The observational method involves watch participants in order to obtain data and gather information about their behaviour.
Overt (or disclosed) observation is when the role of the observer is known to participants.
Covert (or non-disclosed) observation is when the participants are unaware that they are being observed.
Participant observation involves the observer also being a participant in the activity being studied and engaging with the participants as part of a social setting in order to observe and record their behaviours.
STRENGTHS OF PARTICIPANT OBSERVATION:
- the observer can have an insight into the participants real emotions and motives.
- if the observation is covert than the participants may behave as they normally would, eliminating demand characteristics and increasing validity.
- Also participants may reveal more than they would otherwise.
WEAKNESSES OF PARTICIPANT OBSERVATION:
- engaging in the social activities with the group would make it difficult to record data at the same time.
- being involved in the social group can make the observer subjective.
- Covert observations raise practical and ethical issues as participants are not able to give informed concent and distress could also be caused.
Structured observation is where the researchers pre-decide and define behavioural categories in order to record specific behaviours and how often that behaviour appears.
To record these behavioural categories, common techniques include photographs, videos, audio recordings (of spoken descriptions), hand-written notes or direct records inputted into a computer (these can take the form of descriptions, ratings or codings/check lists)
If there is one or more observer they should record the same information when observing the same events. to achieve this they practice the data gathering techniques by simultanious observation, re-watching video recordings, pre-discussing the definitions so that they achieve a high level of agreement before the collection of data for the study begins. this increases inter-rater reliabilty and thus improves the data.
Inter-rater reliability is the extent to which different observers will produce the same records when they watch the same event.
STRUCTURED OBSERVATIONS: EVALUATION
- operational definitions can be developed in a pilot study to be certain that all key actions are included. this increases validity.
- operationally defined behavioural categories agreed between observers are likely to be reliable.
- practicing the use of data collection techniques can improve inter-rater reliabilty.
- photographically/video recorded data collected can be re-analysed, increasing validity as behaviours are less likely to be missed.
- simple definitions of behaviour may not convey sufficient meaning. eg. 'lifting a hand' may be friendly or intimidating etc. this lowers validity.
- total numbers of behaviours within categories may be relatively meaningless without context. this lowers validity.
- predetermined behavioural categories may be limiting if new behaviours become aparent during the study.
- data may be subjective as they are based on the observers perspective.
Event sampling is a data-collection technique which involves a researcher recording an event every time it happens (eg. using a check list). The events are usually pre-determined behavioural categories so the researcher knows specifically what to record.
- It can record every occurance of each behaviour to give a complete record.
- records are easy to obtain and analyse as they are just totals.
- It gives no indication of time spent on each behavioural category.
- it gives no indication of the order in which the recorded events occured.
Time sampling is when a researcher observes and records behaviours that occur within a specific, pre-decided time frame (eg. every 60 seconds) In instantaneous scan sampling only the action performed at the start of each pre-set interval is recorded. In predominant activity sampling the observer records the behaviour that was performed the most during the pre-set time interval. In one-zero sampling the researcher records each behaviour within the time period (like event sampling but divided into time slots)
- gives indication of the order of, and the relative time spent on each behaviour.
- Instantaneous scan sampling can be very reliable
- one-zero scans record the occurance of infrequent behaviours.
- it is difficult to record as many different behaviours as you can in event sampling
- timings have to be precise, 'noisey' alarms may cause demand characteristics