Intra-individual variability on a computer-based working memory task was examined among 25 children/adolescents with ADHD and 24 typically developing peers

This study investigated whether intra-individual variability (IIV)—the moment-to-moment fluctuations in performance—is a hallmark feature of ADHD when applied to working memory (WM) tasks. While IIV is well-documented in response inhibition tasks, the authors sought to determine if these performance fluctuations persist in more complex cognitive tasks and whether they can be better captured using advanced statistical modeling.

The researchers compared 25 children and adolescents with ADHD to 24 typically developing peers using the Visual Serial Addition Task (VSAT). This task required participants to perform serial addition of visually presented digits, manipulating both WM load and event rate (the speed of stimulus presentation). To analyze reaction time (RT) data, the authors employed an ex-Gaussian distribution model. Unlike traditional mean RT, this approach separates RT into three parameters: mu (the mean of the normal component), sigma (variability of the normal component), and tau (the positive skew representing extremely slow responses).

The ADHD group demonstrated significantly higher IIV than the control group, specifically in the sigma and tau parameters of the ex-Gaussian model. This indicates that children with ADHD are not necessarily slower in their modal performance (mu), but rather suffer from more frequent, excessively slow responses (tau) and greater variability in their baseline performance (sigma). Furthermore, ADHD participants showed a linear increase in omission errors as WM load increased, whereas controls showed a different pattern. Interestingly, the event rate did not significantly influence the group differences, and ADHD participants performed with similar accuracy to controls when they did provide a response.

This study reinforces the idea that IIV is a fundamental, ubiquitous feature of ADHD that extends beyond simple inhibition tasks into the domain of working memory. By using ex-Gaussian modeling, the authors demonstrate that ADHD-related deficits are better characterized as fluctuations in attention rather than a generalized slowing of cognitive processing. This suggests that future diagnostic and therapeutic efforts should focus on these performance inconsistencies as a potential endophenotype for the disorder.