observation techniques

• covert

  unaware of being observed
• overt

  aware of being observed
• covert

  :) realistic behavior = high internal + external validity

  :( ethical issues = lack of informed consent

  :( hard to record data (need CCTV/from a far distance) = less detailed info
• internal validity

  whether the study, conduct + analysis answers the research questions without bias
• external validity

  whether the findings can be generalised to other contexts
• overt

  :) less ethical issues (consent + right to withdraw)

  :) easier to record data

  :( demand characteristics = lower internal validity
• controlled

  = manipulated environment
• naturalistic

  = natural behavior in a natural setting
• controlled

  :) higher control over variables

  :) higher replicability

  :( less realistic behavior - demand characteristics
• naturalistic

  :) more realistic behaviour - higher internal + external validity

  :) allowed to study behaviour that may be seen as unethical in a lab

  :( low replicability (maybe)

  :( lack of control over variables
• participant

  =observer is part of the observation
• non-participant

  =observing from a distance
• participant

  :) more in-depth info from being involved

  :) unique insight to behaviours not available to outsiders

  :( can become too involved with ps = bias
• non-participant

  :) remains objective (no bias)

  :) easier to record data

  :( less in depth info as an outsider

  :( more likely to misinterpret behaviour
• structured

  :) more objective measurements of behavior = less bias

  :( less realistic behaviour in controlled settings

  :( more likely to miss unanticipated behaviour
• unstructured

  :) realistic behaviour = high internal validity

  :( low replicability = unique

  :( difficult to record data (missed info)
• time sampling

  = observations occur at specific time intervals (first 10secs of every minute)
• event sampling

  = observations based on previously decided behavioural events/categories and noted when they occur
• behavioural categories

  = list of specific behaviours the researcher is focused on. they must be operationalised, unambiguous, not overlapping and observable.

  -noted with a tally or checklist
• primary data

  = collected first hand by the researcher or researcher's team
• secondary data

  = has been collected by others, including meta-analysis which looks at a collection of various studies