Reply to 'Efficiency and capacity mechanisms can coexist in cognitive training'
Reply to 'Efficiency and capacity mechanisms can coexist in cognitive training'
复制标题
回复“认知训练中效率与能力机制可以共存”
DOI：
10.1038/s44159-022-00147-8
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Von Bastian C
中科院分区：
文献类型：
--
作者：
Von Bastian C
在我们的综述中（von Bastian， C.等）。训练诱发认知改变的机制。心理学报，1,30-41；2022) 1、探讨认知训练和迁移的能力效率模型来解释认知训练的效果。我们很高兴其他作者开始使用我们的框架来系统化和解释研究结果。我们的文章主要关注人类行为和脑成像研究，但我们也意识到，所有来源的证据，包括动物和基因研究，都应该被考虑在内，以建立一门全面的认知变化科学。我们感谢Zhang和Sauce为我们带来了这些领域的证据(Zhang, d.w。效率机制和能力机制在认知训练中可以共存。纳特，精神科牧师。https://doi。org/10.1038/s44159 - 022 - 00146 - 9;2023) 2。Zhang和Sauce强调产能和效率的变化并不一定是相互排斥的，而是可以共存的。事实上，这一说法与我们的假设是一致的，并且在原文章中也有陈述。具体来说，Zhang和Sauce声称一些与训练相关的变化，特别是在工作记忆方面，可以归因于容量机制。然而，目前的人类大脑成像文献并没有提供令人信服的证据证明能力的变化。利用脑成像数据进行的工作记忆训练研究表明，在进行工作记忆训练后，大脑活动既有增加的，也有减少的。然而，元分析证据表明，这些激活模式的变化很可能仅仅反映了相同核心工作记忆网络中激活模式的重新分配，而不是额外资源的补充。关键的是，血氧水平依赖性（BOLD）信号的增加和减少并不直接映射到特定的潜在神经生物学机制。目前仍需要进行研究，为确定大脑激活模式的变化与认知能力和效率的变化之间的关系奠定基础。此外，要确定训练——或者Zhang和Sauce提到的其他可能影响能力的因素，比如儿童时期的发展——如何影响认知能力和效率的神经相关关系，需要的不仅仅是评估整体的行为表现。例如，Zhang和Sauce指出，额顶叶激活的增加与工作记忆的改善有关。然而，工作记忆训练研究几乎只报告了整体的工作记忆表现。整体表现的变化可以反映工作记忆容量、效率或两者的变化。例如，总体性能可以通过提高效率来提高，例如通过改变战略方法或增加基于熟悉度的处理5。计算建模方法可以帮助解开工作记忆表现的组成部分（例如，工作记忆中激活的表征的数量和质量）。通过使用计算方法来形式化训练引起的工作记忆容量和效率的变化，推导出的参数估计可以与BOLD信号的变化相关联。因此，这种方法可以确定激活模式的变化是否反映了容量、效率或两者的变化。我们同意Zhang和Sauce的观点，即多巴胺D1受体密度和敏感性在小鼠认知需求和训练下的变化是有趣的，并且说明了认知系统的显著灵活性。这些发现也强调了动物和基因研究与人类认知训练的相关性。
In our Review (von Bastian, C. et al. Mech-anisms underlying training-induced cognitive change. Nat. Rev. Psychol. 1, 30–41; 2022) 1, we discussed the capacity–efficiency model of cognitive training and transfer to explain effects of cognitive training. We are excited that other authors are starting to use our framework to systematize and interpret findings. Our article focused on human behavioural and brain imaging studies, but we appreciate that evidence from all sources, including animal and genetic studies, should be considered to build a comprehensive science of cognitive change. We thank Zhang and Sauce for bringing evidence from these domains to our attention (Zhang, D.-W. & Sauce, B. Efficiency and capacity mechanisms can coexist in cognitive training. Nat. Rev. Psychol. https://doi. org/10.1038/s44159-022-00146-9; 2023) 2. Zhang and Sauce emphasize that changes in capacity and efficiency are not necessarily mutually exclusive but can coexist. Indeed, this statement is in line with our assumptions and was stated in the original article. Specifically, Zhang and Sauce claim that some training-related changes, specifically in working memory, can be attributed to the capacity mechanism. However, the current human brain imaging literature does not provide convincing evidence for changes in capacity. Working memory training studies with brain imaging data have shown both increases and decreases in brain activation after working memory training 3. Still, metaanalytic evidence suggests that these changes in activation patterns are likely to reflect mere redistributions of activation patterns within the same core working memory networks rather than a recruitment of additional resources 4. Critically, increases and decreases in blood-oxygen level-dependent (BOLD) signal do not map directly onto specific underlying neurobiological mechanisms. Research is still needed to lay the groundwork to identify how changes in brain activation patterns relate to changes in cognitive capacity and efficiency 3.Furthermore, determining how training—or the other factors mentioned by Zhang and Sauce 2 that might influence capacity, such as development in childhood—influence neural correlates of cognitive capacity and efficiency requires moving beyond assessing overall behavioural performance. For example, Zhang and Sauce 2 make the point that increases in frontoparietal activation correlate with improvements in working memory. However, working memory training studies almost exclusively report overall working memory performance. Changes in overall performance can reflect changes in working memory capacity, efficiency or both. For example, overall performance can be boosted by increased efficiency, such as through changes in strategic approach or increased familiarity-based processing 5. Computational modelling approaches can help to disentangle components of working memory performance (for example, the quantity and quality of representations activated in working memory 6). By using computational approaches to formalise training-induced changes in working memory capacity and efficiency, the derived parameter estimates can then be correlated with changes in BOLD signal. This method could therefore determine whether changes in activation patterns reflect changes in capacity, efficiency or both. We agree with Zhang and Sauce 2 that changes in dopamine D1 receptor density and sensitivity in response to cognitive demands 7 and training 8 in mice are intriguing and speak to a remarkable flexibility of the cognitive system. These findings also highlight the relevance of animal and genetic studies for the human cognitive training …