OBSERVATIONAL TECHNIQUES

• Definition of an observation – a type of correlation/ nonexperimental research in which a researcher observes human behaviour. 
• Strengths of observational methods

  S: what people say and do is different, so observations allow us to observe this behaviour. 

  S: researchers can capture spontaneous and unexpected behaviours  

  S: provides information of what people do 
• Weaknesses of observational methods

  W: observer bias – often affected by what people want to see (solution: train the observers or having multiple observers 

  W: allows us to see what people do but not how they think or feel. 
• Naturalistic observation – takes place in the environment where the behaviour would normally occur, and the researcher does not interfere in any way with what is happening e.g. observing animals in a zoo or children in a playground. 

  S: high external validity and able to generalise to real life 

  W: hard to replicate due to uncontrolled environment. 

  W: high levels of extraneous variables. 
• Controlled observation – researcher has some control over the variables and may change but this observation may reduce the naturalness of the environment and participants aware of observation. 

  S: can focus on specific aspects of behaviour 

  S: less extraneous variables 

  W: low external validity so less able to be generalised 

  W: artificial environment means artificial, less natural behaviours. 
• Covert observation – participants not aware they are being studied but likely to be informed after to ensure ethics. 

  S: natural behaviours – less demand characteristics and increased internal validity 

  W: ethical issues to be considered such as informed consent and privacy 
• Overt observation – participants are aware they are being observed and have given informed consent – observers try to be as unobtrusive as possible. 

  S: more ethically acceptable  

  W: Knowledge of observation may lead to demand characteristics 
• Participant observation – observer may become a part of the group they are observing  

  S: provide more insight into the behaviour of the group  

  W: lose objectivity if the observer begins to identify with the group (‘going native’) 
• Non-participant observation – observer is purely watching or listening to the behaviour – keeps a distance and does not interact at all  

  S: more likely to be objective 

  W: more likely to be far removed from their participants and not gain as much insight. 
• ISSUES IN THE DESIGN OF OBSERVATIONS: 

  Things the researcher would have to plan when doing an observational study 

  12 marker – case scenario – designing own observational research 
• Structured vs unstructured

  STRUCTURED: 

  • Involves pre-determined behaviours and sampling methods 
  • Appropriate for larger observations with more people  
  • Makes use of behavioural categories  
• Structured vs unstructured


  Strengths of structured interviews:

  • Recording of data is easier and more systematic 
  • Produces quantitative data – easier to analyse 

  Weaknesses of structured interviews:

  • Data lacks detail 
• Structured vs unstructured

  UNSTRUCTURED: 

  • Researcher writes down everything they see 
  • Produces rich data 
  • Appropriate for smaller observations 
• Structured vs unstructured


  Strengths of unstructured interviews

  • More detailed and rich data 

  Weaknesses of unstructured interviews

  • Produces qualitative data - more difficult to record and analyse 
  • Greater risk of observer bias – only recording certain behaviours if no behaviour categories. 
• Behavioural categories

  Breaking down behaviours into smaller components that are observable and measurable (operationalisation of behaviour) also known as a behaviour checklist. 

  MAKE SPECIFIC EXAMPLES!! 

  For example, aggression in the playground can be broken down into kicks, punches, swears etc 
• Evaluation of behavioural categories

  S: makes data collection more structured and objective 

  W: problematic if categories are ambiguous or unclear 

  W: all target behaviours should be included or may result in ‘dustbin categories’ where many behaviours are deposited 
• Sampling methods

  Event sampling – a target behaviour is established (in a behaviour checklist) and then is recorded (tally marks) every time it occurs 

  S: when the target behaviour is infrequent and may be missed if time sampling is used 

  W: if the target behaviour is too complex, the observer may overlook details 
• Sampling methods

  Time sampling – a target individual (or group) is identified and then the researcher records their behaviour in a fixed time e.g. every 5 minutes. 

  S: effective in reducing no. of observations that must be made 

  W: the times when behaviour is sampled may be unrepresentative of the observation as a whole 
• Inter-observer reliability

  A single observer may miss important details or may only notice events confirming hypothesis – researcher bias 

  Can be reduced by either multiple observers or making a video recording to watch back later.