Abstract：
Proactive interference (PI) occurs when previously learned information impairs memory for more recently learned information. Most PI studies have employed verbal stimuli, while the role of PI in visual working memory (VWM) has had relatively little attention. In the verbal domain, Johansson and colleagues (2018) found that pupil diameter - a real-time neurophysiological index of cognitive effort - reflects the accumulation and resolution of PI. Here we use a novel, naturalistic paradigm to test the behavioral and pupillary correlates of PI resolution for what-was-where item-location bindings in VWM. Importantly, in our paradigm, trials (PI vs. no-PI condition) are mixed in a block, and participants are naïve to the condition until they are tested. This design sidesteps concerns about differences in encoding strategies or generalized effort differences between conditions. Across three experiments (N = 122 total) we assessed PI\'s effect on VWM and whether PI resolution during memory retrieval is associated with greater cognitive effort (as indexed by the phasic, task-evoked pupil response). We found strong support for PI\'s detrimental effect on VWM (even with our spatially distributed stimuli), but no consistent link between interference resolution and effort during memory retrieval (this, even though the pupil was a reliable indicator that higher-performing individuals tried harder during memory encoding). We speculate that when explicit strategies are minimized, and PI resolution relies primarily on implicit processing, the effect may not be sufficient to trigger a robust pupillometric response.
摘要：
当先前学习的信息损害了最近学习的信息的记忆时,发生主动干扰(PI)。大多数PI研究都采用了言语刺激，而PI在视觉工作记忆（VWM）中的作用相对较少受到关注。在语言领域，Johansson及其同事（2018）发现，瞳孔直径-认知努力的实时神经生理指数-反映了PI的积累和分辨率。这里我们用小说,自然范式，用于测试VWM中什么地方的项目位置绑定的PI分辨率的行为和瞳孔相关性。重要的是,在我们的范式中，试验(PI与无PI条件)混合在一个块中，参与者在接受测试之前都对这种情况很幼稚。此设计回避了对编码策略差异或条件之间广义努力差异的关注。在三个实验中（总共N=122），我们评估了PI对VWM的影响，以及在记忆检索过程中PI分辨率是否与更大的认知努力相关（按相位索引，任务诱发的瞳孔反应)。我们发现PI对VWM的有害影响(即使我们的空间分布刺激)得到了强有力的支持，但在内存检索过程中，干扰分辨率和工作量之间没有一致的联系(这，即使瞳孔是一个可靠的指标，表现较高的人在记忆编码期间更加努力地尝试）。我们推测，当显式策略最小化时，PI解析主要依赖于隐式处理，该效应可能不足以触发稳健的瞳孔测量反应.
