Two General Types of Scientific Approaches
In the educational field, it is traditional to consider two general types of scientific approaches. The first, called “hypothetical-deductive”, starts from theories or laws that are as general as possible, drawing predictions from them that are compared with data from experience. The second, called “explanatory”, starts from what is observed and questions its determinants by attempting to trace the causal chain. These two approaches are linked, and there is a back-and-forth flow between them. In both approaches, researchers use complex causal systems involving several explanatory factors revealed through statistical analysis. Both approaches are legitimate in the field of education.
The “hypothetico-deductive” approach is favoured in research in interventional cognitive sciences. This type of research mainly uses the so-called experimental method. Its main objective is to “administer the evidence”, i.e., to show that a factor (for example, a teaching method or learning technique) is indeed the main cause of the appearance of observed behaviour (e.g., better learning performance), “all things being equal” . To acquire the certainty that this causal relationship is univocal, “interventions” (also called training, methods, or pedagogies) must be organized in the classroom to control all other factors likely to influence the observed performance (school level, socio-professional category, etc.) as much or as well as possible. This is a very difficult task given the complexity of the educational system.
The Measure of Effectiveness of Pedagogy: Three Levels of Evidence
To measure the effectiveness of an “intervention” or “pedagogy”, three levels of methodological rigour can be considered in the type of experimental protocol used by researchers . Two main criteria can be used: presence or not of a “control group” (CG) and a random assignment of students (i.e., how students are selected and assigned to the experimental or control groups).
In this perspective, level A corresponds to studies using “intervention-only” protocols. The intervention is conducted with a single group of students, and pre- and post-intervention measures are carried out to evaluate the impact of the intervention but without comparison to a reference control group. This type of protocol makes it possible to test the “feasibility” of the intervention but prevents the interpretation of the results since no control group can ensure that students not benefitting from the intervention do not show the same evolution on the measured variables. In this review, we have chosen not to present the results of the Level A studies given their difficulties of interpretation.
Content Prepared by: Pratheek
Contact no: +9198468 08283
