We consider cases (1) to (3) as indicative of newly discovered TE families because they fulfill the 80/80/80 rule proposed by Wicker et al. 2007 (for a definition of TE family, see Introduction), with case (1) representing the strongest argument for validation as a novel family.
As Schuwirth & Van der Vleuten concluded: "the arguments of validation can only be made if the construct we want to assess is defined clearly enough and when all theoretical notions about it are sufficiently concrete" [ 34].
Finally, the argument against correlational validation is presented, and suggestions as to alternative validation techniques are offered.
What will validation validate?
It can be shown that DIC is an approximation of a penalized loss function based on the deviance, with a penalty from a cross-validation argument.
The concept presented in this debate, however, has provided a rational validation for arguments to preserve the rhetorical power of the descriptor 'global health' for more 'innovative' forms of research, education and practice compared to common 'global health' discourses.
While Walter's study did not adopt an argument-based approach to validation, the results raised issues relevant to the evaluation and explanation inferences of a validity argument for assessing pragmatic comprehension.
Drawing on an argument-based approach to validation, this study reports on the development of an analytic rating scale designed for a Spanish test for academic purposes.
Roever's instrument can be further improved when viewed from the perspective of an argument-based approach to validation.
Fig. 1 Four inferences in Kane's argument-based approach to validation (adapted from Kane, Crooks, & Cohen 1999).
A mix of these methods is of great help for focusing on pertinent candidates, with multiscale arguments for designing experimental validation.
