Valid assessment of students‘ development of professional competence over their course of studies in business and economics - A quasi-experimental longitudinal study (WiWiKom II project)


The WiWiKom II project is based on the competency model and the test instrument developed and tested in WiWiKom I. The aim of WiWiKom II is the objective, reliable, and valid assessment of students‘ development of professional competence in business and economics that allows valid test interpretations. Building on the validation approaches from WiWiKom I, WiWiKom II examines more in-depth validation questions and expands the results gained so far by including higher education and individual change measurement. In order to follow these validation focuses, WiWiKom II implements a longitudinal design over a period of four years and four measuring dates. WiWiKom II follows a quasi-experimental design that examines constructs according to the multitrait-multimethod matrix approach and two group comparisons for convergent, discriminant, incremental, and predicative validity. The quasi-experimental variation is achieved by means of the choice of samples, as not only the two study domains (economics and social studies) but also the level of training (orientation, advanced, and specialization studies in the Bachelor degree course) are varied. In order to examine the discriminant validity, both the WiWiKom test and an abridged version of the Berliner Test zur Erfassung fluider und kristalliner Intelligenz [Berlin test for the assessment of fluid and crystalline intelligence] (BEFKI) is implemented, which in terms of general cognitive abilities assesses fluid as well as crystalline intelligence by means of a general, not subject-related knowledge test. In a longitudinal study, we examine to what extent competence change can be explained by variables relating to economics degree courses (e.g. attended learning opportunities) over the course of studies. Competencies are modeled on the basis of IRT and SEM models. The possibility of group comparisons is tested by means of measurement invariance analyses. Due to the expected missing data in the longitudinal data, MEM as well as growth curve models are estimated according to frequentist and Bayesian approaches. Furthermore, propensity score matching is implemented in order to be able to conduct valid group comparisons of groups with different sizes and distributions of covariates (e.g., different higher education institutions and different individual attributes between the groups).


Principal Investigators
Pant, Hans Anand Prof. Dr. (Details) (Educational Research Methods)

Duration of Project
Start date: 11/2015
End date: 02/2020

Last updated on 2021-04-01 at 17:43