This paper presents a critical analysis, from the point of view of a statistician, of the methodology used by Hattie in "Visible Learning," and explains why it must absolutely be called pseudoscience. We first discuss what appears to be the intentions of Hattie's approach. Then we describe the major mistakes in "Visible Learning" before reviewing the set of questions a researcher should ask when investigating studies and surveys based on data analyses, including meta-analyses. We give concrete examples explaining why Cohen's d (the measure of effect size used in "Visible Learning") simply cannot be used as some sort of universal measure of impact. Finally, we propose solutions to better understand and implement studies and meta-analyses in education. [This paper is a forum contribution that appeared in v51 n2 in French. Due to the "positive buzz" it garnered following its publication, the MJE editorial team has made its translation available to our English readers. Translated by Lysanne Rivard.]