reliability

• what is reliability?


  Consistency of the extent to which the measurement device or test use the same approximate results when utilized repeatedly under similar conditions
• what is assessing?


  Making a decision/judgement about how reliable your data/study is
• how do you improve a study?

  Make it more reliable/consistent
• how do you assess the reliability of observations?


  The observation could be recorded so that behaviour could be watched and coded again, if there is agreement between the data produced in both the observation is said to be reliable however the observer could be biased

  There could be two or more observers making separate recording if there is agreement between the record by all researches (+0.8 correlation) the observation is said to be reliable this is called inter-rater reliability
• what inter-rater reliability?


  when there is agreement between the data recorded by all researchers (+0.8 correlation);
• how do you improve the reliability of observations?


  -Make sure that the behavioural categories were operationalised clearly enough

  -Giving all of the researchers to practice using the behavioural categories that they can respond more quickly the more practice you are the more accurate you are
• how do you assess the reliability of questionnaires?


  Test retest reliability

  inter-interview a reliability
• what is test retest reliability?


  The test/questionnaire is given some to some participants

  The same test/ questionnaire is given to the same participants sometime later, usually there is a short interval

  If there is agreement in scores both times the same test/questionnaire is given if there is at least +0.8 correlation agreement then the test/questionnaire is said to be reliable
• what is inter-interviewer reliability ?


  One interview, I could interview one person twice with a short interval if there was agreement in the answers given by the same same person both times it said that the interview is reliable

  The interview could be filmed and watched twice ( by either the same interviewer or another interviewer to see if they get the same results

  Might as well both interview if there is >/= +0.8 correlation Agreement between the two interviewers, then it said that the interview is reliable
• how do you reduce the ambiguity question asked?


  Questions need to be re-examined and rewritten to ensure no ambiguity
• why are laboratory experiments considered more reliable than other types of experiment?


  Because they allow for tight control over extraneous variables
• why are laboratory experiments considered more reliable than other types of experiment?


  Because they allow for tight control over extraneous variables