Schubert Anna-Lena, Löffler Christoph, Sadus Kathrin, Göttmann Jan, Hein Johanna, Schröer Pauline, Teuber Antonia, Hagemann Dirk
Department of Psychology, University of Mainz, Mainz, Germany.
Department of Psychology, University of Mainz, Mainz, Germany; Institute of Psychology, Heidelberg University, Heidelberg, Germany.
Cognition. 2023 Jul;236:105438. doi: 10.1016/j.cognition.2023.105438. Epub 2023 Apr 12.
There is a broad consensus that individual differences in working memory capacity (WMC) are strongly related to individual differences in intelligence. However, correlational studies do not allow conclusions about the causal nature of the relationship between WMC and fluid intelligence. While research on the cognitive basis of intelligence typically assumes that simpler lower-level cognitive processes contribute to individual differences in higher-order reasoning processes, a reversed causality or a third variable giving rise to two intrinsically uncorrelated variables may exist. In two studies (n = 65, n = 113), we investigated the causal nature of the relationship between WMC and intelligence by assessing the experimental effect of working memory load on intelligence test performance. Moreover, we tested if the effect of working memory load on intelligence test performance increased under time constraints, as previous studies have shown that the association between the two constructs increases if intelligence tests are administered with a strict time limit. We show that working memory load impaired intelligence test performance, but that this experimental effect was not affected by time constraints, which suggests that the experimental manipulations of working memory capacity and processing time did not affect the same underlying cognitive process. Using a computational modeling approach, we demonstrated that external memory load affected both the building and maintenance of relational item bindings and the filtering of irrelevant information in working memory. Our results confirm that WMC causally contributes to higher-order reasoning processes. Moreover, they support the hypothesis that working memory capacity in general and the abilities to maintain arbitrary bindings and to disengage from irrelevant information in particular are intrinsically related to intelligence.
人们普遍认为，工作记忆容量（WMC）的个体差异与智力的个体差异密切相关。然而，相关性研究无法得出关于WMC与流体智力之间关系的因果性质的结论。虽然关于智力认知基础的研究通常假设，更简单的低级认知过程会导致高阶推理过程中的个体差异，但可能存在反向因果关系或一个第三变量导致两个本质上不相关的变量。在两项研究中（n = 65，n = 113），我们通过评估工作记忆负荷对智力测试表现的实验效应，研究了WMC与智力之间关系的因果性质。此外，我们测试了在时间限制下，工作记忆负荷对智力测试表现的影响是否会增加，因为先前的研究表明，如果在严格的时间限制下进行智力测试，这两个构念之间的关联会增加。我们发现，工作记忆负荷会损害智力测试表现，但这种实验效应不受时间限制的影响，这表明对工作记忆容量和处理时间的实验操作并未影响相同的潜在认知过程。使用计算建模方法，我们证明外部记忆负荷会影响关系项绑定的构建和维持以及工作记忆中无关信息的过滤。我们的结果证实，WMC对高阶推理过程有因果贡献。此外，它们支持这样的假设，即一般的工作记忆容量，特别是维持任意绑定和脱离无关信息的能力，与智力本质上相关。
