Observational Design

• Observational design refers to how a psychologist would plan and conduct an observation.

  This includes how they plan to collect their data in the observation.
• There are two types of observational design:

  Structured

  Unstructured
• Structured:

  Preferable to use these observations; they aim to be objective and rigorous.

  Researcher uses a pre-determined list of behaviour categories and sampling methods.
• Unstructured:

  Researcher records all relevant behaviour, but has no system.

  May simply write down everything you see.
• Structured Observation categories are:

  Behavioural categories

  Time Sampling

  Event Sampling
• Behavioural Categories-

  Researcher needs to be very clear on exactly what behaviour they're looking for.

  It is operationalising-breaking up behaviour in a set of components so it can be measured.
• Objective-

  Researcher should not have to make guesses about behaviour.

  Categories have to be observable.
• No waste basket-

  All possible behaviours are covered and avoiding a 'waste basket' category which loads of different behaviours are thrown, unclear where behaviour should be categorised.
• Independent of each other-

  Categories should not overlap, researcher has to mark two categories at one time.
• Time Sampling-

  This method records behaviour within a particular time frame.

  E.g. noting an individual is doing every 30 seconds or some other time frames.
• Event Sampling-

  This involves counting the times a particular behaviour/events occurs in an individual or target group.
• Inter-rater reliability:

  Extent to which observer's are in agreement with their measurements.
• Observer bias->

  more than one observer, one may miss important details- may end up with biased observations.
• Important to make sure that the data is recorded objectively and in an unbiased way.

  Observers have to be trained before hand to ensure that they have inter-rater reliability.
• This includes:

  >Observers should familiarise themselves with behaviour categories.

  >Do a pilot study to check the categories are appropriate/clearly defined.

  >Compare data they have found and discuss differences in the interpretations.

  >Do the real observation.

  >Analyse the data, correlate the findings to get final figure.

  If results are similar enough, we have IRR.