Concept
validity
The degree to which a measurement really measures what it is meant to measure; internal validity looks at the cause of a difference, external validity at whether findings extend to other contexts.
Validity is the degree to which a measurement really measures what it is meant to measure. Turkish sources describe it as an indicator of how correctly the intended characteristic is measured. Reliability seeks consistency, validity seeks accuracy: a survey that gives the same inflated number every time is reliable but may not be valid. Prior showed that news-exposure claims in surveys were on average three times higher than independent viewing measures; Parry and colleagues found that self-reported and logged digital media use are only moderately related.
At the study level it splits in two: internal validity asks whether a difference really comes from the cause under study; external validity asks whether the finding extends to other people, settings and times. Qualitative research developed its own counterparts of these criteria. See reliability.
Example
Measuring news-viewing time only by the person’s own claim.