Negotiating the information-rich sensory world often requires the concurrent management of multiple tasks. Despite this, humans are thought to be poor at multitasking due to the processing limitations of frontoparietal and subcortical (FP-SC) brain regions. Although training is known to improve multitasking performance, it is unknown how the FP-SC system functionally changes to support improved multitasking. To address this, we characterized the FP-SC changes that predict multitasking-training outcomes using an individual differences approach. Participants (N=100) performed single and multiple tasks in pre- and post-training magnetic resonance imaging (fMRI) sessions interspersed by either a multitasking or an active-control training regimen. Multivoxel pattern analyses (MVPA) revealed that training induced multitasking improvements were predicted by divergence in the FP-SC blood-oxygen-level-dependent (BOLD) response patterns to the trained tasks. Importantly, this was only observed for participants who completed multitasking training, and not for the control group. Therefore the FP-SC system supports multitasking behaviour by segregating constituent task representations.

The prospect of enhancing cognition is undoubtedly among the most exciting research questions in current cognitive neuroscience and psychology. Despite the tremendous ramifications of this line of work, cognitive training studies often suffer from limitations or shortcomings, in design or in their statistical analyses. Here, we focus on three pervasive flaws in cognitive interventions: (i) the typical use of correlated gain scores as evidence for transfer; (ii) claims of transfer on the basis of improvements in a single task; and (iii) failures to adequately correct for multiple comparisons in multi-factorial experiments. In each case, we present the mechanisms and magnitude of the issue with repeated stochastic processes. Based on these simulations, we then discuss potential remedies and best practices in the design and analysis of training interventions. Finally, we reflect on the current state of affairs in the field of cognitive enhancement, weighing promises and disappointments, and suggest potential directions to improve our understanding of the plasticity of human cognition.

A biomarker pinpointing who will remit and who will show persistent Attention Deficit/Hyperactivity Disorder (ADHD) symptoms into adulthood remains elusive. The Halperin and Schulz model posits that attenuation of ADHD symptoms is dependent upon the development of the prefrontal cortex during adolescence, to compensate for ongoing ADHD-related dysfunction within the basal ganglia. To test this model, a variable under conscious control - identifying targets in a vigilance task -and a variable under less conscious control - response time variability - in adults with and without a childhood diagnosis of ADHD were measured. Performance on both the Fixed and Random Sustained Attention to Response tasks (SART) were compared between controls (n =26) and adults with either persistent (n = 12) or remitted (n = 15) ADHD symptoms in adulthood stemming from a childhood diagnosis of the disorder. The remitters benefitted from the arousing quality of the Random in contrast with the Fixed SART, as demonstrated by a significant reduction in variability of responding, measured by the standard deviation of response time (RT), slow frequency RT variability and Sigma measures, and in terms of the number of omission errors made. The persisters, in contrast, did not benefit from the arousing nature of the Random SART, but showed a tendency to perform more poorly over the course of the task, as measured by fast frequency RT variability. These findings provide some evidence to support Halperin and Schulz’s sub-cortical theory of ADHD.

Behavioural training can lead to task-specific improvements in performance. Application of transcranial direct current stimulation (tDCS) over relevant areas of cortex during training can further facilitate training benefits. Currently, it is not known whether training combined with tDCS can lead to transferable performance gains for untrained tasks. Nor is it known how tDCS protocols might interact with task demands to influence information processing in the brain. To address these issues, we applied anodal, cathodal or sham tDCS to the left prefrontal cortex of 54 human participants (18 per group) while they trained on a simple response selection task, and assessed its impact on an untrained spatial attention task (visual search). Anodal tDCS applied during training resulted in significantly greater performance gains for both the response selection and visual search tasks than cathodal or sham tDCS. We applied response time modelling to the behavioural data obtained under the three stimulation conditions, using the linear ballistic accumulator framework, which identifies the best model fit via sequentially sampling through multiple possible models of latent components of task performance, namely non-decision processes (stimulus encoding and response execution), response thresholds, and the rate and start point of evidence accumulation. This approach revealed that performance enhancement under anodal stimulation for both the response selection and visual search tasks could be attributed to an increase in the rate of evidence accumulation.