Observational design; behavioural categories; event sampling; time sampling

Observations (Naturalisitic)

AO1, Description:

This refers to the observation of behaviour in its natural setting. The researcher makes no attempt to influence the behaviour of those being observed or manipulate variables. The aim of the research is to observe naturally occurring behaviour.

(1) Disclosed/Overt Observation Participants are aware they’re part of an observation.

(+) A strength of this type of observation is that participants can give consent – therefore this type of observation ca be seen to be ethical.                                                                         (-) A weakness of this type of observation is that participants know that they are being watched and therefore may change their behaviour to suit the experimenter (respond with demand characteristics).

(2) Undisclosed/Covert Observation Participants are not aware they are part of an observation.

(+) Participants are unaware that they are being watched therefore, any behaviour observed will be naturally occuring.                                                                                               (-) Limited to public places (where people expect to be observed) therefore there is a limit to what behaviour can be observed.                                                                                       (-) Undisclosed observations can be argued to be unethical however, researchers believe that as long as only public behaviour is being observed then this type of observation can be deemed to be ethical.

Other observation types:

(1) Participant Observation: Involves the observer becoming actively involved in the activities of the people being studied. Participant observation can be either disclosed (people are told they are being observed) or undisclosed (participant is unaware of being observed).

(2) Non-participant Observation: Non-participant involves the researcher observing the behaviour from a distance; they do not become actively involved in the behaviour to be studies.

Designing a Naturalistic Observation

Remember that observations are not experiments and so there is no IV, instead behaviour is studies in the environment in which it would typically be seen.

You need to decide on the following things;

  • Aim
  • Hypothesis
  • Sample
  • Type of data
  • How data will be recorded


Types of sampling within an observation (ways of collecting data):

(1) Event sampling:

  • Involves counting the number of times that certain behaviours (event) occurs while observing an individual, using a behaviour schedule (quantitative data) A behavioural schedule is like a tally chart/a list of behaviours, everytime a behaviour is observed a ‘tally’ is placed next to that behaviour.

(+) This is useful if you have a lot of behaviours you need to record

(-) However, it does lack details (e.g. no information about preceding events)


(2) Time sampling

  • This means recording behaviours in a given time frame. i.e. you might record what the participant does every 60 seconds.

(+) This is useful if you have a lot of behaviours you need to record

(-) This means that not every behaviour is noted and so the end data can lack detail


Creating and using a behaviour schedule

1) Identify a list of behaviours (categories/events) that you would expect from the observation (e.g. observing in a day care might include sharing, shouting, crying etc).

Behaviour Schedule

2) Conduct a pilot study to ensure that behaviours have not been omitted from your behaviour schedule

3) Create a final list of behaviours (categories/events)

4) Complete the observation tallying whenever one of the categories is observed

Note: the categories must be objective and operationalised (measurable), cover all potential behaviours and be mutually exclusive (the observer should not have to mark 2 categories with one behaviour).

Reliability and Validity in Observations

Reliability: The data produced from an observation can be very subjective- for example, one observer might tick that a behaviour has occurred while another observer might not, making the observation unreliable.

How can reliability be improved? Inter-rater/observer reliability – to use the same behaviour schedule and to ask at least two observers to rate (or code) the behaviours viewed (the observation could be videotaped to make this easier). A correlational analysis is then completed on the results (to test for consistency between the observer’s results) and if there is a significant correlation, then the observation is said to be reliable (i.e., there is inter-rater/observer reliability). One way to objectively assess for inter-rater reliability is to carry out a statistical test (Spearman’s Rho or Peason’s R).

Validity: Ecological validity is likely to be high in a naturalistic observation because it involves studying natural behaviours in a natural environment (however, a natural setting does not guarantee high ecological validity – participants behaviour could be affected by the presence of the researcher).

Internal validity may be at risk if the coding system (behaviour schedule) is flawed. E.g. if an observer has to make an interpretation about a behaviour before coding it, then the behaviour may not be accurately recorded. The validity can also be affected by observer bias – what someone observes can be influenced by their expectations.


Evaluation, AO3 of Observations:


(1) POINT: A strength of naturalistic observations is that they can hold high ecological validity. EXAMPLE: For example, an observation of children’s playing behaviour in the school playground is representative of their real-life behaviour. EVALUATION: This is a strength because participants tend to behave naturally and results can usually be generalised to other real-life settings.


(2) POINT: Astrength of naturalistic observations is that they can have low demand characteristics when an undisclosed observation is used. EXAMPLE: For example, children covertly observed playing in the playground will behave naturally as they are not aware that they are being closely watched by an adult. EVALUATION: This is positive because the findings will represent their real-life behaviour.



(1) POINT: A weaknesses associated with naturalistic observations is that they lack control. EXAMPLE: For example, potential EVs could interfere with the behaviour of the participant, children being observed playing in a playground may behave differently depending on the temperature outdoors. EVALUATION: is problematic because observed behaviour might have occurred due to these uncontrolled variables and therefore a result cause and effect relationship cannot be established.


(2) POINT: A weakness of naturalistic observations is that they can be affected by observer bias. EXAMPLE: For example, observers have to interpret what they see and this can be affected by bias, such as if a researcher has predicted that boys will be more aggressive than girls, they may only report the behaviour that fits in with these expectations. EVALUATION: This is a problem because if researchers are selective in what they notice in this way, their findings are no longer reflective of the truth.