The goal of the present study was to test whether the amount of coaching influenced the results of working memory training on both visual and verbal working memory. Additionally, the effects of the working memory training on the amount of progress after specific training in mathematics were evaluated. In this study, 23 children between 9 and 12 years of age with both attentional and mathematical difficulties participated in a working memory training program with a high amount of coaching, while another 25 children received no working memory training. Results of these groups were compared to 21 children who completed the training with a lower amount of coaching. The quality of working memory, as well as mathematic skills, were measured three times using untrained transfer tasks. Bayesian statistics were used to test informative hypotheses. After receiving working memory training, the highly coached group performed better than the group that received less coaching on visual working memory and mathematics, but not on verbal working memory. The highly coached group retained their advantage in mathematics, even though the effect on visual working memory decreased. However, no added effect of working memory training was found on the learning curve during mathematical training. Moreover, the less-coached group was outperformed by the group that did not receive working memory training, both in visual working memory and mathematics. These results suggest that motivation and proper coaching might be crucial for ensuring compliance and effects of working memory training, and that far transfer might be possible.

The goal of this randomized controlled trial was to investigate whether Jungle Memory working memory training (JM) affects performance on working memory tasks, performance in mathematics and gains made on a mathematics training (MT) in school aged children between 9–12 years old (N = 64) with both difficulties in mathematics, as well as attention and working memory. Children were randomly assigned to three groups and were trained in two periods: (1) JM first, followed by MT, (2) MT first, followed by JM, and (3) a control group that received MT only. Bayesian analyses showed possible short term effects of JM on near transfer measures of verbal working memory, but none on visual working memory. Furthermore, support was found for the hypothesis that children that received JM first, performed better after MT than children who did not follow JM first or did not train with JM at all. However, these effects could be explained at least partly by frequency of training effects, possibly due to motivational issues, and training-specific factors. Furthermore, it remains unclear whether the effects found on improving mathematics were actually mediated by gains in working memory. It is argued that JM might not train the components of working memory involved in mathematics sufficiently. Another possible explanation can be found in the training’s lack of adaptivity, therefore failing to provide the children with tailored instruction and feedback. Finally, it was hypothesized that, since effect sizes are generally small, training effects are bound to a critical period in development.

Number sense and working memory contribute to mathematical development throughout primary school. However, it is still unclear how the contributions of each of these predictors may change across development and whether the cognitive contribution is the same for children with and without mathematical difficulties. The aim of the two studies in this paper was to shed light on these topics. In a cross-sectional design, a typically developing group of children (study 1; N = 459, Grades 1-4) and a group with mathematical difficulties (study 2; N = 61, Grades 4-6) completed a battery of number sense and working memory tests, as well as a measure of arithmetic competence. Results of study 1 indicated that number sense was important in first grade, while working memory gained importance in second grade, before predictive value of both predictors waned. Number sense and working memory supported mathematics development independently from one another from Grade 1. Analysis of task demands showed that typically developing children rely on comprehension and visualization of quantity-tonumber associations in early development. Later in development, pupils rely on comparing larger numerals and working memory until automatization. Children with mathematical difficulties were less able to employ number sense during mathematical operations, and thus might remain dependent on their working memory resources during arithmetic tasks. This suggests that children with mathematical difficulties need aid to employ working memory for mathematics from an early age to be able to automatize mathematical abilities later in development.

Wellicht gaat u ook weleens op vakantie naar Frankrijk. Bij het passeren van de grens is de overgang abrupt: de verkeersborden hebben een andere kleur, het aanbod bij de wegrestaurants is anders, en vooral: de taal is anders. Het klinkt vagelijk bekend, maar toch bevreemdend. Geen zorgen, u bent niet beland in de faits divers van dit tijdschrift. Het punt is dat wij denken dat het met neuropsychologische diagnostiek in de klinische praktijk vergelijkbaar is gesteld: de taal klinkt bekend, maar wat wordt gezegd is tegelijk exotisch en complex. Sommigen verhouden zich wellicht tot neuropsychologische diagnostiek op dezelfde manier als tot het exotische eten op vakantie: ik neem toch maar een potje pindakaas mee. Dat potje pindakaas is dan toch vaak ons vertrouwde instrument: de intelligentietest. Maar met de introductie van de WISC-V lijkt het potje pindakaas zelf een exoot geworden. Allemaal nieuwe indexscores, die ineens lijken op neuropsychologische nomenclatuur -we hebben menig collega horen verzuchten dat het zo ingewikkeld is geworden. Wij denken echter dat de intelligentietest volwassener geworden is en dat de denkbeeldige grens tussen 'intelligentieonderzoek' en 'neuropsychologisch onderzoek' doorbroken kan worden.De nieuwe WISC is geïnspireerd op het Cattell-Horn-Carroll-(CHC-) model. Het voert te ver om dit model hier uit de doeken te doen (zie Bos et al. 2016 voor een overzicht); belangrijk is dat dit model duidelijk maakt dat er geen scheiding is tussen taken in intelligentietests en neuropsychologische taken. Zeker zijn er verschillen tussen het CHC-en het neuropsychologische denkkader, waarbij het