Using working memory intervention to improve math performance: ill-conceived, poorly executed, or just not quite there yet? (Intervenciones en memoria de trabajo para mejorar el rendimiento matemático: ¿mal diseñadas, pobremente ejecutadas, o poco desarrolladas?)

ABSTRACT With over two decades of research showing a close association between working memory (WM) and maths performance, WM training has been suggested as one way to supplement conventional remedial instructions. Although initial findings were promising, recent reviews have found that training typically resulted in WM improvement but no transfer to maths performance. This paper focused on the utility of WM training for maths performance. In addition to a brief review of relevant works, I described my own efforts and insights gained from three iterations of an intervention that produced mixed findings. One of the main questions addressed in this paper was the reasons for robust findings of near but not far transfer. Some of the older studies were beset by poor methodological designs. However, the finding that more recent, better executed studies also found a similar lack of far transfer is troubling. I examined whether the use of WM training for improving maths performance is simply misguided. I also examined the possibility that the extant approaches to intervention may be approaching the problem from the wrong perspective.