The Effect of Computerised Adaptive Cognitive Training on Executive Functioning, Fluid Intelligence, and Academic Achievement in Grade 3 and Grade 5 Students: A Randomised Clinical Trial

Abstract

Adaptive computerised cognitive training programs are used in an effort to improve the cognitive functioning of children. Although the results of these programs have been mixed, there have also been significant results that hold promise for some form of adaptive computerised cognitive training program, especially around the training of executive processes, also known as executive function (EF). This study tested the efficacy of a computerised EF training program. It explored whether a school-based, theoretically-aligned EF training program could improve EF as measured by latent factor scores (near-transfer) and improve academic achievement, fluid intelligence and academic behaviours (far-transfer). A number of ancillary outcomes were considered, including using an improvement expectancy measure and using a measure of implicit theory of intelligence to determine if it was a potential moderator for training gains. Normally-functioning primary school children from grades 3 and 5 (Mean age =10 years) were assigned randomly to the EF training program (n = 50) or an active control group (n = 55). This research program found no evidence for improvement on either near- or far-transfer outcome measures as a result of the EF training program. Although there was some indication of potential moderating effects of the participant’s implicit theory of intelligence, the clear lack of linearity and small sample sizes involved in this three-way interaction analysis reduce the generalisability of this finding. Exploratory results showed some weak evidence that those participants in the intervention who improved more on the working memory measures may have a more malleable view of intelligence than those who did not improve. This study provides a unique contribution to the extant literature by rigorously assessing the efficacy of cognitive training through the use of randomly allocated groups, an active control group, and examining near-transfer using a latent variable approach.