Validity

• what is validity?


  • the extent to which a test measures what it claims to measure

  • how meaningful our results are

  • three main aspects: control, realism, generalisability
• control 


  • how well the experimenter has controlled the experimental situation

  • control it’s important as without it, researchers cannot establish cause and effect relationships

  • without control we cannot state that it was the IV which caused the change in the DV

  • result could have been caused by another variable called an EV or a CV (variables not controlled by the experimenter and which may affect the DV)
• realism

  • if an experiment is too controlled, or the situation is too artificial, participants may act differently than they would in real life, therefore the results may lack validity

  • mundane realism is how well an experiment reflects real life, if an experimental situation has high mundane realism it would be high in external validity - can refer to both the setting and the tasks the participants are asked to do
• conflict between control and realism

  • increasing control often involves simplifying or standardising the experimental environment which may sacrifice

  • enhancing realism may introduce extraneous variables or unpredictability which can reduce control over the experimental conditions

  • needs to be a balance of the two
• generalisability 


  • psychological research is to produce results which can then be generalised beyond the setting of the experiment

  • if an experiment is lacking in realism = unable to generalise

  • even if an experimenter is high in realism we still may not be able to generalise

  • e.g. participants may be all from a small group of similar people = low population validity
• issues with generalisability 


  • limited diversity

  • cultural bias

  • socioeconomic bias

  • educational bias

  • age bias

  • sampling bias
• internal validity 


  • things that happen inside the study

  • internal validity asks whether we actually measured what we intended to measure

  • can we be certain that it was the IV which caused the change in the DV?

  • if aspects lack validity, the results of the study are meaningless and we can make no meaningful conclusions from them

  • can be effected by a lack of mundane realism - could lead the participants to act in a way which is unnatural, thus making the results less valid

  • internal validity can be affected by extraneous/confounding variables
• EV/CV: situational variables


  • anything to do with the environment of the experiment (time of the day, temperature, noise levels etc)

  • affects validity - something about the situation of the experiment could act as an EV if it has an effect on the DV

  • can be overcome by the use of standardised procedures which ensure that all participants are tested under the same conditions
• EV/CV: participants variables 


  • anything to do with differences in the participants (age, gender, intelligence, skill)

  • affects validity - the differences between participants cause the change in the DV, e.g. one group may perform better on a memory test than another because they are younger or more motivated

  • can be overcome by using a repeated measures design, matched pairs could be used too
• type 1: face validity

  • a participant looks at a test and can tell what it is supposed to measure

  • for example, whether the questions on a stress questionnaire are obviously related to stress

  • if a student were to sit a biology paper and see questions in French, the student would doubt the validity
• type 2: concurrent validity

  • when a test will produce the same (or similar) results to a benchmark test

  • an example of this would be when students write mock exams

  • concurrent validity makes sure that similar tests are produced
• type 3: content validity

  • looking at your method of measurement and deciding whether it measures the intended content

  • you could ask an independent expert on the assessment of stress to evaluate the measurement to be used - the expert might suggest improvements, or might approve of the method, thus dealing with content validity
• type 4: construct validity

  • the extent that a test measures the target construct

  • in the case of the stress measurement, we would look at a definition of stress and consider whether the questions were relevant to this construct
• type 5: predictive validity

  • concerned with whether the scores on a test predict what you would expect them to predict

  • we would expect people who score high on a stress questionnaire to have higher blood pressure - therefore we can check this out as a means of assessing the predictive validity of measurement