Melinda Sabo¹, William Narhi-Martinez, Anna C. Nobre; ¹Yale University

A large body of research has established that attentional selection shapes perception and working memory. Yet its consequences for long-term memory remain less clearly defined. Understanding whether and how internal attention impacts long-term memory retrieval requires a framework for manipulating the orienting of attentional selection between competing contents in long-term memory. The present set of three studies addresses this gap by introducing a new experimental framework. The proposed design includes two major phases: a Learning task and a Selective Retrieval task. During Learning, participants study associations between scenes and two embedded objects. After learning, participants complete a five-minute Tetris game. In the Selective Retrieval task, the same scenes reappear. In Experiments 1a and 1b, a cue appears after scene onset (retrocue). The cue either indicates the object to-be-tested (75% validity) or offers no predictive information. In 25% of the trials, informative cues are invalid: they point to the nontarget item. While valid cues enable the study of attentional benefits, invalid cues allow for the examination of attentional costs. Experiment 2 uses 100% informative and neutral cues presented before scene onset (precues). At the end of the Selective Retrieval trials in all experiments, participants are shown a probe array containing two objects and must indicate which one was associated with the scene. Across all experiments, informative valid cues led to response times benefits during long-term memory retrieval. Additionally, invalid cues were associated with slower response times, suggesting attentional costs. These consistent effects demonstrate that orienting internal spatial attention using retrocues or precues influenced memory performance. Additionally, in Experiment 2, the fast timings of the cues and probe arrays suggested that internal attention may operate directly upon long-term memory representations, unmediated by working memory. Our new design and findings introduce a rigorous approach for studying attentional selection in long-term memory.

Acknowledgements: This research was supported by a Walter Benjamin Fellowship offered by the German Research Foundation (Deutsche Forschungsgemeinschaft; SA 5174/1-1)