Reliability and validity, explained simply
From choosing a thesis questionnaire to building a self-test, these two words come up everywhere. What is the difference?
In psychology we measure things we cannot see: anxiety, self-esteem, intelligence. So the trustworthiness of a scale or questionnaire is checked with two questions. The first is reliability, the second is validity.
Reliability
Reliability means consistency of measurement: whether measuring the same person under the same conditions again and again gives nearly the same result.
- Test-retest: how closely scores match when the same scale is given again after some time.
- Internal consistency: whether the items of a scale measure the same thing. Usually checked with Cronbach’s alpha; many studies treat 0.70 or above as acceptable.
- Inter-rater: how similarly two observers score the same behaviour.
Validity
Validity means whether the scale actually measures what it is supposed to measure.
- Content validity: whether all important aspects of the topic are covered by the items. Usually checked by experts.
- Construct validity: whether the score behaves as theory predicts: a strong relationship with scales it should agree with (convergent) and a weak one with scales of different concepts (discriminant).
- Criterion validity: whether the score matches a real-world standard, now (concurrent) or in the future (predictive).
A simple example
Imagine a weighing scale that always shows 2 kg more than the true weight. It is consistent, so it is reliable. But it does not show the correct weight, so it is not valid. In other words, validity is impossible without reliability, but being reliable does not make a measure valid.