Reliability

• Reliability is the extent to which two or more measurements/observations are consistent with each other
• Test-retest reliability is a method of assessing the same person on two separate occasions
• Inter-observer reliability is the extent to which there is an agreement between two more observers
• An Inter-observer reliability score is high when above 0.80
• Split-half reliability is when participants scores on one half of a test are correlated with the other half
• Split-half reliability provides a measure of internal consistency
• Correlation Co-efficients are a method of measuring reliability by correlating the sets of scores using Pearson's rank
• In questionaires reliability can be improved my removing/rewriting ambiguous questions with low test-retest reliability
• In Interviews reliability can be improved by using the same interviewer or training all interviewers to ensure there are no leading questions asked (Use structured interviews with a script)
• In Observations reliability can be improved by making sure all behavioural categories are have been properly operationalised and that categories do not overlap
• In Experiments reliability can be improved by using standardised procedures so that results can be compared accurately