O8-7 Do physical activity, motor skills, and attention capacities predict the academic achievement of primary school children from disadvantaged neighbourhoods? Elaboration of inference conditional trees
Résumé
BackgroundIn the 6-10 year period, physical activity (PA) and motor skillshave a positive impact on cognitive development, which inturn act on academic performance. In order to betterunderstand these links, a study was conducted to explorethese relationships for disadvantaged children between the agesof 6 and 10.MethodsChildren from two primary schools located in a disadvantagedneighbourhood (Tarbes, France) participated in the study.Variables were measured at 5 measurement times over 3academic school years (2016; 2017; 2018). PA was measured byaccelerometry over the whole day and weekend (MVPAF,MVPAWF). Motor skills were also assessed with a shuttle run test (NAV), a standing broad jump test (SBJ) and a tapping test(TT). A cardiorespiratory Shuttle Run test 20m (AR) was alsocarried out. Attentional capacities were measured with acomputer-based Flanker Task: the Total Reaction Rime (RTT)of the correct answers (ms) for each child was collected. Finally,the children’s score academic performance in French Languageand Mathematics were collected. To study the relationshipsbetween the different variables, conditional inference treesincluding all these variables were performed with R software.Two trees were generated having as target variables respectivelyFrench language (FR) and Mathematics (MAT).ResultsFrench language (FR) was predicted first by the age of thechildren (p > 0.001), but as well for the youngest and the oldestchidldren, FR was predicted by performance on the TT(p > 0.001) and RTT (p > 0.001): the children who performbetter on the TT test and have a lower RTT are those whoobtain the best results in French language. Mathematics(MAT) was explained by these same variables as for FR(p > 0.001). The evaluation of these two conditional inferencetree by the Pseudo R-square were respectively 0.13 and 0.11.DiscussionThese two conditional inference trees revealed that Frenchlanguage and Mathematics were predicted by attentionalcapacities (RTT), by motor variables (TT, NAV and SBJ).Finally, these models obtained were non-linear, complex andhighlight different profiles of children. Although theserelationships are documented in the literature (Diamond,2002), this study confirms this from a longitudinal perspective,which is rarely used.