observations

• what is an observation?

  Watching, noticing and recording peoples behaviour.
• why are observations used?

  -to study particular behaviours

  -to show a more realistic picture

  -to research behaviours that we wouldn't otherwise be able to research because of ethical issues (like the effects of a child being separated from their parents - it isn't unethical since it already happened)
• why can observations be better than self-report techniques?

  Participants cannot lie about their behaviour - researchers see the behaviour.
• why can self-report techniques be better than observations?

  Observations do not tell us the why behind behaviours - they just show us the behaviours. Self-report techniques can help understand the why behind behaviours.
• controlled observation

  The environment is controlled as much as it can be - there is a regulation of variables too (controlled observations normally happen in lab settings). Participants are likely to know they are being observed.
• controlled observation - AO3

  strength(s):

  -more likely to establish a cause and effect relationship, this helps predict behaviours

  -high internal validity because of the control of extraneous variables

  weakness(es):

  -low ecological validity

  -hawthorne effect
• naturalistic observation

  When participants are observed in their natural environment. The behaviour is not controlled and the researchers do not interfere.
• naturalistic observation - AO3

  strength(s):

  -increased ecological validity

  -decreased hawthorne effect

  weakness(es):

  -harder to establish cause + effect

  -more extraneous variables
• participant observation

  The observer joins in and takes part in the activity.
• participant observation - AO3

  strength(s):

  -increased internal validity, more insight into behaviour

  -decreased hawthorne effect

  
  

  weakness(es):

  -decreased internal validity as the researcher is doing multiple tasks
• non-participant observation

  Researchers do not get involved in the activity, they are separate from the participants and just observe what they are doing.
• non-participant observation - AO3

  strength(s):

  -increased internal validity, the researcher is focused on observing the behaviours

  
  

  weakness(es):

  -increased hawthorne effect
• overt observation

  Participants know that they are being observed.
• overt observation - AO3

  strength(s):

  -informed consent is given

  
  

  weakness(es):

  -increased hawthorne effect

  -decreased internal validity
• covert observation

  Participants do not know that they are being observed.
• covert observation - AO3

  strength(s):

  -increased internal validity, more genuine behaviour

  -decreased hawthorne effect

  weakness(es):

  -ethical issues, participants cannot give informed consent (but can get retroactive consent, prior general consent or presumptive consent to avoid this)
• structured observation

  When the behaviours researchers are going to observe are preselected. Observers only look for and record these preselected behaviours (ignore all other behaviours).

  Structured observations collect quantitative data, use behavioural categories and use some form of sampling.
• behavioural categories

  The behaviours researchers expect to observe are operationalised - they are broken down and made clear.

  -should be objective.

  -should include all possible behaviours that could happen.

  -should be mutually exclusive (categories should not overlap with each other).
• event sampling

  Recording behaviours every time they happen.

  
  

  -could be hard to record everything that is observed, decreases internal validity.

  -researcher might miss details if the event is complex e.g., lots of detail at one time.
• time sampling

  Recording behaviours according to a specific time frame e.g., recording behaviours every 30 seconds. Ignore all other behaviours and only record the behaviours you seen when the timer goes off.

  
  

  -decrease internal validity as some behaviours will be missed if they do not happen at the right time. Not representative.
• unstructured observation

  Researchers observe however they feel best, they look at all behaviours participants carry out as there are no given categories.

  Normally used in pilot studies to help gain an understanding of the behaviours they need to be aware of.

  Unstructured observations are subjective; could only be looking at some behaviour, not all behaviour but provides rich info + collects qualitative data.
• observer bias

  The observer skews what they observe/record because of their personal views and expectations.

  Can be reduced when using a structured observation, behavioural categories should be objective and can use interobserver reliability to check reliability.
• using interobserver reliability

  -at least 2 observers

  -observe the same behaviour (normally a recording)

  -use the same behavioural categories

  -complete this separately

  -compare results (+.8 or above = reliable)