Superiority: Always and everywhere? On some misconceptions in the validation of dynamic testing

Abstract

Dynamic assessment, as an alternative approach to assessing intellectual capacities, focuses on examinees’ ability to benefit from learning opportunities provided within the assessment process. The level of appreciation of the potential advantages of this assessment concept is not mirrored by the extent of its utilisation in practice. One reason for this constraint might be that, allegedly, the proponents of this approach have not yet succeeded in showing sufficient evidence for the validity of their assessment tools. The diversity within the field of dynamic assessment, characterised by a variety in the goals pursued, the methods employed, and philosophical perspectives on measurement makes it difficult to pass sentence regarding its validity. In the study presented, a set of dynamic tests has been used to exemplify a suggested strategy to evaluate dynamic tests in terms of validity. This strategy emphasised: (1) the explication of the construct dynamic tests are aiming at, and its relationship to other constructs; (2) the definition of construct representative external criteria; and (3) the demonstration of both predictive and incremental validity. Furthermore, it was conceptually argued and empirically demonstrated that a variation of correlations between tests and external criteria across different (sub-) samples or studies can be explained as differential validity of dynamic tests, which is compatible with the conceptual understanding of the construct addressed in such learning tests and would be misinterpreted as an indicator of inconsistencies in validity-related findings in dynamic tests.