This meta-analysis synthesized 35 English studies (130 effect sizes, N = 1,981) that employed online tasks to investigate the processing of multiword sequences (MWSs). We examined (a) to what extent MWSs enjoy a processing advantage over novel word combinations; (b) how such a processing advantage is moderated by statistical regularities (i.e., phrasal frequency, association strength), MWS type, and explicitness of experimental tasks; and (c) whether such moderating patterns differ between L1 speakers and L2 speakers. The results confirmed the processing advantage for most subtypes of MWSs, with effect sizes ranging from small to medium. For L1 speakers and L2 speakers, the processing advantage of MWSs was found across the continuum of phrasal frequency and association strength and varied. Interestingly, task explicitness moderated the processing advantage of MWSs but only for L2 speakers. Taken together, our results shed light on the understanding of MWSs as well as directions for future research.

这项荟萃分析综合了 35 项英语研究（130 个效应大小，N = 1,981），这些研究利用在线任务来研究多词序列 (MWS) 的处理。我们研究了 (a) MWS 在多大程度上比新颖的单词组合享有处理优势； (b) 这种处理优势如何通过统计规律（即短语频率、关联强度）、MWS 类型和实验任务的明确性来调节； (c) L1 说话者和 L2 说话者之间的调节模式是否不同。结果证实了大多数 MWS 亚型的处理优势，效应大小从小到中等不等。对于 L1 说话者和 L2 说话者，MWS 的处理优势在短语频率和关联强度的连续体中被发现并且各不相同。有趣的是，任务明确性调节了 MWS 的处理优势，但仅限于 L2 说话者。总而言之，我们的结果揭示了对 MWS 的理解以及未来研究的方向。