本研究的目的是探讨当今虚拟助手使用的两种重要的社交线索如何影响人们的长时记忆效果。这两种社交线索分别是拟人声音和基于屏幕显示的文本。本研究为组间实验，采用行为实验（N = 61），将被试分为：仅阅读文本组，仅聆听虚拟助手的拟人声音的播报组和边听边看组。运用学习-再认范式进行长时记忆效果的测量。实验结果表明，三组之间拥有显著的差异，通过仅聆听拟人声音播报组的长时记忆效果最好，并且显著优于边听边看组的长时记忆效果，边听边看组的记忆效果最差。该实验一定程度上支持了LC4MP模型。

The aim of this study was to test how social cues used by virtual assistants can influence long-term memory effect. These two social cues refer to the anthropomorphized voice via Siri and text displayed on a screen. The hypotheses were tested in an experiment (N = 61) with between-subjects design. The participants were divided into three groups which are reading text group, listening to Siri group and reading while listening group. The recognize paradigm was used to measure the effect of long-term memory. The experimental results indicate significant differences between the three groups, with the best long-term memory effects achieved by the group that only listening to Siri, which was significantly better than the group that listening and watching simultaneously, and the worst memory effects were observed in the group that listening while watching. This experiment supports the LC4MP model to some extent.