psychlotron.org.uk Why is it important that psychological research be valid and reliable?
psychlotron.org.uk Today’s session You are learning about...You are learning to... Types and ways of assessing reliability and validity Check the reliability and validity of a quantitative questionnaire
psychlotron.org.uk Validity The extent to which something measures what it purports to measure Research studies InternalExternal EcologicalSampling Measurements ConcurrentConstructPredictiveFace
psychlotron.org.uk Reliability The extent to which something produces consistent results Research studies Measurements Split halfTest-retest QuestionnairesInter-observer/ inter-rater
psychlotron.org.uk A while ago you completed a questionnaire to measure stress. We are going to assess its reliability and validity.
psychlotron.org.uk We will use correlations to assess reliability and some aspects of validity. Which statistical test should we use and why?
psychlotron.org.uk Reliability Test-retest – Test-retest reliability is high if a measure yields similar scores on different occasions Split half – Split-half reliability is high if the odd scale items yield a similar score to the even scale items
psychlotron.org.uk Validity Face validity – Does it look OK? (Weak) Concurrent – Does it agree with other measures of the same thing? Predictive – Does it accurately predict what will happen in the future? Construct – Does it capture the entirety of the phenomenon it measures?
psychlotron.org.uk How would you assess the predictive validity of our stress questionnaire?
psychlotron.org.uk Is it necessarily desirable... – For a questionnaire to give consistent results – For a measure to agree with other measures?
psychlotron.org.uk Homework Write a short report about the reliability and validity of the stress questionnaire. Use the data from today’s class to comment on its test-retest and split half reliability and its concurrent validity (assume the other questionnaire is a valid measure of stress).