ANOVA results for manipulation check.

Despite early promise, cognitive training research has failed to deliver consistent real-world benefits and questions have been raised about the experimental rigour of many studies. Several meta-analyses have suggested that there is little to no evidence for transfer of training from computerised tasks to real-world skills. More targeted training approaches that aim to optimise performance on specific tasks have, however, shown more promising effects. In particular, the use of inhibition training for improving shoot/don’t-shoot decision-making has returned positive far transfer effects. In the present work, we tested whether an online inhibition training task could generate near and mid-transfer effects in the context of response inhibition tasks. As there has been relatively little testing of retention effects in the literature to date, we also examined whether any benefits would persist over a 1-month interval. In a pre-registered, randomised-controlled trial, participants (n = 73) were allocated to either an inhibition training programme (six training sessions of a visual search task with singleton distractor) or a closely matched active control task (that omitted the distractor element). We assessed near transfer to a Flanker task, and mid-transfer to a computerised shoot/don’t-shoot task. There was evidence for a near transfer effect, but no evidence for mid-transfer. There was also no evidence that the magnitude of training improvement was related to transfer task performance. This finding adds to the growing body of literature questioning the effectiveness of cognitive training. Given previous positive findings, however, there may still be value in continuing to explore the extent to which cognitive training can capitalise on near or mid-transfer effects for performance optimisation.

尽管认知训练研究早期曾展现出可观前景，但始终未能产出稳定的现实世界效益，且多项研究的实验严谨性备受质疑。多项元分析（meta-analyses）指出，几乎不存在证据表明计算机化训练任务可向现实技能实现训练迁移（transfer of training）。不过，旨在优化特定任务表现的更具针对性的训练方法，已展现出更为积极的效应。其中，利用抑制训练改善射击/不射击决策的相关研究，便取得了正向远迁移效应。本研究旨在验证在线抑制训练任务能否在反应抑制（response inhibition）任务场景下产生近迁移与中度迁移效应。鉴于当前学界对留存效应的相关测试仍较为匮乏，我们同时考察了训练收益是否会在1个月的间隔后依然留存。在一项预注册（pre-registered）的随机对照试验（randomised-controlled trial）中，73名被试被随机分配至两组：抑制训练组（完成6次包含单例干扰项的视觉搜索任务训练），以及条件高度匹配的主动对照组（仅移除了任务中的干扰项元素）。我们分别以侧翼任务（Flanker task）评估近迁移效应，以计算机化射击/不射击任务评估中度迁移效应。结果显示，近迁移效应存在显著证据，但未发现中度迁移效应的相关证据。此外，训练提升幅度与迁移任务的表现并无关联。这一发现进一步充实了质疑认知训练有效性的相关研究成果。不过，结合此前的正向研究结论，继续探索认知训练可在多大程度上利用近/中度迁移效应优化表现，仍具备一定研究价值。