Voormann Anne, Miller Jeff
Department of Psychology, University of Freiburg, DE.
University of Otago, NZ.
J Cogn. 2024 Sep 3;7(1):68. doi: 10.5334/joc.395. eCollection 2024.
A common finding across numerous response time (RT) paradigms is that the mean RT in one trial depends strongly on the characteristics of the immediately preceding trial. Although such sequential effects have usually only been considered within each single paradigm in isolation from the others, there are important similarities across paradigms between the theoretical accounts of these effects. However, so far there has been no systematic comparison of sequential effects across paradigms. To investigate the possible relationships between sequential effects in different paradigms, we conducted an experiment examining sequential effects in visual search, two-choice RT, interference, and task-switching paradigms, using methods designed to maximize the similarity of stimuli and responses across paradigms. Detailed analyses of the observed RT distributions were carried out using both descriptive (e.g., ex-Gaussian) and process-oriented (e.g., diffusion models) methods. The results reveal significant empirical similarities and differences between the sequential effects observed across different paradigms, and in some cases even across different conditions within a single paradigm. Furthermore, the sequential effects are more similar to one another for some pairs of paradigms than for others. These results imply that some cognitive processes eliciting sequential effects are shared across paradigms while others seem to be paradigm-specific.
在众多反应时(RT)范式中一个常见的发现是,一次试验中的平均反应时在很大程度上取决于紧接在前的试验的特征。尽管这种序列效应通常仅在每个单一范式内被单独考虑,与其他范式无关,但在这些效应的理论解释方面,不同范式之间存在重要的相似之处。然而,到目前为止,尚未对不同范式间的序列效应进行系统比较。为了研究不同范式中序列效应之间可能的关系,我们进行了一项实验,使用旨在使不同范式间的刺激和反应最大化相似性的方法,考察了视觉搜索、二选一反应时、干扰和任务切换范式中的序列效应。使用描述性(如前高斯分布)和面向过程(如扩散模型)的方法对观察到的反应时分布进行了详细分析。结果揭示了在不同范式中观察到的序列效应之间存在显著的实证相似性和差异,在某些情况下甚至在单一范式内的不同条件之间也存在差异。此外,某些范式对之间的序列效应比其他范式对之间的序列效应彼此更相似。这些结果意味着,一些引发序列效应的认知过程在不同范式中是共享的,而另一些似乎是特定于范式的。