Validity

• what is validity?


  • the research achieves what it claims to do
  • the research study is legitimate and any research effect can be trusted not to be contaminated with unwanted variables
  • research findings should have external validity where they can be generalised beyond specific context of study
• what are the types of external validity?


  • population validity - when results can be generalised to other groups of people (other populations/samples)
  • ecological validity - when results can be generalised to situations outside the research setting
  • temporal validity - when results can be generalised to other periods of time
• what is face validity?


  • that you measure what you intend to measure
  • to ensure this a panel of experts in the research may be asked to asses the measure for validity
  • the simplest technique to test validity
• what is concurrent validity?


  • how well does the measure agree with existing measures
  • to ensure this we test participant with both the new test and the established test
  • high agreement (+0.8) = high concurrent validity
• what is predictive validity?


  • is measure associated with future behaviour
  • we can investigate this by following up participants to see if future performance is similar to performance on our measure
• what is temporal validity?


  • whether results remain true and valid over a period of time
  • e.g. Freud has been criticised as his ideas (e.g. penis envy) are seen as an outdated reflection of Victorian society in which he lived
• how do you improve the validity of experiments?


  • control group used to make sure IV only affects DV
  • control variables need to be thorough standardisation in order to reduce investigator effects and demand characteristics (single-blinded and double-blinded procedures can be used)
• how do you improve the validity of questionnaires?


  • lie scale used within results (additional questions which assess how truthful someone is in their answers) helps control social desirability bias
  • anonymity can encourage truthfulness, increasing validity
• how do you improve the validity of observations?


  • covert - high ecological validity, behaviour is natural and authentic
  • behavioural categories which are broad/ambiguous/overlap may reduce validity so must be clear and operationalised
• how can you improve the validity of qualitative methods?


  • higher ecological validity due to depth and detail allowing more subjectivity in reflecting a participants reality
  • interpretative validity must be demonstrated - researcher has shown their interpretation matches participants reality using direct quotes
  • triangulation - gathering evidence from multiple sources so interpretations can be validated