Data collection in observation

• Unstructured observations:

  The researchers record all relevant behaviour but have no specific system. ​

  • Get lots of data​
  • But it is qualitative data
  • Risk of observer bias because you are so involved​
  • Subjective as research only records behaviours that catch their eye ​
• Structured observation:

  Fully planned in advance with a coding system and predetermined sections. ​

  • Quantitative data​
  • Observers need to agree on the behaviours
• Behavioural Categories (in structural observations)

  In order to record what the researcher sees the targets behaviour has to be broken down into categories – like a check list. 
• This is similar to operationalising the variables.​

  • Makes data collection easier​
  • Categories have to be really clear, measureable, observable, self evident.
  • No overlap
• Event sampling - The observer decides in advance what types of behavior (events) she is interested in and records all occurrences.

  All other types of behavior are ignored.​
• Time sampling - The observer decides in advance that observation will take place only during specified time periods (e.g. 10 minutes every hour, 1 hour per day) and records the occurrence of the specified behavior during that period only.​
• Instantaneous (target time) sampling - The observer decides in advance the pre-selected moments when observation will take place and records what is happening at that instant. Everything happening before or after is ignored.​

  ​
• Advantages of observations :

  • Can be better than self-report. (Self-report can be unreliable because “what people say they do is often different from what they actually do.”)​
  • High ecological validity
  • Can provide a hypothesis for an experiment.  ​
• Limitations of observation:

  • little or no control of extraneous variables
  • Possibility of observer bias. Researchers might ‘see’ what they were hoping or expecting would happen.​
  • There are ethical and moral issues if participants do not know they are being observed e.g. deception or invasion of privacy. ​

  ​
• ​Inter-observer reliability:

  Researcher do not conduct observations on their own. Bias may occur.​

  ​

  For 2 or more researchers to establish Inter-observer reliability they must be trained to be looking for the same things.​

  ​

  • Familiarise themselves with behavioural categories​
  • Observe together in the pilot study​
  • Compare results and discuss them​
  • Analyse their data and correlate it with each other and make a overall judgement using that data.