Faced with the proliferation of systems aimed at improving the success of students in the first cycle of higher education, research work in education and training sciences has multiplied. However, it remains difficult to have a global view of the effects of these systems: what impact do they have? to what extent? under what conditions? In this article, we discuss the interest of metaanalysis in answering these questions. Widespread in medical research, this method consists of re-exploiting data from previous research in order to propose a critical review as well as an evaluative combination to identify a central trend. In the field of education and training sciences, this is a method that is still little used. However, it has many strengths while not being without its limitations (heterogeneity of the data used, methodological quality of the studies, etc.). If the implementation of such a method requires determining inclusion criteria on the primary studies used, it also requires a strong definition of the object studied, i.e. the devices, their effects and the contexts in which they are deployed. By showing the plurality of possible definitions, effects and contexts in which these devices are deployed, this article proposes a methodological reflection for the implementation of future research programs using meta-analysis to study student success support devices. In response to the growing number of measures aimed at improving student success in the first cycle of higher education, research in the educational sciences has proliferated. However, it remains difficult to gain an overall view of the effects of these schemes: what impact do they have? to what extent? under what conditions? In this article, we discuss the value of metaanalysis in answering these questions. Widely used in medical research, this method involves reusing data from previous research to propose a critical review and an evaluative combination to identify a central trend. In the field of educational science, this method is still little used. It has many strengths, but is not without its limitations (heterogeneity of data mobilized, methodological quality of studies, etc.). If the implementation of such a method requires the determination of inclusion criteria for the primary studies mobilized, it also implies the need for a strong definition of the object studied, i.e. the systems, their effects and the contexts in which they are deployed. By showing the plurality of possible definitions, effects and contexts in which these schemes are deployed, this article proposes a methodological reflection for the implementation of future research programs mobilizing meta-analysis to study student success support schemes. [ABSTRACT FROM AUTHOR]