Reliability

• Reliability 


  consistency in results of a study / research/ test/ procedure. if the researcher replicates their study they will get similar results
• External reliability 


  extent to which a measure is consistent when repeated eg the results of a study are consistent with a exact replication at a different time or different participants
• Internal reliability 


  extent to which different parts of a measure are consistent with its self
• Assessing internal reliability 


  split-half method used in questionnaires and interviews where test is split into 2 parts . participants complete both parts to see if each half of the test correlate a strong correlation indicates internal reliability
• Assessing external reliability 


  test retest - repeat the study using the same procedures on different occasions a high correlation between test scores indicates good external validity inter rater reliability - where 2 or more observers record behaviours during the same observation using same behavioural categories then they test the correlation between the results good reliability if there is a positive correlation of 0.8
• Improving reliability in observations

  observations can be improved by giving training to observers using correct behavioural categories and using a pilot study to identify poorly defined behavioural categories , operationalisation is the process of clearly defining a behavioural category
• Interviews and surveys 


  use structured interviews rather than unstructured. the interview will then include a script the interviewer can follow ensuring participants has similar experiences
• Questionnaires


  used closed questions to reduce the range of possible answers use an established questionnaire than creating a new one
• Experiments


  use standardised procedures for each participant so they all have the same experience and environmental conditions the same . use established tests as measures rather than creating new tests