reliability

• Reliability is a measure of consistency
• In psychology, researchers want to ensure that measures are consistent between participants and over time
• Test-Retest Reliability:

  • Same person or group of people undertake the research measure on different occasions
  • Important to be aware of potential demand characteristics
  • Correlate the scores from the two occasions to determine reliability
• Inter-Observer Reliability:

  • Refers to the consistency of two or more observers in observing and recording behavior
  • Useful in situations where subjectivity is a risk
  • Important in psychology studies with applied behavioral categories, so they are used in the correct way
• Improving Reliability: Questionnaires

  • Identify questions impacting reliability and adjust them
  • Rewrite questions to reduce misinterpretation
  • Change open questions to closed questions to limit ambiguity
• Improving Reliability: Experiments

  • Control over variables influences reliability
  • Laboratory experiments have high reliability due to control over independent variables, making them easier ot replicate by following standardised procedures
  • Take more control over extraneous variables to improve reliabilityand prevent them from become confounding
• Improving Reliability: Observations

  • Observations rely on researcher's interpretations, important they are being objective and not subjective
  • Operationalize behavioral categories to improve reliability, clear and specific on what consititues the behavior
  • Categories should be clear and specific with no overlap, and no room for personal interpretation
• if the correltation is shown to be signifation then ath meaure is deemed ot have agood reliability, a perfect correlations is 1, but +0.8 is a good indication of relaibility
• in observer- reliability the psychologist would obseverce the same sitution or event seperately and then theri observation (scores) would be correlated ot see if they are relaible.
• Improving reliability : interviews

  • ensure same interviewer for all interviews to reduce researcher bias, a different reseachers can pose questions differently which can lead to different answers
  • if same interviewer cannot be used, then another should be trained to reduce bias
  • also changeing interview from unstructued to structed reduces researcher bias