validity

• Validity in psychological research


  Researchers gather data and produce findings in psychological studies. We need to question if it is truthful to say that these findings accurately reflect natural behaviour. Psychologists have to consider both the internal and external validity of their results.
• Internal validity


  Questions the cause and effect relationship between the change the researcher made to the IV and the observed change in the DV. If the change in the DV was influenced by any other factor than the IV, the findings lack internal validity.
• External validity


  Questions if a study's findings can be generalised beyond the study. So from the sample used to the target population and from the experimental set-up to other real world settings and activities.
• Internal validity - social desirability bias


  Participants hide their genuine opinions/behaviours and instead act/respond in a more socially acceptable way to 'look good'.
• Internal validity - demand characteristics


  Participants think they have discovered the aim and behave in a way they believe will produce results supporting the researcher's theory.
• Internal validity - investigator effects


  The researcher's behaviour influences the participants' behaviour, including researcher bias, when the researcher either consciously or unconsciously influences the results. The participant's behaviour could be affected by the researcher using certain body language or tone of voice when participants behave as expected according to theory.
• Internal validity - uncontrolled extraneous variables


  Lack of control - this is not using standardised procedures, identifying and eliminating extraneous variables or not controlling for participant variables and randomly assigning participants to groups.
• External validity - ecological validity


  The extent to which the findings of any particular study can be generalised to alternative environments.
• External validity - mundane realism


  The extent to which the task/materials/activities used in an experimental set-up are similar to the stimuli experienced in the real world.
• External validity - population validity


  The extent to which the sample used in the study is representative of the target population e.g. gender or age.
• External validity - temporal validity


  The extent to which the findings of a study can be generalised to other time periods. Generally asked of older studies, questioning if the findings on topics like social influence or attachment would be the same if researchers conducted the study in modern society.
• Assessing validity - face validity


  Does the test appear to measure what it claims to be measuring?
• Assessing validity - criterion validity


  Confidence in the validity of a test increases if we can compare the data from a test to another measure of the same variable and identify a correlation.
• Assessing validity - concurrent validity


  The extent to which data from the newly created test is similar to an established test of the same variable conducted at the same time. A test of correlation assesses this; there is high concurrent validity if the strength of the correlation is 0.8 or higher.
• Assessing validity - predictive validity


  The extent to which performance on a test can predict future performance/outcomes/behaviour.
• Improving internal validity

  Improved by demonstrating a high level of control over variables.

  Control ensures the researcher is measuring any potential cause and effect relationship between the IV and DV, not the effect of an EV on the DV.

  Random allocation - controls participant variables.

  Standardised procedure - controls extraneous variables.
• Improving external validity

  Improved by demonstrating that findings are generalisable.

  Replication ensures the cause and effect relationship observed in the original study is not just limited to that one experimental/observational set up.

  Replicating findings in multiple settings improves ecological validity.

  Replicating findings with diverse groups of people improves population validity.