Improving Reliability

• Questionnaire 


  -Some questions may be removed or rewritten.

  -If questions are ambiguous they may be interpretated differently on the different occasions doing it in retest.
• Interview 


  -use the same interviewer or one which is fully trained so o leading or ambiguous questions are asked.

  -Use a structured interview where the interviewers behaviour is controlled by fixed questions, unstructured is more free flowing and less reliable.
• Observation


  -Make sure all the behavioural categories have been properly operationalised.

  -Categories should not overlap, all behaviours observed should be covered. e.g. hugging and cuddling are similar therefore only one should be used for the categories.
• Experiment 


  -Need to have standardised procedures.

  -Allows correlations to be made from the behaviours to be consistent so we can replicate and falsify findings.