9.1.7, 9.1.8, 4.2.1, 7.2.1 - Observation and content analysis

Recording actions rather than reports

Observation records behaviour directly instead of asking someone to describe how they behave.

A researcher interested in turn-taking could watch a consented adult group task and record each invitation for another person to speak. This avoids relying only on participants' memory or their wish to appear cooperative. However, an action does not directly reveal its motive. A quiet participant might be listening, uncertain or tired; “quiet” alone does not demonstrate disengagement.

Quantitative observation uses counts, durations or other numerical records. A tally adds one mark for each defined occurrence; a frequency is the resulting total. Qualitative observation uses descriptive field notes about what happened, its sequence and context. Both can be gathered together. “Participant A passed the pen to B and said ‘your turn’” records an observable action; “A is naturally generous” adds an unsupported interpretation.

Define behavioural categories operationally before coding. “Invites a turn” could mean a spoken invitation naming another participant or a clear gesture offering them the shared task item. Specify how a continuous action ends and when a new occurrence begins. Without such a rule, one long invitation might be counted several times. Categories should cover the target behaviour and avoid double counting unless the coding system explicitly permits multiple codes.

In child psychology, the same distinction could guide a hypothetical authorised record of ordinary toy-sharing. A tally of explicitly defined toy offers gives quantitative data; a note describing who offered the toy, what was said and how the other child responded gives qualitative context. Neither alone reveals a child's motive or establishes a stable developmental characteristic. The observer-role choices below and additional child-specific protection are still necessary.

For a practical demonstration, use a supplied staged clip made with consent or an approved harmless adult task, not secret observation of vulnerable people. A written analysis of supplied material develops method skills but does not prove that the learner carried out a real investigation. Topic practical lessons apply these shared decisions to their particular behaviour and recording requirements.

Choosing the observer's role and setting

Observation types describe separate choices about awareness, participation, structure and setting.

In overt observation, participants know they are observed. Meaningful consent and withdrawal are easier, but awareness can cause reactivity: people change their behaviour because they are watched. In covert observation, they do not know. This may reduce that particular awareness effect, but raises serious consent, privacy and withdrawal issues; it does not guarantee unbiased or truly natural behaviour. Never treat “public place” as permission to record identifiable private behaviour.

A participant observer joins the activity or group being studied. Participation can give access to context and meanings an outsider misses, but balancing participation with accurate recording is difficult. Relationships and expectations may influence interpretation or behaviour. A non-participant observer records without joining the activity. This can make systematic recording easier and reduce direct interference, but the observer may miss contextual meaning; visible presence can still change behaviour.

A structured observation uses predefined categories and a systematic recording schedule. It can produce comparable quantitative data but may overlook unexpected behaviour. A naturalistic observation studies behaviour in its ordinary setting without introducing an experimental manipulation. It may reveal realistic behaviour, but the researcher has less control over competing influences. The two labels are not opposites: predefined categories can be used in an ordinary setting.

For example, a researcher openly watching a usual adult discussion from the side of the room and tallying invitations for others to speak uses an overt, non-participant, structured, naturalistic observation. If the researcher instead joins the discussion, only the participation dimension necessarily changes. If the researcher manipulates a prompt to compare conditions, the project includes an experimental intervention even if observation is the outcome measure.

Choose the combination for the aim and ethics. A consented staged adult task may lose some everyday realism but allow clear safe teaching and recording. Research with children needs appropriate adult permission, the child's assent and attention to ongoing willingness and protection; a child's discomfort matters even if an adult signed a form. Child-specific procedures are taught in their separate methods lesson.

Deciding when to record

A recording schedule determines which moments or occurrences become data.

Event sampling records every occurrence of a defined target event during the observation period. For example, tally every invitation for another adult to take a turn during a ten-minute group task. It is useful for counting a particular behaviour, including infrequent events, but a busy group may produce too many simultaneous events for one observer. Focusing on one event can also miss the surrounding behaviour that gives it meaning.

Time sampling records behaviour at predetermined times or intervals. One clear rule is to look at the focal participant exactly every 30 seconds and record whether an invitation is occurring at that instant. This reduces recording burden and gives coverage across the period, but may miss short invitations between checks. If intervals are used instead, state whether any occurrence during the interval, its dominant behaviour or something else counts. These rules produce different data and must not be mixed without explanation.