The choice of gold standard for evaluating tests for caries diagnosis.

All research is intended to provide new information. Studies should be designed in such a way that their results can be generalised to either a population or a process. Researchers function within a common paradigm, first described by Kuhn. 1 A paradigm is a framework containing patterns of thought which are ‘commonly agreed upon’ within a certain discipline. Medical science operates today within the paradigm developed in the last century for the systematic exploration of humans. 22 This so-called ‘critical clinical school’ required that clinical procedures were based on empirical evidence rather than the anecdotal approach. 3 The empiricists used the criterion of verification until Popper 4 formulated criteria for falsification and testing of a theoretical hypothesis. While his approach does not answer all the questions, it is still the basic paradigm for clinical research today. A major part of radiological research is devoted to the evaluation of diagnostic tests for the detection or monitoring of disease. As has been described in several recent textbooks, 5‐9 a new test must be validated against the true diagnosis, usually called the ‘gold standard’, to be scientifically acceptable. Any robust gold standard must fulfil three criteria: 8,9