Test–retest reliability of the N400 component in a sentence-reading paradigm

The N400 component of the event-related potential is considered an index of semantic processing and therefore may be an ideal biomarker of semantic system disorders or individual differences. To this purpose, it is necessary to assess its test–retest reliability. Only one previous study has addressed this question, reporting good test–retest reliability (r = 0.85). However, that study had used a word-pair priming paradigm, which differs in many respects from the more typical and ecologically valid sentence-reading. The present study surveys test–retest reliability of the N400 in a sentence-reading paradigm. The best value obtained was r = 0.63, implying a relatively poor test–retest reliability. Crucial factors for this result may be the long interval between context and critical word as well as more complex contexts in sentence-reading paradigms. These factors might make the N400 effects in sentences more vulnerable to linguistic and non-linguistic factors increasing the variance across sessions.

事件相关电位（event-related potential）的N400成分被视为语义加工的指标，因此或可成为语义系统障碍或个体差异的理想生物标志物。基于此，有必要对其重测信度进行评估。此前仅有一项研究探讨了这一问题，其报告的重测信度良好（r=0.85）。然而，该研究采用的是词对启动范式，该范式在诸多方面与更具典型性且生态效度更高的句子阅读范式存在差异。本研究对句子阅读范式下N400的重测信度进行了考察，得到的最优相关系数为r=0.63，表明其重测信度相对较差。导致这一结果的关键因素可能是句子阅读范式中语境与关键词之间的较长间隔，以及更为复杂的语境设置。这些因素或使句子中的N400效应更易受语言及非语言因素影响，进而增大了不同测试时段间的变异量。