Working-memory updating training improved performance on a trained task and a similar untrained task in abstinent methamphetamine inpatients. It did not demonstrate broader benefits for inhibition or task switching. Learning a task is not the same as improving executive function generally.
Zhao and colleagues studied 32 male inpatients with long-term methamphetamine-use histories who were abstinent. Participants received either repeated working-memory updating training or an active control treatment. The training group completed 20 sessions. Both groups took tests before and after treatment assessing updating, interference control, response inhibition and switching.
The training group improved on the practiced task and performed better after treatment on an untrained running-memory task. The latter result is called near transfer because the new task draws on a closely related ability. The authors found strong Bayesian evidence for the updating transfer-task benefit, rather than relying solely on improvement in the practiced exercise.
The investigators did not find training-related benefits on Stroop, flanker, Go/No-Go or switching outcomes. Improvement on the practiced task also did not significantly correlate with improvement on the transfer task. These findings limit an account in which greater training gains necessarily translate into a generalized executive benefit.
The sample was small, and the Bayesian analyses for several broader outcomes were inconclusive rather than strong evidence of no possible effect. Only a limited set of tasks represented each cognitive domain, and no follow-up established how long the observed benefit lasted. The findings also do not demonstrate improved daily functioning or reduced relapse. This study supports the capacity to learn and generalize to a similar task, while leaving the practical clinical value and mechanism unresolved. When evaluating a cognitive-training claim, identify whether the evidence concerns the practiced activity, a similar test, a different cognitive domain or an actual treatment outcome.
See also why different executive tests should not be treated as interchangeable.