Observational techniques and design

• Types of observation


  Naturalistic/controlled

  Covert/overt

  Participant/non-participant
• Naturalistic observation


  Watching and recording behaviour in the setting within it would normally occur.
• Controlled observation


  Watching and recording behaviour within a structured environment (e.g. Strange situation)
• Covert observation


  Ppts behaviour is watched and recorded without their knowledge/consent
• Overt observation


  Ppts behaviour is watched and recorded with their knowledge/consent
• Participant observation

  The researcher becomes a member of the group whose behaviour they're watching and recording
• Non-participant observation


  The researcher remains outside of the group whos behaviour they are watching and recording
• Evaluate naturalistic/controlled observations


  • NO have high external validity, therefore findings are more likely to be generalised as behaviour occurs in its normal environment
  • NO lack of control, replications are difficult and extraneous variables are more likely, patterns more difficult to derive
  • CO may lack mundane realism, low internal validity meaning generalisations cannot be made to real-life situations
  • CO may have less extraneous variables
• Evaluate covert/overt observations


  • CO removes issue of participant reactivity and decreases demand characteristics, increases validity of data gathered
  • CO has issues of ethics with informed consent as those studied do not know and may not want their behaviours to be noted
  • OO more ethically acceptable but may alter behaviour to appease researchers and appear more socially acceptable
• Evaluate ppt/non-ppt observations


  • PO researcher can experience situation as ppts do, increased insight into situation and therefore may increase validity of findings
  • PO may risk loss of objectivity if researcher identifies too strongly, line between researcher and ppt is blurred
  • NPO allows researcher to maintain objective distance from ppts
  • NPO may lose valuable insight to really understand behaviours if too removed from study
• Types of observational design

  Structured/unstructured observation
• Unstructured observation


  • Researcher may write down everything observed, producing detail-rich accounts of behaviour.
  • May be appropriate for small-scale studies with few ppts
• Structured observation


  • Where target behaviours are simplified and focused on
  • Allows researcher to quantify their observations using a pre-determined list of behaviours and sampling methods
• What are behavioural categories?

  Where a target behaviour is broken up into components that are observable/measurable
• Types of sampling methods used in a structured observation

  Event sampling

  Time sampling
• Time sampling


  When a target individual/group is first established and then behaviour is recorded within a fixed time frame (e.g. every 60s)
• Event sampling


  When a target behaviour/event is first established and recorded every time it occurs
• What is inter-observer reliability

  • When two or more researchers carry out an observation to ensure they are consistent in their judgements to support the observation and inferences from behaviours made
  • Calculated by correlating each pair of observations made
• Evaluate structured vs unstructured observation

  • SO easy to record data, systematic method if using behavioural categories, produces quantitative data which may be easier to compare and derive conclusions from
  • USO produces qualitative data, harder to record and analyse objectively
  • USO benefits from richness of detail and understanding the possible wider context of behaviours rather than reducing it to a number
  • USO may have a high risk of observer bias, may only record eye-catching behaviours that do not contribute to forming a conclusion
• Evaluate the use of behavioural categories

  • :) able to make data collection structured and objective, easy to measure
  • :( if categories are too ambiguous it can be difficult to identify and study
  • :( Categories should not overlap as it is difficult to tell the difference
• Evaluate the use of sampling methods

  • :) ES is useful when the target behaviour/event happen infrequently (could be missed using TS)
  • :( ES, If specified event is too complex, observer may overlook important details
  • :) TS, effective in reducing the number of observations that have to be made, allowing a focus on specific interactions that may be imperitive
  • :( TS, behaviour sampled may be unrepresentative of the observation as a whole