Observations

• observation


  - a non-experimental technique where the researcher watches and records natural behaviour of ppts without manipulating the IV
• controlled observation

  - aspects of environment are controlled, in an attempt to give ppts the same experience (lab setting)

  - controls extraneous variables, standardised procedures; however it is artificial
• naturalistic observation

  - takes place in the 'real world'; places the ppts are likely to spend their time in

  - high realism as ppts more likely to show natural behaviour, external validity; however uncontrolled extraneous variables
• overt observation

  - the ppts can see the researcher, are aware their behaviour is being observed for a study

  - ethical due to informed consent; however demand characteristics/social desirability bias more likely to be present
• covert observation

  - ppts are not aware they are being observed and cannot see someone taking recordings (observer can still be physically present)

  - more likely to be free from demand characteristics/social desirability bias; however more unethical, as cannot give informed consent
• participant observation

  - the researcher joins the group being observed and takes part in the group's activities/conversations

  - can build rapport (more trust/comfort from ppts can lead to more natural behaviour); however researcher can lose objectivity
• non-participant observation

  - the researcher is separate from the ppts, recording observations without taking part in the groups' activities

  - more likely that the researcher remains objective in their interpretation;

  - however lack of trust/rapport can cause researcher to miss out on important insights
• operationalised behavioural categories

  - the behaviours need to be clearly identifiable and measurable

  - eg: aggression = number of pushes, punches and kicks
• time sampling

  - researcher records all relevant behaviours at set intervals, eg: everything for 15 seconds, every minute etc.

  - more flexibility to be able to record unexpected behaviour (as not restricted to categories);

  - can miss behaviour that happens outside the intervals
• event sampling

  - researcher records/tallies every time a behaviour occurs from the list of operationalised behavioural categories

  - as long as behaviour included in the list of categories, it will be recorded at any stage of the observation

  - may miss relevant behaviour if it is not present in the list of behavioural categories
• inter-observer/rater reliability

  - 2 or more trained observers conduct the same observation

  1: agree and use the same checklist of operationalised categories

  2: observation is conducted separately by each observer

  3: compare the 2 independent data sets, a test of correlation (spearman's rho) can assess the correlation strength

  >> a correlation of 0.8 or stronger is generally accepted