空間工作記憶為基的注意力攫取：表徵強度的角色

Abstract

注意力受到維持在工作記憶中的表徵影響，可以藉由工作記憶為基的注意力攫取現象體現。工作記憶為基的注意力攫取研究採用記憶再認與注意力作業的雙作業派典，先是記憶階段維持表徵於工作記憶中，並利用注意力作業偵測受表徵影響的注意力分佈，最後在進行再認記憶作業。回顧過去的研究，一個物體或非空間屬性的工作記憶可以穩定的攫取注意力，但在多於一個記憶項目上則有不一致的結果。空間工作記憶上似乎只有記憶一個位置才能攫取注意力，且此攫取可能需要空間注意力維持在記憶的位置上。本研究驗證是否只有一個空間位置維持於工作記憶時，該位置表徵強度高才能攫取注意力；當四個位置維持於工作記憶時，每一表徵強度弱，必須依賴上而下控制讓空間注意力複誦維持於該表徵才能攫取注意力。採用空間刺激可以降低表徵間的干擾並且排除語音登錄的可能性。本研究以三個實驗探討記憶強度與空間注意力複誦在空間工作記憶為基注意力攫取現象裡扮演的角色。實驗一操弄記憶一個與四個位置，觀察記憶空間的數量影響資源分配的情境下，工作記憶為基的注意力攫取現象。結果發現只有記憶一個位置時可以攫取注意力。實驗二呈現四個位置並利用回溯線索產生不同記憶強度的表徵，以探討記憶強度差異對工作記憶為基注意力攫取的影響。結果回溯線索並無法強化弱表徵，意指弱表徵即使受到空間注意力複誦依然無法攫取注意力。實驗三呈現一個位置，藉由加入無效回溯線索導引空間注意力到非記憶位置上，釐清空間注意力複誦與強表徵是否同時為注意力攫取發生的必要條件。結果顯示強位置表徵攫取注意力，但無法完全排除空間注意力再次複誦記憶位置的可能性。綜合而言，本研究證實只有一個空間位置維持於工作記憶時，強表徵與空間注意力複誦為注意力攫取的重要條件；當記憶四個位置時表徵薄弱即使藉由空間注意力複誦也無法攫取注意力。

The reappearance of the representation stored in working memory captures attention by top-down biasing the competition among external stimuli. However, whether all representations stored in working memory can capture attention remains an issue for investigation. Previous studies show reliable capture effect when only one item is stored in working memory, inconsistent findings when two items are stored in memory, and null results when four items are stored in working memory using objects or non-spatial features. The studies that investigate spatial working memory focus on attention-based rehearsal and the results suggest that one location can capture attention while three locations can be stored in working memory. The goal of this study is to investigate how memory load and attention-based rehearsal can modulate working memory-driven attention capture in the spatial domain. In Experiment 1, the participants were required to remember one or four locations for a later memory test while performing a dot detection task. The results showed attentional capture only in the one-location condition and not in the four-location condition. In Experiment 2, four locations were memorized and two of them were cued sequentially prior to the dot detection display. The results showed null working memory-driven capture effect. That is, a weak representation could not capture attention even after being refreshed by a retro-cue. In Experiment 3, only one location was memorized while the contingency between the retro-cue and the memorized location was decreased by adding the invalid retro-cue condition to examine whether both a strong spatial representation and attention-based rehearsal are necessary to produce the capture effect. The result showed that attention was captured only by the content stored in spatial working memory; the possibility of attention-based rehearsal could not be fully excluded. Taken together, the results from this study suggest that a strong spatial representation stored in working memory can capture attention. When multiple spatial items are encoded for storage in working memory, each item representation is weak so that it cannot capture attention even this representation is reactivated by a spatial retro-cue.