observational techniques

• what are the 4 observational techniques?

  • covert and overt
  • participant and non-participant
  • naturalistic and controlled
  • structured and unstructured
• covert observations

  • 'undisclosed' - people don't know they're being observed e.g. using a one way mirror
  • participants may be informed of their involvement after the study
• evaluating covert observations

  +investigator effects are less likely compared to overt since the investigator is hidden which means behaviours are more natural

  -ethical issues with its deign as participants aren't aware there taking part so can't fully give consent
• overt observations

  • 'open' observations where participants know they're being observed e.g. publicly filming
• evaluating overt observations

  +more ethical than covert as informed consent can be given in advance

  so they have the right to withdrawal

  -possibility of investigator effects - investigator could influence behaviours so participants change them due to demand characteristics (knowing the aim)
• participant observations

  • the person observing also takes part in observation
  • it can be covert - group member quietly observing others without their knowledge
  • or it can be overt - group member using camera to record behaviours and participants being fully aware of this
• evaluating participant observations

  +researcher can obtain detailed data due to close proximity to participants

  -possibility of investigator effects - presence of investigator may influence participants' behaviour through demand characteristics which reduces validity
• non-participant observations

  • observer doesn't take part in the observed activity
  • common in educational settings e.g. someone sat in the back observing a lesson
• evaluating non-participant observations 


  +investigator effects are less likely compared to participant observations as they are often not visible

  -due to a lack of proximity, may miss some behaviours from a distance
• naturalistic observations

  • carried out in an unaltered setting where observer doesn't interfere at all e.g. observing people's behaviour in a shopping centre
• evaluating naturalistic observations

  +high ecological validity as natural behaviours are observed

  -reliability issues- not possible to replicate exact conditions so test-retest method cat be used
• controlled observations

  • under strict conditions e.g. lab setting or observation room where extraneous variables (time of day or noise) can can be controlled
  • can be overt or covert
• evaluating controlled observations


  +easily replicated to check reliability

  -low external validity due to artificial environment so participants may change behaviours accordingly
• structured observations

  • researches uses coded schedules to organise data into behavioural categories which can be overused and measured
• evaluating structured observations

  +researcher can compare behaviours across groups of participants

  -low internal validity - researcher may miss some crucial observations which are linked to the aim
• unstructed observations

  • every instance of observed behaviour is recorded in great detail
  • more usual in naturalistic observations
  • useful if behaviour doesn't occur often
• evaluating unstructured observations

  +rich data obtained - due to detail, researcher is able to obtain.comprehensive view of human behaviour which adds to its high validity

  -observer bias due to lack of behaviour categories meaning there isn't a valid representation of all behaviours so may lead to a problem with inter observer reliability due to lack of consistency
• time sampling

  • recording behaviour at prescribed intervals e.g. every 10 seconds
  • STRENGTH - allows for better use of time as fewer observations are made
  • LIMITATION - some relevant behaviours may be missed if they occur between intervals
• event sampling

  • record behaviour every time it happens
  • STRENGTH - every relevant behaviour will be recorded
  • LIMITATION - may miss some behaviours if everything is recorded at once