Observations: Behavioural Categories & Inter-observer Reliability (AQA GCSE Psychology): Revision Note

Recording behaviour in an observation

• Behaviour categories are the specific, observable behaviours researchers look out for during an observation

• Researchers agree the categories in advance, so all observers are working from the same system before the observation begins

• Categories should be:

  • Clear and unambiguous - described precisely enough that any observer would code the same behaviour the same way

  • Mutually exclusive - categories should not overlap, so a behaviour fits into only one category. If two categories could both apply, observers won't know which to tally

• For example, an observation of children's aggression could use the categories:

  • PA = Physical aggression

  • ABL = Aggressive body language

  • NAB = Non-aggressive behaviour

• Each category must be operationalised into specific sub-behaviours, so it is precise and cannot be confused with anything else, for example:

  • PA: punching, kicking, shoving

  • ABL: pointing, shaking a fist, baring teeth

  • NAB: smiling, arm-linking, hugging

• These categories could then be broken down further, e.g. by whether it is boys or girls being aggressive/non-aggressive, and towards whom the behaviour is directed, for more detailed data

Inter-observer reliability

• Even with clear categories, an observation can still be affected by researcher (observer) bias

• Inter-observer reliability is the level of consistency between two or more trained observers conducting the same observation

• To establish inter-observer reliability:

  • All observers agree on the behaviour categories and how to record them before the observation begins

  • Each observer records the behaviour independently, to avoid one observer being influenced by another

  • After the observation, the two independent data sets (often tally charts) are compared and a correlation is calculated between them

  • A strong positive correlation (+0.8 or more) shows good inter-observer reliability and that the categories worked reliably

• Good inter-observer reliability means there is less chance that observer bias has affected the data

An example of a tally chart with behaviour categories.