Validating a questionnaire demonstrates its internal consistency and reliability. When a questionnaire has been validated as reliable, it should elicit the same answers each time it is given to the same population. Many social sciences use validated questionnaires. In validating a questionnaire, researchers seek to determine that the questions reflect the concept that the questionnaire is intended to measure correctly and comprehensively.
The same population sample should not be used for the validation of the questionnaire and the actual clinical study, but the population used for validation should be representative of the population to be used in the study. Sample size is typically 50 to 100 participants, though a larger sample size is sometimes required for validation of a questionnaire. Validation should be done within two weeks to determine test-retest reliability.
Before validating a questionnaire, hypotheses should be established regarding anticipated differences in scores between various subgroups of the population being studied, as well as regarding any correlations with other studies. Longitudinal pre- and post-testing studies should also be established to determine questionnaire changes. The questionnaire design and change during validation must represent the situation in which the questionnaire is actually to be used. The questionnaire being used should measure what the researchers want to study and should be modifiable to be appropriate for the population being studied.