Validity

• Validity - whether an observed effect is genuine and represents what is actually ‘out there’ in the real world. 

  Data can be reliable, not valid.

  For instance, a test that claims to measure intelligence (IQ test) may produce the same result every time when the same people are tested but not measure what it is designed to do. 
• Ecological validity refers to whether findings can be generalised from one setting to another, most particularly generalised to everyday life. 

  This may not be related to the setting but more to the task that the participants are asked to perform.
• Temporal validity - findings should be consistent over time. 

  For example Asch’s study may lack temporal validity because it was conducted during a conformist era in American history.
• Face validity

  A basic method to test validity - does the test measure what it’s supposed to measure ‘on the face of it’?

  This is achieved by simple ‘eyeballing’ the measuring instrument or by passing it to an expert to check.
• Concurrent validity

  A new intelligence test, for instance, may be administered to a group of participants. 

  Their scores are then compared with performance on a well-established test (correlation should exceed +0.80 for validity).
• Improving validity

  Experiments - control group means that the researcher is more confident that changes in the DV were due to the effect of the IV. 

  Standardised procedures minimise the impact of participant reactivity and investigator effects.

  Questionnaires- lie scales control for the effects of social desirability bias; respondents are assured that all data submitted is confidential. 
• Improving validity

  Observations - behavioural categories that are well-defined, thoroughly operationalised and not ambiguous or overlapping

  Qualitative data - interpretive validity demonstrated through the coherence of the reporting and the inclusion of direct quotes from participants.

  Triangulation involves using a number of different sources as evidence