Many different operationalisations of academic success are used in the literature, but hardly any study or project can be found comparing the outcomes by identifying factors influencing academic success using different definitions of the construct. This study fills this gap by systematically analysing different operationalisations of academic success using secondary data of three bachelor study programmes at the University of Vienna. Both classical regression analysis and random forest models were used to investigate the impact the methodological approach may have on the outcomes of the studies.
Results demonstrate that the individual factors influencing academic success depend on three dimensions: First, they may differ for each study programme. Second, they may differ depending on the chosen definition of academic success. And third, they may differ depending on the applicable statistical model. This article highlights the importance of viewing the construct academic success as a multidimensional construct that has to be investigated within institutions and within study programmes. It contributes to the understanding of institutional study success in Austria and may also have broader implications.