Validity

• Definition - the extent to which the results of something are legitimate.

  Whether the results actually represent what is out there
• Internal validity - when there researcher has measured what they have intended to measure

  Internal validity is good when the results are due to the manipulation of the independent variable. Therefore demand characteristics, order effects can affect internal validity
• External validity - the extent to which findings an be generalised beyond the situation in which it was found.

  External validity is made up of ecological, population and temporal validity
• Ecological validity - is good if the results are generalisable from the setting in which they were gained to other settings
• Population validity - is good if the results are generalisable to other people, other than those used
• Temporal validity - is good if the results are generalisable to other times
• Ways of assessing validity

  • Face validity
  • Concurrent validity
• Face validity - refers to looking at whether the test, scale or measure appears 'on the face of it' to measure what it is supposed to measure.

  Look at the measure and see if it is measuring what it is supposed to
• Concurrent validity - this refers to looking at the results of a test and comparing them to the results of another recognised and well established test

  The results of the two different tests are correlated and a significant positive correlation of at least +0.8 indicates concurrent validity
• Ways of improving validity - experiments

  Using a control group to compare results against

  Standardising procedures (keeping everything the same) can reduce investigator effects

  Single blind technique and double blind technique - reduces demand characteristics and investigator effects
• Ways of improving validity - questionnaire

  Using a lie scale - including a duplicate question which is rephrased to see if the participant is lying

  Telling the participants that results will remain anonymous
• Ways of improving validity - observations

  Using a covert and natural observation which allows higher ecological validity

  Clear behavioural categories
• Ways of improving validity - case studies and interviews

  Triangulation - the results of the interview or case study are combined with other sources of evidence like interviews from family members, personal diaries which corroborate the interview or case study data

  See if they match or support each other