can reveal interruptions, informal rules and differences between what an organisation promises and what happens. Watching behaviour does not automatically reveal what an action means to the person performing it.

Participation answers “Does the researcher join in?” Disclosure answers “Do those studied know this is research?” An overt researcher can participate, and a covert researcher can watch without joining. These original imagined designs show the four combinations; they are examples for classification, not published studies or instructions to conduct covert research.

Participation can give access to shared routines and insider explanations. It can also make observing and recording harder: a person working a busy shift cannot continuously take notes. Non-participation can free attention for systematic recording, but a distant observer may misunderstand an exchange. Neither role guarantees accurate understanding.

Disclosure makes it possible to explain the study and seek informed consent. It may also encourage people to perform an ideal version of their behaviour. Concealment can reduce some reactions to being studied, but removes that opportunity for consent and can create serious privacy and trust problems. Being in public does not settle every ethical question about recording or publishing someone's behaviour.

Recording observational data

The recording procedure determines what observational evidence can show.

Structured observation uses categories decided in advance, such as whether an adviser interrupts a visitor before the visitor finishes speaking. The researcher specifies what counts as an interruption, observes selected periods and records each occurrence. Counts produce quantitative data, which describe amounts or frequencies numerically. In an original invented example, five of ten observed conversations contain an interruption. This is half of those ten conversations; it is not evidence that half of all advisers or all conversations behave this way.

Categories make comparisons possible, but they can impose the researcher's assumptions. “Visitor complied” might combine agreement, confusion and fear of losing support. A tally treats these as the same unless the design distinguishes them. A researcher interested in the meaning of compliance needs more than a count.

Unstructured observation records detailed events without limiting everything to a fixed checklist. Field notes are written records of observations, including setting, sequence, speech, gestures and the researcher's emerging ideas. Such descriptions and accounts of meanings produce qualitative data. They can preserve the context lost in a tally. Unstructured does not mean careless: the researcher still needs a clear purpose, suitable sampling of occasions and careful records.

Consider this original invented field note: “After the adviser mentioned a missed appointment, the visitor lowered their voice and asked whether payments would stop.” The observed speech and change of volume belong in the descriptive record. “The visitor felt intimidated” is an interpretation, not something the observer directly saw. A follow-up conversation could investigate that interpretation; the visitor might have lowered their voice to avoid being overheard. Keeping observation, reported meaning and researcher inference distinguishable makes the explanation more defensible.

Researchers may make brief notes during a visit and expand them promptly afterwards. Delaying risks forgetting sequence and exact wording. Audio or video can preserve detail when appropriate consent and data protection arrangements permit it, but a device may affect behaviour and cannot capture everything the observer experiences. Records need dates, times and context so later analysis does not detach an event from its circumstances.

A single project can generate both forms of data: an observer could count interruptions and describe how they occur. Participant observation is therefore not automatically qualitative, and non-participant observation is not automatically quantitative. The relevant distinction is what is recorded and analysed.

Access and field relationships

Useful observation depends on gaining entry, sustaining trust and recognising how the researcher becomes part of the setting.

A gatekeeper controls access, such as a manager, school leader or community organiser. Formal permission can allow entry, but it does not ensure access to every activity or the willing participation of everyone present. A manager might permit observation only of successful client meetings. The resulting account would omit the difficult encounters most relevant to unequal treatment. Access is therefore also an issue about the quality and selection of evidence.

Rapport is a relationship of trust and understanding. Listening, respecting agreed boundaries and explaining how records will be used can help an overt researcher maintain access. Trust may make people more willing to explain unfamiliar routines or allow observation of less formal situations. It takes time, and some participants may remain uncomfortable or refuse particular visits. Losing access halfway through can leave important processes unseen.

The researcher effect, or reactivity, occurs when the researcher's presence or role changes behaviour. Staff may become unusually polite when an observer is nearby. A participant researcher also changes the interaction by contributing to the activity: another person helping at a desk can shorten queues. Repeated visits may make the researcher less novel, but familiarity does not prove the effect has disappeared. Covert observation can still affect behaviour through the observer's participation, appearance or relationships.

Researcher characteristics can influence what is disclosed and how it is interpreted. Someone perceived as connected to management may hear fewer complaints; someone sharing a participant's background may receive a different account. Reflexivity means examining how one's position, actions and assumptions shape the evidence. A reflexive record might note that complaints stopped when a supervisor joined the room, then compare encounters under different conditions rather than treating silence as satisfaction.

There is also a risk of going native: becoming so identified with the group that critical distance is lost. Belonging can make members' explanations seem obvious and discourage recording evidence that challenges them. This differs from empathy, which helps understand another person's viewpoint without requiring the researcher to accept it as the whole explanation. Keeping analytical notes, discussing interpretations and actively seeking contradictory cases can help retain that distinction.